Hospital Medicine

I have been in the business of medicine for more than 15 years and I will never forget the initial surge of the COVID-19 pandemic in Massachusetts.

As a hospitalist, I admitted patients infected with COVID-19, followed them on the floor, and, since I had some experience working in an intensive care unit (ICU), was assigned to cover a “COVID ICU.” This wing of the hospital used to be a fancy orthopedic floor that our institution was lucky enough to have. So began the most life-changing experience in my career as a physician.

In this role, we witness death more than any of us would care to discuss. It comes with the territory, and we never expected this to change once COVID hit. However, so many patients succumbed to this disease, especially during the first surge, which made it difficult to handle emotionally. Patients that fell ill initially stayed isolated at home, optimistic they would turn the corner only to enter the hospital a week later after their conditioned worsened. After requiring a couple of liters of supplemental oxygen in the emergency room, they eventually ended up on a high flow nasal cannula in just a matter of hours.

Patients slowly got sicker and felt more helpless as the days passed, leading us to prescribe drugs that eventually proved to have no benefit. We checked countless inflammatory markers, most of which we were not even sure what to do with. Many times, we hosted a family meeting via FaceTime, holding a patient’s hand in one hand and an iPad in the other to discuss goals of care. Too often, a dark cloud hung over these discussions, a realization that there was not much else we could do.

I have always felt that helping someone have a decent and peaceful death is important, especially when the prognosis is grim, and that patient is suffering. But the sheer number of times this happened during the initial surge of the pandemic was difficult to handle. It felt like I had more of those discussions in 3 months than I did during my entire career as a hospitalist.

We helped plenty of people get better, with some heading home in a week. They thanked us, painted rocks and the sidewalks in front of the hospital displaying messages of gratitude, and sent lunches. Others, though, left the hospital 2 months later with a tube in their stomach so they could receive some form of nutrition and another in their neck to help them breathe.

These struggles were by no means special to me; other hospitalists around the world faced similar situations at one point or another during the pandemic. Working overtime, coming home late, exhausted, undressing in the garage, trying to be there for my 3 kids who were full of energy after a whole day of Zoom and doing the usual kid stuff. My house used to have strict rules about screen time. No more.

The summer months provided a bit of a COVID break, with only 1 or 2 infected patients entering my care. We went to outdoor restaurants and tried to get our lives back to “normal.” As the weather turned cold, however, things went south again. This time no more hydroxychloroquine, a drug used to fight malaria but also treat other autoimmune diseases, as it was proven eventually over many studies that it is not helpful and was potentially harmful. We instead shifted our focus to remdesivir—an antiviral drug that displayed some benefits—tocilizumab, and dexamethasone, anti-inflammatory drugs with the latter providing some positive outcomes on mortality.

Patient survival rates improved slightly, likely due to a combination of factors. We were more experienced at fighting the disease, which led to things in the hospital not being as chaotic and more time available to spend with the patients. Personal protective equipment (PPE) and tests were more readily available, and the population getting hit by the disease changed slightly with fewer elderly people from nursing homes falling ill because of social distancing, other safety measures, or having already fought the disease. Our attention turned instead to more young people that had returned to work and their social lives.

The arrival of the vaccines brought considerable relief. I remember a few decades ago debating and sometimes fighting with friends and family over who was better: Iron Man or Spider-Man. Now I found myself having the same conversation about the Pfizer and Moderna COVID vaccines.

Summer 2021 holds significantly more promise. Most of the adult population is getting vaccinated, and I am very hopeful that we are approaching the end of this nightmare. In June, our office received word that we could remove our masks if we were fully vaccinated. It felt weird, but represented another sign that things are improving. I took my kids to the mall and removed my mask. It felt odd considering how that little blue thing became part of me during the pandemic. It also felt strange to not prescribe a single dose of remdesivir for an entire month.

It feels good—and normal—to care for the patients that we neglected for a year. It has been a needed boost to see patients return to their health care providers for their colonoscopy screenings, mammograms, and managing chronic problems like coronary artery disease, congestive heart failure, or receiving chemotherapy.

I learned plenty from this pandemic and hope I am not alone. I learned to be humble. We started with a drug that was harmful, moved on to a drug that is probably neutral and eventually were able to come up with a drug that seems to decrease mortality at least in some COVID patients. I learned it is fine to try new therapies based on the best data in the hope they result in positive clinical outcomes. However, it is critical that we all keep an eye on the rapidly evolving literature and adjust our behavior accordingly.

I also learned, or relearned, that if people are desperate enough, they will drink bleach to see if it works. Others are convinced that the purpose of vaccination is to inject a microchip allowing ourselves to be tracked by some higher power. I learned that we must take the first step to prepare for the next pandemic by having a decent reserve of PPE.

It is clear synthetic messenger RNA (mRNA) technology is here to stay, and I believe it has a huge potential to change many areas of medicine. mRNA vaccines proved to be much faster to develop and probably much easier to change as the pathogen, in this case coronavirus, changes.

The technology could be used against a variety of infectious diseases to make vaccines against malaria, tuberculosis, HIV, or hepatitis. It can also be very useful for faster vaccine development needed in future possible pandemics such as influenza, Ebola, or severe acute respiratory syndrome. It may also be used for cancer treatment.

As John P. Cooke, MD, PhD, the medical director for the Center of RNA Therapeutics Program at the Houston Methodist Research Institute, said, “Most vaccines today are still viral vaccines – they are inactivated virus, so it’s potentially infectious and you have to have virus on hand. With mRNA, you’re just writing code which is going to tell the cell to make a viral protein – one part of a viral protein to stimulate an immune response. And, here’s the wonderful thing, you don’t even need the virus in hand, just its DNA code.”¹

Corresponding author: Dragos Vesbianu, MD, Attending Hospitalist, Newton-Wellesley Hospital, 2014 Washington St, Newton, MA 02462; [email protected].

Financial dislosures: None.

I have been in the business of medicine for more than 15 years and I will never forget the initial surge of the COVID-19 pandemic in Massachusetts.

As a hospitalist, I admitted patients infected with COVID-19, followed them on the floor, and, since I had some experience working in an intensive care unit (ICU), was assigned to cover a “COVID ICU.” This wing of the hospital used to be a fancy orthopedic floor that our institution was lucky enough to have. So began the most life-changing experience in my career as a physician.

In this role, we witness death more than any of us would care to discuss. It comes with the territory, and we never expected this to change once COVID hit. However, so many patients succumbed to this disease, especially during the first surge, which made it difficult to handle emotionally. Patients that fell ill initially stayed isolated at home, optimistic they would turn the corner only to enter the hospital a week later after their conditioned worsened. After requiring a couple of liters of supplemental oxygen in the emergency room, they eventually ended up on a high flow nasal cannula in just a matter of hours.

Patients slowly got sicker and felt more helpless as the days passed, leading us to prescribe drugs that eventually proved to have no benefit. We checked countless inflammatory markers, most of which we were not even sure what to do with. Many times, we hosted a family meeting via FaceTime, holding a patient’s hand in one hand and an iPad in the other to discuss goals of care. Too often, a dark cloud hung over these discussions, a realization that there was not much else we could do.

I have always felt that helping someone have a decent and peaceful death is important, especially when the prognosis is grim, and that patient is suffering. But the sheer number of times this happened during the initial surge of the pandemic was difficult to handle. It felt like I had more of those discussions in 3 months than I did during my entire career as a hospitalist.

We helped plenty of people get better, with some heading home in a week. They thanked us, painted rocks and the sidewalks in front of the hospital displaying messages of gratitude, and sent lunches. Others, though, left the hospital 2 months later with a tube in their stomach so they could receive some form of nutrition and another in their neck to help them breathe.

These struggles were by no means special to me; other hospitalists around the world faced similar situations at one point or another during the pandemic. Working overtime, coming home late, exhausted, undressing in the garage, trying to be there for my 3 kids who were full of energy after a whole day of Zoom and doing the usual kid stuff. My house used to have strict rules about screen time. No more.

The summer months provided a bit of a COVID break, with only 1 or 2 infected patients entering my care. We went to outdoor restaurants and tried to get our lives back to “normal.” As the weather turned cold, however, things went south again. This time no more hydroxychloroquine, a drug used to fight malaria but also treat other autoimmune diseases, as it was proven eventually over many studies that it is not helpful and was potentially harmful. We instead shifted our focus to remdesivir—an antiviral drug that displayed some benefits—tocilizumab, and dexamethasone, anti-inflammatory drugs with the latter providing some positive outcomes on mortality.

Patient survival rates improved slightly, likely due to a combination of factors. We were more experienced at fighting the disease, which led to things in the hospital not being as chaotic and more time available to spend with the patients. Personal protective equipment (PPE) and tests were more readily available, and the population getting hit by the disease changed slightly with fewer elderly people from nursing homes falling ill because of social distancing, other safety measures, or having already fought the disease. Our attention turned instead to more young people that had returned to work and their social lives.

The arrival of the vaccines brought considerable relief. I remember a few decades ago debating and sometimes fighting with friends and family over who was better: Iron Man or Spider-Man. Now I found myself having the same conversation about the Pfizer and Moderna COVID vaccines.

Summer 2021 holds significantly more promise. Most of the adult population is getting vaccinated, and I am very hopeful that we are approaching the end of this nightmare. In June, our office received word that we could remove our masks if we were fully vaccinated. It felt weird, but represented another sign that things are improving. I took my kids to the mall and removed my mask. It felt odd considering how that little blue thing became part of me during the pandemic. It also felt strange to not prescribe a single dose of remdesivir for an entire month.

It feels good—and normal—to care for the patients that we neglected for a year. It has been a needed boost to see patients return to their health care providers for their colonoscopy screenings, mammograms, and managing chronic problems like coronary artery disease, congestive heart failure, or receiving chemotherapy.

I learned plenty from this pandemic and hope I am not alone. I learned to be humble. We started with a drug that was harmful, moved on to a drug that is probably neutral and eventually were able to come up with a drug that seems to decrease mortality at least in some COVID patients. I learned it is fine to try new therapies based on the best data in the hope they result in positive clinical outcomes. However, it is critical that we all keep an eye on the rapidly evolving literature and adjust our behavior accordingly.

I also learned, or relearned, that if people are desperate enough, they will drink bleach to see if it works. Others are convinced that the purpose of vaccination is to inject a microchip allowing ourselves to be tracked by some higher power. I learned that we must take the first step to prepare for the next pandemic by having a decent reserve of PPE.

It is clear synthetic messenger RNA (mRNA) technology is here to stay, and I believe it has a huge potential to change many areas of medicine. mRNA vaccines proved to be much faster to develop and probably much easier to change as the pathogen, in this case coronavirus, changes.

The technology could be used against a variety of infectious diseases to make vaccines against malaria, tuberculosis, HIV, or hepatitis. It can also be very useful for faster vaccine development needed in future possible pandemics such as influenza, Ebola, or severe acute respiratory syndrome. It may also be used for cancer treatment.

As John P. Cooke, MD, PhD, the medical director for the Center of RNA Therapeutics Program at the Houston Methodist Research Institute, said, “Most vaccines today are still viral vaccines – they are inactivated virus, so it’s potentially infectious and you have to have virus on hand. With mRNA, you’re just writing code which is going to tell the cell to make a viral protein – one part of a viral protein to stimulate an immune response. And, here’s the wonderful thing, you don’t even need the virus in hand, just its DNA code.”¹

Corresponding author: Dragos Vesbianu, MD, Attending Hospitalist, Newton-Wellesley Hospital, 2014 Washington St, Newton, MA 02462; [email protected].

Financial dislosures: None.

I have been in the business of medicine for more than 15 years and I will never forget the initial surge of the COVID-19 pandemic in Massachusetts.

As a hospitalist, I admitted patients infected with COVID-19, followed them on the floor, and, since I had some experience working in an intensive care unit (ICU), was assigned to cover a “COVID ICU.” This wing of the hospital used to be a fancy orthopedic floor that our institution was lucky enough to have. So began the most life-changing experience in my career as a physician.

In this role, we witness death more than any of us would care to discuss. It comes with the territory, and we never expected this to change once COVID hit. However, so many patients succumbed to this disease, especially during the first surge, which made it difficult to handle emotionally. Patients that fell ill initially stayed isolated at home, optimistic they would turn the corner only to enter the hospital a week later after their conditioned worsened. After requiring a couple of liters of supplemental oxygen in the emergency room, they eventually ended up on a high flow nasal cannula in just a matter of hours.

Patients slowly got sicker and felt more helpless as the days passed, leading us to prescribe drugs that eventually proved to have no benefit. We checked countless inflammatory markers, most of which we were not even sure what to do with. Many times, we hosted a family meeting via FaceTime, holding a patient’s hand in one hand and an iPad in the other to discuss goals of care. Too often, a dark cloud hung over these discussions, a realization that there was not much else we could do.

I have always felt that helping someone have a decent and peaceful death is important, especially when the prognosis is grim, and that patient is suffering. But the sheer number of times this happened during the initial surge of the pandemic was difficult to handle. It felt like I had more of those discussions in 3 months than I did during my entire career as a hospitalist.

We helped plenty of people get better, with some heading home in a week. They thanked us, painted rocks and the sidewalks in front of the hospital displaying messages of gratitude, and sent lunches. Others, though, left the hospital 2 months later with a tube in their stomach so they could receive some form of nutrition and another in their neck to help them breathe.

These struggles were by no means special to me; other hospitalists around the world faced similar situations at one point or another during the pandemic. Working overtime, coming home late, exhausted, undressing in the garage, trying to be there for my 3 kids who were full of energy after a whole day of Zoom and doing the usual kid stuff. My house used to have strict rules about screen time. No more.

The summer months provided a bit of a COVID break, with only 1 or 2 infected patients entering my care. We went to outdoor restaurants and tried to get our lives back to “normal.” As the weather turned cold, however, things went south again. This time no more hydroxychloroquine, a drug used to fight malaria but also treat other autoimmune diseases, as it was proven eventually over many studies that it is not helpful and was potentially harmful. We instead shifted our focus to remdesivir—an antiviral drug that displayed some benefits—tocilizumab, and dexamethasone, anti-inflammatory drugs with the latter providing some positive outcomes on mortality.

Patient survival rates improved slightly, likely due to a combination of factors. We were more experienced at fighting the disease, which led to things in the hospital not being as chaotic and more time available to spend with the patients. Personal protective equipment (PPE) and tests were more readily available, and the population getting hit by the disease changed slightly with fewer elderly people from nursing homes falling ill because of social distancing, other safety measures, or having already fought the disease. Our attention turned instead to more young people that had returned to work and their social lives.

The arrival of the vaccines brought considerable relief. I remember a few decades ago debating and sometimes fighting with friends and family over who was better: Iron Man or Spider-Man. Now I found myself having the same conversation about the Pfizer and Moderna COVID vaccines.

Summer 2021 holds significantly more promise. Most of the adult population is getting vaccinated, and I am very hopeful that we are approaching the end of this nightmare. In June, our office received word that we could remove our masks if we were fully vaccinated. It felt weird, but represented another sign that things are improving. I took my kids to the mall and removed my mask. It felt odd considering how that little blue thing became part of me during the pandemic. It also felt strange to not prescribe a single dose of remdesivir for an entire month.

It feels good—and normal—to care for the patients that we neglected for a year. It has been a needed boost to see patients return to their health care providers for their colonoscopy screenings, mammograms, and managing chronic problems like coronary artery disease, congestive heart failure, or receiving chemotherapy.

I learned plenty from this pandemic and hope I am not alone. I learned to be humble. We started with a drug that was harmful, moved on to a drug that is probably neutral and eventually were able to come up with a drug that seems to decrease mortality at least in some COVID patients. I learned it is fine to try new therapies based on the best data in the hope they result in positive clinical outcomes. However, it is critical that we all keep an eye on the rapidly evolving literature and adjust our behavior accordingly.

I also learned, or relearned, that if people are desperate enough, they will drink bleach to see if it works. Others are convinced that the purpose of vaccination is to inject a microchip allowing ourselves to be tracked by some higher power. I learned that we must take the first step to prepare for the next pandemic by having a decent reserve of PPE.

It is clear synthetic messenger RNA (mRNA) technology is here to stay, and I believe it has a huge potential to change many areas of medicine. mRNA vaccines proved to be much faster to develop and probably much easier to change as the pathogen, in this case coronavirus, changes.

The technology could be used against a variety of infectious diseases to make vaccines against malaria, tuberculosis, HIV, or hepatitis. It can also be very useful for faster vaccine development needed in future possible pandemics such as influenza, Ebola, or severe acute respiratory syndrome. It may also be used for cancer treatment.

As John P. Cooke, MD, PhD, the medical director for the Center of RNA Therapeutics Program at the Houston Methodist Research Institute, said, “Most vaccines today are still viral vaccines – they are inactivated virus, so it’s potentially infectious and you have to have virus on hand. With mRNA, you’re just writing code which is going to tell the cell to make a viral protein – one part of a viral protein to stimulate an immune response. And, here’s the wonderful thing, you don’t even need the virus in hand, just its DNA code.”¹

Corresponding author: Dragos Vesbianu, MD, Attending Hospitalist, Newton-Wellesley Hospital, 2014 Washington St, Newton, MA 02462; [email protected].

Financial dislosures: None.

From the Department of Pharmaceutical and Health Economics, University of Southern California, Los Angeles, CA, (Drs. Sangha and McCombs), Department of Pediatrics, Keck School of Medicine, and Department of Clinical Pharmacy, School of Pharmacy, University of Southern California, Los Angeles, CA, (Dr. Steinberg), and Leonard Schaeffer Center for Health Policy and Economics, University of Southern California, Los Angeles, CA (Dr. McCombs).

Objective: The recommended treatment for children and adolescents under 18 years of age who have a positive test for group A Streptococcus (GAS) are antibiotics using the “test and treat” strategy to detect and treat GAS for pediatric pharyngitis. This study used paid claims data to document the extent to which real-world treatment patterns are consistent with these recommendations. We document the factors correlated with testing and treatment, then examine the effects of receiving a GAS test and being treated with an antibiotic impact the likelihood of a revisit for an acute respiratory tract infection within 28 days.

Methods: This retrospective cohort study used Optum Insight Clinformatics data for medical and pharmacy claims from 2011-2013 to identify episodes of care for children and adolescents with pharyngitis around their index visit (± 6 months). The sample population included children and adolescents under 18 years of age with a diagnosis of pharyngitis. Multivariable logistic regression analyses were used to document factors associated with receipt of GAS test and antibiotic treatment. Next, we used logistic regression models to estimate the impact of test and treat recommendation on revisit risk.

Results: There were 24 685 treatment episodes for children and adolescents diagnosed with pharyngitis. Nearly 47% of these episodes included a GAS test and 48% of tested patients were prescribed an antibiotic prescription. Failing to perform a GAS test increased the risk of a revisit within 28 days by 44%. The use of antibiotics by tested and untested patients had no impact on revisit risk.

Conclusion: While the judicious use of antibiotics is important in managing pharyngitis infections and managing complications, the use of rapid diagnostic tools was found to be the determining factor in reducing revisits for pediatric patients with pharyngitis.

Keywords: pediatrics; pharyngitis; respiratory infections; acute infections; diagnostic tests; group A Streptococcus; antibiotics; revisits.

Acute pharyngitis is a common acute respiratory tract infection (ARTI) in children. Group A β-hemolytic streptococci (GABHS) is the most common bacterial etiology for pediatric pharyngitis, accounting for 15% to 30% of cases.¹

Beyond clinical assessment, laboratory diagnostic testing generally plays a limited role in guiding appropriate antibiotic prescribing for patients with an ARTI.^2,3 Most diagnostic tests require 2 or 3 days to result, incur additional costs, and may delay treatment.⁴ While these tests do not provide clear and timely guidance on which specific antibiotic is appropriate for ARTI patients, this is not the case for patients with pharyngitis.^5,6,7 A rapid diagnostic test exists to identify pharyngitis patients with GABHS which accounts for 1 in 4 children with acute sore throat.^1,4,6 Both the American Academy of Pediatrics and the Infectious Diseases Society of America recommend antibiotic treatment for children and adolescents under 18 years of age who have a positive test for group A Streptococcus (GAS).^8,9 This “test and treat” protocol has been consistently included in the Healthcare Effectiveness Data and Information Set (HEDIS) standards over time for pediatric pharyngitis patients aged 3 to 18 years before dispensing an antibiotic.¹⁰

Sinusitis, pneumonia, and acute otitis media are considered ARTIs where antibiotic treatment is justified. Therefore, pharyngitis of unclear etiology seen with these comorbid infections may not always undergo GAS testing but move directly to the patient being prescribed antibiotics. This analysis enumerates ARTI-related comorbidities present together with the initial coded pharyngitis diagnosis to evaluate their impact on the provider’s decision to test and treat, and on revisit risk.

Antibiotic treatment for GAS patients is likely to eradicate the acute GABHS infection within 10 days. Penicillin and amoxicillin are commonly recommended because of their narrow spectrum of activity, few adverse effects, established efficacy, and modest cost. Alternative antibiotics for patients with penicillin allergy, or with polymicrobial infection seen on culture results, include a first-generation cephalosporin, clindamycin, clarithromycin (Biaxin), or azithromycin (Zithromax).^1,8,11 However, while compliance with these HEDIS guidelines has been evaluated, the outcome effects of following the HEDIS “test and treat” recommendations for children with pharyngitis have not been adequately evaluated.

These outcome evaluations have increasing importance as the latest HEDIS survey has shown testing rates in commercial Preferred Provider Organizations (PPO) falling from 86.4% in 2018 to 75.9% in 2019, the lowest rate of testing since 2009, with similar reductions under 80% for Health Maintenance Organizations (HMO).¹⁰ While health plans may execute cost-benefit analyses and algorithms to forge best practices for GAS testing in children and adolescents presenting with symptoms of pharyngitis, it is important to regard the wasteful resource utilization and additional cost of revisits that may offset any gains accrued by more focused GAS testing outside the existing clinical guidelines and HEDIS measures. This may be of particular importance in documenting infection and sparing antibiotic therapy in toddlers and younger.

The objective of this study was to investigate the correlation between testing and antibiotic use on the likelihood of a revisit for an acute respiratory tract infection within 28 days. To achieve this objective, this investigation consists of 3 sequential analyses. First, we document the factors associated with the decision to test the patient for a GABHS infection using the GAS test. Next, we document the factors associated with the decision to use an antibiotic to treat the patient as a function of having tested the patient. Finally, we investigate the impact of the testing and treatment decisions on the likelihood of a revisit within 28 days.

Methods

Study design

This was a retrospective cohort study of episodes of treatment for pediatric patients with pharyngitis. Episodes were identified using data derived from the Optum Insight Clinformatics claims database provided to the University of Southern California to facilitate the training of graduate students. These data cover commercially insured patients with both medical and pharmacy benefits. Data were retrieved from the 3-year period spanning 2011-2013. An episode of care was identified based on date of the first (index) outpatient visit for a pharyngitis diagnosis (International Classification of Diseases, Ninth Revision [ICD-9]: 462, 463, 034.0). Outpatient visits were defined by visit setting: ambulatory clinics, physician offices, emergency rooms, and urgent care facilities. Each pharyngitis treatment episode was then screened for at least a 6-month enrollment in a health insurance plan prior and subsequent to the index visit using Optum enrollment data. Finally, eligible treatment episodes were restricted to children and adolescents under 18 years of age, who had an index outpatient visit for a primary diagnosis of acute pharyngitis.

A diagnostic profile was created for each episode using the diagnoses recorded for the index visit. Up to 3 diagnoses may be recorded for any outpatient visit and the first recorded diagnosis was assumed to be the primary diagnosis for that episode. Any secondary diagnoses recorded on the index visit were used to define comorbidities present at the index visit. ARTI-related comorbidities included: acute otitis media (AOM), bronchitis, sinusitis, pneumonia, and upper respiratory infection (URI). Other comorbid medical diagnoses were documented using diagnostic data from the pre-index period. Dichotomous variables for the following categories were created: mental disorders, nervous system disorders, respiratory symptoms, fever, injury and poisoning, other, or no diseases.

Prior visits for other respiratory infections in the previous 90 days were also identified for patients based on their index visit for pharyngitis. Similarly, any subsequent visits, within 28 days of the index visit, were also recorded to measure the health outcome for analysis. Practice settings include physician offices and federally qualified health centers, state and local health clinics, outpatient hospitals facilities, emergency departments, and other outpatient settings such as walk-in retail health clinic or ambulatory centers. Providers include primary care physicians (family practice, pediatricians, internal medicine), specialty care physicians (emergency medicine, preventive medicine), nonphysician providers (nurse practitioners, physician assistants) and other providers (urgent care, acute outpatient care, ambulatory care centers). Seasons of the year were determined based on the index date of the episode to account for possible seasonality in pharyngitis treatment. Lastly, a previous visits variable was created to identify whether the child had nonpharyngitis ARTI visits in the 3 months prior to the index visit.

Demographic variables were created based on enrollment and the socioeconomic data available in the Optum socioeconomic status file. These variables include patient age, race, sex, household income, geographic location, practice setting type, provider specialty, and type of insurance. An estimate of patient household income was based on algorithms using census block groups. Income categories were informed by the federal guidelines for a family of 4. A low-income family was defined as earning less than $50 000; a middle-income family earned between $50 000 and $75 000, and a high-income family earned $75 000 and above.¹² Patient insurance type was categorized as HMO, Exclusive Provider Organization (EPO), Point of Service (POS), and PPO. Race was identified as White, Black, Hispanic, and Asian. Patient location was defined according to national census regions.

Outcomes

GAS test

The HEDIS measures for pharyngitis recommend using the GAS test to identify the bacterial etiology of the pharyngitis infection. Patients who received the test were identified based on Current Procedural Terminology (CPT) codes 87070-71, 87081, 87430, 87650-52, and 87880.¹⁰

The pharmacy administrative claims dataset was used to identify study patients who filled a prescription for an antibiotic during their pharyngitis treatment episode. Optum pharmacy data identify the medications received, specifies the date of prescription filling, National Drug Codes, and American Hospital Formulary Service (AHFS) Classification System codes for each medication. We used the AHFS Pharmacologic-Therapeutic classification of antibiotics to create dichotomous variables documenting the antibacterial used by each patient.¹³ These are categorized under antibacterial including penicillins, cephalosporins (first, second, third, fourth generation cephalosporins), macrolides (first generation and others), tetracyclines, sulfonamides, fluoroquinolones (ciprofloxacin, levofloxacin, moxifloxacin), cephamycin, carbapenems, and β-lactam antibiotics (amoxicillin, amoxicillin/clavulanate, cephalexin, cefuroxime, cefdinir).

Revisits to physician or other provider

Revisits within 28 days were used as the measure of patient outcomes related to testing and filling of an antibiotic prescription for acute pharyngitis. Revisits may also be due to a patient returning for a follow-up, alternative treatment, worsening pharyngitis, or for another ARTI. An ARTI-related revisit also increases total resources used to treat pediatric pharyngitis patients.

Statistical analysis

Logistic regression was used for all 3 analyses conducted in this study. First, we determined the patient and treating physician characteristics that impact the decision to use GAS testing for pharyngitis. Second, we identified those factors that impact the decision to use antibiotic prescriptions among children who were diagnosed with pharyngitis adding in the dichotomous variable indicating if the patient had received a GAS test. Third, we used a logit regression analysis to document if receiving a GAS test and/or an antibiotic impacted the likelihood of a revisit by comparing revisit risk. To estimate the effect of testing and/or antibiotic use, we divided patients into 4 groups based on whether the patient received a GAS test and/or an antibiotic prescription. This specification of the analysis of revisits as an outcome focuses on adherence to HEDIS “test and treat” guidelines¹⁰:

1. Patients who were not tested yet filled an antibiotic prescription. This decision was likely based on the clinician’s judgment of the patient’s signs and symptoms, and confirmational testing not performed.
2. Patients who were not tested and did not fill an antibiotic prescription. Apparently, in the clinician’s judgment the patient’s signs and symptoms were such that the infection did not warrant treatment and the clinical presentation did not necessitate the GAS test to confirm the recorded diagnosis of pharyngitis.
3. Patients who were tested and received antibiotic prescription, likely because the test was positive for GABHS.
4. Patients who were tested and did not receive antibiotic prescription.

We tested for statistically significant differences in baseline characteristics across these 4 patient groups using t tests for continuous variables and χ² tests for categorical variables. Odds ratios (OR) and CI were computed for the influential variables included the regression analyses.

We conducted a sensitivity analysis using a model specification which included the dichotomous variables for testing and for treatment, and the interaction term between these variables to assess if treatment effects varied in tested and untested patients. We also estimated this model of revisit risk using revisits within 7 days as the outcome variable.

All analyses were completed using STATA/IC 13 (StataCorp, College Station, TX).

Results

There were 24 685 treatment episodes for children diagnosed with pharyngitis. Nearly 47% of these episodes included GAS testing and 47% of the tested patients filled an antibiotic prescription. Similarly, 53% of patients were not tested and 49% of untested patients filled an antibiotic prescription. As a result, the 4 groups identified for analysis were evenly distributed: untested and no prescription (26.9%), untested and prescription (26.3%), tested and prescription (21.9%), and tested and no prescription (24.9%) (Figure).

Table 1 presents the descriptive statistics for these 4 patient groups. Note first that the rate of revisits within 28 days is under 5% across all groups. Second, the 2 tested groups have a lower revisit rate than the untested groups: the tested and treated have a revisit rate of 3.3%, and the tested and untreated have a revisit rate of 2.4%, while both the untested groups have a revisit rate of nearly 5%. These small absolute differences in revisit rates across groups were statistically significant.

Factors associated with receiving GAS test

Several factors were found to impact the decision to test (Table 2). Only 9.7% of children were reported to have any ARTI coinfection. As expected, these comorbidities resulted in a significantly lower likelihood of receiving the GAS test: AOM, bronchitis, sinusitis, pneumonia, and URI as comorbid infections had a 48%, 41%, 37%, 63%, and 13% lower likelihood of receiving the GAS test, respectively, than those with no comorbidities. Similarly, children with fever and respiratory symptoms were 35% and 45% less likely to receiving the GAS test, respectively. This is consistent with our expectation that comorbid ARTI infections will lead many providers to forgo testing.

Provider type and patient age also plays a role in receipt of the GAS test. Relative to outpatient facility providers, primary care physicians were 24% more likely and specialty physicians were 38% less likely of employing the GAS test. The child’s age played a significant role in receipt of the GAS test. Children aged 1 to 5 years and 5 to 12 years were 15% and 14% more likely to receive the test compared to children older than 12 years.

Pharyngitis patients have disproportionately higher odds of receiving a GAS test in most regions of the country compared to the Pacific region. For instance, children in the Mid-Atlantic region have 51% higher odds of receiving a GAS test while children in New England have 80% higher odds of receiving the same test.

Black children have 11% lower odds of receiving the GAS test compared to White children. Both middle-income and high-income children have 12% and 32% higher odds of receiving the test compared to low-income children. Compared to office-based visits, children visiting a clinic were twice as likely to receive a GAS test while those seen in the emergency room have 43% lower odds of receiving a GAS test. Hospital outpatient departments, which account for less than 1% of all visits, rarely used a GAS test which could be a statistical artifact due to small sample size. Lastly, insurance and season of the year had no significant impact of receipt of a GAS test.

Factors associated with receiving antibiotic prescription

Surprisingly, receiving the GAS test has a small but insignificant impact on the likelihood that the patient will receive an antibiotic prescription (Table 3) (Adjusted OR = 1.055, P = .07). After controlling for receipt of a GAS test, children with AOM and sinusitis comorbidities have an increased likelihood of being prescribed an antibiotic. Children with URI have a lower likelihood of being prescribed an antibiotic. Additionally, relative to primary care physicians, children visiting nonphysician providers for pharyngitis were more likely to be prescribed an antibiotic.

Children under 12 years of age were more likely to use an antibiotic compared to children 12 years and older. Geographically, there is some evidence of regional variation in antibiotic use as well. Children in the south Atlantic, west-south central, and southeast central regions had a significantly lower odds of being prescribed an antibiotic respectively than pharyngitis patients in the Pacific region. Black children had a 10% lower likelihood of being prescribed an antibiotic compared to White children, possibly related to their lower rate of GAS testing. Compared to office-based visits, children visiting a clinic were less likely to use an antibiotic. Household income, insurance type, and season had no significant impact on revisit risk.

Effects of GAS test and antibiotic prescriptions on likelihood of revisits

The multivariate analysis of the risk of a revisit within 28 days is presented in Table 4. Children with pharyngitis who tested and did not receive an antibiotic serve as the reference comparison group for this analysis to illustrate the impact of using the GAS test and treatment with an antibiotic. The results in Table 4 are quite clear: patients who receive the GAS test were significantly less likely to have a revisit within 28 days. Moreover, within the group of patients who were tested, those not receiving an antibiotic, presumedly because their GAS test was negative, experienced the lowest risk of a revisit. This result is consistent with the data in Table 1. Moreover, using an antibiotic had no impact on the likelihood of a revisit in patients not receiving the GAS test. This result is also consistent with Table 1.

Other results from the analysis of revisit risk may be of interest to clinicians. Pharyngitis patients with a prior episode of treatment within 90 days for an acute respiratory tract infection were more than 7 times more likely to experience a revisit within 28 days of the pharyngitis diagnosis than patients without a history of recent ARTI infections. Age is also a risk factor in likelihood of initiating a revisit. Children under 1 year and children aged 1 to 5 years were more likely to have a revisit than children aged more than 12 years. Compared to White children, Black children were 25% (P = .04) less likely to have a revisit. The care setting also has a significant impact on revisit risk. Children visiting outpatient hospital and other care settings had a significantly higher revisit risk than those visiting a physician’s office. Lastly, household income, geographic region, season, medical comorbidities, gender, and insurance type have no significant impact on revisit risk.

Sensitivity analysis

The results from the analysis of 7-day and 28-day revisit risk are summarized in Table 5. These results indicate that patients who were tested had a more significant decrease in revisit risk at 7 days (72%) than was evident at 28 days (47% reduction). Receiving an antibiotic, with or without the test, had no impact on revisit risk.

Discussion

Published data on revisits for pharyngitis are lacking with the concentration of prior research focused more on systemic complications of undertreated GABHS disease or on identifying carrier status. Our study results suggest that GAS testing is the most important factor in reducing revisit risk. Being prescribed an antibiotic, on its own, does not have a significant impact on the risk of a revisit. However, once the GAS test is used, the decision not to use an antibiotic was correlated with the lowest revisit rate, likely because the source of the pharyngitis infection was viral and more likely to resolve without a revisit. Prior studies have reported variable rates of testing among children with pharyngitis prescribed an antibiotic, ranging from 23% to 91%,^14,15 with testing important toward more appropriate antibiotic use.¹⁶ More recently, among more than 67 000 patients aged 3 to 21 years presenting with sore throat and receiving a GAS test, 32.6% were positive.¹⁷

Our analysis found that more than 46% of pediatric pharyngitis patients were given the rapid GAS test. While this testing rate is substantially lower than HEDIS recommendations and lower than testing rates achieved by several health maintenance organizations,¹⁰ it is similar to the 53% of children receiving such testing in a recent National Ambulatory Medical Care Survey.¹⁸ Furthermore, we found that when antibiotics are prescribed following a GAS test, the revisit risk is not significantly reduced, possibly because antibiotics lower revisit risk when informed by diagnostic testing tools that determine the infectious organism. This is supported by a similar population analysis in which we observed reduced revisit rates in children with AOM managed with antibiotics within 3 days of index diagnosis.¹⁹

Several other factors also affect the likelihood of a child receiving the GAS test. Children aged 1 to 12 years were significantly more likely to receive the GAS test than children over the age of 12. This included children in the 1 to 5 years old bracket who had a 15% higher likelihood of undergoing a GAS test, despite children less than 3 years of age as not recommended targets for GAS testing.²⁰ As expected, children with reported ARTI-associated comorbidities were also less likely to receive a GAS test. Additionally, specialty care physicians were less inclined to implement the GAS test, possibly because of diagnostic confidence without testing or referral after GAS was ruled out. Black and low-income children had statistically lower odds of receiving the test, even after controlling for other factors, and yet were less likely to consume a revisit. As the overall data suggested more revisits in those not tested, further study is needed to examine if race or income discrepancies are equity based. Finally, children in the Pacific region, compared to the rest of the nation, were the least likely to receive a GAS test and yet there were no significant differences in revisit rates by region. Regional differences in antibiotic use were also observed in our study, as has been seen by others.²¹

After statistically controlling for having received the diagnostic GAS test and filled a prescription for an antibiotic, there are multitude of factors that independently affect the revisit risk, the most important of which if which was a history of an ARTI infection in the prior 90 days. While prior visit history had no impact on the likelihood of being tested or filling an antibiotic, patients with prior visits were more than 7 times more likely to consume a revisit. This was not reflected in nor related to comorbid ARTIs as these patients did not have statistically higher revisits than those with pharyngitis as the sole-coded diagnosis. Moreover, speculation for bacterial etiology of primary or superinfection based on a recent history of ARTI accounting for revisits seems unlikely as it did not yield greater antibiotic use in that group. Further analysis is required to determine the clinical and behavioral factors that promote for prior ARTI history as a major factor in revisit risk after an index visit for pharyngitis.

Children aged between 1 and 5 years, though 15% more likely to be tested than those aged 12 through 17 years, were also 39% more likely to initiate a revisit compared to older children when statistically controlling for other covariates. This perhaps suggests longer illness, wrong diagnosis, delay in appropriate treatment, or more caution by parents and providers in this age group. Justification for testing children less than 3 years of age who are outside of the HEDIS suggested age group, when clinical judgement does not point to another infection source, can result in positivity rates between 22% and 30% as previously observed.^22,23 Patients visiting nonphysician providers and outpatient facility providers were less likely to have a revisit than those visiting primary and specialty care physicians, though slightly higher propensity for antibiotic prescriptions was seen for nonphysician providers. Pediatricians have been noted to be less likely to prescribe antibiotics without GAS testing than nonpediatric providers, and more guidelines-compliant in prescribing.²⁴

Recommendations to not test children under 3 years of age are based on the lack of acute rheumatic fever and other complications in this age group together with more frequent viral syndromes. Selectivity in applying clinical criteria to testing can be attempted to separate bacterial from viral illness. Postnasal drainage/rhinorrhea, hoarse voice, and cough have been used successfully to identify those with viral illness and less need for testing, with greater certainty of low risk for GABHS in those over 11 years of age without tonsillar exudates, cervical adenopathy, or fever.¹⁷ However, the marginal benefits of those who have all 3 features of viral illness versus none in identifying GAS positivity was 23.3% vs 37.6% - helpful, but certainly not diminishing the need for testing. These constitutional findings of viral URI also do not exclude the GAS carrier state that features these symptoms.²⁵ Others have reinforced the doubt of pharyngeal exudates as the premier diagnostic finding for test-positive GAS.²⁶

This study had several limitations. The Optum claims dataset only contains ICD-9 codes for diagnoses. It does not include data on infection severity and clinical findings related to symptoms, thus empiric treatment warranted based in clinical severity is not assessed. Antibiotics are commonly available as generics and very inexpensive. Patients may fill and pay for these prescriptions directly, in which case, a claim for payment may not be filed with Optum. This could result in an undercount of treated patients in our study.

There is no corresponding problem of missing medical claims for GAS testing which were obtained from the CPT codes within the Optum claims data set. However, we elected not to verify the test results due to these data being missing for 75% of the study population. Nevertheless, this study’s focus was less about justifying antibiotic treatment, but dealt with the outcomes generated by testing and treatment. Toward that end, we used CPT codes to identify a revisit, and while those can at times be affected by financial reimbursement incentives, differences related to revisits in the 4 patient groups should not be subject to bias.

Conclusion

This study used data from real world practices to document the patterns of GAS testing and antibiotic use in pediatric pharyngitis patients. Revisit rates were under 5% for all patient groups and the use of rapid diagnostic tools were found to be the determining factor in further reducing the risk of revisits. This supports the need for compliance with the HEDIS quality metric for pharyngitis to the recommended levels of rapid testing which have been falling in recent years. Use of more accurate antigen and newer molecular detection testing methods may help further delineate important factors in determining pediatric pharyngitis treatment and need for revisits.²⁷

Corresponding author: Jeffrey McCombs, MD, University of Southern California School of Pharmacy, Department of Pharmaceutical and Health Economics, Leonard D. Schaeffer Center for Health Policy & Economics, 635 Downey Way, Verna & Peter Dauterive Hall 310, Los Angeles, CA 90089-3333; [email protected].

Financial disclosures: None.

From the Department of Pharmaceutical and Health Economics, University of Southern California, Los Angeles, CA, (Drs. Sangha and McCombs), Department of Pediatrics, Keck School of Medicine, and Department of Clinical Pharmacy, School of Pharmacy, University of Southern California, Los Angeles, CA, (Dr. Steinberg), and Leonard Schaeffer Center for Health Policy and Economics, University of Southern California, Los Angeles, CA (Dr. McCombs).

Objective: The recommended treatment for children and adolescents under 18 years of age who have a positive test for group A Streptococcus (GAS) are antibiotics using the “test and treat” strategy to detect and treat GAS for pediatric pharyngitis. This study used paid claims data to document the extent to which real-world treatment patterns are consistent with these recommendations. We document the factors correlated with testing and treatment, then examine the effects of receiving a GAS test and being treated with an antibiotic impact the likelihood of a revisit for an acute respiratory tract infection within 28 days.

Methods: This retrospective cohort study used Optum Insight Clinformatics data for medical and pharmacy claims from 2011-2013 to identify episodes of care for children and adolescents with pharyngitis around their index visit (± 6 months). The sample population included children and adolescents under 18 years of age with a diagnosis of pharyngitis. Multivariable logistic regression analyses were used to document factors associated with receipt of GAS test and antibiotic treatment. Next, we used logistic regression models to estimate the impact of test and treat recommendation on revisit risk.

Results: There were 24 685 treatment episodes for children and adolescents diagnosed with pharyngitis. Nearly 47% of these episodes included a GAS test and 48% of tested patients were prescribed an antibiotic prescription. Failing to perform a GAS test increased the risk of a revisit within 28 days by 44%. The use of antibiotics by tested and untested patients had no impact on revisit risk.

Conclusion: While the judicious use of antibiotics is important in managing pharyngitis infections and managing complications, the use of rapid diagnostic tools was found to be the determining factor in reducing revisits for pediatric patients with pharyngitis.

Keywords: pediatrics; pharyngitis; respiratory infections; acute infections; diagnostic tests; group A Streptococcus; antibiotics; revisits.

Acute pharyngitis is a common acute respiratory tract infection (ARTI) in children. Group A β-hemolytic streptococci (GABHS) is the most common bacterial etiology for pediatric pharyngitis, accounting for 15% to 30% of cases.¹

Beyond clinical assessment, laboratory diagnostic testing generally plays a limited role in guiding appropriate antibiotic prescribing for patients with an ARTI.^2,3 Most diagnostic tests require 2 or 3 days to result, incur additional costs, and may delay treatment.⁴ While these tests do not provide clear and timely guidance on which specific antibiotic is appropriate for ARTI patients, this is not the case for patients with pharyngitis.^5,6,7 A rapid diagnostic test exists to identify pharyngitis patients with GABHS which accounts for 1 in 4 children with acute sore throat.^1,4,6 Both the American Academy of Pediatrics and the Infectious Diseases Society of America recommend antibiotic treatment for children and adolescents under 18 years of age who have a positive test for group A Streptococcus (GAS).^8,9 This “test and treat” protocol has been consistently included in the Healthcare Effectiveness Data and Information Set (HEDIS) standards over time for pediatric pharyngitis patients aged 3 to 18 years before dispensing an antibiotic.¹⁰

Sinusitis, pneumonia, and acute otitis media are considered ARTIs where antibiotic treatment is justified. Therefore, pharyngitis of unclear etiology seen with these comorbid infections may not always undergo GAS testing but move directly to the patient being prescribed antibiotics. This analysis enumerates ARTI-related comorbidities present together with the initial coded pharyngitis diagnosis to evaluate their impact on the provider’s decision to test and treat, and on revisit risk.

Antibiotic treatment for GAS patients is likely to eradicate the acute GABHS infection within 10 days. Penicillin and amoxicillin are commonly recommended because of their narrow spectrum of activity, few adverse effects, established efficacy, and modest cost. Alternative antibiotics for patients with penicillin allergy, or with polymicrobial infection seen on culture results, include a first-generation cephalosporin, clindamycin, clarithromycin (Biaxin), or azithromycin (Zithromax).^1,8,11 However, while compliance with these HEDIS guidelines has been evaluated, the outcome effects of following the HEDIS “test and treat” recommendations for children with pharyngitis have not been adequately evaluated.

These outcome evaluations have increasing importance as the latest HEDIS survey has shown testing rates in commercial Preferred Provider Organizations (PPO) falling from 86.4% in 2018 to 75.9% in 2019, the lowest rate of testing since 2009, with similar reductions under 80% for Health Maintenance Organizations (HMO).¹⁰ While health plans may execute cost-benefit analyses and algorithms to forge best practices for GAS testing in children and adolescents presenting with symptoms of pharyngitis, it is important to regard the wasteful resource utilization and additional cost of revisits that may offset any gains accrued by more focused GAS testing outside the existing clinical guidelines and HEDIS measures. This may be of particular importance in documenting infection and sparing antibiotic therapy in toddlers and younger.

The objective of this study was to investigate the correlation between testing and antibiotic use on the likelihood of a revisit for an acute respiratory tract infection within 28 days. To achieve this objective, this investigation consists of 3 sequential analyses. First, we document the factors associated with the decision to test the patient for a GABHS infection using the GAS test. Next, we document the factors associated with the decision to use an antibiotic to treat the patient as a function of having tested the patient. Finally, we investigate the impact of the testing and treatment decisions on the likelihood of a revisit within 28 days.

Methods

Study design

This was a retrospective cohort study of episodes of treatment for pediatric patients with pharyngitis. Episodes were identified using data derived from the Optum Insight Clinformatics claims database provided to the University of Southern California to facilitate the training of graduate students. These data cover commercially insured patients with both medical and pharmacy benefits. Data were retrieved from the 3-year period spanning 2011-2013. An episode of care was identified based on date of the first (index) outpatient visit for a pharyngitis diagnosis (International Classification of Diseases, Ninth Revision [ICD-9]: 462, 463, 034.0). Outpatient visits were defined by visit setting: ambulatory clinics, physician offices, emergency rooms, and urgent care facilities. Each pharyngitis treatment episode was then screened for at least a 6-month enrollment in a health insurance plan prior and subsequent to the index visit using Optum enrollment data. Finally, eligible treatment episodes were restricted to children and adolescents under 18 years of age, who had an index outpatient visit for a primary diagnosis of acute pharyngitis.

A diagnostic profile was created for each episode using the diagnoses recorded for the index visit. Up to 3 diagnoses may be recorded for any outpatient visit and the first recorded diagnosis was assumed to be the primary diagnosis for that episode. Any secondary diagnoses recorded on the index visit were used to define comorbidities present at the index visit. ARTI-related comorbidities included: acute otitis media (AOM), bronchitis, sinusitis, pneumonia, and upper respiratory infection (URI). Other comorbid medical diagnoses were documented using diagnostic data from the pre-index period. Dichotomous variables for the following categories were created: mental disorders, nervous system disorders, respiratory symptoms, fever, injury and poisoning, other, or no diseases.

Prior visits for other respiratory infections in the previous 90 days were also identified for patients based on their index visit for pharyngitis. Similarly, any subsequent visits, within 28 days of the index visit, were also recorded to measure the health outcome for analysis. Practice settings include physician offices and federally qualified health centers, state and local health clinics, outpatient hospitals facilities, emergency departments, and other outpatient settings such as walk-in retail health clinic or ambulatory centers. Providers include primary care physicians (family practice, pediatricians, internal medicine), specialty care physicians (emergency medicine, preventive medicine), nonphysician providers (nurse practitioners, physician assistants) and other providers (urgent care, acute outpatient care, ambulatory care centers). Seasons of the year were determined based on the index date of the episode to account for possible seasonality in pharyngitis treatment. Lastly, a previous visits variable was created to identify whether the child had nonpharyngitis ARTI visits in the 3 months prior to the index visit.

Demographic variables were created based on enrollment and the socioeconomic data available in the Optum socioeconomic status file. These variables include patient age, race, sex, household income, geographic location, practice setting type, provider specialty, and type of insurance. An estimate of patient household income was based on algorithms using census block groups. Income categories were informed by the federal guidelines for a family of 4. A low-income family was defined as earning less than $50 000; a middle-income family earned between $50 000 and $75 000, and a high-income family earned $75 000 and above.¹² Patient insurance type was categorized as HMO, Exclusive Provider Organization (EPO), Point of Service (POS), and PPO. Race was identified as White, Black, Hispanic, and Asian. Patient location was defined according to national census regions.

Outcomes

GAS test

The HEDIS measures for pharyngitis recommend using the GAS test to identify the bacterial etiology of the pharyngitis infection. Patients who received the test were identified based on Current Procedural Terminology (CPT) codes 87070-71, 87081, 87430, 87650-52, and 87880.¹⁰

The pharmacy administrative claims dataset was used to identify study patients who filled a prescription for an antibiotic during their pharyngitis treatment episode. Optum pharmacy data identify the medications received, specifies the date of prescription filling, National Drug Codes, and American Hospital Formulary Service (AHFS) Classification System codes for each medication. We used the AHFS Pharmacologic-Therapeutic classification of antibiotics to create dichotomous variables documenting the antibacterial used by each patient.¹³ These are categorized under antibacterial including penicillins, cephalosporins (first, second, third, fourth generation cephalosporins), macrolides (first generation and others), tetracyclines, sulfonamides, fluoroquinolones (ciprofloxacin, levofloxacin, moxifloxacin), cephamycin, carbapenems, and β-lactam antibiotics (amoxicillin, amoxicillin/clavulanate, cephalexin, cefuroxime, cefdinir).

Revisits to physician or other provider

Revisits within 28 days were used as the measure of patient outcomes related to testing and filling of an antibiotic prescription for acute pharyngitis. Revisits may also be due to a patient returning for a follow-up, alternative treatment, worsening pharyngitis, or for another ARTI. An ARTI-related revisit also increases total resources used to treat pediatric pharyngitis patients.

Statistical analysis

Logistic regression was used for all 3 analyses conducted in this study. First, we determined the patient and treating physician characteristics that impact the decision to use GAS testing for pharyngitis. Second, we identified those factors that impact the decision to use antibiotic prescriptions among children who were diagnosed with pharyngitis adding in the dichotomous variable indicating if the patient had received a GAS test. Third, we used a logit regression analysis to document if receiving a GAS test and/or an antibiotic impacted the likelihood of a revisit by comparing revisit risk. To estimate the effect of testing and/or antibiotic use, we divided patients into 4 groups based on whether the patient received a GAS test and/or an antibiotic prescription. This specification of the analysis of revisits as an outcome focuses on adherence to HEDIS “test and treat” guidelines¹⁰:

1. Patients who were not tested yet filled an antibiotic prescription. This decision was likely based on the clinician’s judgment of the patient’s signs and symptoms, and confirmational testing not performed.
2. Patients who were not tested and did not fill an antibiotic prescription. Apparently, in the clinician’s judgment the patient’s signs and symptoms were such that the infection did not warrant treatment and the clinical presentation did not necessitate the GAS test to confirm the recorded diagnosis of pharyngitis.
3. Patients who were tested and received antibiotic prescription, likely because the test was positive for GABHS.
4. Patients who were tested and did not receive antibiotic prescription.

We tested for statistically significant differences in baseline characteristics across these 4 patient groups using t tests for continuous variables and χ² tests for categorical variables. Odds ratios (OR) and CI were computed for the influential variables included the regression analyses.

We conducted a sensitivity analysis using a model specification which included the dichotomous variables for testing and for treatment, and the interaction term between these variables to assess if treatment effects varied in tested and untested patients. We also estimated this model of revisit risk using revisits within 7 days as the outcome variable.

All analyses were completed using STATA/IC 13 (StataCorp, College Station, TX).

Results

There were 24 685 treatment episodes for children diagnosed with pharyngitis. Nearly 47% of these episodes included GAS testing and 47% of the tested patients filled an antibiotic prescription. Similarly, 53% of patients were not tested and 49% of untested patients filled an antibiotic prescription. As a result, the 4 groups identified for analysis were evenly distributed: untested and no prescription (26.9%), untested and prescription (26.3%), tested and prescription (21.9%), and tested and no prescription (24.9%) (Figure).

Table 1 presents the descriptive statistics for these 4 patient groups. Note first that the rate of revisits within 28 days is under 5% across all groups. Second, the 2 tested groups have a lower revisit rate than the untested groups: the tested and treated have a revisit rate of 3.3%, and the tested and untreated have a revisit rate of 2.4%, while both the untested groups have a revisit rate of nearly 5%. These small absolute differences in revisit rates across groups were statistically significant.

Factors associated with receiving GAS test

Several factors were found to impact the decision to test (Table 2). Only 9.7% of children were reported to have any ARTI coinfection. As expected, these comorbidities resulted in a significantly lower likelihood of receiving the GAS test: AOM, bronchitis, sinusitis, pneumonia, and URI as comorbid infections had a 48%, 41%, 37%, 63%, and 13% lower likelihood of receiving the GAS test, respectively, than those with no comorbidities. Similarly, children with fever and respiratory symptoms were 35% and 45% less likely to receiving the GAS test, respectively. This is consistent with our expectation that comorbid ARTI infections will lead many providers to forgo testing.

Provider type and patient age also plays a role in receipt of the GAS test. Relative to outpatient facility providers, primary care physicians were 24% more likely and specialty physicians were 38% less likely of employing the GAS test. The child’s age played a significant role in receipt of the GAS test. Children aged 1 to 5 years and 5 to 12 years were 15% and 14% more likely to receive the test compared to children older than 12 years.

Pharyngitis patients have disproportionately higher odds of receiving a GAS test in most regions of the country compared to the Pacific region. For instance, children in the Mid-Atlantic region have 51% higher odds of receiving a GAS test while children in New England have 80% higher odds of receiving the same test.

Black children have 11% lower odds of receiving the GAS test compared to White children. Both middle-income and high-income children have 12% and 32% higher odds of receiving the test compared to low-income children. Compared to office-based visits, children visiting a clinic were twice as likely to receive a GAS test while those seen in the emergency room have 43% lower odds of receiving a GAS test. Hospital outpatient departments, which account for less than 1% of all visits, rarely used a GAS test which could be a statistical artifact due to small sample size. Lastly, insurance and season of the year had no significant impact of receipt of a GAS test.

Factors associated with receiving antibiotic prescription

Surprisingly, receiving the GAS test has a small but insignificant impact on the likelihood that the patient will receive an antibiotic prescription (Table 3) (Adjusted OR = 1.055, P = .07). After controlling for receipt of a GAS test, children with AOM and sinusitis comorbidities have an increased likelihood of being prescribed an antibiotic. Children with URI have a lower likelihood of being prescribed an antibiotic. Additionally, relative to primary care physicians, children visiting nonphysician providers for pharyngitis were more likely to be prescribed an antibiotic.

Children under 12 years of age were more likely to use an antibiotic compared to children 12 years and older. Geographically, there is some evidence of regional variation in antibiotic use as well. Children in the south Atlantic, west-south central, and southeast central regions had a significantly lower odds of being prescribed an antibiotic respectively than pharyngitis patients in the Pacific region. Black children had a 10% lower likelihood of being prescribed an antibiotic compared to White children, possibly related to their lower rate of GAS testing. Compared to office-based visits, children visiting a clinic were less likely to use an antibiotic. Household income, insurance type, and season had no significant impact on revisit risk.

Effects of GAS test and antibiotic prescriptions on likelihood of revisits

The multivariate analysis of the risk of a revisit within 28 days is presented in Table 4. Children with pharyngitis who tested and did not receive an antibiotic serve as the reference comparison group for this analysis to illustrate the impact of using the GAS test and treatment with an antibiotic. The results in Table 4 are quite clear: patients who receive the GAS test were significantly less likely to have a revisit within 28 days. Moreover, within the group of patients who were tested, those not receiving an antibiotic, presumedly because their GAS test was negative, experienced the lowest risk of a revisit. This result is consistent with the data in Table 1. Moreover, using an antibiotic had no impact on the likelihood of a revisit in patients not receiving the GAS test. This result is also consistent with Table 1.

Other results from the analysis of revisit risk may be of interest to clinicians. Pharyngitis patients with a prior episode of treatment within 90 days for an acute respiratory tract infection were more than 7 times more likely to experience a revisit within 28 days of the pharyngitis diagnosis than patients without a history of recent ARTI infections. Age is also a risk factor in likelihood of initiating a revisit. Children under 1 year and children aged 1 to 5 years were more likely to have a revisit than children aged more than 12 years. Compared to White children, Black children were 25% (P = .04) less likely to have a revisit. The care setting also has a significant impact on revisit risk. Children visiting outpatient hospital and other care settings had a significantly higher revisit risk than those visiting a physician’s office. Lastly, household income, geographic region, season, medical comorbidities, gender, and insurance type have no significant impact on revisit risk.

Sensitivity analysis

The results from the analysis of 7-day and 28-day revisit risk are summarized in Table 5. These results indicate that patients who were tested had a more significant decrease in revisit risk at 7 days (72%) than was evident at 28 days (47% reduction). Receiving an antibiotic, with or without the test, had no impact on revisit risk.

Discussion

Published data on revisits for pharyngitis are lacking with the concentration of prior research focused more on systemic complications of undertreated GABHS disease or on identifying carrier status. Our study results suggest that GAS testing is the most important factor in reducing revisit risk. Being prescribed an antibiotic, on its own, does not have a significant impact on the risk of a revisit. However, once the GAS test is used, the decision not to use an antibiotic was correlated with the lowest revisit rate, likely because the source of the pharyngitis infection was viral and more likely to resolve without a revisit. Prior studies have reported variable rates of testing among children with pharyngitis prescribed an antibiotic, ranging from 23% to 91%,^14,15 with testing important toward more appropriate antibiotic use.¹⁶ More recently, among more than 67 000 patients aged 3 to 21 years presenting with sore throat and receiving a GAS test, 32.6% were positive.¹⁷

Our analysis found that more than 46% of pediatric pharyngitis patients were given the rapid GAS test. While this testing rate is substantially lower than HEDIS recommendations and lower than testing rates achieved by several health maintenance organizations,¹⁰ it is similar to the 53% of children receiving such testing in a recent National Ambulatory Medical Care Survey.¹⁸ Furthermore, we found that when antibiotics are prescribed following a GAS test, the revisit risk is not significantly reduced, possibly because antibiotics lower revisit risk when informed by diagnostic testing tools that determine the infectious organism. This is supported by a similar population analysis in which we observed reduced revisit rates in children with AOM managed with antibiotics within 3 days of index diagnosis.¹⁹

Several other factors also affect the likelihood of a child receiving the GAS test. Children aged 1 to 12 years were significantly more likely to receive the GAS test than children over the age of 12. This included children in the 1 to 5 years old bracket who had a 15% higher likelihood of undergoing a GAS test, despite children less than 3 years of age as not recommended targets for GAS testing.²⁰ As expected, children with reported ARTI-associated comorbidities were also less likely to receive a GAS test. Additionally, specialty care physicians were less inclined to implement the GAS test, possibly because of diagnostic confidence without testing or referral after GAS was ruled out. Black and low-income children had statistically lower odds of receiving the test, even after controlling for other factors, and yet were less likely to consume a revisit. As the overall data suggested more revisits in those not tested, further study is needed to examine if race or income discrepancies are equity based. Finally, children in the Pacific region, compared to the rest of the nation, were the least likely to receive a GAS test and yet there were no significant differences in revisit rates by region. Regional differences in antibiotic use were also observed in our study, as has been seen by others.²¹

After statistically controlling for having received the diagnostic GAS test and filled a prescription for an antibiotic, there are multitude of factors that independently affect the revisit risk, the most important of which if which was a history of an ARTI infection in the prior 90 days. While prior visit history had no impact on the likelihood of being tested or filling an antibiotic, patients with prior visits were more than 7 times more likely to consume a revisit. This was not reflected in nor related to comorbid ARTIs as these patients did not have statistically higher revisits than those with pharyngitis as the sole-coded diagnosis. Moreover, speculation for bacterial etiology of primary or superinfection based on a recent history of ARTI accounting for revisits seems unlikely as it did not yield greater antibiotic use in that group. Further analysis is required to determine the clinical and behavioral factors that promote for prior ARTI history as a major factor in revisit risk after an index visit for pharyngitis.

Children aged between 1 and 5 years, though 15% more likely to be tested than those aged 12 through 17 years, were also 39% more likely to initiate a revisit compared to older children when statistically controlling for other covariates. This perhaps suggests longer illness, wrong diagnosis, delay in appropriate treatment, or more caution by parents and providers in this age group. Justification for testing children less than 3 years of age who are outside of the HEDIS suggested age group, when clinical judgement does not point to another infection source, can result in positivity rates between 22% and 30% as previously observed.^22,23 Patients visiting nonphysician providers and outpatient facility providers were less likely to have a revisit than those visiting primary and specialty care physicians, though slightly higher propensity for antibiotic prescriptions was seen for nonphysician providers. Pediatricians have been noted to be less likely to prescribe antibiotics without GAS testing than nonpediatric providers, and more guidelines-compliant in prescribing.²⁴

Recommendations to not test children under 3 years of age are based on the lack of acute rheumatic fever and other complications in this age group together with more frequent viral syndromes. Selectivity in applying clinical criteria to testing can be attempted to separate bacterial from viral illness. Postnasal drainage/rhinorrhea, hoarse voice, and cough have been used successfully to identify those with viral illness and less need for testing, with greater certainty of low risk for GABHS in those over 11 years of age without tonsillar exudates, cervical adenopathy, or fever.¹⁷ However, the marginal benefits of those who have all 3 features of viral illness versus none in identifying GAS positivity was 23.3% vs 37.6% - helpful, but certainly not diminishing the need for testing. These constitutional findings of viral URI also do not exclude the GAS carrier state that features these symptoms.²⁵ Others have reinforced the doubt of pharyngeal exudates as the premier diagnostic finding for test-positive GAS.²⁶

This study had several limitations. The Optum claims dataset only contains ICD-9 codes for diagnoses. It does not include data on infection severity and clinical findings related to symptoms, thus empiric treatment warranted based in clinical severity is not assessed. Antibiotics are commonly available as generics and very inexpensive. Patients may fill and pay for these prescriptions directly, in which case, a claim for payment may not be filed with Optum. This could result in an undercount of treated patients in our study.

There is no corresponding problem of missing medical claims for GAS testing which were obtained from the CPT codes within the Optum claims data set. However, we elected not to verify the test results due to these data being missing for 75% of the study population. Nevertheless, this study’s focus was less about justifying antibiotic treatment, but dealt with the outcomes generated by testing and treatment. Toward that end, we used CPT codes to identify a revisit, and while those can at times be affected by financial reimbursement incentives, differences related to revisits in the 4 patient groups should not be subject to bias.

Conclusion

This study used data from real world practices to document the patterns of GAS testing and antibiotic use in pediatric pharyngitis patients. Revisit rates were under 5% for all patient groups and the use of rapid diagnostic tools were found to be the determining factor in further reducing the risk of revisits. This supports the need for compliance with the HEDIS quality metric for pharyngitis to the recommended levels of rapid testing which have been falling in recent years. Use of more accurate antigen and newer molecular detection testing methods may help further delineate important factors in determining pediatric pharyngitis treatment and need for revisits.²⁷

Corresponding author: Jeffrey McCombs, MD, University of Southern California School of Pharmacy, Department of Pharmaceutical and Health Economics, Leonard D. Schaeffer Center for Health Policy & Economics, 635 Downey Way, Verna & Peter Dauterive Hall 310, Los Angeles, CA 90089-3333; [email protected].

Financial disclosures: None.

From the Department of Pharmaceutical and Health Economics, University of Southern California, Los Angeles, CA, (Drs. Sangha and McCombs), Department of Pediatrics, Keck School of Medicine, and Department of Clinical Pharmacy, School of Pharmacy, University of Southern California, Los Angeles, CA, (Dr. Steinberg), and Leonard Schaeffer Center for Health Policy and Economics, University of Southern California, Los Angeles, CA (Dr. McCombs).

Objective: The recommended treatment for children and adolescents under 18 years of age who have a positive test for group A Streptococcus (GAS) are antibiotics using the “test and treat” strategy to detect and treat GAS for pediatric pharyngitis. This study used paid claims data to document the extent to which real-world treatment patterns are consistent with these recommendations. We document the factors correlated with testing and treatment, then examine the effects of receiving a GAS test and being treated with an antibiotic impact the likelihood of a revisit for an acute respiratory tract infection within 28 days.

Methods: This retrospective cohort study used Optum Insight Clinformatics data for medical and pharmacy claims from 2011-2013 to identify episodes of care for children and adolescents with pharyngitis around their index visit (± 6 months). The sample population included children and adolescents under 18 years of age with a diagnosis of pharyngitis. Multivariable logistic regression analyses were used to document factors associated with receipt of GAS test and antibiotic treatment. Next, we used logistic regression models to estimate the impact of test and treat recommendation on revisit risk.

Results: There were 24 685 treatment episodes for children and adolescents diagnosed with pharyngitis. Nearly 47% of these episodes included a GAS test and 48% of tested patients were prescribed an antibiotic prescription. Failing to perform a GAS test increased the risk of a revisit within 28 days by 44%. The use of antibiotics by tested and untested patients had no impact on revisit risk.

Conclusion: While the judicious use of antibiotics is important in managing pharyngitis infections and managing complications, the use of rapid diagnostic tools was found to be the determining factor in reducing revisits for pediatric patients with pharyngitis.

Keywords: pediatrics; pharyngitis; respiratory infections; acute infections; diagnostic tests; group A Streptococcus; antibiotics; revisits.

Acute pharyngitis is a common acute respiratory tract infection (ARTI) in children. Group A β-hemolytic streptococci (GABHS) is the most common bacterial etiology for pediatric pharyngitis, accounting for 15% to 30% of cases.¹

Beyond clinical assessment, laboratory diagnostic testing generally plays a limited role in guiding appropriate antibiotic prescribing for patients with an ARTI.^2,3 Most diagnostic tests require 2 or 3 days to result, incur additional costs, and may delay treatment.⁴ While these tests do not provide clear and timely guidance on which specific antibiotic is appropriate for ARTI patients, this is not the case for patients with pharyngitis.^5,6,7 A rapid diagnostic test exists to identify pharyngitis patients with GABHS which accounts for 1 in 4 children with acute sore throat.^1,4,6 Both the American Academy of Pediatrics and the Infectious Diseases Society of America recommend antibiotic treatment for children and adolescents under 18 years of age who have a positive test for group A Streptococcus (GAS).^8,9 This “test and treat” protocol has been consistently included in the Healthcare Effectiveness Data and Information Set (HEDIS) standards over time for pediatric pharyngitis patients aged 3 to 18 years before dispensing an antibiotic.¹⁰

Sinusitis, pneumonia, and acute otitis media are considered ARTIs where antibiotic treatment is justified. Therefore, pharyngitis of unclear etiology seen with these comorbid infections may not always undergo GAS testing but move directly to the patient being prescribed antibiotics. This analysis enumerates ARTI-related comorbidities present together with the initial coded pharyngitis diagnosis to evaluate their impact on the provider’s decision to test and treat, and on revisit risk.

Antibiotic treatment for GAS patients is likely to eradicate the acute GABHS infection within 10 days. Penicillin and amoxicillin are commonly recommended because of their narrow spectrum of activity, few adverse effects, established efficacy, and modest cost. Alternative antibiotics for patients with penicillin allergy, or with polymicrobial infection seen on culture results, include a first-generation cephalosporin, clindamycin, clarithromycin (Biaxin), or azithromycin (Zithromax).^1,8,11 However, while compliance with these HEDIS guidelines has been evaluated, the outcome effects of following the HEDIS “test and treat” recommendations for children with pharyngitis have not been adequately evaluated.

These outcome evaluations have increasing importance as the latest HEDIS survey has shown testing rates in commercial Preferred Provider Organizations (PPO) falling from 86.4% in 2018 to 75.9% in 2019, the lowest rate of testing since 2009, with similar reductions under 80% for Health Maintenance Organizations (HMO).¹⁰ While health plans may execute cost-benefit analyses and algorithms to forge best practices for GAS testing in children and adolescents presenting with symptoms of pharyngitis, it is important to regard the wasteful resource utilization and additional cost of revisits that may offset any gains accrued by more focused GAS testing outside the existing clinical guidelines and HEDIS measures. This may be of particular importance in documenting infection and sparing antibiotic therapy in toddlers and younger.

The objective of this study was to investigate the correlation between testing and antibiotic use on the likelihood of a revisit for an acute respiratory tract infection within 28 days. To achieve this objective, this investigation consists of 3 sequential analyses. First, we document the factors associated with the decision to test the patient for a GABHS infection using the GAS test. Next, we document the factors associated with the decision to use an antibiotic to treat the patient as a function of having tested the patient. Finally, we investigate the impact of the testing and treatment decisions on the likelihood of a revisit within 28 days.

Methods

Study design

This was a retrospective cohort study of episodes of treatment for pediatric patients with pharyngitis. Episodes were identified using data derived from the Optum Insight Clinformatics claims database provided to the University of Southern California to facilitate the training of graduate students. These data cover commercially insured patients with both medical and pharmacy benefits. Data were retrieved from the 3-year period spanning 2011-2013. An episode of care was identified based on date of the first (index) outpatient visit for a pharyngitis diagnosis (International Classification of Diseases, Ninth Revision [ICD-9]: 462, 463, 034.0). Outpatient visits were defined by visit setting: ambulatory clinics, physician offices, emergency rooms, and urgent care facilities. Each pharyngitis treatment episode was then screened for at least a 6-month enrollment in a health insurance plan prior and subsequent to the index visit using Optum enrollment data. Finally, eligible treatment episodes were restricted to children and adolescents under 18 years of age, who had an index outpatient visit for a primary diagnosis of acute pharyngitis.

A diagnostic profile was created for each episode using the diagnoses recorded for the index visit. Up to 3 diagnoses may be recorded for any outpatient visit and the first recorded diagnosis was assumed to be the primary diagnosis for that episode. Any secondary diagnoses recorded on the index visit were used to define comorbidities present at the index visit. ARTI-related comorbidities included: acute otitis media (AOM), bronchitis, sinusitis, pneumonia, and upper respiratory infection (URI). Other comorbid medical diagnoses were documented using diagnostic data from the pre-index period. Dichotomous variables for the following categories were created: mental disorders, nervous system disorders, respiratory symptoms, fever, injury and poisoning, other, or no diseases.

Prior visits for other respiratory infections in the previous 90 days were also identified for patients based on their index visit for pharyngitis. Similarly, any subsequent visits, within 28 days of the index visit, were also recorded to measure the health outcome for analysis. Practice settings include physician offices and federally qualified health centers, state and local health clinics, outpatient hospitals facilities, emergency departments, and other outpatient settings such as walk-in retail health clinic or ambulatory centers. Providers include primary care physicians (family practice, pediatricians, internal medicine), specialty care physicians (emergency medicine, preventive medicine), nonphysician providers (nurse practitioners, physician assistants) and other providers (urgent care, acute outpatient care, ambulatory care centers). Seasons of the year were determined based on the index date of the episode to account for possible seasonality in pharyngitis treatment. Lastly, a previous visits variable was created to identify whether the child had nonpharyngitis ARTI visits in the 3 months prior to the index visit.

Demographic variables were created based on enrollment and the socioeconomic data available in the Optum socioeconomic status file. These variables include patient age, race, sex, household income, geographic location, practice setting type, provider specialty, and type of insurance. An estimate of patient household income was based on algorithms using census block groups. Income categories were informed by the federal guidelines for a family of 4. A low-income family was defined as earning less than $50 000; a middle-income family earned between $50 000 and $75 000, and a high-income family earned $75 000 and above.¹² Patient insurance type was categorized as HMO, Exclusive Provider Organization (EPO), Point of Service (POS), and PPO. Race was identified as White, Black, Hispanic, and Asian. Patient location was defined according to national census regions.

Outcomes

GAS test

The HEDIS measures for pharyngitis recommend using the GAS test to identify the bacterial etiology of the pharyngitis infection. Patients who received the test were identified based on Current Procedural Terminology (CPT) codes 87070-71, 87081, 87430, 87650-52, and 87880.¹⁰

The pharmacy administrative claims dataset was used to identify study patients who filled a prescription for an antibiotic during their pharyngitis treatment episode. Optum pharmacy data identify the medications received, specifies the date of prescription filling, National Drug Codes, and American Hospital Formulary Service (AHFS) Classification System codes for each medication. We used the AHFS Pharmacologic-Therapeutic classification of antibiotics to create dichotomous variables documenting the antibacterial used by each patient.¹³ These are categorized under antibacterial including penicillins, cephalosporins (first, second, third, fourth generation cephalosporins), macrolides (first generation and others), tetracyclines, sulfonamides, fluoroquinolones (ciprofloxacin, levofloxacin, moxifloxacin), cephamycin, carbapenems, and β-lactam antibiotics (amoxicillin, amoxicillin/clavulanate, cephalexin, cefuroxime, cefdinir).

Revisits to physician or other provider

Revisits within 28 days were used as the measure of patient outcomes related to testing and filling of an antibiotic prescription for acute pharyngitis. Revisits may also be due to a patient returning for a follow-up, alternative treatment, worsening pharyngitis, or for another ARTI. An ARTI-related revisit also increases total resources used to treat pediatric pharyngitis patients.

Statistical analysis

Logistic regression was used for all 3 analyses conducted in this study. First, we determined the patient and treating physician characteristics that impact the decision to use GAS testing for pharyngitis. Second, we identified those factors that impact the decision to use antibiotic prescriptions among children who were diagnosed with pharyngitis adding in the dichotomous variable indicating if the patient had received a GAS test. Third, we used a logit regression analysis to document if receiving a GAS test and/or an antibiotic impacted the likelihood of a revisit by comparing revisit risk. To estimate the effect of testing and/or antibiotic use, we divided patients into 4 groups based on whether the patient received a GAS test and/or an antibiotic prescription. This specification of the analysis of revisits as an outcome focuses on adherence to HEDIS “test and treat” guidelines¹⁰:

1. Patients who were not tested yet filled an antibiotic prescription. This decision was likely based on the clinician’s judgment of the patient’s signs and symptoms, and confirmational testing not performed.
2. Patients who were not tested and did not fill an antibiotic prescription. Apparently, in the clinician’s judgment the patient’s signs and symptoms were such that the infection did not warrant treatment and the clinical presentation did not necessitate the GAS test to confirm the recorded diagnosis of pharyngitis.
3. Patients who were tested and received antibiotic prescription, likely because the test was positive for GABHS.
4. Patients who were tested and did not receive antibiotic prescription.

We tested for statistically significant differences in baseline characteristics across these 4 patient groups using t tests for continuous variables and χ² tests for categorical variables. Odds ratios (OR) and CI were computed for the influential variables included the regression analyses.

We conducted a sensitivity analysis using a model specification which included the dichotomous variables for testing and for treatment, and the interaction term between these variables to assess if treatment effects varied in tested and untested patients. We also estimated this model of revisit risk using revisits within 7 days as the outcome variable.

All analyses were completed using STATA/IC 13 (StataCorp, College Station, TX).

Results

There were 24 685 treatment episodes for children diagnosed with pharyngitis. Nearly 47% of these episodes included GAS testing and 47% of the tested patients filled an antibiotic prescription. Similarly, 53% of patients were not tested and 49% of untested patients filled an antibiotic prescription. As a result, the 4 groups identified for analysis were evenly distributed: untested and no prescription (26.9%), untested and prescription (26.3%), tested and prescription (21.9%), and tested and no prescription (24.9%) (Figure).

Table 1 presents the descriptive statistics for these 4 patient groups. Note first that the rate of revisits within 28 days is under 5% across all groups. Second, the 2 tested groups have a lower revisit rate than the untested groups: the tested and treated have a revisit rate of 3.3%, and the tested and untreated have a revisit rate of 2.4%, while both the untested groups have a revisit rate of nearly 5%. These small absolute differences in revisit rates across groups were statistically significant.

Factors associated with receiving GAS test

Several factors were found to impact the decision to test (Table 2). Only 9.7% of children were reported to have any ARTI coinfection. As expected, these comorbidities resulted in a significantly lower likelihood of receiving the GAS test: AOM, bronchitis, sinusitis, pneumonia, and URI as comorbid infections had a 48%, 41%, 37%, 63%, and 13% lower likelihood of receiving the GAS test, respectively, than those with no comorbidities. Similarly, children with fever and respiratory symptoms were 35% and 45% less likely to receiving the GAS test, respectively. This is consistent with our expectation that comorbid ARTI infections will lead many providers to forgo testing.

Provider type and patient age also plays a role in receipt of the GAS test. Relative to outpatient facility providers, primary care physicians were 24% more likely and specialty physicians were 38% less likely of employing the GAS test. The child’s age played a significant role in receipt of the GAS test. Children aged 1 to 5 years and 5 to 12 years were 15% and 14% more likely to receive the test compared to children older than 12 years.

Pharyngitis patients have disproportionately higher odds of receiving a GAS test in most regions of the country compared to the Pacific region. For instance, children in the Mid-Atlantic region have 51% higher odds of receiving a GAS test while children in New England have 80% higher odds of receiving the same test.

Black children have 11% lower odds of receiving the GAS test compared to White children. Both middle-income and high-income children have 12% and 32% higher odds of receiving the test compared to low-income children. Compared to office-based visits, children visiting a clinic were twice as likely to receive a GAS test while those seen in the emergency room have 43% lower odds of receiving a GAS test. Hospital outpatient departments, which account for less than 1% of all visits, rarely used a GAS test which could be a statistical artifact due to small sample size. Lastly, insurance and season of the year had no significant impact of receipt of a GAS test.

Factors associated with receiving antibiotic prescription

Surprisingly, receiving the GAS test has a small but insignificant impact on the likelihood that the patient will receive an antibiotic prescription (Table 3) (Adjusted OR = 1.055, P = .07). After controlling for receipt of a GAS test, children with AOM and sinusitis comorbidities have an increased likelihood of being prescribed an antibiotic. Children with URI have a lower likelihood of being prescribed an antibiotic. Additionally, relative to primary care physicians, children visiting nonphysician providers for pharyngitis were more likely to be prescribed an antibiotic.

Children under 12 years of age were more likely to use an antibiotic compared to children 12 years and older. Geographically, there is some evidence of regional variation in antibiotic use as well. Children in the south Atlantic, west-south central, and southeast central regions had a significantly lower odds of being prescribed an antibiotic respectively than pharyngitis patients in the Pacific region. Black children had a 10% lower likelihood of being prescribed an antibiotic compared to White children, possibly related to their lower rate of GAS testing. Compared to office-based visits, children visiting a clinic were less likely to use an antibiotic. Household income, insurance type, and season had no significant impact on revisit risk.

Effects of GAS test and antibiotic prescriptions on likelihood of revisits

The multivariate analysis of the risk of a revisit within 28 days is presented in Table 4. Children with pharyngitis who tested and did not receive an antibiotic serve as the reference comparison group for this analysis to illustrate the impact of using the GAS test and treatment with an antibiotic. The results in Table 4 are quite clear: patients who receive the GAS test were significantly less likely to have a revisit within 28 days. Moreover, within the group of patients who were tested, those not receiving an antibiotic, presumedly because their GAS test was negative, experienced the lowest risk of a revisit. This result is consistent with the data in Table 1. Moreover, using an antibiotic had no impact on the likelihood of a revisit in patients not receiving the GAS test. This result is also consistent with Table 1.

Other results from the analysis of revisit risk may be of interest to clinicians. Pharyngitis patients with a prior episode of treatment within 90 days for an acute respiratory tract infection were more than 7 times more likely to experience a revisit within 28 days of the pharyngitis diagnosis than patients without a history of recent ARTI infections. Age is also a risk factor in likelihood of initiating a revisit. Children under 1 year and children aged 1 to 5 years were more likely to have a revisit than children aged more than 12 years. Compared to White children, Black children were 25% (P = .04) less likely to have a revisit. The care setting also has a significant impact on revisit risk. Children visiting outpatient hospital and other care settings had a significantly higher revisit risk than those visiting a physician’s office. Lastly, household income, geographic region, season, medical comorbidities, gender, and insurance type have no significant impact on revisit risk.

Sensitivity analysis

The results from the analysis of 7-day and 28-day revisit risk are summarized in Table 5. These results indicate that patients who were tested had a more significant decrease in revisit risk at 7 days (72%) than was evident at 28 days (47% reduction). Receiving an antibiotic, with or without the test, had no impact on revisit risk.

Discussion

Published data on revisits for pharyngitis are lacking with the concentration of prior research focused more on systemic complications of undertreated GABHS disease or on identifying carrier status. Our study results suggest that GAS testing is the most important factor in reducing revisit risk. Being prescribed an antibiotic, on its own, does not have a significant impact on the risk of a revisit. However, once the GAS test is used, the decision not to use an antibiotic was correlated with the lowest revisit rate, likely because the source of the pharyngitis infection was viral and more likely to resolve without a revisit. Prior studies have reported variable rates of testing among children with pharyngitis prescribed an antibiotic, ranging from 23% to 91%,^14,15 with testing important toward more appropriate antibiotic use.¹⁶ More recently, among more than 67 000 patients aged 3 to 21 years presenting with sore throat and receiving a GAS test, 32.6% were positive.¹⁷

Our analysis found that more than 46% of pediatric pharyngitis patients were given the rapid GAS test. While this testing rate is substantially lower than HEDIS recommendations and lower than testing rates achieved by several health maintenance organizations,¹⁰ it is similar to the 53% of children receiving such testing in a recent National Ambulatory Medical Care Survey.¹⁸ Furthermore, we found that when antibiotics are prescribed following a GAS test, the revisit risk is not significantly reduced, possibly because antibiotics lower revisit risk when informed by diagnostic testing tools that determine the infectious organism. This is supported by a similar population analysis in which we observed reduced revisit rates in children with AOM managed with antibiotics within 3 days of index diagnosis.¹⁹

Several other factors also affect the likelihood of a child receiving the GAS test. Children aged 1 to 12 years were significantly more likely to receive the GAS test than children over the age of 12. This included children in the 1 to 5 years old bracket who had a 15% higher likelihood of undergoing a GAS test, despite children less than 3 years of age as not recommended targets for GAS testing.²⁰ As expected, children with reported ARTI-associated comorbidities were also less likely to receive a GAS test. Additionally, specialty care physicians were less inclined to implement the GAS test, possibly because of diagnostic confidence without testing or referral after GAS was ruled out. Black and low-income children had statistically lower odds of receiving the test, even after controlling for other factors, and yet were less likely to consume a revisit. As the overall data suggested more revisits in those not tested, further study is needed to examine if race or income discrepancies are equity based. Finally, children in the Pacific region, compared to the rest of the nation, were the least likely to receive a GAS test and yet there were no significant differences in revisit rates by region. Regional differences in antibiotic use were also observed in our study, as has been seen by others.²¹

After statistically controlling for having received the diagnostic GAS test and filled a prescription for an antibiotic, there are multitude of factors that independently affect the revisit risk, the most important of which if which was a history of an ARTI infection in the prior 90 days. While prior visit history had no impact on the likelihood of being tested or filling an antibiotic, patients with prior visits were more than 7 times more likely to consume a revisit. This was not reflected in nor related to comorbid ARTIs as these patients did not have statistically higher revisits than those with pharyngitis as the sole-coded diagnosis. Moreover, speculation for bacterial etiology of primary or superinfection based on a recent history of ARTI accounting for revisits seems unlikely as it did not yield greater antibiotic use in that group. Further analysis is required to determine the clinical and behavioral factors that promote for prior ARTI history as a major factor in revisit risk after an index visit for pharyngitis.

Children aged between 1 and 5 years, though 15% more likely to be tested than those aged 12 through 17 years, were also 39% more likely to initiate a revisit compared to older children when statistically controlling for other covariates. This perhaps suggests longer illness, wrong diagnosis, delay in appropriate treatment, or more caution by parents and providers in this age group. Justification for testing children less than 3 years of age who are outside of the HEDIS suggested age group, when clinical judgement does not point to another infection source, can result in positivity rates between 22% and 30% as previously observed.^22,23 Patients visiting nonphysician providers and outpatient facility providers were less likely to have a revisit than those visiting primary and specialty care physicians, though slightly higher propensity for antibiotic prescriptions was seen for nonphysician providers. Pediatricians have been noted to be less likely to prescribe antibiotics without GAS testing than nonpediatric providers, and more guidelines-compliant in prescribing.²⁴

Recommendations to not test children under 3 years of age are based on the lack of acute rheumatic fever and other complications in this age group together with more frequent viral syndromes. Selectivity in applying clinical criteria to testing can be attempted to separate bacterial from viral illness. Postnasal drainage/rhinorrhea, hoarse voice, and cough have been used successfully to identify those with viral illness and less need for testing, with greater certainty of low risk for GABHS in those over 11 years of age without tonsillar exudates, cervical adenopathy, or fever.¹⁷ However, the marginal benefits of those who have all 3 features of viral illness versus none in identifying GAS positivity was 23.3% vs 37.6% - helpful, but certainly not diminishing the need for testing. These constitutional findings of viral URI also do not exclude the GAS carrier state that features these symptoms.²⁵ Others have reinforced the doubt of pharyngeal exudates as the premier diagnostic finding for test-positive GAS.²⁶

This study had several limitations. The Optum claims dataset only contains ICD-9 codes for diagnoses. It does not include data on infection severity and clinical findings related to symptoms, thus empiric treatment warranted based in clinical severity is not assessed. Antibiotics are commonly available as generics and very inexpensive. Patients may fill and pay for these prescriptions directly, in which case, a claim for payment may not be filed with Optum. This could result in an undercount of treated patients in our study.

There is no corresponding problem of missing medical claims for GAS testing which were obtained from the CPT codes within the Optum claims data set. However, we elected not to verify the test results due to these data being missing for 75% of the study population. Nevertheless, this study’s focus was less about justifying antibiotic treatment, but dealt with the outcomes generated by testing and treatment. Toward that end, we used CPT codes to identify a revisit, and while those can at times be affected by financial reimbursement incentives, differences related to revisits in the 4 patient groups should not be subject to bias.

Conclusion

This study used data from real world practices to document the patterns of GAS testing and antibiotic use in pediatric pharyngitis patients. Revisit rates were under 5% for all patient groups and the use of rapid diagnostic tools were found to be the determining factor in further reducing the risk of revisits. This supports the need for compliance with the HEDIS quality metric for pharyngitis to the recommended levels of rapid testing which have been falling in recent years. Use of more accurate antigen and newer molecular detection testing methods may help further delineate important factors in determining pediatric pharyngitis treatment and need for revisits.²⁷

Corresponding author: Jeffrey McCombs, MD, University of Southern California School of Pharmacy, Department of Pharmaceutical and Health Economics, Leonard D. Schaeffer Center for Health Policy & Economics, 635 Downey Way, Verna & Peter Dauterive Hall 310, Los Angeles, CA 90089-3333; [email protected].

Financial disclosures: None.

From the Department of Anesthesiology, Thomas Jefferson University and Hospitals, Sidney Kimmel Medical College, Philadelphia, PA, and Sidney Kimmel Medical College at Thomas Jefferson University, Philadelphia, PA.

Objective: Retrospective study examining hospital cost information of patients requiring endotracheal intubation with video laryngoscopy. Provide a practical cost assessment on use of the McGRATH and GlideScope video laryngoscopes (VLs).

Methods: This study examined 52 hospital locations within a single, large university hospital, with most of those locations being hospital operating rooms. A total of 34 600 endotracheal intubations performed over 24 months, of which 11 345 were video laryngoscopies. Electronic medical records containing demographic data and information related to endotracheal intubation procedures, with monthly breakdowns between GlideScope and McGRATH intubations, were reviewed. Cost information calculated for equipment, blades, batteries, repairs, and subsequent analysis performed to determine cost differences between those 2 instruments during the COVID-19 period.

Results: A total of 5501 video laryngoscopy procedures were performed using the McGRATH VL and 5305 were performed using the GlideScope VL. Costs over 24 months were $181 093 lower (55.5%) for McGRATH compared to GlideScope. The mean (SD) monthly costs for GlideScope blades were $3837 ($1050) and $3236 ($538) for years 1 and 2, respectively, vs $1652 ($663) and $2933 ($585) for McGRATH blades (P < .001). Most total cost differences were attributed to equipment and blade purchases, which were $202 595 (65.0%) higher for GlideScope. During the COVID-19 period, the use of the McGRATH increased to 61% of all video laryngoscopy cases, compared to 37% for GlideScope (P < .001). Blade cost difference for the COVID-19 period was $128 higher for the McGRATH even though 293 more intubations were performed with that device.

Conclusions: Use of the McGRATH resulted in a cost savings of 55% compared to the GlideScope, and its use was highest during the COVID-19 period, which may be explained by its more portable and practical features.

Keywords: video laryngoscope; McGRATH; GlideScope; endotracheal intubation; hospital costs; COVID-19.

Hospitals have come to rely on video laryngoscopes (VLs) for tracheal intubation as necessary tools for better visualization of airways. Modern video laryngoscopy developed in the 2000s¹ as a progression from direct laryngoscopy, which began in 1852 when Horace Green used a bent tongue spatula and sunlight to examine a child.² VLs have seen many improvements and adaptations of their own, resulting in many different styles and types circulating around hospitals. The GlideScope (Verathon Inc, Bothell, WA) and the McGRATH (Medtronic, Minneapolis, MN) are examples of such instruments, which are now widely used in the US and are the 2 VLs of choice at our institution.

A few studies have compared VLs to direct laryngoscopes. In their systematic review, Lewis et al have shown the numerous benefits of using a VL over a direct laryngoscope. Some general conclusions were that the use of video laryngoscopy reduced the number of failed intubations, decreased laryngeal trauma, and provided improved visualizations.³ Other studies have compared the different types of VLs, including the McGRATH and the GlideScope, examining factors such as intubation time and display quality of the image. Two studies found that medical students were equally successful at using both the McGRATH and the GlideScope,^4,5 while another study found that care providers using the GlideScope had quicker intubation times.⁶ Lastly, Savoldelli et al concluded that more providers preferred the McGRATH, which provided better laryngeal views,⁷ while their subsequent study showed more favorable learning curves of the Airtraq compared to the McGRATH and other VLs.⁸

Although there have been no reported differences in safety and effectiveness of the McGRATH and GlideScope devices, cost data on the use of these 2 popular laryngoscopes are lacking. Such information is important considering the increasing costs of medical technologies and the significant financial losses experienced by health care systems due to the COVID-19 crisis. The purpose of this retrospective cohort study was to compare the cost efficiency of the McGRATH MAC and GlideScope Core VLs at a large academic center.

Methods

This retrospective study was performed under exemption from the Thomas Jefferson University Institutional Review Board. The primary data sources consisted of hospital electronic patient records (EPIC) and cost information from the device manufacturers and hospital staff. The electronic patient data were provided by the EPIC Enterprise Analytics Business Intelligence group at Thomas Jefferson University Hospital (Center City Campus, Philadelphia, PA), while device costs were obtained from Verathon, Medtronic, and departmental staff responsible for purchasing equipment. Monthly data were obtained over a 24-month period (June 2018 through May 2020) when the McGRATH VL was placed into use in the department of anesthesiology. The 2 types of VLs were made available for use in a total of 52 locations, with the majority being hospital operating rooms.

The following variables were recorded: number of endotracheal intubations performed each month with breakdown between video laryngoscopy and flexible bronchoscopy airways, frequency of use for each type of laryngoscope, blades used, and equipment costs for use of each laryngoscope. Hospital cost estimates for both the McGRATH and GlideScope laryngoscopes included batteries, handles, blades, and the devices themselves. Cost data were also collected on frequency of device failure, maintenance, and replacement of parts and lost equipment.

Analysis

De-identified electronic medical records consisted of nominal and quantitative variables, with demographic data and information related to the endotracheal intubation procedure. All data were in chronological order and sorted by date after which coding was applied, to identify device type and allocate pertinent cost information. Descriptive statistics were reported as mean (SD) and sum for costs; frequency tables were generated for intubation procedures according to device type and time periods. Data were analyzed using the χ² test, the student t test, and the Wilcoxon Mann-Whitney U test, with a P value set at .05 for statistical significance. SPSS version 26 and GraphPad Prism version 6 were used for all statistical analyses.

A total of 34 600 endotracheal intubations were performed over the 24-month study period, and 11 345 (32.8%) were video laryngoscopy procedures. Out of all video laryngoscopy procedures, 5501 (48.5%) were performed using the McGRATH VL and 5305 (46.8%) were conducted using the GlideScope VL. The difference of 539 (4.8%) cases accounts for flexible bronchoscopy procedures and endotracheal intubations using other video laryngoscopy equipment. The mean (SD) monthly number of video laryngoscopy procedures for the 24 months was 221 (54) and 229 (89) for the GlideScope and McGRATH devices, respectively. Monthly endotracheal intubation distributions over 24 months trended upward for the McGRATH VL and downward for the GlideScope, but there was no statistically significant (P = .71) difference in overall use between the 2 instruments (Figure 1).

To examine the observed usage trends between the 2 VL during the first and last 12 months, a univariate ANOVA was conducted with the 2 time periods entered as predictors in the model. Video laryngoscopy intubations were performed (P = .001) more frequently with the GlideScope during the first 12 months; however, use of the McGRATH VL increased (P < .001) during the following 12 months compared to GlideScope. The GlideScope accounted for 54% of all VL intubations during the first 12 months, with the McGRATH accounting for 58% of all video laryngoscopy procedures for months 12 to 24. Additionally, the increase in video laryngoscopy procedures with the McGRATH during the last 3 months of the study period was despite an overall reduction in surgical volume due to the COVID-19 crisis, defined for this study as March 1, 2020, to May 31, 2020 (Figure 1). There was a statistically significant (P < .001) difference in the case distribution between use of the McGRATH and GlideScope VL for that period. The anesthesia personnel’s use of the McGRATH VL increased to 61% of all video laryngoscopy cases, compared to 37% for the GlideScope (Figure 2).

The total costs calculated for equipment, blades, and repairs are presented in Table 1 and yearly total costs are shown in Figure 3. Overall costs were $181 093 lower (55.5%) for the McGRATH VL compared to the GlideScope over the 24-month period. The mean (SD) monthly costs for GlideScope VL blades were $3837 ($1050) and $3236 ($538) for years 1 and 2, respectively, vs $1652 ($663) and $2933 ($585) for the McGRATH VL blades. Most of the total cost differences were attributed to equipment and blade purchases, which were $202 595 (65.0%) higher for the GlideScope compared to the McGRATH VL. The monthly blade costs alone were higher (P < .001) for the GlideScope over the 2-year period; however, the McGRATH VL required use of disposable stylets at a cost of $10 177 for all endotracheal intubations, compared to $700 for the GlideScope device.

An analysis was performed to determine whether costs differed between those 2 instruments during the COVID-19 period. There was a statistically significant (P < .001) difference in the case distribution between use of the McGRATH and GlideScope VLs during that period. The calculated blade cost difference for the COVID period was $128 higher for the McGRATH even though 293 more intubations were performed with that device (Table 2).

Discussion

We attempted to provide useful cost estimates by presenting pricing data reflecting the approximate cost that most large institutional anesthesia practices would incur for using those 2 specific devices and related peripherals. The main findings of our analysis showed that use of the McGRATH MAC VL resulted in a 55% cost savings compared to the GlideScope, with a similar number of cases performed with each device over the 24-month study period. We believe this represents a substantial savings to the department and institution, which has prompted internal review on the use of video laryngoscopy equipment. None of the McGRATH units failed; however, the GlideScope required 3 baton replacements.

Of note, use of the McGRATH MAC increased during the COVID-19 period, which may be explained by the fact that the operators found it to be a more portable device. Several physicians in the department commented that its smaller size made the McGRATH MAC more practical during the time when a plexiglass box was being used around the patient’s head to shield the intubator from aerosolized viral particles.

Although this study demonstrated the cost-saving value of the McGRATH over the GlideScope, a suggested next step would be to examine resource utilization related to video laryngoscopy use. The more dynamic tracking of the use of these devices should facilitate the assessment of existing related resources and decision making, to optimize the benefits of this initiative. We would anticipate reduced use of anesthesia personnel, such as technicians to assist with the management of this device which could be significant. As new respiratory viruses are appearing each year, video laryngoscopy will continue to gain increasing use in operating rooms and acute care locations. The adding of protective barriers between patients and providers calls for use of the most practical and effective VL devices, to protect personnel who are at high risk of contamination from airway secretions and aerosolized particles.^9,10

The COVID-19 pandemic has demonstrated the value of anesthesiology in regards to analyzing and finding solutions to effectively manage infected patients or those suspected of infection in the perioperative environment. Inexpensive products are often avoided because cheaper devices are associated with being of lower quality. However, the association with cost and quality—and the assumption that a higher price is positively correlated with higher quality—is overall inconsistent in the medical literature.¹¹ A more effective or higher quality treatment does not necessarily cost more and may actually end up costing less,¹² as was the case in this study. We have been able to directly cut departmental expenses by using a more efficient and cost-effective device for intubations, without compromising safety and efficacy. Future studies should determine whether this significant reduction in costs from video laryngoscopy intubations with the McGRATH VL will be sustained across anesthesiology departments in the Jefferson Health Enterprise Hospitals, or other health systems, as well as its impact on workflow and personnel resources.

This analysis was restricted to one of the campuses of the Jefferson Health Enterprise. However, this is the largest anesthesia practice, encompassing several locations, which should reflect the general practice patterns across other anesthesiology departments in this large institution. The costs for the devices and peripherals may vary across anesthesia practices depending on volume and contracts negotiated with the suppliers. It was not possible to estimate this variability, which could change the total costs by a few percentage points. We recognize that there may be other costs associated with securing the McGRATH VL to prevent loss from theft or misplacement, which were not included in the study. Lastly, the inability to obtain randomized samples for the 2 groups treated with each device opens up the possibility of selection bias. There were, however, multiple intubators who were free to select 1 of the devices for endotracheal intubation, which may have reduced the effect of selection bias.

Conclusion

This study demonstrated that over a 24-month period use of the McGRATH MAC VL resulted in a cost reduction of around 55% compared to using the GlideScope for endotracheal intubation procedures performed at a major academic center. Over the first 3 months of the COVID-19 crisis, which our study included, use of the McGRATH VL increased while GlideScope use decreased. This was most likely related to the portability and smaller size of the McGRATH, which better facilitated intubations of COVID-19 patients.

Acknowledgements: The authors thank Craig Smith, Senior Anesthesia Technician, for his assistance with the cost information and excellent record-keeping related to the use of video laryngoscopes.

Corresponding author: Marc C. Torjman, PhD, Professor, Department of Anesthesiology, Sidney Kimmel Medical College at Thomas Jefferson University, 111 South 11th St, Suite G-8290, Philadelphia, PA 19107; [email protected].

Financial disclosures: Dr. Thaler has served as a consultant for Medtronic since September 2020. He has participated in 2 webinars on the routine use of video laryngoscopy.

Funding: This study was supported by the Department of Anesthesiology at Thomas Jefferson University.

From the Department of Anesthesiology, Thomas Jefferson University and Hospitals, Sidney Kimmel Medical College, Philadelphia, PA, and Sidney Kimmel Medical College at Thomas Jefferson University, Philadelphia, PA.

Objective: Retrospective study examining hospital cost information of patients requiring endotracheal intubation with video laryngoscopy. Provide a practical cost assessment on use of the McGRATH and GlideScope video laryngoscopes (VLs).

Methods: This study examined 52 hospital locations within a single, large university hospital, with most of those locations being hospital operating rooms. A total of 34 600 endotracheal intubations performed over 24 months, of which 11 345 were video laryngoscopies. Electronic medical records containing demographic data and information related to endotracheal intubation procedures, with monthly breakdowns between GlideScope and McGRATH intubations, were reviewed. Cost information calculated for equipment, blades, batteries, repairs, and subsequent analysis performed to determine cost differences between those 2 instruments during the COVID-19 period.

Results: A total of 5501 video laryngoscopy procedures were performed using the McGRATH VL and 5305 were performed using the GlideScope VL. Costs over 24 months were $181 093 lower (55.5%) for McGRATH compared to GlideScope. The mean (SD) monthly costs for GlideScope blades were $3837 ($1050) and $3236 ($538) for years 1 and 2, respectively, vs $1652 ($663) and $2933 ($585) for McGRATH blades (P < .001). Most total cost differences were attributed to equipment and blade purchases, which were $202 595 (65.0%) higher for GlideScope. During the COVID-19 period, the use of the McGRATH increased to 61% of all video laryngoscopy cases, compared to 37% for GlideScope (P < .001). Blade cost difference for the COVID-19 period was $128 higher for the McGRATH even though 293 more intubations were performed with that device.

Conclusions: Use of the McGRATH resulted in a cost savings of 55% compared to the GlideScope, and its use was highest during the COVID-19 period, which may be explained by its more portable and practical features.

Keywords: video laryngoscope; McGRATH; GlideScope; endotracheal intubation; hospital costs; COVID-19.

Hospitals have come to rely on video laryngoscopes (VLs) for tracheal intubation as necessary tools for better visualization of airways. Modern video laryngoscopy developed in the 2000s¹ as a progression from direct laryngoscopy, which began in 1852 when Horace Green used a bent tongue spatula and sunlight to examine a child.² VLs have seen many improvements and adaptations of their own, resulting in many different styles and types circulating around hospitals. The GlideScope (Verathon Inc, Bothell, WA) and the McGRATH (Medtronic, Minneapolis, MN) are examples of such instruments, which are now widely used in the US and are the 2 VLs of choice at our institution.

A few studies have compared VLs to direct laryngoscopes. In their systematic review, Lewis et al have shown the numerous benefits of using a VL over a direct laryngoscope. Some general conclusions were that the use of video laryngoscopy reduced the number of failed intubations, decreased laryngeal trauma, and provided improved visualizations.³ Other studies have compared the different types of VLs, including the McGRATH and the GlideScope, examining factors such as intubation time and display quality of the image. Two studies found that medical students were equally successful at using both the McGRATH and the GlideScope,^4,5 while another study found that care providers using the GlideScope had quicker intubation times.⁶ Lastly, Savoldelli et al concluded that more providers preferred the McGRATH, which provided better laryngeal views,⁷ while their subsequent study showed more favorable learning curves of the Airtraq compared to the McGRATH and other VLs.⁸

Although there have been no reported differences in safety and effectiveness of the McGRATH and GlideScope devices, cost data on the use of these 2 popular laryngoscopes are lacking. Such information is important considering the increasing costs of medical technologies and the significant financial losses experienced by health care systems due to the COVID-19 crisis. The purpose of this retrospective cohort study was to compare the cost efficiency of the McGRATH MAC and GlideScope Core VLs at a large academic center.

Methods

This retrospective study was performed under exemption from the Thomas Jefferson University Institutional Review Board. The primary data sources consisted of hospital electronic patient records (EPIC) and cost information from the device manufacturers and hospital staff. The electronic patient data were provided by the EPIC Enterprise Analytics Business Intelligence group at Thomas Jefferson University Hospital (Center City Campus, Philadelphia, PA), while device costs were obtained from Verathon, Medtronic, and departmental staff responsible for purchasing equipment. Monthly data were obtained over a 24-month period (June 2018 through May 2020) when the McGRATH VL was placed into use in the department of anesthesiology. The 2 types of VLs were made available for use in a total of 52 locations, with the majority being hospital operating rooms.

The following variables were recorded: number of endotracheal intubations performed each month with breakdown between video laryngoscopy and flexible bronchoscopy airways, frequency of use for each type of laryngoscope, blades used, and equipment costs for use of each laryngoscope. Hospital cost estimates for both the McGRATH and GlideScope laryngoscopes included batteries, handles, blades, and the devices themselves. Cost data were also collected on frequency of device failure, maintenance, and replacement of parts and lost equipment.

Analysis

De-identified electronic medical records consisted of nominal and quantitative variables, with demographic data and information related to the endotracheal intubation procedure. All data were in chronological order and sorted by date after which coding was applied, to identify device type and allocate pertinent cost information. Descriptive statistics were reported as mean (SD) and sum for costs; frequency tables were generated for intubation procedures according to device type and time periods. Data were analyzed using the χ² test, the student t test, and the Wilcoxon Mann-Whitney U test, with a P value set at .05 for statistical significance. SPSS version 26 and GraphPad Prism version 6 were used for all statistical analyses.

A total of 34 600 endotracheal intubations were performed over the 24-month study period, and 11 345 (32.8%) were video laryngoscopy procedures. Out of all video laryngoscopy procedures, 5501 (48.5%) were performed using the McGRATH VL and 5305 (46.8%) were conducted using the GlideScope VL. The difference of 539 (4.8%) cases accounts for flexible bronchoscopy procedures and endotracheal intubations using other video laryngoscopy equipment. The mean (SD) monthly number of video laryngoscopy procedures for the 24 months was 221 (54) and 229 (89) for the GlideScope and McGRATH devices, respectively. Monthly endotracheal intubation distributions over 24 months trended upward for the McGRATH VL and downward for the GlideScope, but there was no statistically significant (P = .71) difference in overall use between the 2 instruments (Figure 1).

To examine the observed usage trends between the 2 VL during the first and last 12 months, a univariate ANOVA was conducted with the 2 time periods entered as predictors in the model. Video laryngoscopy intubations were performed (P = .001) more frequently with the GlideScope during the first 12 months; however, use of the McGRATH VL increased (P < .001) during the following 12 months compared to GlideScope. The GlideScope accounted for 54% of all VL intubations during the first 12 months, with the McGRATH accounting for 58% of all video laryngoscopy procedures for months 12 to 24. Additionally, the increase in video laryngoscopy procedures with the McGRATH during the last 3 months of the study period was despite an overall reduction in surgical volume due to the COVID-19 crisis, defined for this study as March 1, 2020, to May 31, 2020 (Figure 1). There was a statistically significant (P < .001) difference in the case distribution between use of the McGRATH and GlideScope VL for that period. The anesthesia personnel’s use of the McGRATH VL increased to 61% of all video laryngoscopy cases, compared to 37% for the GlideScope (Figure 2).

The total costs calculated for equipment, blades, and repairs are presented in Table 1 and yearly total costs are shown in Figure 3. Overall costs were $181 093 lower (55.5%) for the McGRATH VL compared to the GlideScope over the 24-month period. The mean (SD) monthly costs for GlideScope VL blades were $3837 ($1050) and $3236 ($538) for years 1 and 2, respectively, vs $1652 ($663) and $2933 ($585) for the McGRATH VL blades. Most of the total cost differences were attributed to equipment and blade purchases, which were $202 595 (65.0%) higher for the GlideScope compared to the McGRATH VL. The monthly blade costs alone were higher (P < .001) for the GlideScope over the 2-year period; however, the McGRATH VL required use of disposable stylets at a cost of $10 177 for all endotracheal intubations, compared to $700 for the GlideScope device.

An analysis was performed to determine whether costs differed between those 2 instruments during the COVID-19 period. There was a statistically significant (P < .001) difference in the case distribution between use of the McGRATH and GlideScope VLs during that period. The calculated blade cost difference for the COVID period was $128 higher for the McGRATH even though 293 more intubations were performed with that device (Table 2).

Discussion

We attempted to provide useful cost estimates by presenting pricing data reflecting the approximate cost that most large institutional anesthesia practices would incur for using those 2 specific devices and related peripherals. The main findings of our analysis showed that use of the McGRATH MAC VL resulted in a 55% cost savings compared to the GlideScope, with a similar number of cases performed with each device over the 24-month study period. We believe this represents a substantial savings to the department and institution, which has prompted internal review on the use of video laryngoscopy equipment. None of the McGRATH units failed; however, the GlideScope required 3 baton replacements.

Of note, use of the McGRATH MAC increased during the COVID-19 period, which may be explained by the fact that the operators found it to be a more portable device. Several physicians in the department commented that its smaller size made the McGRATH MAC more practical during the time when a plexiglass box was being used around the patient’s head to shield the intubator from aerosolized viral particles.

Although this study demonstrated the cost-saving value of the McGRATH over the GlideScope, a suggested next step would be to examine resource utilization related to video laryngoscopy use. The more dynamic tracking of the use of these devices should facilitate the assessment of existing related resources and decision making, to optimize the benefits of this initiative. We would anticipate reduced use of anesthesia personnel, such as technicians to assist with the management of this device which could be significant. As new respiratory viruses are appearing each year, video laryngoscopy will continue to gain increasing use in operating rooms and acute care locations. The adding of protective barriers between patients and providers calls for use of the most practical and effective VL devices, to protect personnel who are at high risk of contamination from airway secretions and aerosolized particles.^9,10

The COVID-19 pandemic has demonstrated the value of anesthesiology in regards to analyzing and finding solutions to effectively manage infected patients or those suspected of infection in the perioperative environment. Inexpensive products are often avoided because cheaper devices are associated with being of lower quality. However, the association with cost and quality—and the assumption that a higher price is positively correlated with higher quality—is overall inconsistent in the medical literature.¹¹ A more effective or higher quality treatment does not necessarily cost more and may actually end up costing less,¹² as was the case in this study. We have been able to directly cut departmental expenses by using a more efficient and cost-effective device for intubations, without compromising safety and efficacy. Future studies should determine whether this significant reduction in costs from video laryngoscopy intubations with the McGRATH VL will be sustained across anesthesiology departments in the Jefferson Health Enterprise Hospitals, or other health systems, as well as its impact on workflow and personnel resources.

This analysis was restricted to one of the campuses of the Jefferson Health Enterprise. However, this is the largest anesthesia practice, encompassing several locations, which should reflect the general practice patterns across other anesthesiology departments in this large institution. The costs for the devices and peripherals may vary across anesthesia practices depending on volume and contracts negotiated with the suppliers. It was not possible to estimate this variability, which could change the total costs by a few percentage points. We recognize that there may be other costs associated with securing the McGRATH VL to prevent loss from theft or misplacement, which were not included in the study. Lastly, the inability to obtain randomized samples for the 2 groups treated with each device opens up the possibility of selection bias. There were, however, multiple intubators who were free to select 1 of the devices for endotracheal intubation, which may have reduced the effect of selection bias.

Conclusion

This study demonstrated that over a 24-month period use of the McGRATH MAC VL resulted in a cost reduction of around 55% compared to using the GlideScope for endotracheal intubation procedures performed at a major academic center. Over the first 3 months of the COVID-19 crisis, which our study included, use of the McGRATH VL increased while GlideScope use decreased. This was most likely related to the portability and smaller size of the McGRATH, which better facilitated intubations of COVID-19 patients.

Acknowledgements: The authors thank Craig Smith, Senior Anesthesia Technician, for his assistance with the cost information and excellent record-keeping related to the use of video laryngoscopes.

Corresponding author: Marc C. Torjman, PhD, Professor, Department of Anesthesiology, Sidney Kimmel Medical College at Thomas Jefferson University, 111 South 11th St, Suite G-8290, Philadelphia, PA 19107; [email protected].

Financial disclosures: Dr. Thaler has served as a consultant for Medtronic since September 2020. He has participated in 2 webinars on the routine use of video laryngoscopy.

Funding: This study was supported by the Department of Anesthesiology at Thomas Jefferson University.

From the Department of Anesthesiology, Thomas Jefferson University and Hospitals, Sidney Kimmel Medical College, Philadelphia, PA, and Sidney Kimmel Medical College at Thomas Jefferson University, Philadelphia, PA.

Objective: Retrospective study examining hospital cost information of patients requiring endotracheal intubation with video laryngoscopy. Provide a practical cost assessment on use of the McGRATH and GlideScope video laryngoscopes (VLs).

Methods: This study examined 52 hospital locations within a single, large university hospital, with most of those locations being hospital operating rooms. A total of 34 600 endotracheal intubations performed over 24 months, of which 11 345 were video laryngoscopies. Electronic medical records containing demographic data and information related to endotracheal intubation procedures, with monthly breakdowns between GlideScope and McGRATH intubations, were reviewed. Cost information calculated for equipment, blades, batteries, repairs, and subsequent analysis performed to determine cost differences between those 2 instruments during the COVID-19 period.

Results: A total of 5501 video laryngoscopy procedures were performed using the McGRATH VL and 5305 were performed using the GlideScope VL. Costs over 24 months were $181 093 lower (55.5%) for McGRATH compared to GlideScope. The mean (SD) monthly costs for GlideScope blades were $3837 ($1050) and $3236 ($538) for years 1 and 2, respectively, vs $1652 ($663) and $2933 ($585) for McGRATH blades (P < .001). Most total cost differences were attributed to equipment and blade purchases, which were $202 595 (65.0%) higher for GlideScope. During the COVID-19 period, the use of the McGRATH increased to 61% of all video laryngoscopy cases, compared to 37% for GlideScope (P < .001). Blade cost difference for the COVID-19 period was $128 higher for the McGRATH even though 293 more intubations were performed with that device.

Conclusions: Use of the McGRATH resulted in a cost savings of 55% compared to the GlideScope, and its use was highest during the COVID-19 period, which may be explained by its more portable and practical features.

Keywords: video laryngoscope; McGRATH; GlideScope; endotracheal intubation; hospital costs; COVID-19.

Hospitals have come to rely on video laryngoscopes (VLs) for tracheal intubation as necessary tools for better visualization of airways. Modern video laryngoscopy developed in the 2000s¹ as a progression from direct laryngoscopy, which began in 1852 when Horace Green used a bent tongue spatula and sunlight to examine a child.² VLs have seen many improvements and adaptations of their own, resulting in many different styles and types circulating around hospitals. The GlideScope (Verathon Inc, Bothell, WA) and the McGRATH (Medtronic, Minneapolis, MN) are examples of such instruments, which are now widely used in the US and are the 2 VLs of choice at our institution.

A few studies have compared VLs to direct laryngoscopes. In their systematic review, Lewis et al have shown the numerous benefits of using a VL over a direct laryngoscope. Some general conclusions were that the use of video laryngoscopy reduced the number of failed intubations, decreased laryngeal trauma, and provided improved visualizations.³ Other studies have compared the different types of VLs, including the McGRATH and the GlideScope, examining factors such as intubation time and display quality of the image. Two studies found that medical students were equally successful at using both the McGRATH and the GlideScope,^4,5 while another study found that care providers using the GlideScope had quicker intubation times.⁶ Lastly, Savoldelli et al concluded that more providers preferred the McGRATH, which provided better laryngeal views,⁷ while their subsequent study showed more favorable learning curves of the Airtraq compared to the McGRATH and other VLs.⁸

Although there have been no reported differences in safety and effectiveness of the McGRATH and GlideScope devices, cost data on the use of these 2 popular laryngoscopes are lacking. Such information is important considering the increasing costs of medical technologies and the significant financial losses experienced by health care systems due to the COVID-19 crisis. The purpose of this retrospective cohort study was to compare the cost efficiency of the McGRATH MAC and GlideScope Core VLs at a large academic center.

Methods

This retrospective study was performed under exemption from the Thomas Jefferson University Institutional Review Board. The primary data sources consisted of hospital electronic patient records (EPIC) and cost information from the device manufacturers and hospital staff. The electronic patient data were provided by the EPIC Enterprise Analytics Business Intelligence group at Thomas Jefferson University Hospital (Center City Campus, Philadelphia, PA), while device costs were obtained from Verathon, Medtronic, and departmental staff responsible for purchasing equipment. Monthly data were obtained over a 24-month period (June 2018 through May 2020) when the McGRATH VL was placed into use in the department of anesthesiology. The 2 types of VLs were made available for use in a total of 52 locations, with the majority being hospital operating rooms.

The following variables were recorded: number of endotracheal intubations performed each month with breakdown between video laryngoscopy and flexible bronchoscopy airways, frequency of use for each type of laryngoscope, blades used, and equipment costs for use of each laryngoscope. Hospital cost estimates for both the McGRATH and GlideScope laryngoscopes included batteries, handles, blades, and the devices themselves. Cost data were also collected on frequency of device failure, maintenance, and replacement of parts and lost equipment.

Analysis

De-identified electronic medical records consisted of nominal and quantitative variables, with demographic data and information related to the endotracheal intubation procedure. All data were in chronological order and sorted by date after which coding was applied, to identify device type and allocate pertinent cost information. Descriptive statistics were reported as mean (SD) and sum for costs; frequency tables were generated for intubation procedures according to device type and time periods. Data were analyzed using the χ² test, the student t test, and the Wilcoxon Mann-Whitney U test, with a P value set at .05 for statistical significance. SPSS version 26 and GraphPad Prism version 6 were used for all statistical analyses.

A total of 34 600 endotracheal intubations were performed over the 24-month study period, and 11 345 (32.8%) were video laryngoscopy procedures. Out of all video laryngoscopy procedures, 5501 (48.5%) were performed using the McGRATH VL and 5305 (46.8%) were conducted using the GlideScope VL. The difference of 539 (4.8%) cases accounts for flexible bronchoscopy procedures and endotracheal intubations using other video laryngoscopy equipment. The mean (SD) monthly number of video laryngoscopy procedures for the 24 months was 221 (54) and 229 (89) for the GlideScope and McGRATH devices, respectively. Monthly endotracheal intubation distributions over 24 months trended upward for the McGRATH VL and downward for the GlideScope, but there was no statistically significant (P = .71) difference in overall use between the 2 instruments (Figure 1).

To examine the observed usage trends between the 2 VL during the first and last 12 months, a univariate ANOVA was conducted with the 2 time periods entered as predictors in the model. Video laryngoscopy intubations were performed (P = .001) more frequently with the GlideScope during the first 12 months; however, use of the McGRATH VL increased (P < .001) during the following 12 months compared to GlideScope. The GlideScope accounted for 54% of all VL intubations during the first 12 months, with the McGRATH accounting for 58% of all video laryngoscopy procedures for months 12 to 24. Additionally, the increase in video laryngoscopy procedures with the McGRATH during the last 3 months of the study period was despite an overall reduction in surgical volume due to the COVID-19 crisis, defined for this study as March 1, 2020, to May 31, 2020 (Figure 1). There was a statistically significant (P < .001) difference in the case distribution between use of the McGRATH and GlideScope VL for that period. The anesthesia personnel’s use of the McGRATH VL increased to 61% of all video laryngoscopy cases, compared to 37% for the GlideScope (Figure 2).

The total costs calculated for equipment, blades, and repairs are presented in Table 1 and yearly total costs are shown in Figure 3. Overall costs were $181 093 lower (55.5%) for the McGRATH VL compared to the GlideScope over the 24-month period. The mean (SD) monthly costs for GlideScope VL blades were $3837 ($1050) and $3236 ($538) for years 1 and 2, respectively, vs $1652 ($663) and $2933 ($585) for the McGRATH VL blades. Most of the total cost differences were attributed to equipment and blade purchases, which were $202 595 (65.0%) higher for the GlideScope compared to the McGRATH VL. The monthly blade costs alone were higher (P < .001) for the GlideScope over the 2-year period; however, the McGRATH VL required use of disposable stylets at a cost of $10 177 for all endotracheal intubations, compared to $700 for the GlideScope device.

An analysis was performed to determine whether costs differed between those 2 instruments during the COVID-19 period. There was a statistically significant (P < .001) difference in the case distribution between use of the McGRATH and GlideScope VLs during that period. The calculated blade cost difference for the COVID period was $128 higher for the McGRATH even though 293 more intubations were performed with that device (Table 2).

Discussion

We attempted to provide useful cost estimates by presenting pricing data reflecting the approximate cost that most large institutional anesthesia practices would incur for using those 2 specific devices and related peripherals. The main findings of our analysis showed that use of the McGRATH MAC VL resulted in a 55% cost savings compared to the GlideScope, with a similar number of cases performed with each device over the 24-month study period. We believe this represents a substantial savings to the department and institution, which has prompted internal review on the use of video laryngoscopy equipment. None of the McGRATH units failed; however, the GlideScope required 3 baton replacements.

Of note, use of the McGRATH MAC increased during the COVID-19 period, which may be explained by the fact that the operators found it to be a more portable device. Several physicians in the department commented that its smaller size made the McGRATH MAC more practical during the time when a plexiglass box was being used around the patient’s head to shield the intubator from aerosolized viral particles.

Although this study demonstrated the cost-saving value of the McGRATH over the GlideScope, a suggested next step would be to examine resource utilization related to video laryngoscopy use. The more dynamic tracking of the use of these devices should facilitate the assessment of existing related resources and decision making, to optimize the benefits of this initiative. We would anticipate reduced use of anesthesia personnel, such as technicians to assist with the management of this device which could be significant. As new respiratory viruses are appearing each year, video laryngoscopy will continue to gain increasing use in operating rooms and acute care locations. The adding of protective barriers between patients and providers calls for use of the most practical and effective VL devices, to protect personnel who are at high risk of contamination from airway secretions and aerosolized particles.^9,10

The COVID-19 pandemic has demonstrated the value of anesthesiology in regards to analyzing and finding solutions to effectively manage infected patients or those suspected of infection in the perioperative environment. Inexpensive products are often avoided because cheaper devices are associated with being of lower quality. However, the association with cost and quality—and the assumption that a higher price is positively correlated with higher quality—is overall inconsistent in the medical literature.¹¹ A more effective or higher quality treatment does not necessarily cost more and may actually end up costing less,¹² as was the case in this study. We have been able to directly cut departmental expenses by using a more efficient and cost-effective device for intubations, without compromising safety and efficacy. Future studies should determine whether this significant reduction in costs from video laryngoscopy intubations with the McGRATH VL will be sustained across anesthesiology departments in the Jefferson Health Enterprise Hospitals, or other health systems, as well as its impact on workflow and personnel resources.

This analysis was restricted to one of the campuses of the Jefferson Health Enterprise. However, this is the largest anesthesia practice, encompassing several locations, which should reflect the general practice patterns across other anesthesiology departments in this large institution. The costs for the devices and peripherals may vary across anesthesia practices depending on volume and contracts negotiated with the suppliers. It was not possible to estimate this variability, which could change the total costs by a few percentage points. We recognize that there may be other costs associated with securing the McGRATH VL to prevent loss from theft or misplacement, which were not included in the study. Lastly, the inability to obtain randomized samples for the 2 groups treated with each device opens up the possibility of selection bias. There were, however, multiple intubators who were free to select 1 of the devices for endotracheal intubation, which may have reduced the effect of selection bias.

Conclusion

This study demonstrated that over a 24-month period use of the McGRATH MAC VL resulted in a cost reduction of around 55% compared to using the GlideScope for endotracheal intubation procedures performed at a major academic center. Over the first 3 months of the COVID-19 crisis, which our study included, use of the McGRATH VL increased while GlideScope use decreased. This was most likely related to the portability and smaller size of the McGRATH, which better facilitated intubations of COVID-19 patients.

Acknowledgements: The authors thank Craig Smith, Senior Anesthesia Technician, for his assistance with the cost information and excellent record-keeping related to the use of video laryngoscopes.

Corresponding author: Marc C. Torjman, PhD, Professor, Department of Anesthesiology, Sidney Kimmel Medical College at Thomas Jefferson University, 111 South 11th St, Suite G-8290, Philadelphia, PA 19107; [email protected].

Financial disclosures: Dr. Thaler has served as a consultant for Medtronic since September 2020. He has participated in 2 webinars on the routine use of video laryngoscopy.

Funding: This study was supported by the Department of Anesthesiology at Thomas Jefferson University.

From Children’s Health System of Texas, Division of Pediatric Anesthesiology, Dallas, TX (Drs. Sakhai, Bocanegra, Chandran, Kimatian, and Kiss), UT Southwestern Medical Center, Department of Anesthesiology and Pain Management, Dallas, TX (Drs. Bocanegra, Chandran, Kimatian, and Kiss), and UT Southwestern Medical Center, Department of Population and Data Sciences, Dallas, TX (Dr. Reisch).

Objective: Policy-driven changes in medical practice have long been the norm. Seldom are changes in clinical practice sought to be brought about by a person’s tendency toward growth or self‐actualization. Many hospitals have instituted hypothermia bundles to help reduce the incidence of unanticipated postoperative hypothermia. Although successful in the short-term, sustained changes are difficult to maintain. We implemented a quality-improvement project focused on addressing the affective components of self-determination theory (SDT) to create sustainable behavioral change while satisfying providers’ basic psychological needs for autonomy, competence, and relatedness.

Methods: A total of 3 Plan-Do-Study-Act (PDSA) cycles were enacted over the span of 14 months at a major tertiary care pediatric hospital to recruit and motivate anesthesia providers and perioperative team members to reduce the percentage of hypothermic postsurgical patients by 50%. As an optional initial incentive for participation, anesthesiologists would qualify for American Board of Anesthesiology Maintenance of Certification in Anesthesiology (MOCA) Part 4 Quality Improvement credits for monitoring their own temperature data and participating in project-related meetings. Providers were given autonomy to develop a personal plan for achieving the desired goals.

Results: The median rate of hypothermia was reduced from 6.9% to 1.6% in July 2019 and was reduced again in July 2020 to 1.3%, an 81% reduction overall. A low hypothermia rate was successfully maintained for at least 21 subsequent months after participants received their MOCA credits in July 2019.

Conclusions: Using an approach that focused on the elements of competency, autonomy, and relatedness central to the principles of SDT, we observed the development of a new culture of vigilance for prevention of hypothermia that successfully endured beyond the project end date.

Keywords: postoperative hypothermia; self-determination theory; motivation; quality improvement.

Perioperative hypothermia, generally accepted as a core temperature less than 36 °C in clinical practice, is a common complication in the pediatric surgical population and is associated with poor postoperative outcomes.¹ Hypothermic patients may develop respiratory depression, hypoglycemia, and metabolic acidosis that may lead to decreased oxygen delivery and end organ tissue hypoxia.^2-4 Other potential detrimental effects of failing to maintain normal body temperature are impaired clotting factor enzyme function and platelet dysfunction, increasing the risk for postoperative bleeding.^5,6 In addition, there are financial implications when hypothermic patients require care and resources postoperatively because of delayed emergence or shivering.⁷

The American Society of Anesthesiologists recommends intraoperative temperature monitoring for procedures when clinically significant changes in body temperature are anticipated.⁸ Maintenance of normothermia in the pediatric population is especially challenging owing to a larger skin-surface area compared with body mass ratio and less subcutaneous fat content than in adults. Preventing postoperative hypothermia starts preoperatively with parental education and can be as simple as covering the child with a blanket and setting the preoperative room to an acceptably warm temperature.^9,10 Intraoperatively, maintaining operating room (OR) temperatures at or above 21.1 °C and using active warming devices and radiant warmers when appropriate are important techniques to preserve the child’s body temperature.^11,12

Despite the knowledge of these risks and vigilant avoidance of hypothermia, unplanned perioperative hypothermia can occur in up to 70% of surgical patients.¹ Beyond the clinical benefits, as health care marches toward a value-based payment methodology, quality indicators such as avoiding hypothermia may be linked directly to payment.

Self-determination theory (SDT) was first developed in 1980 by Deci and Ryan.¹³ The central premise of the theory states that people develop their full potential if circumstances allow them to satisfy their basic psychological needs: autonomy, competence, and relatedness. Under these conditions, people’s natural inclination toward growth can be realized, and they are more likely to internalize external goals. Under an extrinsic reward system, motivation can waver, as people may perceive rewards as controlling.

Many institutions have implemented hypothermia bundles to help decrease the rate of hypothermic patients, but while initially successful, the effectiveness of these interventions tends to fade over time as participants settle into old, comfortable routines.¹⁴ With SDT in mind, we designed our quality-improvement (QI) project with interventions to allow clinicians autonomy without instituting rigid guidelines or punitive actions. We aimed to directly address the affective components central to motivation and engagement so that we could bring about long-term meaningful changes in our practice.

Methods

Setting

The hypothermia QI intervention was instituted at a major tertiary care children’s hospital that performs more than 40 000 pediatric general anesthetics annually. Our division of pediatric anesthesiology consists of 66 fellowship-trained pediatric anesthesiologists, 15 or more rotating trainees per month, 13 anesthesiology assistants, 15 anesthesia technicians, and more than 50 perioperative nurses.

The most frequent pediatric surgeries include, but are not limited to, general surgery, otolaryngology, urology, gastroenterology, plastic surgery, neurosurgery, and dentistry. The surgeries are conducted in the hospital’s main operative floor, which consists of 15 ORs and 2 gastroenterology procedure rooms. Although the implementation of the QI project included several operating sites, we focused on collecting temperature data from surgical patients at our main campus recovery unit. We obtained the patients’ initial temperatures upon arrival to the recovery unit from a retrospective electronic health record review of all patients who underwent anesthesia from January 2016 through April 2021.

Postoperative hypothermia was identified as an area of potential improvement after several patients were reported to be hypothermic upon arrival to the recovery unit in the later part of 2018. Further review revealed significant heterogeneity of practices and lack of standardization of patient-warming methods. By comparing the temperatures pre- and postintervention, we could measure the effectiveness of the QI initiative. Prior to the start of our project, the hypothermia rate in our patient population was not actively tracked, and the effectiveness of our variable practice was not measured.

The cutoff for hypothermia for our QI project was defined as body temperature below 36 °C, since this value has been previously used in the literature and is commonly accepted in anesthesia practice as the delineation for hypothermia in patients undergoing general anesthesia.¹

Interventions

This QI project was designed and modeled after the Institute for Healthcare Improvement Model for Improvement.¹⁵ Three cycles of Plan-Do-Study-Act (PDSA) were developed and instituted over a 14-month period until December 2019 (Table 1).

A retrospective review was conducted to determine the percentage of surgical patients arriving to our recovery units with an initial temperature reading of less than 36 °C. A project key driver diagram and smart aim were created and approved by the hospital’s continuing medical education (CME) committee for credit via the American Board of Medical Specialties (ABMS) Multi-Specialty Portfolio Program, Maintenance of Certification in Anesthesiology (MOCA) Part 4.

The first PDSA cycle involved introducing the QI project and sharing the aims of the project at a department grand rounds in the latter part of October 2018. Enrollment to participate in the project was open to all anesthesiologists in the division, and participants could earn up to 20 hours of MOCA Part 4 credits. A spreadsheet was developed and maintained to track each anesthesiologist’s monthly percentage of hypothermic patients. The de-identified patient data were shared with the division via monthly emails. In addition, individual providers with a hypothermic patient in the recovery room received a notification email.

The anesthesiologists participated in the QI project by reviewing their personal percentage of hypothermic patients on an ongoing basis to earn the credit. There was no explicit requirement to decrease their own rate of patients with body temperature less than 36 °C or expectation to achieve a predetermined goal, so the participants could not “fail.”

Because of the large interest in this project, a hypothermia committee was formed that consisted of 36 anesthesiologists. This group reviewed the data and exchanged ideas for improvement in November 2018 as part of the first PDSA cycle. The committee met monthly and was responsible for actively engaging other members of the department and perioperative staff to help in this multidisciplinary effort of combating hypothermia in our surgical pediatric population.

PDSA cycle 2 involved several major initiatives, including direct incorporation of the rest of the perioperative team. The perioperative nursing team was educated on the risks of hypothermia and engaged to take an active role by maintaining the operating suite temperature at 21.1 °C and turning on the Bair Hugger (3M) blanket to 43 °C on the OR bed prior to patient arrival to the OR. Additionally, anesthesia technicians (ATs) were tasked with ensuring an adequate supply of Bair Hugger drapes for all cases of the day. The facility’s engineering team was engaged to move the preoperative room temperature controls away from families (who frequently made the rooms cold) and instead set it at a consistent temperature of 23.9 °C. ATs were also asked to place axillary and nasal temperature probes on the anesthesia workstations as a visual reminder to facilitate temperature monitoring closer to the start of anesthesia (instead of the anesthesia provider having to remember to retrieve a temperature probe out of a drawer and place it on the patient). Furthermore, anesthesiologists were instructed via the aforementioned monthly emails and at monthly department meetings to place the temperature probes as early as possible in order to recognize and respond to intraoperative hypothermia in a timelier manner. Finally, supply chain leaders were informed of our expected increase in the use of the blankets and probes and proportionally increased ordering of these supplies to make sure availability would not present an obstacle.

In PDSA cycle 3, trainees (anesthesia assistant students, anesthesia residents and fellows) and advanced practice providers (APPs) (certified registered nurse-anesthetists [CRNAs] and certified anesthesia assistants [C-AAs]) were informed of the QI project. This initiative was guided toward improving vigilance for hypothermia in the rest of the anesthesia team members. The trainees and APPs usually set up the anesthesia area prior to patient arrival, so their recruitment in support of this effort would ensure appropriate OR temperature, active warming device deployment, and the availability and early placement of the correct temperature probe for the case. To facilitate personal accountability, the trainees and APPs were also emailed their own patients’ rate of hypothermia.

Along the course of the project, quarterly committee meetings and departmental monthly meetings served as venues to express concerns and look for areas of improvement, such as specific patterns or trends leading to hypothermic patients. One specific example was the identification of the gastrointestinal endoscopic patients having a rate of hypothermia that was 2% higher than average. Directed education on the importance of Bair Hugger blankets and using warm intravenous fluids worked well to decrease the rate of hypothermia in these patients. This collection of data was shared at regular intervals during monthly department meetings as well and more frequently using departmental emails. The hospital’s secure intranet SharePoint (Microsoft) site was used to share the data among providers.

Study of the interventions and measures

To study the effectiveness and impact of the project to motivate our anesthesiologists and other team members, we compared the first temperatures obtained in the recovery unit prior to the start of the intervention with those collected after the start of the QI project in November 2018. Because of the variability of temperature monitoring intraoperatively (nasal, axillary, rectal), we decided to use the temperature obtained by the nurse in the recovery room upon the patient’s arrival. Over the years analyzed, the nurse’s technique of measuring the temperature remained consistent. All patient temperature measurements were performed using the TAT-5000 (Exergen Corporation). This temporal artery thermometer has been previously shown to correlate well with bladder temperatures (70% of measurements differ by no more than 0.5 °C, as reported by Langham et al¹⁶).

Admittedly, we could not measure the degree of motivation or internalization of the project goals by our cohort, but we could measure the reduction in the rate of hypothermia and subjectively gauge engagement in the project by the various groups of participants and the sustainability of the results. In addition, all participating anesthesiologists received MOCA Part 4 credits in July 2019. We continued our data collection until April 2021 to determine if our project had brought about sustainable changes in practice that would continue past the initial motivator of obtaining CME credit.

Analysis

Data analysis was performed using Excel (Microsoft) and SAS, version 9.4 (SAS Institute).

The median of the monthly percentage of patients with a temperature of less than 36.0 °C was also determined for the preintervention time frame. This served as our baseline hypothermia rate, and we aimed to lower it by 50%. Run charts, a well-described methodology to gauge the effectiveness of the QI project, were constructed with the collected data.¹⁷

We performed additional analysis to adjust for different time periods throughout the year. The time period between January 2016 and October 2018 was considered preintervention. We considered November 2018 the start of our intervention, or more specifically, the start of our PDSA cycles. October 2018 was analyzed as part of the preintervention data. To account for seasonal temperature variations, the statistical analysis focused on the comparisons of the same calendar quarters for before and after starting intervention using Wilcoxon Mann-Whitney U tests. To reach an overall conclusion, the probabilities for the 4 quarters were combined for each criterion separately utilizing the Fisher χ² combined probability method.

The hypothermia QI project was reviewed by the institutional review board and determined to be exempt.

Results

The temperatures of 40 875 patients were available for analysis for the preintervention period between January 2016 and October 2018. The median percentage of patients with temperatures less than 36.0 °C was 6.9% (interquartile range [IQR], 5.8%-8.4%). The highest percentage was in February 2016 (9.9%), and the lowest was in March 2018 (3.4%). Following the start of the first PDSA cycle, the next 6 consecutive rates of hypothermia were below the median preintervention value, and a new median for these percentages was calculated at 3.4% (IQR, 2.6%-4.3%). In July 2019, the proportion of hypothermic patients decreased once more for 6 consecutive months, yielding a new median of 1.6% (IQR, 1.2%-1.8%) and again in July 2020, to yield a median of 1.3% (IQR, 1.2%-1.5%) (Figure). In all, 33 799 patients were analyzed after the start of the project from November 2018 to the end of the data collection period through April 2021.

The preintervention monthly rates of hypothermia were compared, quarter to quarter, with those starting in November 2018 using the Wilcoxon Mann-Whitney U test. The decrease in proportion of hypothermic patients after the start of the intervention was statistically significant (P < .001). In addition, the percentage of patients with temperatures greater than 38 °C was not significantly different between the pre- and postintervention time periods (P < .25) (Table 2). The decrease in the number of patients available for analysis from March 2020 to May 2020 was due to the COVID-19 pandemic.

Subjectively, we did not experience any notable resistance to our efforts, and the experience was largely positive for everyone involved. Clinicians identified as having high monthly rates of hypothermia (5% or higher) corrected their numbers the following month after being notified via email or in person.

To achieve changes in practice, the health care industry has relied on instituting guidelines, regulations, and policies, often with punitive consequences. We call into question this long-standing framework and propose a novel approach to help evolve the field of QI. Studies in human psychology have long demonstrated the demotivation power of a reward system and the negative response to attempts by authority to use incentives to control or coerce. In our QI project, we instituted 3 PDSA cycles and applied elements from SDT to motivate people’s behaviors. We demonstrate how a new culture focused on maintaining intraoperative normothermia was developed and brought about a measurable and significant decrease in the rate of hypothermia. The relevance of SDT, a widely accepted unifying theory that bridges and links social and personality psychology, should not be understated in health care. Authorities wishing to have long-standing influence should consider a person’s right to make their own decisions and, if possible, a unique way of doing things.

Positively reinforcing behavior has been shown to have a paradoxical effect by dampening an individual’s intrinsic motivation or desire to perform certain tasks.¹⁸ Deadlines, surveillance, and authoritative commands are also deterrents.^19,20 We focused on providing the tools and information to the clinicians and relied on their innate need for autonomy, growth, and self-actualization to bring about change in clinical practice.²¹ Group meetings served as a construct for exchanging ideas and to encourage participation, but without the implementation of rigid guidelines or policies. Intraoperative active warming devices and temperature probes were made available, but their use was not mandated. The use of these devices was intentionally not audited to avoid any overbearing control. Providers were, however, given monthly temperature data to help individually assess the effectiveness of their interventions. We did not impose any negative or punitive actions for those clinicians who had high rates of hypothermic patients, and we did not reward those who had low rates of hypothermia. We wanted the participants to feel that the inner self was the source of their behavior, and this was in parallel with their own interests and values. If providers could feel their need for competency could be realized, we hoped they would continue to adhere to the measures we provided to maintain a low rate of hypothermia.

The effectiveness of our efforts was demonstrated by a decrease in the prevalence of postoperative hypothermia in our surgical patients. The initial decrease of the median rate of hypothermia from 6.9% to 3.4% occurred shortly into the start of the first PDSA cycle. The second PDSA cycle started in January 2019 with a multimodal approach and included almost all parties involved in the perioperative care of our surgical patients. Not only was this intervention responsible for a continued downward trend in the percentage of hypothermic patients, but it set the stage for the third and final PDSA cycle, which started in July 2019. The architecture was in place to integrate trainees and APPs to reinforce our initiative. Subsequently, the new median percentage of hypothermic patients was further decreased to an all-time low of 1.6% per month, satisfying and surpassing the goal of the QI project of decreasing the rate of hypothermia by only 50%. Our organization thereafter maintained a monthly hypothermia rate below 2%, except for April 2020, when it reached 2.5%. Our lowest median percentage was obtained after July 2020, reaching 1.3%.

To account for seasonal variations in temperatures and types of surgeries performed, we compared the percentage of hypothermic patients before and after the start of intervention, quarter by quarter. The decrease in the proportion of hypothermic patients after the start of intervention was statistically significant (P < .001). In addition, the data failed to prove any statistical difference for temperatures above 38 °C between the 2 periods, indicating that our interventions did not result in significant overwarming of patients. The clinical implications of decreasing the percentage of hypothermic patients from 6.9% to 1.3% is likely clinically important when considering the large number of patients who undergo surgery at large tertiary care pediatric centers. Even if simple interventions reduce hypothermia in only a handful of patients, routine applications of simple measures to keep patients normothermic is likely best clinical practice.

Anesthesiologists who participated in the hypothermia QI project by tracking the incidence of hypothermia in their patients were able to collect MOCA Part 4 credits in July 2019. There was no requirement for the individual anesthesiologist to reduce the rate of hypothermia or apply any of the encouraged strategies to obtain credit. As previously stated, there were also no rewards for obtaining low hypothermia rates for the providers. The temperature data continued to be collected through April 2021, 21 months after the credits were distributed, to demonstrate a continued, meaningful change, at least in the short-term. While the MOCA Part 4 credits likely served as an initial motivating factor to encourage participation in the QI project, they certainly were not responsible for the sustained low hypothermia rate after July 2019. We showed that the low rate of hypothermia was successfully maintained, indicating that the change in providers’ behavior was independent of the external motivator of obtaining the credit hours. Mere participation in the project by reviewing one’s temperature data was all that was required to obtain the credit. The Organismic Integration Theory, a mini-theory within SDT, best explains this phenomenon by describing any motivated behavior on a continuum ranging from controlled to autonomous.²² Do people perform the task resentfully, on their own volition because they believe it is the correct action, or somewhere in between? We explain the sustained low rates of hypothermia after the MOCA credits were distributed due to a shift to the autonomous end of the continuum with the clinician’s active willingness to meet the challenges and apply intrinsically motivated behaviors to lower the rate of hypothermia. The internalization of external motivators is difficult to prove, but the evidence supports that the methods we used to motivate individuals were effective and have resulted in a significant downward trend in our hypothermia rate.

There are several limitations to our QI project. The first involves the measuring of postoperative temperature in the recovery units. The temperatures were obtained using the same medical-grade infrared thermometer for all the patients, but other variables, such as timing and techniques, were not standardized. Secondly, overall surgical outcomes related to hypothermia were not tracked because we were unable to control for other confounding variables in our large cohort of patients, so we cannot say if the drop in the hypothermia rate had a clinically significant outcome. Thirdly, we propose that SDT offers a compellingly fitting explanation of the psychology of motivation in our efforts, but it may be possible that other theories may offer equally fitting explanations. The ability to measure the degree of motivation is lacking, and we did not explicitly ask participants what their specific source of motivation was. Aside from SDT, the reduction in hypothermia rate could also be attributed to the ease and availability of warming equipment that was made in each OR. This QI project was successfully applied to only 1 institution, so its ability to be widely applicable remains uncertain. In addition, data collection continued during the COVID-19 pandemic when case volumes decreased. However, by June 2020, the number of surgical cases at our institution had largely returned to prepandemic levels. Additional data collection beyond April 2021 would be helpful to determine if the reduction in hypothermia rates is truly sustained.

Conclusion

Overall, the importance of maintaining perioperative normothermia was well disseminated and agreed upon by all departments involved. Despite the limitations of the project, there was a significant reduction in rates of hypothermia, and sustainability of outcomes was consistently demonstrated in the poststudy period.

Using 3 cycles of the PDSA method, we successfully decreased the median rate of postoperative hypothermia in our pediatric surgical population from a preintervention value of 6.9% to 1.3%—a reduction of more than 81.2%. We provided motivation for members of our anesthesiology staff to participate by offering MOCA 2.0 Part 4 credits, but the lower rate of hypothermic patients was maintained for 15 months after the credits were distributed. Over the course of the project, there was a shift in culture, and extra vigilance was given to temperature monitoring and assessment. We attribute this sustained cultural change to the deliberate incorporation of the principles of competency, autonomy, and relatedness central to SDT to the structure of the interventions, avoiding rigid guidelines and pathways in favor of affective engagement to establish intrinsic motivation.

Acknowledgements: The authors thank Joan Reisch, PhD, for her assistance with the statistical analysis.

Corresponding author: Edgar Erold Kiss, MD, 1935 Medical District Dr, Dallas, TX 75235; [email protected].

Financial disclosures: None.

From Children’s Health System of Texas, Division of Pediatric Anesthesiology, Dallas, TX (Drs. Sakhai, Bocanegra, Chandran, Kimatian, and Kiss), UT Southwestern Medical Center, Department of Anesthesiology and Pain Management, Dallas, TX (Drs. Bocanegra, Chandran, Kimatian, and Kiss), and UT Southwestern Medical Center, Department of Population and Data Sciences, Dallas, TX (Dr. Reisch).

Objective: Policy-driven changes in medical practice have long been the norm. Seldom are changes in clinical practice sought to be brought about by a person’s tendency toward growth or self‐actualization. Many hospitals have instituted hypothermia bundles to help reduce the incidence of unanticipated postoperative hypothermia. Although successful in the short-term, sustained changes are difficult to maintain. We implemented a quality-improvement project focused on addressing the affective components of self-determination theory (SDT) to create sustainable behavioral change while satisfying providers’ basic psychological needs for autonomy, competence, and relatedness.

Methods: A total of 3 Plan-Do-Study-Act (PDSA) cycles were enacted over the span of 14 months at a major tertiary care pediatric hospital to recruit and motivate anesthesia providers and perioperative team members to reduce the percentage of hypothermic postsurgical patients by 50%. As an optional initial incentive for participation, anesthesiologists would qualify for American Board of Anesthesiology Maintenance of Certification in Anesthesiology (MOCA) Part 4 Quality Improvement credits for monitoring their own temperature data and participating in project-related meetings. Providers were given autonomy to develop a personal plan for achieving the desired goals.

Results: The median rate of hypothermia was reduced from 6.9% to 1.6% in July 2019 and was reduced again in July 2020 to 1.3%, an 81% reduction overall. A low hypothermia rate was successfully maintained for at least 21 subsequent months after participants received their MOCA credits in July 2019.

Conclusions: Using an approach that focused on the elements of competency, autonomy, and relatedness central to the principles of SDT, we observed the development of a new culture of vigilance for prevention of hypothermia that successfully endured beyond the project end date.

Keywords: postoperative hypothermia; self-determination theory; motivation; quality improvement.

Perioperative hypothermia, generally accepted as a core temperature less than 36 °C in clinical practice, is a common complication in the pediatric surgical population and is associated with poor postoperative outcomes.¹ Hypothermic patients may develop respiratory depression, hypoglycemia, and metabolic acidosis that may lead to decreased oxygen delivery and end organ tissue hypoxia.^2-4 Other potential detrimental effects of failing to maintain normal body temperature are impaired clotting factor enzyme function and platelet dysfunction, increasing the risk for postoperative bleeding.^5,6 In addition, there are financial implications when hypothermic patients require care and resources postoperatively because of delayed emergence or shivering.⁷

The American Society of Anesthesiologists recommends intraoperative temperature monitoring for procedures when clinically significant changes in body temperature are anticipated.⁸ Maintenance of normothermia in the pediatric population is especially challenging owing to a larger skin-surface area compared with body mass ratio and less subcutaneous fat content than in adults. Preventing postoperative hypothermia starts preoperatively with parental education and can be as simple as covering the child with a blanket and setting the preoperative room to an acceptably warm temperature.^9,10 Intraoperatively, maintaining operating room (OR) temperatures at or above 21.1 °C and using active warming devices and radiant warmers when appropriate are important techniques to preserve the child’s body temperature.^11,12

Despite the knowledge of these risks and vigilant avoidance of hypothermia, unplanned perioperative hypothermia can occur in up to 70% of surgical patients.¹ Beyond the clinical benefits, as health care marches toward a value-based payment methodology, quality indicators such as avoiding hypothermia may be linked directly to payment.

Self-determination theory (SDT) was first developed in 1980 by Deci and Ryan.¹³ The central premise of the theory states that people develop their full potential if circumstances allow them to satisfy their basic psychological needs: autonomy, competence, and relatedness. Under these conditions, people’s natural inclination toward growth can be realized, and they are more likely to internalize external goals. Under an extrinsic reward system, motivation can waver, as people may perceive rewards as controlling.

Many institutions have implemented hypothermia bundles to help decrease the rate of hypothermic patients, but while initially successful, the effectiveness of these interventions tends to fade over time as participants settle into old, comfortable routines.¹⁴ With SDT in mind, we designed our quality-improvement (QI) project with interventions to allow clinicians autonomy without instituting rigid guidelines or punitive actions. We aimed to directly address the affective components central to motivation and engagement so that we could bring about long-term meaningful changes in our practice.

Methods

Setting

The hypothermia QI intervention was instituted at a major tertiary care children’s hospital that performs more than 40 000 pediatric general anesthetics annually. Our division of pediatric anesthesiology consists of 66 fellowship-trained pediatric anesthesiologists, 15 or more rotating trainees per month, 13 anesthesiology assistants, 15 anesthesia technicians, and more than 50 perioperative nurses.

The most frequent pediatric surgeries include, but are not limited to, general surgery, otolaryngology, urology, gastroenterology, plastic surgery, neurosurgery, and dentistry. The surgeries are conducted in the hospital’s main operative floor, which consists of 15 ORs and 2 gastroenterology procedure rooms. Although the implementation of the QI project included several operating sites, we focused on collecting temperature data from surgical patients at our main campus recovery unit. We obtained the patients’ initial temperatures upon arrival to the recovery unit from a retrospective electronic health record review of all patients who underwent anesthesia from January 2016 through April 2021.

Postoperative hypothermia was identified as an area of potential improvement after several patients were reported to be hypothermic upon arrival to the recovery unit in the later part of 2018. Further review revealed significant heterogeneity of practices and lack of standardization of patient-warming methods. By comparing the temperatures pre- and postintervention, we could measure the effectiveness of the QI initiative. Prior to the start of our project, the hypothermia rate in our patient population was not actively tracked, and the effectiveness of our variable practice was not measured.

The cutoff for hypothermia for our QI project was defined as body temperature below 36 °C, since this value has been previously used in the literature and is commonly accepted in anesthesia practice as the delineation for hypothermia in patients undergoing general anesthesia.¹

Interventions

This QI project was designed and modeled after the Institute for Healthcare Improvement Model for Improvement.¹⁵ Three cycles of Plan-Do-Study-Act (PDSA) were developed and instituted over a 14-month period until December 2019 (Table 1).

A retrospective review was conducted to determine the percentage of surgical patients arriving to our recovery units with an initial temperature reading of less than 36 °C. A project key driver diagram and smart aim were created and approved by the hospital’s continuing medical education (CME) committee for credit via the American Board of Medical Specialties (ABMS) Multi-Specialty Portfolio Program, Maintenance of Certification in Anesthesiology (MOCA) Part 4.

The first PDSA cycle involved introducing the QI project and sharing the aims of the project at a department grand rounds in the latter part of October 2018. Enrollment to participate in the project was open to all anesthesiologists in the division, and participants could earn up to 20 hours of MOCA Part 4 credits. A spreadsheet was developed and maintained to track each anesthesiologist’s monthly percentage of hypothermic patients. The de-identified patient data were shared with the division via monthly emails. In addition, individual providers with a hypothermic patient in the recovery room received a notification email.

The anesthesiologists participated in the QI project by reviewing their personal percentage of hypothermic patients on an ongoing basis to earn the credit. There was no explicit requirement to decrease their own rate of patients with body temperature less than 36 °C or expectation to achieve a predetermined goal, so the participants could not “fail.”

Because of the large interest in this project, a hypothermia committee was formed that consisted of 36 anesthesiologists. This group reviewed the data and exchanged ideas for improvement in November 2018 as part of the first PDSA cycle. The committee met monthly and was responsible for actively engaging other members of the department and perioperative staff to help in this multidisciplinary effort of combating hypothermia in our surgical pediatric population.

PDSA cycle 2 involved several major initiatives, including direct incorporation of the rest of the perioperative team. The perioperative nursing team was educated on the risks of hypothermia and engaged to take an active role by maintaining the operating suite temperature at 21.1 °C and turning on the Bair Hugger (3M) blanket to 43 °C on the OR bed prior to patient arrival to the OR. Additionally, anesthesia technicians (ATs) were tasked with ensuring an adequate supply of Bair Hugger drapes for all cases of the day. The facility’s engineering team was engaged to move the preoperative room temperature controls away from families (who frequently made the rooms cold) and instead set it at a consistent temperature of 23.9 °C. ATs were also asked to place axillary and nasal temperature probes on the anesthesia workstations as a visual reminder to facilitate temperature monitoring closer to the start of anesthesia (instead of the anesthesia provider having to remember to retrieve a temperature probe out of a drawer and place it on the patient). Furthermore, anesthesiologists were instructed via the aforementioned monthly emails and at monthly department meetings to place the temperature probes as early as possible in order to recognize and respond to intraoperative hypothermia in a timelier manner. Finally, supply chain leaders were informed of our expected increase in the use of the blankets and probes and proportionally increased ordering of these supplies to make sure availability would not present an obstacle.

In PDSA cycle 3, trainees (anesthesia assistant students, anesthesia residents and fellows) and advanced practice providers (APPs) (certified registered nurse-anesthetists [CRNAs] and certified anesthesia assistants [C-AAs]) were informed of the QI project. This initiative was guided toward improving vigilance for hypothermia in the rest of the anesthesia team members. The trainees and APPs usually set up the anesthesia area prior to patient arrival, so their recruitment in support of this effort would ensure appropriate OR temperature, active warming device deployment, and the availability and early placement of the correct temperature probe for the case. To facilitate personal accountability, the trainees and APPs were also emailed their own patients’ rate of hypothermia.

Along the course of the project, quarterly committee meetings and departmental monthly meetings served as venues to express concerns and look for areas of improvement, such as specific patterns or trends leading to hypothermic patients. One specific example was the identification of the gastrointestinal endoscopic patients having a rate of hypothermia that was 2% higher than average. Directed education on the importance of Bair Hugger blankets and using warm intravenous fluids worked well to decrease the rate of hypothermia in these patients. This collection of data was shared at regular intervals during monthly department meetings as well and more frequently using departmental emails. The hospital’s secure intranet SharePoint (Microsoft) site was used to share the data among providers.

Study of the interventions and measures

To study the effectiveness and impact of the project to motivate our anesthesiologists and other team members, we compared the first temperatures obtained in the recovery unit prior to the start of the intervention with those collected after the start of the QI project in November 2018. Because of the variability of temperature monitoring intraoperatively (nasal, axillary, rectal), we decided to use the temperature obtained by the nurse in the recovery room upon the patient’s arrival. Over the years analyzed, the nurse’s technique of measuring the temperature remained consistent. All patient temperature measurements were performed using the TAT-5000 (Exergen Corporation). This temporal artery thermometer has been previously shown to correlate well with bladder temperatures (70% of measurements differ by no more than 0.5 °C, as reported by Langham et al¹⁶).

Admittedly, we could not measure the degree of motivation or internalization of the project goals by our cohort, but we could measure the reduction in the rate of hypothermia and subjectively gauge engagement in the project by the various groups of participants and the sustainability of the results. In addition, all participating anesthesiologists received MOCA Part 4 credits in July 2019. We continued our data collection until April 2021 to determine if our project had brought about sustainable changes in practice that would continue past the initial motivator of obtaining CME credit.

Analysis

Data analysis was performed using Excel (Microsoft) and SAS, version 9.4 (SAS Institute).

The median of the monthly percentage of patients with a temperature of less than 36.0 °C was also determined for the preintervention time frame. This served as our baseline hypothermia rate, and we aimed to lower it by 50%. Run charts, a well-described methodology to gauge the effectiveness of the QI project, were constructed with the collected data.¹⁷

We performed additional analysis to adjust for different time periods throughout the year. The time period between January 2016 and October 2018 was considered preintervention. We considered November 2018 the start of our intervention, or more specifically, the start of our PDSA cycles. October 2018 was analyzed as part of the preintervention data. To account for seasonal temperature variations, the statistical analysis focused on the comparisons of the same calendar quarters for before and after starting intervention using Wilcoxon Mann-Whitney U tests. To reach an overall conclusion, the probabilities for the 4 quarters were combined for each criterion separately utilizing the Fisher χ² combined probability method.

The hypothermia QI project was reviewed by the institutional review board and determined to be exempt.

Results

The temperatures of 40 875 patients were available for analysis for the preintervention period between January 2016 and October 2018. The median percentage of patients with temperatures less than 36.0 °C was 6.9% (interquartile range [IQR], 5.8%-8.4%). The highest percentage was in February 2016 (9.9%), and the lowest was in March 2018 (3.4%). Following the start of the first PDSA cycle, the next 6 consecutive rates of hypothermia were below the median preintervention value, and a new median for these percentages was calculated at 3.4% (IQR, 2.6%-4.3%). In July 2019, the proportion of hypothermic patients decreased once more for 6 consecutive months, yielding a new median of 1.6% (IQR, 1.2%-1.8%) and again in July 2020, to yield a median of 1.3% (IQR, 1.2%-1.5%) (Figure). In all, 33 799 patients were analyzed after the start of the project from November 2018 to the end of the data collection period through April 2021.

The preintervention monthly rates of hypothermia were compared, quarter to quarter, with those starting in November 2018 using the Wilcoxon Mann-Whitney U test. The decrease in proportion of hypothermic patients after the start of the intervention was statistically significant (P < .001). In addition, the percentage of patients with temperatures greater than 38 °C was not significantly different between the pre- and postintervention time periods (P < .25) (Table 2). The decrease in the number of patients available for analysis from March 2020 to May 2020 was due to the COVID-19 pandemic.

Subjectively, we did not experience any notable resistance to our efforts, and the experience was largely positive for everyone involved. Clinicians identified as having high monthly rates of hypothermia (5% or higher) corrected their numbers the following month after being notified via email or in person.

To achieve changes in practice, the health care industry has relied on instituting guidelines, regulations, and policies, often with punitive consequences. We call into question this long-standing framework and propose a novel approach to help evolve the field of QI. Studies in human psychology have long demonstrated the demotivation power of a reward system and the negative response to attempts by authority to use incentives to control or coerce. In our QI project, we instituted 3 PDSA cycles and applied elements from SDT to motivate people’s behaviors. We demonstrate how a new culture focused on maintaining intraoperative normothermia was developed and brought about a measurable and significant decrease in the rate of hypothermia. The relevance of SDT, a widely accepted unifying theory that bridges and links social and personality psychology, should not be understated in health care. Authorities wishing to have long-standing influence should consider a person’s right to make their own decisions and, if possible, a unique way of doing things.

Positively reinforcing behavior has been shown to have a paradoxical effect by dampening an individual’s intrinsic motivation or desire to perform certain tasks.¹⁸ Deadlines, surveillance, and authoritative commands are also deterrents.^19,20 We focused on providing the tools and information to the clinicians and relied on their innate need for autonomy, growth, and self-actualization to bring about change in clinical practice.²¹ Group meetings served as a construct for exchanging ideas and to encourage participation, but without the implementation of rigid guidelines or policies. Intraoperative active warming devices and temperature probes were made available, but their use was not mandated. The use of these devices was intentionally not audited to avoid any overbearing control. Providers were, however, given monthly temperature data to help individually assess the effectiveness of their interventions. We did not impose any negative or punitive actions for those clinicians who had high rates of hypothermic patients, and we did not reward those who had low rates of hypothermia. We wanted the participants to feel that the inner self was the source of their behavior, and this was in parallel with their own interests and values. If providers could feel their need for competency could be realized, we hoped they would continue to adhere to the measures we provided to maintain a low rate of hypothermia.

The effectiveness of our efforts was demonstrated by a decrease in the prevalence of postoperative hypothermia in our surgical patients. The initial decrease of the median rate of hypothermia from 6.9% to 3.4% occurred shortly into the start of the first PDSA cycle. The second PDSA cycle started in January 2019 with a multimodal approach and included almost all parties involved in the perioperative care of our surgical patients. Not only was this intervention responsible for a continued downward trend in the percentage of hypothermic patients, but it set the stage for the third and final PDSA cycle, which started in July 2019. The architecture was in place to integrate trainees and APPs to reinforce our initiative. Subsequently, the new median percentage of hypothermic patients was further decreased to an all-time low of 1.6% per month, satisfying and surpassing the goal of the QI project of decreasing the rate of hypothermia by only 50%. Our organization thereafter maintained a monthly hypothermia rate below 2%, except for April 2020, when it reached 2.5%. Our lowest median percentage was obtained after July 2020, reaching 1.3%.

To account for seasonal variations in temperatures and types of surgeries performed, we compared the percentage of hypothermic patients before and after the start of intervention, quarter by quarter. The decrease in the proportion of hypothermic patients after the start of intervention was statistically significant (P < .001). In addition, the data failed to prove any statistical difference for temperatures above 38 °C between the 2 periods, indicating that our interventions did not result in significant overwarming of patients. The clinical implications of decreasing the percentage of hypothermic patients from 6.9% to 1.3% is likely clinically important when considering the large number of patients who undergo surgery at large tertiary care pediatric centers. Even if simple interventions reduce hypothermia in only a handful of patients, routine applications of simple measures to keep patients normothermic is likely best clinical practice.

Anesthesiologists who participated in the hypothermia QI project by tracking the incidence of hypothermia in their patients were able to collect MOCA Part 4 credits in July 2019. There was no requirement for the individual anesthesiologist to reduce the rate of hypothermia or apply any of the encouraged strategies to obtain credit. As previously stated, there were also no rewards for obtaining low hypothermia rates for the providers. The temperature data continued to be collected through April 2021, 21 months after the credits were distributed, to demonstrate a continued, meaningful change, at least in the short-term. While the MOCA Part 4 credits likely served as an initial motivating factor to encourage participation in the QI project, they certainly were not responsible for the sustained low hypothermia rate after July 2019. We showed that the low rate of hypothermia was successfully maintained, indicating that the change in providers’ behavior was independent of the external motivator of obtaining the credit hours. Mere participation in the project by reviewing one’s temperature data was all that was required to obtain the credit. The Organismic Integration Theory, a mini-theory within SDT, best explains this phenomenon by describing any motivated behavior on a continuum ranging from controlled to autonomous.²² Do people perform the task resentfully, on their own volition because they believe it is the correct action, or somewhere in between? We explain the sustained low rates of hypothermia after the MOCA credits were distributed due to a shift to the autonomous end of the continuum with the clinician’s active willingness to meet the challenges and apply intrinsically motivated behaviors to lower the rate of hypothermia. The internalization of external motivators is difficult to prove, but the evidence supports that the methods we used to motivate individuals were effective and have resulted in a significant downward trend in our hypothermia rate.

There are several limitations to our QI project. The first involves the measuring of postoperative temperature in the recovery units. The temperatures were obtained using the same medical-grade infrared thermometer for all the patients, but other variables, such as timing and techniques, were not standardized. Secondly, overall surgical outcomes related to hypothermia were not tracked because we were unable to control for other confounding variables in our large cohort of patients, so we cannot say if the drop in the hypothermia rate had a clinically significant outcome. Thirdly, we propose that SDT offers a compellingly fitting explanation of the psychology of motivation in our efforts, but it may be possible that other theories may offer equally fitting explanations. The ability to measure the degree of motivation is lacking, and we did not explicitly ask participants what their specific source of motivation was. Aside from SDT, the reduction in hypothermia rate could also be attributed to the ease and availability of warming equipment that was made in each OR. This QI project was successfully applied to only 1 institution, so its ability to be widely applicable remains uncertain. In addition, data collection continued during the COVID-19 pandemic when case volumes decreased. However, by June 2020, the number of surgical cases at our institution had largely returned to prepandemic levels. Additional data collection beyond April 2021 would be helpful to determine if the reduction in hypothermia rates is truly sustained.

Conclusion

Overall, the importance of maintaining perioperative normothermia was well disseminated and agreed upon by all departments involved. Despite the limitations of the project, there was a significant reduction in rates of hypothermia, and sustainability of outcomes was consistently demonstrated in the poststudy period.

Using 3 cycles of the PDSA method, we successfully decreased the median rate of postoperative hypothermia in our pediatric surgical population from a preintervention value of 6.9% to 1.3%—a reduction of more than 81.2%. We provided motivation for members of our anesthesiology staff to participate by offering MOCA 2.0 Part 4 credits, but the lower rate of hypothermic patients was maintained for 15 months after the credits were distributed. Over the course of the project, there was a shift in culture, and extra vigilance was given to temperature monitoring and assessment. We attribute this sustained cultural change to the deliberate incorporation of the principles of competency, autonomy, and relatedness central to SDT to the structure of the interventions, avoiding rigid guidelines and pathways in favor of affective engagement to establish intrinsic motivation.

Acknowledgements: The authors thank Joan Reisch, PhD, for her assistance with the statistical analysis.

Corresponding author: Edgar Erold Kiss, MD, 1935 Medical District Dr, Dallas, TX 75235; [email protected].

Financial disclosures: None.

From Children’s Health System of Texas, Division of Pediatric Anesthesiology, Dallas, TX (Drs. Sakhai, Bocanegra, Chandran, Kimatian, and Kiss), UT Southwestern Medical Center, Department of Anesthesiology and Pain Management, Dallas, TX (Drs. Bocanegra, Chandran, Kimatian, and Kiss), and UT Southwestern Medical Center, Department of Population and Data Sciences, Dallas, TX (Dr. Reisch).

Objective: Policy-driven changes in medical practice have long been the norm. Seldom are changes in clinical practice sought to be brought about by a person’s tendency toward growth or self‐actualization. Many hospitals have instituted hypothermia bundles to help reduce the incidence of unanticipated postoperative hypothermia. Although successful in the short-term, sustained changes are difficult to maintain. We implemented a quality-improvement project focused on addressing the affective components of self-determination theory (SDT) to create sustainable behavioral change while satisfying providers’ basic psychological needs for autonomy, competence, and relatedness.

Methods: A total of 3 Plan-Do-Study-Act (PDSA) cycles were enacted over the span of 14 months at a major tertiary care pediatric hospital to recruit and motivate anesthesia providers and perioperative team members to reduce the percentage of hypothermic postsurgical patients by 50%. As an optional initial incentive for participation, anesthesiologists would qualify for American Board of Anesthesiology Maintenance of Certification in Anesthesiology (MOCA) Part 4 Quality Improvement credits for monitoring their own temperature data and participating in project-related meetings. Providers were given autonomy to develop a personal plan for achieving the desired goals.

Results: The median rate of hypothermia was reduced from 6.9% to 1.6% in July 2019 and was reduced again in July 2020 to 1.3%, an 81% reduction overall. A low hypothermia rate was successfully maintained for at least 21 subsequent months after participants received their MOCA credits in July 2019.

Conclusions: Using an approach that focused on the elements of competency, autonomy, and relatedness central to the principles of SDT, we observed the development of a new culture of vigilance for prevention of hypothermia that successfully endured beyond the project end date.

Keywords: postoperative hypothermia; self-determination theory; motivation; quality improvement.

Perioperative hypothermia, generally accepted as a core temperature less than 36 °C in clinical practice, is a common complication in the pediatric surgical population and is associated with poor postoperative outcomes.¹ Hypothermic patients may develop respiratory depression, hypoglycemia, and metabolic acidosis that may lead to decreased oxygen delivery and end organ tissue hypoxia.^2-4 Other potential detrimental effects of failing to maintain normal body temperature are impaired clotting factor enzyme function and platelet dysfunction, increasing the risk for postoperative bleeding.^5,6 In addition, there are financial implications when hypothermic patients require care and resources postoperatively because of delayed emergence or shivering.⁷

The American Society of Anesthesiologists recommends intraoperative temperature monitoring for procedures when clinically significant changes in body temperature are anticipated.⁸ Maintenance of normothermia in the pediatric population is especially challenging owing to a larger skin-surface area compared with body mass ratio and less subcutaneous fat content than in adults. Preventing postoperative hypothermia starts preoperatively with parental education and can be as simple as covering the child with a blanket and setting the preoperative room to an acceptably warm temperature.^9,10 Intraoperatively, maintaining operating room (OR) temperatures at or above 21.1 °C and using active warming devices and radiant warmers when appropriate are important techniques to preserve the child’s body temperature.^11,12

Despite the knowledge of these risks and vigilant avoidance of hypothermia, unplanned perioperative hypothermia can occur in up to 70% of surgical patients.¹ Beyond the clinical benefits, as health care marches toward a value-based payment methodology, quality indicators such as avoiding hypothermia may be linked directly to payment.

Self-determination theory (SDT) was first developed in 1980 by Deci and Ryan.¹³ The central premise of the theory states that people develop their full potential if circumstances allow them to satisfy their basic psychological needs: autonomy, competence, and relatedness. Under these conditions, people’s natural inclination toward growth can be realized, and they are more likely to internalize external goals. Under an extrinsic reward system, motivation can waver, as people may perceive rewards as controlling.

Many institutions have implemented hypothermia bundles to help decrease the rate of hypothermic patients, but while initially successful, the effectiveness of these interventions tends to fade over time as participants settle into old, comfortable routines.¹⁴ With SDT in mind, we designed our quality-improvement (QI) project with interventions to allow clinicians autonomy without instituting rigid guidelines or punitive actions. We aimed to directly address the affective components central to motivation and engagement so that we could bring about long-term meaningful changes in our practice.

Methods

Setting

The hypothermia QI intervention was instituted at a major tertiary care children’s hospital that performs more than 40 000 pediatric general anesthetics annually. Our division of pediatric anesthesiology consists of 66 fellowship-trained pediatric anesthesiologists, 15 or more rotating trainees per month, 13 anesthesiology assistants, 15 anesthesia technicians, and more than 50 perioperative nurses.

The most frequent pediatric surgeries include, but are not limited to, general surgery, otolaryngology, urology, gastroenterology, plastic surgery, neurosurgery, and dentistry. The surgeries are conducted in the hospital’s main operative floor, which consists of 15 ORs and 2 gastroenterology procedure rooms. Although the implementation of the QI project included several operating sites, we focused on collecting temperature data from surgical patients at our main campus recovery unit. We obtained the patients’ initial temperatures upon arrival to the recovery unit from a retrospective electronic health record review of all patients who underwent anesthesia from January 2016 through April 2021.

Postoperative hypothermia was identified as an area of potential improvement after several patients were reported to be hypothermic upon arrival to the recovery unit in the later part of 2018. Further review revealed significant heterogeneity of practices and lack of standardization of patient-warming methods. By comparing the temperatures pre- and postintervention, we could measure the effectiveness of the QI initiative. Prior to the start of our project, the hypothermia rate in our patient population was not actively tracked, and the effectiveness of our variable practice was not measured.

The cutoff for hypothermia for our QI project was defined as body temperature below 36 °C, since this value has been previously used in the literature and is commonly accepted in anesthesia practice as the delineation for hypothermia in patients undergoing general anesthesia.¹

Interventions

This QI project was designed and modeled after the Institute for Healthcare Improvement Model for Improvement.¹⁵ Three cycles of Plan-Do-Study-Act (PDSA) were developed and instituted over a 14-month period until December 2019 (Table 1).

A retrospective review was conducted to determine the percentage of surgical patients arriving to our recovery units with an initial temperature reading of less than 36 °C. A project key driver diagram and smart aim were created and approved by the hospital’s continuing medical education (CME) committee for credit via the American Board of Medical Specialties (ABMS) Multi-Specialty Portfolio Program, Maintenance of Certification in Anesthesiology (MOCA) Part 4.

The first PDSA cycle involved introducing the QI project and sharing the aims of the project at a department grand rounds in the latter part of October 2018. Enrollment to participate in the project was open to all anesthesiologists in the division, and participants could earn up to 20 hours of MOCA Part 4 credits. A spreadsheet was developed and maintained to track each anesthesiologist’s monthly percentage of hypothermic patients. The de-identified patient data were shared with the division via monthly emails. In addition, individual providers with a hypothermic patient in the recovery room received a notification email.

The anesthesiologists participated in the QI project by reviewing their personal percentage of hypothermic patients on an ongoing basis to earn the credit. There was no explicit requirement to decrease their own rate of patients with body temperature less than 36 °C or expectation to achieve a predetermined goal, so the participants could not “fail.”

Because of the large interest in this project, a hypothermia committee was formed that consisted of 36 anesthesiologists. This group reviewed the data and exchanged ideas for improvement in November 2018 as part of the first PDSA cycle. The committee met monthly and was responsible for actively engaging other members of the department and perioperative staff to help in this multidisciplinary effort of combating hypothermia in our surgical pediatric population.

PDSA cycle 2 involved several major initiatives, including direct incorporation of the rest of the perioperative team. The perioperative nursing team was educated on the risks of hypothermia and engaged to take an active role by maintaining the operating suite temperature at 21.1 °C and turning on the Bair Hugger (3M) blanket to 43 °C on the OR bed prior to patient arrival to the OR. Additionally, anesthesia technicians (ATs) were tasked with ensuring an adequate supply of Bair Hugger drapes for all cases of the day. The facility’s engineering team was engaged to move the preoperative room temperature controls away from families (who frequently made the rooms cold) and instead set it at a consistent temperature of 23.9 °C. ATs were also asked to place axillary and nasal temperature probes on the anesthesia workstations as a visual reminder to facilitate temperature monitoring closer to the start of anesthesia (instead of the anesthesia provider having to remember to retrieve a temperature probe out of a drawer and place it on the patient). Furthermore, anesthesiologists were instructed via the aforementioned monthly emails and at monthly department meetings to place the temperature probes as early as possible in order to recognize and respond to intraoperative hypothermia in a timelier manner. Finally, supply chain leaders were informed of our expected increase in the use of the blankets and probes and proportionally increased ordering of these supplies to make sure availability would not present an obstacle.

In PDSA cycle 3, trainees (anesthesia assistant students, anesthesia residents and fellows) and advanced practice providers (APPs) (certified registered nurse-anesthetists [CRNAs] and certified anesthesia assistants [C-AAs]) were informed of the QI project. This initiative was guided toward improving vigilance for hypothermia in the rest of the anesthesia team members. The trainees and APPs usually set up the anesthesia area prior to patient arrival, so their recruitment in support of this effort would ensure appropriate OR temperature, active warming device deployment, and the availability and early placement of the correct temperature probe for the case. To facilitate personal accountability, the trainees and APPs were also emailed their own patients’ rate of hypothermia.

Along the course of the project, quarterly committee meetings and departmental monthly meetings served as venues to express concerns and look for areas of improvement, such as specific patterns or trends leading to hypothermic patients. One specific example was the identification of the gastrointestinal endoscopic patients having a rate of hypothermia that was 2% higher than average. Directed education on the importance of Bair Hugger blankets and using warm intravenous fluids worked well to decrease the rate of hypothermia in these patients. This collection of data was shared at regular intervals during monthly department meetings as well and more frequently using departmental emails. The hospital’s secure intranet SharePoint (Microsoft) site was used to share the data among providers.

Study of the interventions and measures

To study the effectiveness and impact of the project to motivate our anesthesiologists and other team members, we compared the first temperatures obtained in the recovery unit prior to the start of the intervention with those collected after the start of the QI project in November 2018. Because of the variability of temperature monitoring intraoperatively (nasal, axillary, rectal), we decided to use the temperature obtained by the nurse in the recovery room upon the patient’s arrival. Over the years analyzed, the nurse’s technique of measuring the temperature remained consistent. All patient temperature measurements were performed using the TAT-5000 (Exergen Corporation). This temporal artery thermometer has been previously shown to correlate well with bladder temperatures (70% of measurements differ by no more than 0.5 °C, as reported by Langham et al¹⁶).

Admittedly, we could not measure the degree of motivation or internalization of the project goals by our cohort, but we could measure the reduction in the rate of hypothermia and subjectively gauge engagement in the project by the various groups of participants and the sustainability of the results. In addition, all participating anesthesiologists received MOCA Part 4 credits in July 2019. We continued our data collection until April 2021 to determine if our project had brought about sustainable changes in practice that would continue past the initial motivator of obtaining CME credit.

Analysis

Data analysis was performed using Excel (Microsoft) and SAS, version 9.4 (SAS Institute).

The median of the monthly percentage of patients with a temperature of less than 36.0 °C was also determined for the preintervention time frame. This served as our baseline hypothermia rate, and we aimed to lower it by 50%. Run charts, a well-described methodology to gauge the effectiveness of the QI project, were constructed with the collected data.¹⁷

We performed additional analysis to adjust for different time periods throughout the year. The time period between January 2016 and October 2018 was considered preintervention. We considered November 2018 the start of our intervention, or more specifically, the start of our PDSA cycles. October 2018 was analyzed as part of the preintervention data. To account for seasonal temperature variations, the statistical analysis focused on the comparisons of the same calendar quarters for before and after starting intervention using Wilcoxon Mann-Whitney U tests. To reach an overall conclusion, the probabilities for the 4 quarters were combined for each criterion separately utilizing the Fisher χ² combined probability method.

The hypothermia QI project was reviewed by the institutional review board and determined to be exempt.

Results

The temperatures of 40 875 patients were available for analysis for the preintervention period between January 2016 and October 2018. The median percentage of patients with temperatures less than 36.0 °C was 6.9% (interquartile range [IQR], 5.8%-8.4%). The highest percentage was in February 2016 (9.9%), and the lowest was in March 2018 (3.4%). Following the start of the first PDSA cycle, the next 6 consecutive rates of hypothermia were below the median preintervention value, and a new median for these percentages was calculated at 3.4% (IQR, 2.6%-4.3%). In July 2019, the proportion of hypothermic patients decreased once more for 6 consecutive months, yielding a new median of 1.6% (IQR, 1.2%-1.8%) and again in July 2020, to yield a median of 1.3% (IQR, 1.2%-1.5%) (Figure). In all, 33 799 patients were analyzed after the start of the project from November 2018 to the end of the data collection period through April 2021.

The preintervention monthly rates of hypothermia were compared, quarter to quarter, with those starting in November 2018 using the Wilcoxon Mann-Whitney U test. The decrease in proportion of hypothermic patients after the start of the intervention was statistically significant (P < .001). In addition, the percentage of patients with temperatures greater than 38 °C was not significantly different between the pre- and postintervention time periods (P < .25) (Table 2). The decrease in the number of patients available for analysis from March 2020 to May 2020 was due to the COVID-19 pandemic.

Subjectively, we did not experience any notable resistance to our efforts, and the experience was largely positive for everyone involved. Clinicians identified as having high monthly rates of hypothermia (5% or higher) corrected their numbers the following month after being notified via email or in person.

To achieve changes in practice, the health care industry has relied on instituting guidelines, regulations, and policies, often with punitive consequences. We call into question this long-standing framework and propose a novel approach to help evolve the field of QI. Studies in human psychology have long demonstrated the demotivation power of a reward system and the negative response to attempts by authority to use incentives to control or coerce. In our QI project, we instituted 3 PDSA cycles and applied elements from SDT to motivate people’s behaviors. We demonstrate how a new culture focused on maintaining intraoperative normothermia was developed and brought about a measurable and significant decrease in the rate of hypothermia. The relevance of SDT, a widely accepted unifying theory that bridges and links social and personality psychology, should not be understated in health care. Authorities wishing to have long-standing influence should consider a person’s right to make their own decisions and, if possible, a unique way of doing things.

Positively reinforcing behavior has been shown to have a paradoxical effect by dampening an individual’s intrinsic motivation or desire to perform certain tasks.¹⁸ Deadlines, surveillance, and authoritative commands are also deterrents.^19,20 We focused on providing the tools and information to the clinicians and relied on their innate need for autonomy, growth, and self-actualization to bring about change in clinical practice.²¹ Group meetings served as a construct for exchanging ideas and to encourage participation, but without the implementation of rigid guidelines or policies. Intraoperative active warming devices and temperature probes were made available, but their use was not mandated. The use of these devices was intentionally not audited to avoid any overbearing control. Providers were, however, given monthly temperature data to help individually assess the effectiveness of their interventions. We did not impose any negative or punitive actions for those clinicians who had high rates of hypothermic patients, and we did not reward those who had low rates of hypothermia. We wanted the participants to feel that the inner self was the source of their behavior, and this was in parallel with their own interests and values. If providers could feel their need for competency could be realized, we hoped they would continue to adhere to the measures we provided to maintain a low rate of hypothermia.

The effectiveness of our efforts was demonstrated by a decrease in the prevalence of postoperative hypothermia in our surgical patients. The initial decrease of the median rate of hypothermia from 6.9% to 3.4% occurred shortly into the start of the first PDSA cycle. The second PDSA cycle started in January 2019 with a multimodal approach and included almost all parties involved in the perioperative care of our surgical patients. Not only was this intervention responsible for a continued downward trend in the percentage of hypothermic patients, but it set the stage for the third and final PDSA cycle, which started in July 2019. The architecture was in place to integrate trainees and APPs to reinforce our initiative. Subsequently, the new median percentage of hypothermic patients was further decreased to an all-time low of 1.6% per month, satisfying and surpassing the goal of the QI project of decreasing the rate of hypothermia by only 50%. Our organization thereafter maintained a monthly hypothermia rate below 2%, except for April 2020, when it reached 2.5%. Our lowest median percentage was obtained after July 2020, reaching 1.3%.

To account for seasonal variations in temperatures and types of surgeries performed, we compared the percentage of hypothermic patients before and after the start of intervention, quarter by quarter. The decrease in the proportion of hypothermic patients after the start of intervention was statistically significant (P < .001). In addition, the data failed to prove any statistical difference for temperatures above 38 °C between the 2 periods, indicating that our interventions did not result in significant overwarming of patients. The clinical implications of decreasing the percentage of hypothermic patients from 6.9% to 1.3% is likely clinically important when considering the large number of patients who undergo surgery at large tertiary care pediatric centers. Even if simple interventions reduce hypothermia in only a handful of patients, routine applications of simple measures to keep patients normothermic is likely best clinical practice.

Anesthesiologists who participated in the hypothermia QI project by tracking the incidence of hypothermia in their patients were able to collect MOCA Part 4 credits in July 2019. There was no requirement for the individual anesthesiologist to reduce the rate of hypothermia or apply any of the encouraged strategies to obtain credit. As previously stated, there were also no rewards for obtaining low hypothermia rates for the providers. The temperature data continued to be collected through April 2021, 21 months after the credits were distributed, to demonstrate a continued, meaningful change, at least in the short-term. While the MOCA Part 4 credits likely served as an initial motivating factor to encourage participation in the QI project, they certainly were not responsible for the sustained low hypothermia rate after July 2019. We showed that the low rate of hypothermia was successfully maintained, indicating that the change in providers’ behavior was independent of the external motivator of obtaining the credit hours. Mere participation in the project by reviewing one’s temperature data was all that was required to obtain the credit. The Organismic Integration Theory, a mini-theory within SDT, best explains this phenomenon by describing any motivated behavior on a continuum ranging from controlled to autonomous.²² Do people perform the task resentfully, on their own volition because they believe it is the correct action, or somewhere in between? We explain the sustained low rates of hypothermia after the MOCA credits were distributed due to a shift to the autonomous end of the continuum with the clinician’s active willingness to meet the challenges and apply intrinsically motivated behaviors to lower the rate of hypothermia. The internalization of external motivators is difficult to prove, but the evidence supports that the methods we used to motivate individuals were effective and have resulted in a significant downward trend in our hypothermia rate.

There are several limitations to our QI project. The first involves the measuring of postoperative temperature in the recovery units. The temperatures were obtained using the same medical-grade infrared thermometer for all the patients, but other variables, such as timing and techniques, were not standardized. Secondly, overall surgical outcomes related to hypothermia were not tracked because we were unable to control for other confounding variables in our large cohort of patients, so we cannot say if the drop in the hypothermia rate had a clinically significant outcome. Thirdly, we propose that SDT offers a compellingly fitting explanation of the psychology of motivation in our efforts, but it may be possible that other theories may offer equally fitting explanations. The ability to measure the degree of motivation is lacking, and we did not explicitly ask participants what their specific source of motivation was. Aside from SDT, the reduction in hypothermia rate could also be attributed to the ease and availability of warming equipment that was made in each OR. This QI project was successfully applied to only 1 institution, so its ability to be widely applicable remains uncertain. In addition, data collection continued during the COVID-19 pandemic when case volumes decreased. However, by June 2020, the number of surgical cases at our institution had largely returned to prepandemic levels. Additional data collection beyond April 2021 would be helpful to determine if the reduction in hypothermia rates is truly sustained.

Conclusion

Overall, the importance of maintaining perioperative normothermia was well disseminated and agreed upon by all departments involved. Despite the limitations of the project, there was a significant reduction in rates of hypothermia, and sustainability of outcomes was consistently demonstrated in the poststudy period.

Using 3 cycles of the PDSA method, we successfully decreased the median rate of postoperative hypothermia in our pediatric surgical population from a preintervention value of 6.9% to 1.3%—a reduction of more than 81.2%. We provided motivation for members of our anesthesiology staff to participate by offering MOCA 2.0 Part 4 credits, but the lower rate of hypothermic patients was maintained for 15 months after the credits were distributed. Over the course of the project, there was a shift in culture, and extra vigilance was given to temperature monitoring and assessment. We attribute this sustained cultural change to the deliberate incorporation of the principles of competency, autonomy, and relatedness central to SDT to the structure of the interventions, avoiding rigid guidelines and pathways in favor of affective engagement to establish intrinsic motivation.

Acknowledgements: The authors thank Joan Reisch, PhD, for her assistance with the statistical analysis.

Corresponding author: Edgar Erold Kiss, MD, 1935 Medical District Dr, Dallas, TX 75235; [email protected].

Financial disclosures: None.

Population-based hospital payments provide incentives to reduce unnecessary healthcare use and a mechanism to finance population health investments. For hospitals, these payments provide stable revenue and flexibility in exchange for increased financial risk. The COVID-19 pandemic significantly reduced fee-for-service revenues, which has spurred provider interest in population-based payments, particularly from cash-strapped rural hospitals.

The Centers for Medicare & Medicaid Services (CMS) recently announced the launch of the Community Health Access and Rural Transformation (CHART) Model to test whether up-front, population-based payments improve access to high-quality care in rural communities and protect the financial stability of rural providers. This model follows the ongoing Pennsylvania Rural Health Model (PARHM), which offers similar payments to Pennsylvania’s rural hospitals. Prospective population-based hospital reimbursement appears to have helped Maryland’s hospitals survive the financial stress of the COVID-19 pandemic,¹ and it is likely that the PARHM did the same for rural hospitals in Pennsylvania. Both the PARHM and the CHART Model place quality measurement and improvement at the core of payment reform, and for good reason. Capitation generates incentives for care stinting; linking prospective payments to quality measurement helps to ensure accountability. However, measuring the quality of rural healthcare is challenging. Rural health is different: Hospital size, payment mechanisms, and community health priorities are all distinct from those of metropolitan areas, which is why CMS exempts Critical Access Hospitals from Medicare’s core quality programs. Rural quality reporting programs could be established that address the unique aspects of rural healthcare.

As designers (JEF, DTL) of, and an advisor (ALS) for, a proposed pay-for-performance (P4P) program for the PARHM,² we identified three central challenges in constructing and implementing P4P programs for rural hospitals, along with potential solutions. We hope that the lessons we learned can inform similar policy efforts.

First, many rural hospitals serve as stewards of community health resources. While metropolitan hospital systems can make targeted investments in population health, assigning accountability for health outcomes is challenging in cities where geographically overlapping provider systems compete for patients. In contrast, a rural hospital system with few or no competing providers is more naturally accountable for community health outcomes, especially if it owns most ambulatory clinics in its community. P4P programs could therefore reward rural hospitals for improving healthcare quality or health outcomes within their catchment areas. Like an accountable care organization (ACO), a rural hospital or hospital-based health system could be held accountable for appropriate screening for, and treatment of, uncontrolled hypertension, diabetes, or asthma, even without a network of community-based primary care providers that ACOs usually possess. Participants in the CHART Model’s Community Transformation Track, for example, select three community-level population health measures from four domains: substance use, chronic conditions, maternal health, and prevention. Accountability for community health outcomes is increasingly feasible because many larger rural hospitals have merged or been acquired.³

Second, small rural hospital patient volumes obscure the signal of true quality with statistical noise. Many common quality indicators, like risk-standardized mortality rates, are unreliable in rural settings with low patient volumes; in 2012-2013, the mean rural hospital daily census was seven inpatients.^4,5 Payers and regulators have addressed this challenge by exempting rural hospitals from quality-reporting programs or by employing statistical techniques that diminish incentives to invest in improvement. CMS, for example, uses “shrinkage” estimators that adjust a hospital’s quality score toward a program-wide average, which makes it difficult to detect and reward performance improvement.⁴ Instead, rural P4P programs should use measures that are resistant to low patient volumes, such as the Measure Application Partnership’s (MAP) Core Set of Rural-Relevant Measures.⁶ Low volume–resistant measures include process and population-health outcome measures with naturally large denominators (eg, medication reconciliation), structural measures for which sample size is irrelevant (eg, nurse staffing ratios), and qualitative assessments of hospital adherence to best practices. CMS and other measure developers should also prioritize the creation of other rural-relevant, cross-cutting, low volume–resistant measures, like avoidance of deliriogenic medications in the elderly or initiation of treatment for substance use disorders, in consultation with rural stakeholders and the MAP Rural Health Workgroup. When extensive measurement noise is inevitable, public and private policymakers should eschew downside risk in rural P4P contracts.

Third, many rural hospitals have limited resources for measurement and improvement.⁷ While many well-resourced community hospitals have dedicated quality departments, quality directors in rural hospitals often have at least one other full-time job. Well-intentioned exemptions from P4P programs have left rural hospitals with limited experience with basic data collection and reporting, a handicap compounded by redundant and misaligned payor quality reporting requirements. To engage rural hospitals in quality improvement work, payors should coordinate to make participation in rural P4P programs as easy as possible. The adoption of a locally aligned set of healthcare quality measures by all payors in a region, like the PARHM’s proposed “all-payer quality program,” could substantially reduce administrative burden and motivate rural hospitals to enhance patient care and improve community health. In the CHART Model’s Community Transformation Track, for example, all public and private participating payers in each region must report on six quality measures: inpatient and emergency department visits for ambulatory care sensitive conditions, hospital-wide all-cause unplanned readmissions, and the Hospital Consumer Assessment of Health Care survey, as well as three community-chosen measures from the domains of substance use, maternal health, and prevention.⁸ As with all P4P programs, rural P4P programs should focus on a small number of meaningful measures, such as functional and clinical outcomes, complications, and patient experience, and feature relatively large rewards for improvement.⁹ The National Quality Forum recommends that rural programs avoid downside risk, reward improvement as well as achievement, and permit virtual provider groups.¹⁰ We would add that programs in rural communities ought to pair economic rewards with social recognition and comparison, offer technical assistance and opportunities for shared learning, and account for social as well as medical risk.

Many challenges to the adoption of rural P4P programs have been targeted through multi-stakeholder collaborations like the PARHM. Careful allocation of technical assistance resources may help address barriers such as comparing the performance of heterogeneous rural hospitals that vary in characteristics like size, affiliation with large health systems, or integration of ambulatory care services, which may affect hospital measurement capabilities and performance. Quality improvement efforts could be further bolstered through direct allocation of funds to the creation of virtual shared learning platforms, and by providing performance bonuses to groups of small hospitals that elect to engage in shared reporting.

The stakes are high for designing robust quality programs for rural hospitals. Although one in five Americans rely on them for healthcare, their rate of closure has accelerated in the past decade.¹¹ CMS has made it clear that a sustainable system for financing rural health must be built around a commitment to quality measurement and improvement. While some rural provider organizations might be best served by participating in voluntary rural health networks and preexisting federal programs like the Medicare Beneficiary Quality Improvement Project, they should also have the opportunity to accept payments tied to quality, especially as growing numbers of rural hospitals are absorbed into larger healthcare systems. Adopting aligned sets of reliable and meaningful quality measures alongside population-based payments will help to create a sustainable future for rural hospitals.

We thank Mark Friedberg, MD, MPP, for his helpful comments on an earlier draft of this manuscript.

Population-based hospital payments provide incentives to reduce unnecessary healthcare use and a mechanism to finance population health investments. For hospitals, these payments provide stable revenue and flexibility in exchange for increased financial risk. The COVID-19 pandemic significantly reduced fee-for-service revenues, which has spurred provider interest in population-based payments, particularly from cash-strapped rural hospitals.

The Centers for Medicare & Medicaid Services (CMS) recently announced the launch of the Community Health Access and Rural Transformation (CHART) Model to test whether up-front, population-based payments improve access to high-quality care in rural communities and protect the financial stability of rural providers. This model follows the ongoing Pennsylvania Rural Health Model (PARHM), which offers similar payments to Pennsylvania’s rural hospitals. Prospective population-based hospital reimbursement appears to have helped Maryland’s hospitals survive the financial stress of the COVID-19 pandemic,¹ and it is likely that the PARHM did the same for rural hospitals in Pennsylvania. Both the PARHM and the CHART Model place quality measurement and improvement at the core of payment reform, and for good reason. Capitation generates incentives for care stinting; linking prospective payments to quality measurement helps to ensure accountability. However, measuring the quality of rural healthcare is challenging. Rural health is different: Hospital size, payment mechanisms, and community health priorities are all distinct from those of metropolitan areas, which is why CMS exempts Critical Access Hospitals from Medicare’s core quality programs. Rural quality reporting programs could be established that address the unique aspects of rural healthcare.

As designers (JEF, DTL) of, and an advisor (ALS) for, a proposed pay-for-performance (P4P) program for the PARHM,² we identified three central challenges in constructing and implementing P4P programs for rural hospitals, along with potential solutions. We hope that the lessons we learned can inform similar policy efforts.

First, many rural hospitals serve as stewards of community health resources. While metropolitan hospital systems can make targeted investments in population health, assigning accountability for health outcomes is challenging in cities where geographically overlapping provider systems compete for patients. In contrast, a rural hospital system with few or no competing providers is more naturally accountable for community health outcomes, especially if it owns most ambulatory clinics in its community. P4P programs could therefore reward rural hospitals for improving healthcare quality or health outcomes within their catchment areas. Like an accountable care organization (ACO), a rural hospital or hospital-based health system could be held accountable for appropriate screening for, and treatment of, uncontrolled hypertension, diabetes, or asthma, even without a network of community-based primary care providers that ACOs usually possess. Participants in the CHART Model’s Community Transformation Track, for example, select three community-level population health measures from four domains: substance use, chronic conditions, maternal health, and prevention. Accountability for community health outcomes is increasingly feasible because many larger rural hospitals have merged or been acquired.³

Second, small rural hospital patient volumes obscure the signal of true quality with statistical noise. Many common quality indicators, like risk-standardized mortality rates, are unreliable in rural settings with low patient volumes; in 2012-2013, the mean rural hospital daily census was seven inpatients.^4,5 Payers and regulators have addressed this challenge by exempting rural hospitals from quality-reporting programs or by employing statistical techniques that diminish incentives to invest in improvement. CMS, for example, uses “shrinkage” estimators that adjust a hospital’s quality score toward a program-wide average, which makes it difficult to detect and reward performance improvement.⁴ Instead, rural P4P programs should use measures that are resistant to low patient volumes, such as the Measure Application Partnership’s (MAP) Core Set of Rural-Relevant Measures.⁶ Low volume–resistant measures include process and population-health outcome measures with naturally large denominators (eg, medication reconciliation), structural measures for which sample size is irrelevant (eg, nurse staffing ratios), and qualitative assessments of hospital adherence to best practices. CMS and other measure developers should also prioritize the creation of other rural-relevant, cross-cutting, low volume–resistant measures, like avoidance of deliriogenic medications in the elderly or initiation of treatment for substance use disorders, in consultation with rural stakeholders and the MAP Rural Health Workgroup. When extensive measurement noise is inevitable, public and private policymakers should eschew downside risk in rural P4P contracts.

Third, many rural hospitals have limited resources for measurement and improvement.⁷ While many well-resourced community hospitals have dedicated quality departments, quality directors in rural hospitals often have at least one other full-time job. Well-intentioned exemptions from P4P programs have left rural hospitals with limited experience with basic data collection and reporting, a handicap compounded by redundant and misaligned payor quality reporting requirements. To engage rural hospitals in quality improvement work, payors should coordinate to make participation in rural P4P programs as easy as possible. The adoption of a locally aligned set of healthcare quality measures by all payors in a region, like the PARHM’s proposed “all-payer quality program,” could substantially reduce administrative burden and motivate rural hospitals to enhance patient care and improve community health. In the CHART Model’s Community Transformation Track, for example, all public and private participating payers in each region must report on six quality measures: inpatient and emergency department visits for ambulatory care sensitive conditions, hospital-wide all-cause unplanned readmissions, and the Hospital Consumer Assessment of Health Care survey, as well as three community-chosen measures from the domains of substance use, maternal health, and prevention.⁸ As with all P4P programs, rural P4P programs should focus on a small number of meaningful measures, such as functional and clinical outcomes, complications, and patient experience, and feature relatively large rewards for improvement.⁹ The National Quality Forum recommends that rural programs avoid downside risk, reward improvement as well as achievement, and permit virtual provider groups.¹⁰ We would add that programs in rural communities ought to pair economic rewards with social recognition and comparison, offer technical assistance and opportunities for shared learning, and account for social as well as medical risk.

Many challenges to the adoption of rural P4P programs have been targeted through multi-stakeholder collaborations like the PARHM. Careful allocation of technical assistance resources may help address barriers such as comparing the performance of heterogeneous rural hospitals that vary in characteristics like size, affiliation with large health systems, or integration of ambulatory care services, which may affect hospital measurement capabilities and performance. Quality improvement efforts could be further bolstered through direct allocation of funds to the creation of virtual shared learning platforms, and by providing performance bonuses to groups of small hospitals that elect to engage in shared reporting.

The stakes are high for designing robust quality programs for rural hospitals. Although one in five Americans rely on them for healthcare, their rate of closure has accelerated in the past decade.¹¹ CMS has made it clear that a sustainable system for financing rural health must be built around a commitment to quality measurement and improvement. While some rural provider organizations might be best served by participating in voluntary rural health networks and preexisting federal programs like the Medicare Beneficiary Quality Improvement Project, they should also have the opportunity to accept payments tied to quality, especially as growing numbers of rural hospitals are absorbed into larger healthcare systems. Adopting aligned sets of reliable and meaningful quality measures alongside population-based payments will help to create a sustainable future for rural hospitals.

We thank Mark Friedberg, MD, MPP, for his helpful comments on an earlier draft of this manuscript.

Population-based hospital payments provide incentives to reduce unnecessary healthcare use and a mechanism to finance population health investments. For hospitals, these payments provide stable revenue and flexibility in exchange for increased financial risk. The COVID-19 pandemic significantly reduced fee-for-service revenues, which has spurred provider interest in population-based payments, particularly from cash-strapped rural hospitals.

The Centers for Medicare & Medicaid Services (CMS) recently announced the launch of the Community Health Access and Rural Transformation (CHART) Model to test whether up-front, population-based payments improve access to high-quality care in rural communities and protect the financial stability of rural providers. This model follows the ongoing Pennsylvania Rural Health Model (PARHM), which offers similar payments to Pennsylvania’s rural hospitals. Prospective population-based hospital reimbursement appears to have helped Maryland’s hospitals survive the financial stress of the COVID-19 pandemic,¹ and it is likely that the PARHM did the same for rural hospitals in Pennsylvania. Both the PARHM and the CHART Model place quality measurement and improvement at the core of payment reform, and for good reason. Capitation generates incentives for care stinting; linking prospective payments to quality measurement helps to ensure accountability. However, measuring the quality of rural healthcare is challenging. Rural health is different: Hospital size, payment mechanisms, and community health priorities are all distinct from those of metropolitan areas, which is why CMS exempts Critical Access Hospitals from Medicare’s core quality programs. Rural quality reporting programs could be established that address the unique aspects of rural healthcare.

As designers (JEF, DTL) of, and an advisor (ALS) for, a proposed pay-for-performance (P4P) program for the PARHM,² we identified three central challenges in constructing and implementing P4P programs for rural hospitals, along with potential solutions. We hope that the lessons we learned can inform similar policy efforts.

First, many rural hospitals serve as stewards of community health resources. While metropolitan hospital systems can make targeted investments in population health, assigning accountability for health outcomes is challenging in cities where geographically overlapping provider systems compete for patients. In contrast, a rural hospital system with few or no competing providers is more naturally accountable for community health outcomes, especially if it owns most ambulatory clinics in its community. P4P programs could therefore reward rural hospitals for improving healthcare quality or health outcomes within their catchment areas. Like an accountable care organization (ACO), a rural hospital or hospital-based health system could be held accountable for appropriate screening for, and treatment of, uncontrolled hypertension, diabetes, or asthma, even without a network of community-based primary care providers that ACOs usually possess. Participants in the CHART Model’s Community Transformation Track, for example, select three community-level population health measures from four domains: substance use, chronic conditions, maternal health, and prevention. Accountability for community health outcomes is increasingly feasible because many larger rural hospitals have merged or been acquired.³

Second, small rural hospital patient volumes obscure the signal of true quality with statistical noise. Many common quality indicators, like risk-standardized mortality rates, are unreliable in rural settings with low patient volumes; in 2012-2013, the mean rural hospital daily census was seven inpatients.^4,5 Payers and regulators have addressed this challenge by exempting rural hospitals from quality-reporting programs or by employing statistical techniques that diminish incentives to invest in improvement. CMS, for example, uses “shrinkage” estimators that adjust a hospital’s quality score toward a program-wide average, which makes it difficult to detect and reward performance improvement.⁴ Instead, rural P4P programs should use measures that are resistant to low patient volumes, such as the Measure Application Partnership’s (MAP) Core Set of Rural-Relevant Measures.⁶ Low volume–resistant measures include process and population-health outcome measures with naturally large denominators (eg, medication reconciliation), structural measures for which sample size is irrelevant (eg, nurse staffing ratios), and qualitative assessments of hospital adherence to best practices. CMS and other measure developers should also prioritize the creation of other rural-relevant, cross-cutting, low volume–resistant measures, like avoidance of deliriogenic medications in the elderly or initiation of treatment for substance use disorders, in consultation with rural stakeholders and the MAP Rural Health Workgroup. When extensive measurement noise is inevitable, public and private policymakers should eschew downside risk in rural P4P contracts.

Third, many rural hospitals have limited resources for measurement and improvement.⁷ While many well-resourced community hospitals have dedicated quality departments, quality directors in rural hospitals often have at least one other full-time job. Well-intentioned exemptions from P4P programs have left rural hospitals with limited experience with basic data collection and reporting, a handicap compounded by redundant and misaligned payor quality reporting requirements. To engage rural hospitals in quality improvement work, payors should coordinate to make participation in rural P4P programs as easy as possible. The adoption of a locally aligned set of healthcare quality measures by all payors in a region, like the PARHM’s proposed “all-payer quality program,” could substantially reduce administrative burden and motivate rural hospitals to enhance patient care and improve community health. In the CHART Model’s Community Transformation Track, for example, all public and private participating payers in each region must report on six quality measures: inpatient and emergency department visits for ambulatory care sensitive conditions, hospital-wide all-cause unplanned readmissions, and the Hospital Consumer Assessment of Health Care survey, as well as three community-chosen measures from the domains of substance use, maternal health, and prevention.⁸ As with all P4P programs, rural P4P programs should focus on a small number of meaningful measures, such as functional and clinical outcomes, complications, and patient experience, and feature relatively large rewards for improvement.⁹ The National Quality Forum recommends that rural programs avoid downside risk, reward improvement as well as achievement, and permit virtual provider groups.¹⁰ We would add that programs in rural communities ought to pair economic rewards with social recognition and comparison, offer technical assistance and opportunities for shared learning, and account for social as well as medical risk.

Many challenges to the adoption of rural P4P programs have been targeted through multi-stakeholder collaborations like the PARHM. Careful allocation of technical assistance resources may help address barriers such as comparing the performance of heterogeneous rural hospitals that vary in characteristics like size, affiliation with large health systems, or integration of ambulatory care services, which may affect hospital measurement capabilities and performance. Quality improvement efforts could be further bolstered through direct allocation of funds to the creation of virtual shared learning platforms, and by providing performance bonuses to groups of small hospitals that elect to engage in shared reporting.

The stakes are high for designing robust quality programs for rural hospitals. Although one in five Americans rely on them for healthcare, their rate of closure has accelerated in the past decade.¹¹ CMS has made it clear that a sustainable system for financing rural health must be built around a commitment to quality measurement and improvement. While some rural provider organizations might be best served by participating in voluntary rural health networks and preexisting federal programs like the Medicare Beneficiary Quality Improvement Project, they should also have the opportunity to accept payments tied to quality, especially as growing numbers of rural hospitals are absorbed into larger healthcare systems. Adopting aligned sets of reliable and meaningful quality measures alongside population-based payments will help to create a sustainable future for rural hospitals.

We thank Mark Friedberg, MD, MPP, for his helpful comments on an earlier draft of this manuscript.

A 61-year-old man presented to the emergency department (ED) for persistent headache that began after he fell in his bathroom 4 days earlier. He described the headache as generalized and constant, rating the severity as a 5 on a scale of 0 to 10. The patient denied any associated neck pain or changes in headache quality with position change. He reported a 3-day history of nausea and four episodes of vomiting.

Headache after a fall raises concern for intracranial hemorrhage, particularly if this patient is on anticoagulant or antiplatelet medications. Subdural hematoma (SDH) would be more likely than epidural or subarachnoid hematoma (SAH) given the duration of days without progression. While nausea and vomiting are nonspecific, persistent vomiting may indicate increased intracranial pressure (eg, from an intracranial mass or SDH), particularly if provoked by positional changes. Without a history of fever or neck stiffness, meningitis is less likely unless the patient has a history of immunosuppression. Secondary causes of headache include vascular etiologies (eg, hemorrhagic cerebrovascular accident [CVA], arterial dissection, aneurysm, vasculitis), systemic causes (eg, chronic hypoxia/hypercapnia, hypertension), or medication overuse or withdrawal. In this patient, traumatic head injury with resultant postconcussive symptoms, though a diagnosis of exclusion, should also be considered. If the patient has a history of migraines, it is essential to obtain a history of typical migraine symptoms. More information regarding the mechanism of the fall is also essential to help elucidate a potential cause.

The patient had a 1-year history of recurrent loss of consciousness resulting in falls. After each fall, he quickly regained consciousness and exhibited no residual deficits or confusion. These episodes occurred suddenly when the patient was performing normal daily activities such as walking, driving, doing light chores, and standing up from a seated position. Immediately before this most recent fall, the patient stood up from a chair, walked toward the bathroom and, without any warning signs, lost consciousness. He denied dizziness, lightheadedness, nausea, or diaphoresis immediately before or after the fall. He also reported experiencing intermittent palpitations, but these did not appear to be related to the syncopal episodes. He denied experiencing chest pain, shortness of breath, or seizures.

The differential diagnosis for syncope is broad; therefore, it is important to identify features that suggest an etiology requiring urgent management. In this patient, cardiac etiologies such as arrhythmia (eg, atrial fibrillation [AF], ventricular tachycardia, heart block), ischemia, heart failure, and structural heart disease (eg, valvular abnormalities, cardiomyopathies) must be considered. His complaints of intermittent palpitations could suggest arrhythmia; however, the absence of a correlation to the syncopal episodes and other associated cardiac symptoms makes arrhythmias such as AF less likely. Medication side effects provoking cardiac conduction disturbances, heart block, or hypotension should be considered. Ischemic heart disease and heart failure are possible causes despite the absence of chest pain and dyspnea. While the exertional nature of the patient’s symptoms could support cardiac etiologies, it could also be indicative of recurrent pulmonary embolism  or right ventricular dysfunction/strain, such as chronic thromboembolic pulmonary hypertension (CTEPH).

Neurologic causes of syncope should also be included in the differential diagnosis. Seizure is less likely the underlying cause in this case since the patient regained consciousness quickly after each episode and reported no residual deficits, confusion, incontinence, or oral trauma. While less likely, other neurovascular causes can be considered, including transient ischemic attack (TIA), CVA, SAH, or vertebrobasilar insufficiency.

Neurocardiogenic syncope is less likely due to lack of a clear trigger or classical prodromal symptoms. Without a history of volume loss, orthostatic syncope is also unlikely. Other possibilities include adrenal insufficiency or an autonomic dysfunction resulting from diabetic neuropathy, chronic kidney disease, amyloidosis, spinal cord injury, or neurologic diseases (eg, Parkinson disease, Lewy body dementia). Thus far, the provided history is not suggestive of these etiologies. Other causes for loss of consciousness include hypoglycemia, sleep disorders (eg, narcolepsy), or psychiatric causes.

About 10 months prior to this presentation, the patient had presented to the hospital for evaluation of headache and was found to have bilateral SDH requiring burr hole evacuation. At that time, he was on anticoagulation therapy for a history of left superficial femoral vein thrombosis with negative workup for hypercoagulability. Warfarin was discontinued after the SDH was diagnosed. Regarding the patient’s social history, although he reported drinking two glasses of wine with dinner each night and smoking marijuana afterward, all syncopal events occurred during the daytime.

The history of prior SDH should raise suspicion for recurrent SDH, particularly considering the patient’s ongoing alcohol use. History of deep vein thrombosis (DVT) and possible exertional syncope might suggest recurrent pulmonary embolism or CTEPH as an etiology. DVT and TIA/CVA secondary to paradoxical embolism are also possible. Depending on extent of alcohol use, intoxication and cardiomyopathy with secondary arrhythmias are possibilities.

The basic workup should focus on identifying any acute intracranial processes that may explain the patient’s presentation and evaluating for syncope. This includes a complete blood count with differential, electrolytes, hepatic panel (based on patient’s history of alcohol use), and coagulation studies. Troponins and B-type natriuretic peptide would help assess for cardiac disease, and a urine/serum drug test would be beneficial to screen for substance use. Considering the patient’s prior history of SDH, head imaging should be obtained. If the patient were to exhibit focal neurologic deficits or persistent alterations in consciousness (thereby raising the index of suspicion for TIA or CVA), perfusion/diffusion-weighted magnetic resonance imaging (MRI) studies should be obtained. If obtaining a brain MRI is not practical, then a computed tomography angiogram (CTA) of the head and neck should be obtained. A noncontrast head CT would be sufficient to reveal the presence of SDH. An electroencephalogram (EEG) to assess for seizure should be performed if the patient is noted to have any focal neurologic findings or complaints consistent with seizure. With possible exertional syncope, an electrocardiogram (ECG) and transthoracic echocardiogram (with bubble study to assess for patent foramen ovale) should be obtained urgently.

The patient had a history of hypertension and irritable bowel syndrome, for which he took metoprolol and duloxetine, respectively. Eight months prior to the current ED presentation, he was admitted to the hospital for a syncope workup after falling and sustaining a fractured jaw and torn rotator cuff. ECG and continuous telemetry monitoring showed normal sinus rhythm, normal intervals, and rare episodes of sinus tachycardia, but no evidence of arrhythmia. An echocardiogram demonstrated normal ejection fraction and chamber sizes; CT and MRI of the brain showed no residual SDH; and EEG monitoring showed no seizure activity. It was determined that the patient’s syncopal episodes were multifactorial; possible etiologies included episodic hypotension from irritable bowel syndrome—related diarrhea, paroxysmal arrhythmias, and ongoing substance use.

The patient was discharged home with a 14-day Holter monitor. Rare episodes of AF (total burden 0.4%) were detected, and dronedarone was prescribed for rhythm control; he remained off anticoagulation therapy due to the history of SDH. Over the next few months, cardiology, electrophysiology, and neurology consultants concluded that paroxysmal AF was the likely etiology of the patient’s syncopal episodes. The patient was considered high risk for CVA, but the risk of bleeding from syncope-related falls was too high to resume anticoagulation therapy.

One month prior to the current ED presentation, the patient underwent a left atrial appendage closure with a WATCHMAN implant to avoid long-term anticoagulation. After the procedure, he was started on warfarin with plans to permanently discontinue anticoagulation after 6 to 8 weeks of completed therapy. He had been on warfarin for 3 weeks prior the most recent fall and current ED visit.At the time of this presentation, the patient was on dronedarone, duloxetine, metoprolol, and warfarin. On exam, he was alert and in no distress. His temperature was 36.8 °C, heart rate 98 beats per minute , blood pressure (BP) 110/75 mm Hg (with no orthostatic changes), respiratory rate 18 breaths per minute, and oxygen saturation 95% on room air. He had a regular heart rate and rhythm, clear lung fields, and a benign abdominal exam. He was oriented to time, place, and person. His pupils were equal in size and reactive to light, and sensation and strength were equal bilaterally with no focal neurologic deficits. His neck was supple, and head movements did not cause any symptoms. His musculoskeletal exam was notable for right supraspinatus weakness upon abduction of arm to 90° and a positive impingement sign. ECG showed normal sinus rhythm with normal intervals. Laboratory findings were notable only for an international normalized ratio of 4.9. CT of the head did not show any pathology. The patient was admitted to the medicine floor for further evaluation.

At this point in his clinical course, the patient has had a thorough workup—one that has largely been unrevealing aside from paroxysmal AF. With his current presentation, acute intracranial causes remain on the differential, but the normal CT scan essentially excludes hemorrhage or mass. Although previous MRI studies have been negative and no focal neurologic findings have been described throughout his course, given the patient’s repeated presentations for syncope, intracranial vessel imaging should be obtained to exclude anatomical abnormalities or focal stenosis that could cause recurrent TIAs.

Seizure is also a consideration, but prior EEG and normal neurologic exam makes this less likely. While cardiac workup for syncope has been reassuring, the patient’s history of AF should continue to remain a consideration even though this is less likely the underlying cause since he is now taking dronedarone. He should be placed on telemetry upon admission. While negative orthostatic vital signs make orthostatic syncope less likely, this could be confounded by use of beta-blockers. Overall, the patient’s case remains a challenging one, with the etiology of his syncope remaining unclear at this time.

During this hospitalization, possible etiologies for recurrent syncope and falls were reviewed. The burden of verifiable AF was too low to explain the patient’s recurrent syncopal episodes. Further review of his medical record revealed that a carotid ultrasound was obtained a year earlier in the course of a previous hospitalization. The ultrasound report described patent carotid arteries and retrograde flow in the left vertebral artery consistent with ipsilateral subclavian stenosis. At the time, the ultrasound was interpreted as reassuring based on the lack of significant carotid stenosis; the findings were thought to be unrelated to the patient’s syncopal episodes. On further questioning, the patient noted that minimal exertion such as unloading a few items from the dishwasher caused left arm pain and paresthesia, accompanied by headache and lightheadedness. He also reported using his left arm more frequently following a right shoulder injury. Repeat physical exam found an inter-arm systolic BP difference (IASBPD) >40 mm Hg and left-arm claudication. CT-angiogram of the neck was obtained and showed total occlusion of the left proximal subclavian artery, patent bilateral internal carotid arteries, and retrograde flow in the left vertebral artery (Figure 1).

Subclavian steal syndrome (SSS) results from compromised flow to the distal arm or brainstem circulation due to a proximal subclavian artery occlusion or stenosis (prior to the origin of the vertebral artery).^1,2 Subclavian stenosis may cause lowered pressure in the distal subclavian artery, creating a gradient for blood flow from the contralateral vertebral artery through the basilar artery to the ipsilateral vertebral artery, ultimately supplying blood flow to the affected subclavian artery distal to the occlusion (subclavian steal phenomenon). Flow reversal in the vertebrobasilar system can result in hypoperfusion of the brainstem (ie, vertebrobasilar insufficiency), which can cause a variety of neurologic symptoms, including SSS. While atherosclerosis is the most common cause of subclavian steal, it may be due to other conditions (eg, Takayasu arteritis, thoracic outlet syndrome, congenital heart disease).

Clinically, although many patients with proximal subclavian stenosis are asymptomatic (even in cases wherein angiographic flow reversal is detected), it is critical that clinicians be familiar with common symptoms associated with the diagnosis. Symptoms may include arm claudication related to hypoperfusion of the extremity, particularly when performing activities, as was observed in this patient. Neurologic symptoms are less common but include symptoms consistent with compromised posterior circulation such as dizziness, vertigo, ataxia, diplopia, nystagmus, impaired vision (blurring of vision, hemianopia), otologic symptoms (tinnitus, hearing loss), and/or syncope (ie, “drop attacks”). The patient’s initial complaints of sudden syncope are consistent with this presentation, as are his history of headache and lightheadedness upon use of his left arm.

Diagnostically, a gradient in upper extremity BP >15 mm Hg (as seen in this patient) or findings of arterial insufficiency would suggest subclavian stenosis. Duplex ultrasound is a reliable imaging modality and can demonstrate proximal subclavian stenosis (sensitivity of 90.9% and specificity of 82.5% for predicting >70% of stenosis cases) and ipsilateral vertebral artery flow reversal.¹ Transcranial Doppler studies can be obtained to assess for basilar artery flow reversal as well. CTA/MRA can help delineate location, severity, and cause of stenosis. However, detection of vertebral or basilar artery flow reversal does not always correlate with the development of neurologic symptoms.

For patients with asymptomatic subclavian stenosis, medical management with aspirin, beta-blockers, angiotensin-converting enzyme inhibitors, and statins should be considered given the high likelihood for other atherosclerotic disease. Management of SSS may include percutaneous/surgical intervention in combination with medical therapy, particularly for patients with severe symptomatic disease (arm claudication, posterior circulation deficits, or coronary ischemia in patients with history of coronary bypass utilizing the left internal mammary artery).

The patient was diagnosed with SSS. Cardiovascular medicine and vascular surgery services were asked to evaluate the patient for a revascularization procedure. Because the patient’s anterior circulation was intact, several specialists remained skeptical of SSS as the cause of his syncope. As such, further evaluation for arrhythmia was recommended. The patient’s arm claudication was thought to be due to SSS; however, the well-established retrograde flow via the vertebral artery made a revascularization procedure nonurgent. Moreover, continuation of warfarin was necessary in the setting of his recent left atrial appendage closure and prior history of DVT. It was determined that the risks of discontinuing anticoagulation in order to surgically treat his subclavian stenosis outweighed the benefits. In the meantime, brachial-radial index measurement and a 30-day event monitor were ordered to further assess for arrhythmias. The patient reported being overwhelmed by diagnostic testing without resolution of his syncopal episodes and missed some of his scheduled appointments. One month later, he fell again and sustained vertebral fractures at C1, C4, and L1, and a subsequent SDH requiring craniotomy with a bone flap followed by clot evacuation. The 30-day event monitor report did not reveal any arrhythmias before, during, or after multiple syncopal events that occurred in the period leading up to this fall. The patient later died in a neurology intensive care unit.

SSS often stems from atherosclerotic arterial disease that leads to stenosis or occlusion of the proximal subclavian artery, causing decreased pressure distal to the lesion. The left subclavian artery is affected more often than the right because of its acute angle of origin, which presumably causes turbulence and predisposes to atherosclerosis.³ Compromised blood flow to the arm causes exertional arm claudication and paresthesia. The compensatory retrograde flow in the ipsilateral vertebral artery causes symptoms of vertebrobasilar insufficiency such as dizziness, vertigo, and syncope (Figure 2). This conglomerate of symptoms from subclavian steal, by definition, comprises SSS. The most remarkable signs of SSS are IASBPD >20 mm Hg and, less commonly, reproducible arm claudication.

Diagnosis of SSS requires a careful correlation of clinical history, physical examination, and radiologic findings. Over 80% of patients with subclavian disease have concomitant lesions (eg, in carotid arteries) that can affect collateral circulation.⁴ While symptoms of SSS may vary depending on the adequacy of collaterals, patent anterior circulation does not, by default, prevent SSS in patients with subclavian stenosis.³ In one study, neurologic symptoms were found in 36% of individuals with subclavian stenosis and concomitant carotid atherosclerotic lesions, and in only 5% in patients without carotid lesions.⁵

A key step in diagnosis is measurement of bilateral arm BP as elevated IASBPDs are highly sensitive for subclavian steal. More than 80% of patients with IASBPD >20 mm Hg have evidence of this condition on Doppler ultrasound.³ Higher differentials in BP correlate with occurrence of symptoms (~30% of patients with IASBPD 40-50 mm Hg, and ~40% of those with IASBPD >50 mm Hg).⁶

The severity of subclavian stenosis is traditionally classified by imaging into three separate grades or stages based on the direction of blood flow in vertebral arteries. Grade I involves no retrograde flow; grade II involves cardiac cycle dependent alternating antegrade and retrograde flow; and grade III involves permanent retrograde flow (complete steal).⁷ Our patient’s care was impacted by an unsupported conventional belief that grade II SSS may involve more hemodynamic instability and produce more severe symptoms compared to permanent retrograde flow (grade III), which would result in more stability with a reset of hemodynamics in posterior circulation.⁷ This hypothesis has been disproven in the past, and our patient’s tragic outcome also demonstrates that complete steal is not harmless.⁸ Our patient had permanent retrograde flow in the left vertebral artery, and he suffered classic symptoms of SSS, with devastating consequences. Moreover, increased demand or exertion can enhance the retrograde flow even in grade III stenosis and can precipitate neurologic symptoms of SSS, including syncope. This case provides an important lesson: Management of patients with SSS should depend on the severity of symptoms, not solely on radiologic grading.

Management of SSS is often medical for atherosclerosis and hypertension, especially if symptoms are mild and infrequent. Less than 10% patients with radiologic evidence of subclavian stenosis are symptomatic, and <20% patients with symptomatic SSS require revascularization.³ Percutaneous transluminal angioplasty (PTA) and stenting have become the most favored surgical approach rather than extra-anatomic revascularization techniques.⁷ Both endovascular interventions and open revascularization carry an excellent success rate with low morbidity. Patients undergoing PTA have a combined rate of 3.6% for CVA and death⁹ and a 5-year primary patency rate¹⁰ of 82%. Bypass surgery appears similarly well tolerated, with low perioperative CVA/mortality, and a 10-year primary patency rate of 92%.¹¹ For patients with SSS and coexisting disease in the anterior circulation, carotid endarterectomy is prioritized over subclavian revascularization as repair of the anterior circulation often resolves symptoms of SSS.¹²

In our patient, SSS presented with classic vertebrobasilar and brachial symptoms, but several features of his presentation made the diagnosis a challenge. First, his history suggested several potential causes of syncope, including arrhythmia, orthostatic hypotension, and substance use. Second, he reported arm paresthesia and claudication only when specifically prompted and after a targeted history was obtained. Third, there were no consistent triggers for his syncopal episodes. The patient noted that he lost consciousness when walking, driving, doing light chores, and arising from a seated position. These atypical triggers of syncope were not consistent with any of the illnesses considered during the initial workup, and therefore resulted in a broad differential, delaying the targeted workup for SSS. The wisdom of parsimony may also have played an unintended role: In clinical practice, common things are common, and explanation of most or all symptoms with a known diagnosis is often correct rather than addition of uncommon disorders.

Unfortunately, this patient kept falling through the cracks. Several providers believed that AF and alcohol use were the likely causes of his syncope. This assumption enabled a less than rigorous appraisal of the critical ultrasound report. If SSS had been on the differential, assessing the patient for the associated signs and symptoms might have led to an earlier diagnosis.

• SSS should be included in the differential diagnosis of patients with syncope, especially when common diagnoses have been ruled out.
• Incidentally detected retrograde vertebral flow on ultrasound should never be dismissed, and the patients should be assessed for signs and symptoms of subclavian steal.
• A difference in inter-arm systolic blood pressure >20 mm Hg is highly suggestive of subclavian stenosis.
• SSS has excellent prognosis with appropriate medical treatment or revascularization.

A 61-year-old man presented to the emergency department (ED) for persistent headache that began after he fell in his bathroom 4 days earlier. He described the headache as generalized and constant, rating the severity as a 5 on a scale of 0 to 10. The patient denied any associated neck pain or changes in headache quality with position change. He reported a 3-day history of nausea and four episodes of vomiting.

Headache after a fall raises concern for intracranial hemorrhage, particularly if this patient is on anticoagulant or antiplatelet medications. Subdural hematoma (SDH) would be more likely than epidural or subarachnoid hematoma (SAH) given the duration of days without progression. While nausea and vomiting are nonspecific, persistent vomiting may indicate increased intracranial pressure (eg, from an intracranial mass or SDH), particularly if provoked by positional changes. Without a history of fever or neck stiffness, meningitis is less likely unless the patient has a history of immunosuppression. Secondary causes of headache include vascular etiologies (eg, hemorrhagic cerebrovascular accident [CVA], arterial dissection, aneurysm, vasculitis), systemic causes (eg, chronic hypoxia/hypercapnia, hypertension), or medication overuse or withdrawal. In this patient, traumatic head injury with resultant postconcussive symptoms, though a diagnosis of exclusion, should also be considered. If the patient has a history of migraines, it is essential to obtain a history of typical migraine symptoms. More information regarding the mechanism of the fall is also essential to help elucidate a potential cause.

The patient had a 1-year history of recurrent loss of consciousness resulting in falls. After each fall, he quickly regained consciousness and exhibited no residual deficits or confusion. These episodes occurred suddenly when the patient was performing normal daily activities such as walking, driving, doing light chores, and standing up from a seated position. Immediately before this most recent fall, the patient stood up from a chair, walked toward the bathroom and, without any warning signs, lost consciousness. He denied dizziness, lightheadedness, nausea, or diaphoresis immediately before or after the fall. He also reported experiencing intermittent palpitations, but these did not appear to be related to the syncopal episodes. He denied experiencing chest pain, shortness of breath, or seizures.

The differential diagnosis for syncope is broad; therefore, it is important to identify features that suggest an etiology requiring urgent management. In this patient, cardiac etiologies such as arrhythmia (eg, atrial fibrillation [AF], ventricular tachycardia, heart block), ischemia, heart failure, and structural heart disease (eg, valvular abnormalities, cardiomyopathies) must be considered. His complaints of intermittent palpitations could suggest arrhythmia; however, the absence of a correlation to the syncopal episodes and other associated cardiac symptoms makes arrhythmias such as AF less likely. Medication side effects provoking cardiac conduction disturbances, heart block, or hypotension should be considered. Ischemic heart disease and heart failure are possible causes despite the absence of chest pain and dyspnea. While the exertional nature of the patient’s symptoms could support cardiac etiologies, it could also be indicative of recurrent pulmonary embolism  or right ventricular dysfunction/strain, such as chronic thromboembolic pulmonary hypertension (CTEPH).

Neurologic causes of syncope should also be included in the differential diagnosis. Seizure is less likely the underlying cause in this case since the patient regained consciousness quickly after each episode and reported no residual deficits, confusion, incontinence, or oral trauma. While less likely, other neurovascular causes can be considered, including transient ischemic attack (TIA), CVA, SAH, or vertebrobasilar insufficiency.

Neurocardiogenic syncope is less likely due to lack of a clear trigger or classical prodromal symptoms. Without a history of volume loss, orthostatic syncope is also unlikely. Other possibilities include adrenal insufficiency or an autonomic dysfunction resulting from diabetic neuropathy, chronic kidney disease, amyloidosis, spinal cord injury, or neurologic diseases (eg, Parkinson disease, Lewy body dementia). Thus far, the provided history is not suggestive of these etiologies. Other causes for loss of consciousness include hypoglycemia, sleep disorders (eg, narcolepsy), or psychiatric causes.

About 10 months prior to this presentation, the patient had presented to the hospital for evaluation of headache and was found to have bilateral SDH requiring burr hole evacuation. At that time, he was on anticoagulation therapy for a history of left superficial femoral vein thrombosis with negative workup for hypercoagulability. Warfarin was discontinued after the SDH was diagnosed. Regarding the patient’s social history, although he reported drinking two glasses of wine with dinner each night and smoking marijuana afterward, all syncopal events occurred during the daytime.

The history of prior SDH should raise suspicion for recurrent SDH, particularly considering the patient’s ongoing alcohol use. History of deep vein thrombosis (DVT) and possible exertional syncope might suggest recurrent pulmonary embolism or CTEPH as an etiology. DVT and TIA/CVA secondary to paradoxical embolism are also possible. Depending on extent of alcohol use, intoxication and cardiomyopathy with secondary arrhythmias are possibilities.

The basic workup should focus on identifying any acute intracranial processes that may explain the patient’s presentation and evaluating for syncope. This includes a complete blood count with differential, electrolytes, hepatic panel (based on patient’s history of alcohol use), and coagulation studies. Troponins and B-type natriuretic peptide would help assess for cardiac disease, and a urine/serum drug test would be beneficial to screen for substance use. Considering the patient’s prior history of SDH, head imaging should be obtained. If the patient were to exhibit focal neurologic deficits or persistent alterations in consciousness (thereby raising the index of suspicion for TIA or CVA), perfusion/diffusion-weighted magnetic resonance imaging (MRI) studies should be obtained. If obtaining a brain MRI is not practical, then a computed tomography angiogram (CTA) of the head and neck should be obtained. A noncontrast head CT would be sufficient to reveal the presence of SDH. An electroencephalogram (EEG) to assess for seizure should be performed if the patient is noted to have any focal neurologic findings or complaints consistent with seizure. With possible exertional syncope, an electrocardiogram (ECG) and transthoracic echocardiogram (with bubble study to assess for patent foramen ovale) should be obtained urgently.

The patient had a history of hypertension and irritable bowel syndrome, for which he took metoprolol and duloxetine, respectively. Eight months prior to the current ED presentation, he was admitted to the hospital for a syncope workup after falling and sustaining a fractured jaw and torn rotator cuff. ECG and continuous telemetry monitoring showed normal sinus rhythm, normal intervals, and rare episodes of sinus tachycardia, but no evidence of arrhythmia. An echocardiogram demonstrated normal ejection fraction and chamber sizes; CT and MRI of the brain showed no residual SDH; and EEG monitoring showed no seizure activity. It was determined that the patient’s syncopal episodes were multifactorial; possible etiologies included episodic hypotension from irritable bowel syndrome—related diarrhea, paroxysmal arrhythmias, and ongoing substance use.

The patient was discharged home with a 14-day Holter monitor. Rare episodes of AF (total burden 0.4%) were detected, and dronedarone was prescribed for rhythm control; he remained off anticoagulation therapy due to the history of SDH. Over the next few months, cardiology, electrophysiology, and neurology consultants concluded that paroxysmal AF was the likely etiology of the patient’s syncopal episodes. The patient was considered high risk for CVA, but the risk of bleeding from syncope-related falls was too high to resume anticoagulation therapy.

One month prior to the current ED presentation, the patient underwent a left atrial appendage closure with a WATCHMAN implant to avoid long-term anticoagulation. After the procedure, he was started on warfarin with plans to permanently discontinue anticoagulation after 6 to 8 weeks of completed therapy. He had been on warfarin for 3 weeks prior the most recent fall and current ED visit.At the time of this presentation, the patient was on dronedarone, duloxetine, metoprolol, and warfarin. On exam, he was alert and in no distress. His temperature was 36.8 °C, heart rate 98 beats per minute , blood pressure (BP) 110/75 mm Hg (with no orthostatic changes), respiratory rate 18 breaths per minute, and oxygen saturation 95% on room air. He had a regular heart rate and rhythm, clear lung fields, and a benign abdominal exam. He was oriented to time, place, and person. His pupils were equal in size and reactive to light, and sensation and strength were equal bilaterally with no focal neurologic deficits. His neck was supple, and head movements did not cause any symptoms. His musculoskeletal exam was notable for right supraspinatus weakness upon abduction of arm to 90° and a positive impingement sign. ECG showed normal sinus rhythm with normal intervals. Laboratory findings were notable only for an international normalized ratio of 4.9. CT of the head did not show any pathology. The patient was admitted to the medicine floor for further evaluation.

At this point in his clinical course, the patient has had a thorough workup—one that has largely been unrevealing aside from paroxysmal AF. With his current presentation, acute intracranial causes remain on the differential, but the normal CT scan essentially excludes hemorrhage or mass. Although previous MRI studies have been negative and no focal neurologic findings have been described throughout his course, given the patient’s repeated presentations for syncope, intracranial vessel imaging should be obtained to exclude anatomical abnormalities or focal stenosis that could cause recurrent TIAs.

Seizure is also a consideration, but prior EEG and normal neurologic exam makes this less likely. While cardiac workup for syncope has been reassuring, the patient’s history of AF should continue to remain a consideration even though this is less likely the underlying cause since he is now taking dronedarone. He should be placed on telemetry upon admission. While negative orthostatic vital signs make orthostatic syncope less likely, this could be confounded by use of beta-blockers. Overall, the patient’s case remains a challenging one, with the etiology of his syncope remaining unclear at this time.

During this hospitalization, possible etiologies for recurrent syncope and falls were reviewed. The burden of verifiable AF was too low to explain the patient’s recurrent syncopal episodes. Further review of his medical record revealed that a carotid ultrasound was obtained a year earlier in the course of a previous hospitalization. The ultrasound report described patent carotid arteries and retrograde flow in the left vertebral artery consistent with ipsilateral subclavian stenosis. At the time, the ultrasound was interpreted as reassuring based on the lack of significant carotid stenosis; the findings were thought to be unrelated to the patient’s syncopal episodes. On further questioning, the patient noted that minimal exertion such as unloading a few items from the dishwasher caused left arm pain and paresthesia, accompanied by headache and lightheadedness. He also reported using his left arm more frequently following a right shoulder injury. Repeat physical exam found an inter-arm systolic BP difference (IASBPD) >40 mm Hg and left-arm claudication. CT-angiogram of the neck was obtained and showed total occlusion of the left proximal subclavian artery, patent bilateral internal carotid arteries, and retrograde flow in the left vertebral artery (Figure 1).

Subclavian steal syndrome (SSS) results from compromised flow to the distal arm or brainstem circulation due to a proximal subclavian artery occlusion or stenosis (prior to the origin of the vertebral artery).^1,2 Subclavian stenosis may cause lowered pressure in the distal subclavian artery, creating a gradient for blood flow from the contralateral vertebral artery through the basilar artery to the ipsilateral vertebral artery, ultimately supplying blood flow to the affected subclavian artery distal to the occlusion (subclavian steal phenomenon). Flow reversal in the vertebrobasilar system can result in hypoperfusion of the brainstem (ie, vertebrobasilar insufficiency), which can cause a variety of neurologic symptoms, including SSS. While atherosclerosis is the most common cause of subclavian steal, it may be due to other conditions (eg, Takayasu arteritis, thoracic outlet syndrome, congenital heart disease).

Clinically, although many patients with proximal subclavian stenosis are asymptomatic (even in cases wherein angiographic flow reversal is detected), it is critical that clinicians be familiar with common symptoms associated with the diagnosis. Symptoms may include arm claudication related to hypoperfusion of the extremity, particularly when performing activities, as was observed in this patient. Neurologic symptoms are less common but include symptoms consistent with compromised posterior circulation such as dizziness, vertigo, ataxia, diplopia, nystagmus, impaired vision (blurring of vision, hemianopia), otologic symptoms (tinnitus, hearing loss), and/or syncope (ie, “drop attacks”). The patient’s initial complaints of sudden syncope are consistent with this presentation, as are his history of headache and lightheadedness upon use of his left arm.

Diagnostically, a gradient in upper extremity BP >15 mm Hg (as seen in this patient) or findings of arterial insufficiency would suggest subclavian stenosis. Duplex ultrasound is a reliable imaging modality and can demonstrate proximal subclavian stenosis (sensitivity of 90.9% and specificity of 82.5% for predicting >70% of stenosis cases) and ipsilateral vertebral artery flow reversal.¹ Transcranial Doppler studies can be obtained to assess for basilar artery flow reversal as well. CTA/MRA can help delineate location, severity, and cause of stenosis. However, detection of vertebral or basilar artery flow reversal does not always correlate with the development of neurologic symptoms.

For patients with asymptomatic subclavian stenosis, medical management with aspirin, beta-blockers, angiotensin-converting enzyme inhibitors, and statins should be considered given the high likelihood for other atherosclerotic disease. Management of SSS may include percutaneous/surgical intervention in combination with medical therapy, particularly for patients with severe symptomatic disease (arm claudication, posterior circulation deficits, or coronary ischemia in patients with history of coronary bypass utilizing the left internal mammary artery).

The patient was diagnosed with SSS. Cardiovascular medicine and vascular surgery services were asked to evaluate the patient for a revascularization procedure. Because the patient’s anterior circulation was intact, several specialists remained skeptical of SSS as the cause of his syncope. As such, further evaluation for arrhythmia was recommended. The patient’s arm claudication was thought to be due to SSS; however, the well-established retrograde flow via the vertebral artery made a revascularization procedure nonurgent. Moreover, continuation of warfarin was necessary in the setting of his recent left atrial appendage closure and prior history of DVT. It was determined that the risks of discontinuing anticoagulation in order to surgically treat his subclavian stenosis outweighed the benefits. In the meantime, brachial-radial index measurement and a 30-day event monitor were ordered to further assess for arrhythmias. The patient reported being overwhelmed by diagnostic testing without resolution of his syncopal episodes and missed some of his scheduled appointments. One month later, he fell again and sustained vertebral fractures at C1, C4, and L1, and a subsequent SDH requiring craniotomy with a bone flap followed by clot evacuation. The 30-day event monitor report did not reveal any arrhythmias before, during, or after multiple syncopal events that occurred in the period leading up to this fall. The patient later died in a neurology intensive care unit.

SSS often stems from atherosclerotic arterial disease that leads to stenosis or occlusion of the proximal subclavian artery, causing decreased pressure distal to the lesion. The left subclavian artery is affected more often than the right because of its acute angle of origin, which presumably causes turbulence and predisposes to atherosclerosis.³ Compromised blood flow to the arm causes exertional arm claudication and paresthesia. The compensatory retrograde flow in the ipsilateral vertebral artery causes symptoms of vertebrobasilar insufficiency such as dizziness, vertigo, and syncope (Figure 2). This conglomerate of symptoms from subclavian steal, by definition, comprises SSS. The most remarkable signs of SSS are IASBPD >20 mm Hg and, less commonly, reproducible arm claudication.

Diagnosis of SSS requires a careful correlation of clinical history, physical examination, and radiologic findings. Over 80% of patients with subclavian disease have concomitant lesions (eg, in carotid arteries) that can affect collateral circulation.⁴ While symptoms of SSS may vary depending on the adequacy of collaterals, patent anterior circulation does not, by default, prevent SSS in patients with subclavian stenosis.³ In one study, neurologic symptoms were found in 36% of individuals with subclavian stenosis and concomitant carotid atherosclerotic lesions, and in only 5% in patients without carotid lesions.⁵

A key step in diagnosis is measurement of bilateral arm BP as elevated IASBPDs are highly sensitive for subclavian steal. More than 80% of patients with IASBPD >20 mm Hg have evidence of this condition on Doppler ultrasound.³ Higher differentials in BP correlate with occurrence of symptoms (~30% of patients with IASBPD 40-50 mm Hg, and ~40% of those with IASBPD >50 mm Hg).⁶

The severity of subclavian stenosis is traditionally classified by imaging into three separate grades or stages based on the direction of blood flow in vertebral arteries. Grade I involves no retrograde flow; grade II involves cardiac cycle dependent alternating antegrade and retrograde flow; and grade III involves permanent retrograde flow (complete steal).⁷ Our patient’s care was impacted by an unsupported conventional belief that grade II SSS may involve more hemodynamic instability and produce more severe symptoms compared to permanent retrograde flow (grade III), which would result in more stability with a reset of hemodynamics in posterior circulation.⁷ This hypothesis has been disproven in the past, and our patient’s tragic outcome also demonstrates that complete steal is not harmless.⁸ Our patient had permanent retrograde flow in the left vertebral artery, and he suffered classic symptoms of SSS, with devastating consequences. Moreover, increased demand or exertion can enhance the retrograde flow even in grade III stenosis and can precipitate neurologic symptoms of SSS, including syncope. This case provides an important lesson: Management of patients with SSS should depend on the severity of symptoms, not solely on radiologic grading.

Management of SSS is often medical for atherosclerosis and hypertension, especially if symptoms are mild and infrequent. Less than 10% patients with radiologic evidence of subclavian stenosis are symptomatic, and <20% patients with symptomatic SSS require revascularization.³ Percutaneous transluminal angioplasty (PTA) and stenting have become the most favored surgical approach rather than extra-anatomic revascularization techniques.⁷ Both endovascular interventions and open revascularization carry an excellent success rate with low morbidity. Patients undergoing PTA have a combined rate of 3.6% for CVA and death⁹ and a 5-year primary patency rate¹⁰ of 82%. Bypass surgery appears similarly well tolerated, with low perioperative CVA/mortality, and a 10-year primary patency rate of 92%.¹¹ For patients with SSS and coexisting disease in the anterior circulation, carotid endarterectomy is prioritized over subclavian revascularization as repair of the anterior circulation often resolves symptoms of SSS.¹²

In our patient, SSS presented with classic vertebrobasilar and brachial symptoms, but several features of his presentation made the diagnosis a challenge. First, his history suggested several potential causes of syncope, including arrhythmia, orthostatic hypotension, and substance use. Second, he reported arm paresthesia and claudication only when specifically prompted and after a targeted history was obtained. Third, there were no consistent triggers for his syncopal episodes. The patient noted that he lost consciousness when walking, driving, doing light chores, and arising from a seated position. These atypical triggers of syncope were not consistent with any of the illnesses considered during the initial workup, and therefore resulted in a broad differential, delaying the targeted workup for SSS. The wisdom of parsimony may also have played an unintended role: In clinical practice, common things are common, and explanation of most or all symptoms with a known diagnosis is often correct rather than addition of uncommon disorders.

Unfortunately, this patient kept falling through the cracks. Several providers believed that AF and alcohol use were the likely causes of his syncope. This assumption enabled a less than rigorous appraisal of the critical ultrasound report. If SSS had been on the differential, assessing the patient for the associated signs and symptoms might have led to an earlier diagnosis.

• SSS should be included in the differential diagnosis of patients with syncope, especially when common diagnoses have been ruled out.
• Incidentally detected retrograde vertebral flow on ultrasound should never be dismissed, and the patients should be assessed for signs and symptoms of subclavian steal.
• A difference in inter-arm systolic blood pressure >20 mm Hg is highly suggestive of subclavian stenosis.
• SSS has excellent prognosis with appropriate medical treatment or revascularization.

A 61-year-old man presented to the emergency department (ED) for persistent headache that began after he fell in his bathroom 4 days earlier. He described the headache as generalized and constant, rating the severity as a 5 on a scale of 0 to 10. The patient denied any associated neck pain or changes in headache quality with position change. He reported a 3-day history of nausea and four episodes of vomiting.

Headache after a fall raises concern for intracranial hemorrhage, particularly if this patient is on anticoagulant or antiplatelet medications. Subdural hematoma (SDH) would be more likely than epidural or subarachnoid hematoma (SAH) given the duration of days without progression. While nausea and vomiting are nonspecific, persistent vomiting may indicate increased intracranial pressure (eg, from an intracranial mass or SDH), particularly if provoked by positional changes. Without a history of fever or neck stiffness, meningitis is less likely unless the patient has a history of immunosuppression. Secondary causes of headache include vascular etiologies (eg, hemorrhagic cerebrovascular accident [CVA], arterial dissection, aneurysm, vasculitis), systemic causes (eg, chronic hypoxia/hypercapnia, hypertension), or medication overuse or withdrawal. In this patient, traumatic head injury with resultant postconcussive symptoms, though a diagnosis of exclusion, should also be considered. If the patient has a history of migraines, it is essential to obtain a history of typical migraine symptoms. More information regarding the mechanism of the fall is also essential to help elucidate a potential cause.

The patient had a 1-year history of recurrent loss of consciousness resulting in falls. After each fall, he quickly regained consciousness and exhibited no residual deficits or confusion. These episodes occurred suddenly when the patient was performing normal daily activities such as walking, driving, doing light chores, and standing up from a seated position. Immediately before this most recent fall, the patient stood up from a chair, walked toward the bathroom and, without any warning signs, lost consciousness. He denied dizziness, lightheadedness, nausea, or diaphoresis immediately before or after the fall. He also reported experiencing intermittent palpitations, but these did not appear to be related to the syncopal episodes. He denied experiencing chest pain, shortness of breath, or seizures.

The differential diagnosis for syncope is broad; therefore, it is important to identify features that suggest an etiology requiring urgent management. In this patient, cardiac etiologies such as arrhythmia (eg, atrial fibrillation [AF], ventricular tachycardia, heart block), ischemia, heart failure, and structural heart disease (eg, valvular abnormalities, cardiomyopathies) must be considered. His complaints of intermittent palpitations could suggest arrhythmia; however, the absence of a correlation to the syncopal episodes and other associated cardiac symptoms makes arrhythmias such as AF less likely. Medication side effects provoking cardiac conduction disturbances, heart block, or hypotension should be considered. Ischemic heart disease and heart failure are possible causes despite the absence of chest pain and dyspnea. While the exertional nature of the patient’s symptoms could support cardiac etiologies, it could also be indicative of recurrent pulmonary embolism  or right ventricular dysfunction/strain, such as chronic thromboembolic pulmonary hypertension (CTEPH).

Neurologic causes of syncope should also be included in the differential diagnosis. Seizure is less likely the underlying cause in this case since the patient regained consciousness quickly after each episode and reported no residual deficits, confusion, incontinence, or oral trauma. While less likely, other neurovascular causes can be considered, including transient ischemic attack (TIA), CVA, SAH, or vertebrobasilar insufficiency.

Neurocardiogenic syncope is less likely due to lack of a clear trigger or classical prodromal symptoms. Without a history of volume loss, orthostatic syncope is also unlikely. Other possibilities include adrenal insufficiency or an autonomic dysfunction resulting from diabetic neuropathy, chronic kidney disease, amyloidosis, spinal cord injury, or neurologic diseases (eg, Parkinson disease, Lewy body dementia). Thus far, the provided history is not suggestive of these etiologies. Other causes for loss of consciousness include hypoglycemia, sleep disorders (eg, narcolepsy), or psychiatric causes.

About 10 months prior to this presentation, the patient had presented to the hospital for evaluation of headache and was found to have bilateral SDH requiring burr hole evacuation. At that time, he was on anticoagulation therapy for a history of left superficial femoral vein thrombosis with negative workup for hypercoagulability. Warfarin was discontinued after the SDH was diagnosed. Regarding the patient’s social history, although he reported drinking two glasses of wine with dinner each night and smoking marijuana afterward, all syncopal events occurred during the daytime.

The history of prior SDH should raise suspicion for recurrent SDH, particularly considering the patient’s ongoing alcohol use. History of deep vein thrombosis (DVT) and possible exertional syncope might suggest recurrent pulmonary embolism or CTEPH as an etiology. DVT and TIA/CVA secondary to paradoxical embolism are also possible. Depending on extent of alcohol use, intoxication and cardiomyopathy with secondary arrhythmias are possibilities.

The basic workup should focus on identifying any acute intracranial processes that may explain the patient’s presentation and evaluating for syncope. This includes a complete blood count with differential, electrolytes, hepatic panel (based on patient’s history of alcohol use), and coagulation studies. Troponins and B-type natriuretic peptide would help assess for cardiac disease, and a urine/serum drug test would be beneficial to screen for substance use. Considering the patient’s prior history of SDH, head imaging should be obtained. If the patient were to exhibit focal neurologic deficits or persistent alterations in consciousness (thereby raising the index of suspicion for TIA or CVA), perfusion/diffusion-weighted magnetic resonance imaging (MRI) studies should be obtained. If obtaining a brain MRI is not practical, then a computed tomography angiogram (CTA) of the head and neck should be obtained. A noncontrast head CT would be sufficient to reveal the presence of SDH. An electroencephalogram (EEG) to assess for seizure should be performed if the patient is noted to have any focal neurologic findings or complaints consistent with seizure. With possible exertional syncope, an electrocardiogram (ECG) and transthoracic echocardiogram (with bubble study to assess for patent foramen ovale) should be obtained urgently.

The patient had a history of hypertension and irritable bowel syndrome, for which he took metoprolol and duloxetine, respectively. Eight months prior to the current ED presentation, he was admitted to the hospital for a syncope workup after falling and sustaining a fractured jaw and torn rotator cuff. ECG and continuous telemetry monitoring showed normal sinus rhythm, normal intervals, and rare episodes of sinus tachycardia, but no evidence of arrhythmia. An echocardiogram demonstrated normal ejection fraction and chamber sizes; CT and MRI of the brain showed no residual SDH; and EEG monitoring showed no seizure activity. It was determined that the patient’s syncopal episodes were multifactorial; possible etiologies included episodic hypotension from irritable bowel syndrome—related diarrhea, paroxysmal arrhythmias, and ongoing substance use.

The patient was discharged home with a 14-day Holter monitor. Rare episodes of AF (total burden 0.4%) were detected, and dronedarone was prescribed for rhythm control; he remained off anticoagulation therapy due to the history of SDH. Over the next few months, cardiology, electrophysiology, and neurology consultants concluded that paroxysmal AF was the likely etiology of the patient’s syncopal episodes. The patient was considered high risk for CVA, but the risk of bleeding from syncope-related falls was too high to resume anticoagulation therapy.

One month prior to the current ED presentation, the patient underwent a left atrial appendage closure with a WATCHMAN implant to avoid long-term anticoagulation. After the procedure, he was started on warfarin with plans to permanently discontinue anticoagulation after 6 to 8 weeks of completed therapy. He had been on warfarin for 3 weeks prior the most recent fall and current ED visit.At the time of this presentation, the patient was on dronedarone, duloxetine, metoprolol, and warfarin. On exam, he was alert and in no distress. His temperature was 36.8 °C, heart rate 98 beats per minute , blood pressure (BP) 110/75 mm Hg (with no orthostatic changes), respiratory rate 18 breaths per minute, and oxygen saturation 95% on room air. He had a regular heart rate and rhythm, clear lung fields, and a benign abdominal exam. He was oriented to time, place, and person. His pupils were equal in size and reactive to light, and sensation and strength were equal bilaterally with no focal neurologic deficits. His neck was supple, and head movements did not cause any symptoms. His musculoskeletal exam was notable for right supraspinatus weakness upon abduction of arm to 90° and a positive impingement sign. ECG showed normal sinus rhythm with normal intervals. Laboratory findings were notable only for an international normalized ratio of 4.9. CT of the head did not show any pathology. The patient was admitted to the medicine floor for further evaluation.

At this point in his clinical course, the patient has had a thorough workup—one that has largely been unrevealing aside from paroxysmal AF. With his current presentation, acute intracranial causes remain on the differential, but the normal CT scan essentially excludes hemorrhage or mass. Although previous MRI studies have been negative and no focal neurologic findings have been described throughout his course, given the patient’s repeated presentations for syncope, intracranial vessel imaging should be obtained to exclude anatomical abnormalities or focal stenosis that could cause recurrent TIAs.

Seizure is also a consideration, but prior EEG and normal neurologic exam makes this less likely. While cardiac workup for syncope has been reassuring, the patient’s history of AF should continue to remain a consideration even though this is less likely the underlying cause since he is now taking dronedarone. He should be placed on telemetry upon admission. While negative orthostatic vital signs make orthostatic syncope less likely, this could be confounded by use of beta-blockers. Overall, the patient’s case remains a challenging one, with the etiology of his syncope remaining unclear at this time.

During this hospitalization, possible etiologies for recurrent syncope and falls were reviewed. The burden of verifiable AF was too low to explain the patient’s recurrent syncopal episodes. Further review of his medical record revealed that a carotid ultrasound was obtained a year earlier in the course of a previous hospitalization. The ultrasound report described patent carotid arteries and retrograde flow in the left vertebral artery consistent with ipsilateral subclavian stenosis. At the time, the ultrasound was interpreted as reassuring based on the lack of significant carotid stenosis; the findings were thought to be unrelated to the patient’s syncopal episodes. On further questioning, the patient noted that minimal exertion such as unloading a few items from the dishwasher caused left arm pain and paresthesia, accompanied by headache and lightheadedness. He also reported using his left arm more frequently following a right shoulder injury. Repeat physical exam found an inter-arm systolic BP difference (IASBPD) >40 mm Hg and left-arm claudication. CT-angiogram of the neck was obtained and showed total occlusion of the left proximal subclavian artery, patent bilateral internal carotid arteries, and retrograde flow in the left vertebral artery (Figure 1).

Subclavian steal syndrome (SSS) results from compromised flow to the distal arm or brainstem circulation due to a proximal subclavian artery occlusion or stenosis (prior to the origin of the vertebral artery).^1,2 Subclavian stenosis may cause lowered pressure in the distal subclavian artery, creating a gradient for blood flow from the contralateral vertebral artery through the basilar artery to the ipsilateral vertebral artery, ultimately supplying blood flow to the affected subclavian artery distal to the occlusion (subclavian steal phenomenon). Flow reversal in the vertebrobasilar system can result in hypoperfusion of the brainstem (ie, vertebrobasilar insufficiency), which can cause a variety of neurologic symptoms, including SSS. While atherosclerosis is the most common cause of subclavian steal, it may be due to other conditions (eg, Takayasu arteritis, thoracic outlet syndrome, congenital heart disease).

Clinically, although many patients with proximal subclavian stenosis are asymptomatic (even in cases wherein angiographic flow reversal is detected), it is critical that clinicians be familiar with common symptoms associated with the diagnosis. Symptoms may include arm claudication related to hypoperfusion of the extremity, particularly when performing activities, as was observed in this patient. Neurologic symptoms are less common but include symptoms consistent with compromised posterior circulation such as dizziness, vertigo, ataxia, diplopia, nystagmus, impaired vision (blurring of vision, hemianopia), otologic symptoms (tinnitus, hearing loss), and/or syncope (ie, “drop attacks”). The patient’s initial complaints of sudden syncope are consistent with this presentation, as are his history of headache and lightheadedness upon use of his left arm.

Diagnostically, a gradient in upper extremity BP >15 mm Hg (as seen in this patient) or findings of arterial insufficiency would suggest subclavian stenosis. Duplex ultrasound is a reliable imaging modality and can demonstrate proximal subclavian stenosis (sensitivity of 90.9% and specificity of 82.5% for predicting >70% of stenosis cases) and ipsilateral vertebral artery flow reversal.¹ Transcranial Doppler studies can be obtained to assess for basilar artery flow reversal as well. CTA/MRA can help delineate location, severity, and cause of stenosis. However, detection of vertebral or basilar artery flow reversal does not always correlate with the development of neurologic symptoms.

For patients with asymptomatic subclavian stenosis, medical management with aspirin, beta-blockers, angiotensin-converting enzyme inhibitors, and statins should be considered given the high likelihood for other atherosclerotic disease. Management of SSS may include percutaneous/surgical intervention in combination with medical therapy, particularly for patients with severe symptomatic disease (arm claudication, posterior circulation deficits, or coronary ischemia in patients with history of coronary bypass utilizing the left internal mammary artery).

The patient was diagnosed with SSS. Cardiovascular medicine and vascular surgery services were asked to evaluate the patient for a revascularization procedure. Because the patient’s anterior circulation was intact, several specialists remained skeptical of SSS as the cause of his syncope. As such, further evaluation for arrhythmia was recommended. The patient’s arm claudication was thought to be due to SSS; however, the well-established retrograde flow via the vertebral artery made a revascularization procedure nonurgent. Moreover, continuation of warfarin was necessary in the setting of his recent left atrial appendage closure and prior history of DVT. It was determined that the risks of discontinuing anticoagulation in order to surgically treat his subclavian stenosis outweighed the benefits. In the meantime, brachial-radial index measurement and a 30-day event monitor were ordered to further assess for arrhythmias. The patient reported being overwhelmed by diagnostic testing without resolution of his syncopal episodes and missed some of his scheduled appointments. One month later, he fell again and sustained vertebral fractures at C1, C4, and L1, and a subsequent SDH requiring craniotomy with a bone flap followed by clot evacuation. The 30-day event monitor report did not reveal any arrhythmias before, during, or after multiple syncopal events that occurred in the period leading up to this fall. The patient later died in a neurology intensive care unit.

SSS often stems from atherosclerotic arterial disease that leads to stenosis or occlusion of the proximal subclavian artery, causing decreased pressure distal to the lesion. The left subclavian artery is affected more often than the right because of its acute angle of origin, which presumably causes turbulence and predisposes to atherosclerosis.³ Compromised blood flow to the arm causes exertional arm claudication and paresthesia. The compensatory retrograde flow in the ipsilateral vertebral artery causes symptoms of vertebrobasilar insufficiency such as dizziness, vertigo, and syncope (Figure 2). This conglomerate of symptoms from subclavian steal, by definition, comprises SSS. The most remarkable signs of SSS are IASBPD >20 mm Hg and, less commonly, reproducible arm claudication.

Diagnosis of SSS requires a careful correlation of clinical history, physical examination, and radiologic findings. Over 80% of patients with subclavian disease have concomitant lesions (eg, in carotid arteries) that can affect collateral circulation.⁴ While symptoms of SSS may vary depending on the adequacy of collaterals, patent anterior circulation does not, by default, prevent SSS in patients with subclavian stenosis.³ In one study, neurologic symptoms were found in 36% of individuals with subclavian stenosis and concomitant carotid atherosclerotic lesions, and in only 5% in patients without carotid lesions.⁵

A key step in diagnosis is measurement of bilateral arm BP as elevated IASBPDs are highly sensitive for subclavian steal. More than 80% of patients with IASBPD >20 mm Hg have evidence of this condition on Doppler ultrasound.³ Higher differentials in BP correlate with occurrence of symptoms (~30% of patients with IASBPD 40-50 mm Hg, and ~40% of those with IASBPD >50 mm Hg).⁶

The severity of subclavian stenosis is traditionally classified by imaging into three separate grades or stages based on the direction of blood flow in vertebral arteries. Grade I involves no retrograde flow; grade II involves cardiac cycle dependent alternating antegrade and retrograde flow; and grade III involves permanent retrograde flow (complete steal).⁷ Our patient’s care was impacted by an unsupported conventional belief that grade II SSS may involve more hemodynamic instability and produce more severe symptoms compared to permanent retrograde flow (grade III), which would result in more stability with a reset of hemodynamics in posterior circulation.⁷ This hypothesis has been disproven in the past, and our patient’s tragic outcome also demonstrates that complete steal is not harmless.⁸ Our patient had permanent retrograde flow in the left vertebral artery, and he suffered classic symptoms of SSS, with devastating consequences. Moreover, increased demand or exertion can enhance the retrograde flow even in grade III stenosis and can precipitate neurologic symptoms of SSS, including syncope. This case provides an important lesson: Management of patients with SSS should depend on the severity of symptoms, not solely on radiologic grading.

Management of SSS is often medical for atherosclerosis and hypertension, especially if symptoms are mild and infrequent. Less than 10% patients with radiologic evidence of subclavian stenosis are symptomatic, and <20% patients with symptomatic SSS require revascularization.³ Percutaneous transluminal angioplasty (PTA) and stenting have become the most favored surgical approach rather than extra-anatomic revascularization techniques.⁷ Both endovascular interventions and open revascularization carry an excellent success rate with low morbidity. Patients undergoing PTA have a combined rate of 3.6% for CVA and death⁹ and a 5-year primary patency rate¹⁰ of 82%. Bypass surgery appears similarly well tolerated, with low perioperative CVA/mortality, and a 10-year primary patency rate of 92%.¹¹ For patients with SSS and coexisting disease in the anterior circulation, carotid endarterectomy is prioritized over subclavian revascularization as repair of the anterior circulation often resolves symptoms of SSS.¹²

In our patient, SSS presented with classic vertebrobasilar and brachial symptoms, but several features of his presentation made the diagnosis a challenge. First, his history suggested several potential causes of syncope, including arrhythmia, orthostatic hypotension, and substance use. Second, he reported arm paresthesia and claudication only when specifically prompted and after a targeted history was obtained. Third, there were no consistent triggers for his syncopal episodes. The patient noted that he lost consciousness when walking, driving, doing light chores, and arising from a seated position. These atypical triggers of syncope were not consistent with any of the illnesses considered during the initial workup, and therefore resulted in a broad differential, delaying the targeted workup for SSS. The wisdom of parsimony may also have played an unintended role: In clinical practice, common things are common, and explanation of most or all symptoms with a known diagnosis is often correct rather than addition of uncommon disorders.

Unfortunately, this patient kept falling through the cracks. Several providers believed that AF and alcohol use were the likely causes of his syncope. This assumption enabled a less than rigorous appraisal of the critical ultrasound report. If SSS had been on the differential, assessing the patient for the associated signs and symptoms might have led to an earlier diagnosis.

• SSS should be included in the differential diagnosis of patients with syncope, especially when common diagnoses have been ruled out.
• Incidentally detected retrograde vertebral flow on ultrasound should never be dismissed, and the patients should be assessed for signs and symptoms of subclavian steal.
• A difference in inter-arm systolic blood pressure >20 mm Hg is highly suggestive of subclavian stenosis.
• SSS has excellent prognosis with appropriate medical treatment or revascularization.

This icon represents the patient’s case. Each paragraph that follows represents the discussant’s thoughts.

A 56-year-old-woman with a history of HIV and locally invasive ductal carcinoma recently treated with mastectomy and adjuvant doxorubicin and cyclophosphamide, now on paclitaxel, was transferred from another hospital with worsening nausea, epigastric pain, and dyspnea. She had been admitted multiple times to both this hospital and another hospital and had extensive workup over the previous 2 months for gastrointestinal (GI) bleeding and progressive dyspnea with orthopnea and paroxysmal nocturnal dyspnea in the setting of a documented 43-lb weight loss.

The patient presents with two clusters of symptoms in the context of chemotherapy regimens following mastectomy for local breast cancer. The first cluster of nausea, epigastric pain, and weight loss indicates upper-GI dysfunction/pathology, such as gastroparesis from paclitaxel-induced autonomic neuropathy given onset of symptoms following paclitaxel chemotherapy. The pretest probability of this is fairly high; however, in light of the GI bleeding, it is prudent to evaluate for other etiologies. The absence of vomiting does not exclude gastroparesis. Other considerations would be peptic ulcer disease or, less likely, the development of a gastric malignancy. A mechanical cause of gastric outlet obstruction from her known cancer is unlikely in this patient with localized breast cancer and recently completed chemotherapy, as breast cancer is unlikely to metastasize to the stomach. Preliminary workup would consist of computed tomography (CT) scan of the abdomen and an esophagogastroduodenoscopy (EGD) to further evaluate these etiologies. The second cluster of dyspnea, paroxysmal nocturnal dyspnea, and orthopnea suggests heart failure, given recent treatment with doxorubicin, and prompts consideration of anthracycline-induced cardiomyopathy. Work-up would include obtaining a transthoracic echocardiogram (TTE) to look for a decrease in her ejection fraction.

Her past medical history was otherwise significant only for the events of the previous few months. Eight months earlier, she was diagnosed with grade 3 triple-negative (estrogen receptor, progesterone receptor, and human epidermal growth factor receptor 2) invasive ductal carcinoma and underwent mastectomy with negative sentinel lymph node biopsy. She completed four cycles of adjuvant doxorubicin and cyclophosphamide and most recently completed cycle three of paclitaxel chemotherapy.

Her HIV disease was controlled with an antiretroviral regimen of dolutegravir/rilpivirine. She had an undetectable viral load for 20 years (CD4, 239 cells/μL 2 weeks prior to transfer).

Her social history included a 1-pack-year smoking history. She denied alcohol or illicit drug use. Family history included pancreatic cancer in her father and endometrial cancer in her paternal grandmother. She was originally from Mexico but moved to Illinois 27 years earlier.

Work-up for her dyspnea was initiated 7 weeks earlier: noncontrast CT of the chest showed extensive diffuse interstitial thickening and ground-glass opacities bilaterally. Bronchoscopy showed no gross abnormalities, and bronchial washings were negative for bacteria, fungi, Pneumocystis jirovecii , acid-fast bacilli, and cancer. She also had a TTE, which showed an ejection fraction of 65% to 70% and was only significant for a pulmonary artery systolic pressure of 45 mm Hg . She was diagnosed with paclitaxel-induced pneumonitis and was discharged home with prednisone 50 mg daily, dapsone, pantoprazole, and 2 L oxygen via nasal cannula.

Two weeks later, she was admitted for coffee-ground emesis and epigastric pain. Her hemoglobin was 5.9 g/dL, for which she was transfused 3 units of packed red blood cells. EGD showed bleeding from diffuse duodenitis, which was treated with argon plasma coagulation. She was also found to have bilateral pulmonary emboli and lower-extremity deep venous thromboses. An inferior vena cava filter was placed, and she was discharged. One week later, she was readmitted with melena, and repeat EGD showed multiple duodenal ulcers with no active bleeding. Colonoscopy was normal. She was continued on prednisone 40 mg daily, as any attempts at tapering the dose resulted in hypotension.

At the time of transfer, she had presented to the outside hospital with worsening nausea and epigastric pain, increasing postprandial abdominal pain, ongoing weight loss, worsening dyspnea on exertion, paroxysmal nocturnal dyspnea, and orthopnea. She denied symptoms of GI bleeding at that time.

Her imaging is consistent with, albeit not specific for, paclitaxel-induced acute pneumonitis. Her persistent dyspnea may be due to worsening of this pneumonitis. Given her GI bleeding and continued epigastric pain, worsening peptic ulcer disease is the likely culprit. The patient has had multiple episodes of bleeding from multiple duodenal ulcers while on proton pump inhibitors, so we should consider reasons for this progression, such as use of nonsteroidal anti-inflammatory drugs (NSAIDs); Helicobacter pylori, herpes simplex virus (HSV), Epstein-Barr virus (EBV), or cytomegalovirus (CMV) infection; Crohn disease, and antral G-cell hyperfunction, as well as lymphoproliferative disease. I would start with histopathology, urease testing, and viral cultures from biopsy specimens at EGD to help eliminate these possibilities for progression of her peptic ulcer disease. Another consideration, albeit rare for progression of peptic ulcer disease and specifically duodenal ulcers, is a gastrinoma-driven acid hypersecretion state, as in the Zollinger-Ellison syndrome. If the above studies do not reveal a cause for her continued peptic ulcer disease, I would check a fasting serum gastrin level and gastric pH to further evaluate this possibility.

Upon arrival on physical exam, her temperature was 35.4° C, heart rate 112 beats per minute, blood pressure 135/96 mm Hg, respiratory rate 34 breaths per minute, and oxygen saturation 97% on room air. She was ill- appearing and in mild respiratory distress with severe muscle wasting. Cervical and supraclavicular lymphadenopathy were not detected. Heart sounds were normal without murmurs. Her jugular venous pressure was approximately 7 cm H₂O. She had no lower-extremity edema. On lung exam, diffuse rhonchi were audible bilaterally with no crackles or wheezing. There was no accessory muscle use. No clubbing was present. Her abdomen was soft and mildly tender in the epigastrium with normal bowel sounds.

Her labs revealed a white blood cell (WBC) count of 5,050/μL (neutrophils, 3,600/μL; lymphocytes, 560/μL; eosinophils, 560/μL; hemoglobin, 8.7 g/dL; mean corpuscular volume, 89.3 fL; and platelets, 402,000/μL). Her CD4 count was 235 cells/μL. Her comprehensive metabolic panel demonstrated a sodium of 127 mmol/L; potassium, 4.0 mmol/L; albumin, 2.0 g/dL; calcium, 8.6 mg/dL; creatinine, 0.41 mg/dL; aspartate aminotransferase (AST), 11 U/L; alanine aminotransferase (ALT), 17 U/L; and serum osmolarity, 258 mOs/kg. Her lipase was 30 U/L, and lactate was 0.8 mmol/L. Urine studies showed creatinine 41 mg/dL, osmolality 503 mOs/kg, and sodium 53 mmol/L.

At this point, the patient has been diagnosed with multiple pulmonary emboli and recurrent GI bleeding from duodenal ulcers with chest imaging suggestive of taxane-induced pulmonary toxicity. She now presents with worsening dyspnea and upper-GI symptoms.

Her dyspnea may represent worsening of her taxane-induced lung disease. However, she may have developed a superimposed infection, heart failure, or further pulmonary emboli. A CT pulmonary angiogram to evaluate for infection, recurrent pulmonary emboli, and worsening of her known taxane-induced pulmonary toxicity should be performed. If the CT demonstrates new areas of consolidation suggestive of a superimposed infection, a bronchoscopy would be warranted to obtain specimens for bacterial, fungal, and viral cultures.

On exam, she is in respiratory distress, almost mildly hypothermic and tachycardic with rhonchi on auscultation. This combination of findings could reflect worsening of her pulmonary disease and/or infection on the background of her cachectic state. Her epigastric tenderness, upper-GI symptoms, and anemia have continued to cause concern for persistent duodenal ulcers. Repeat EGD is warranted to evaluate for ulcers, recent bleed stigmata, and/or complications of ulcers, such as stricture formation, obstruction, or perforation. An abdominal CT scan should be done if the EGD does not show worsening of her peptic ulcer disease or obvious complications. The CT scan is more sensitive for certain complications of peptic ulcer disease, such as perforation or proximal small bowel obstruction due to strictures.

Her anemia could represent ongoing blood loss since her last EGD or an inflammatory state due to infection. Also of concern is her use of dapsone, which can lead to hemolysis with or without glucose-6-phosphate dehydrogenase deficiency (G6PD), and this should be excluded.

She has hypotonic hyponatremia and apparent euvolemia with a high urine sodium and osmolality; this suggests syndrome of inappropriate antidiuretic hormone secretion, which may be due to her ongoing pulmonary disease process.

On day 3 of her hospitalization, her abdominal pain became more diffuse and colicky, with two episodes of associated nonbloody bilious vomiting. During the next 48 hours, her abdominal pain and tenderness worsened diffusely but without rigidity or peritoneal signs. She developed mild abdominal distention. An abdominal X-ray showed moderate to large stool burden and increased bowel dilation concerning for small bowel obstruction. A nasogastric tube was placed, with initial improvement of her abdominal pain and distention. On the morning of day six of hospitalization, she had approximately 100 mL of hematemesis. She immediately became hypotensive to the 50s/20s, and roughly 400 mL of sanguineous fluid was suctioned from her nasogastric tube. She was promptly given intravenous (IV) fluids and 2 units of cross-matched packed red blood cells with normalization of her blood pressure and was transferred to the medical intensive care unit (MICU).

Later that day, she had an EGD that showed copious clots and a severely friable duodenum with duodenal narrowing. Duodenal biopsies were taken.

The duodenal ulcers have led to a complication of stricture formation and obstruction resulting in some degree of small bowel obstruction. EGD with biopsies can shed light on the etiology of these ulcers and can specifically exclude viral, fungal, protozoal, or mycobacterial infection; infiltrative diseases (lymphoma, sarcoidosis, amyloidosis); cancer; and inflammatory noninfectious diseases such as vasculitis/connective tissue disorder. Biopsy specimens should undergo light and electron microscopy (for protozoa-like Cryptosporidium); stains for fungal infections such as histoplasmosis, Candida, and Cryptococcus; and stains for mycobacterium. Immunohistochemistry and polymerase chain reaction (PCR) testing can identify CMV, HIV, HSV, and EBV within the duodenal tissue.

She remained on methylprednisolone 30 mg IV because of her known history of pneumonitis and concern for adrenal insufficiency in the setting of acute illness. Over the next 3 days, she remained normotensive with a stable hemoglobin and had no further episodes of hematemesis. She was transferred to the general medical floor.

One day later, she required an additional unit of cross-matched red blood cells because of a hemoglobin decrease to 6.4 g/dL. The next day, she developed acute-onset respiratory distress and was intubated for hypoxemic respiratory failure and readmitted to the MICU.

Her drop in hemoglobin may reflect ongoing bleeding from the duodenum or may be due to diffuse alveolar hemorrhage (DAH) complicating her pneumonitis. The deterioration in the patient’s respiratory status could represent worsening of her taxane pneumonitis (possibly complicated by DAH or acute respiratory distress syndrome), as fatalities have been reported despite steroid treatment. However, as stated earlier, it is prudent to exclude superimposed pulmonary infection or recurrent pulmonary embolism. Broad-spectrum antibiotics should be provided to cover hospital-acquired pneumonia. Transfusion-related acute lung injury (TRALI) as a cause of her respiratory distress is much less likely given onset after 24 hours from transfusion. Symptoms of TRALI almost always develop within 1 to 2 hours of starting a transfusion, with most starting within minutes. The timing of respiratory distress after 24 hours of transfusion also makes transfusion-associated circulatory overload unlikely, as this presents within 6 to 12 hours of a transfusion being completed and generally in patients receiving large transfusion volumes who have underlying cardiac or renal disease.

Her duodenal pathology revealed Strongyloides stercoralis infection (Figure 1), and she was placed on ivermectin. Steroids were continued due to concern for adrenal insufficiency in the setting of critical illness and later septic shock. Bronchoscopy was also performed, and a specimen grew S stercoralis. She developed septic shock from disseminated S stercoralis infection that required vasopressors. Her sanguineous orogastric output increased, and her abdominal distension worsened, concerning for an intra-abdominal bleed or possible duodenal perforation. As attempts were made to stabilize the patient, ultimately, she experienced cardiac arrest and died.

The patient succumbed to hyperinfection/dissemination of strongyloidiasis. Her risk factors for superinfection included chemotherapy and high-dose steroids, which led to an unchecked autoinfection.

A high index of suspicion remains the most effective overall diagnostic tool for superinfection, which carries a mortality rate of up to 85% even with treatment. Therefore, prevention is the best treatment. Asymptomatic patients with epidemiological exposure or from endemic areas should be evaluated for empiric treatment of S stercoralis prior to initiation of immunosuppressive treatment.

Strongyloides stercoralis is a helminth responsible for one of the most overlooked tropical diseases worldwide.¹ It is estimated that 370 million individuals are infected with S stercoralis globally, and prevalence in the endemic tropics and subtropics is 10% to 40%.^2,3Strongyloides stercoralis infection is characterized by typically nonspecific cutaneous, pulmonary, and GI symptoms, and chronic infection can often be asymptomatic. Once the infection is established, the entirety of the S stercoralis unique life cycle can occur inside the human host, forming a cycle of endogenous autoinfection that can keep the host chronically infected and infectious for decades (Figure 2⁴). While our patient was likely chronically infected for 27 years, cases of patients being infected for up to 75 years have been reported.⁵ Though mostly identified in societies where fecal contamination of soil and poor sanitation are common, S stercoralis should be considered among populations who have traveled to endemic areas and are immunocompromised.

Most chronic S stercoralis infections are asymptomatic, but infection can progress to the life-threatening hyperinfection phase, which has a mortality rate of approximately 85%.⁶ Hyperinfection and disseminated disease occur when there is a rapid proliferation of larvae within the pulmonary and GI tracts, but in the case of disseminated disease, may travel to the liver, brain, and kidneys.^7,8 Typically, this is caused by decreased cellular immunity, often due to preexisting conditions such as human T-cell leukemia virus type 1 (HTLV-1) or medications that allow larvae proliferation to go unchecked.^6,7 One common class of medications known to increase risk of progression to hyperinfection is corticosteroids, which are thought to both depress immunity and directly increase larvae population growth.^6,9 Our patient had been on a prolonged course of steroids for her pulmonary symptoms, with increased doses during her acute illness because of concern for adrenal insufficiency; this likely further contributed to her progression to hyperinfection syndrome. Furthermore, the patient was also immunocompromised from chemotherapy. In addition, she had HIV, which has a controversial association with S stercoralis infection. While previously an AIDS-defining illness, prevalence data indicate a significant co-infection rate between S stercoralis and HIV, but it is unlikely that HIV increases progression to hyperinfection.³

Diagnosing chronic S stercoralis infection is difficult given the lack of a widely accepted gold standard for diagnosis. Traditionally, diagnosis relied on direct visualization of larvae with stool microscopy studies. However, to obtain adequate sensitivity from this method, up to seven serial stool samples must be examined, which is impractical from patient, cost, and efficiency standpoints.¹⁰ While other stool-based techniques, such as enriching the stool sample, stool agar plate culture, or PCR-based stool analysis, improve sensitivity, all stool-based studies are limited by intermittent larvae shedding and low worm burden associated with chronic infection.¹¹ Conversely, serologic studies have higher sensitivity, but concerns exist about lower specificity due to potential cross-reactions with other helminths and the persistence of antibodies even after larvae eradication.^11,12 Patients with suspected S stercoralis infection and pulmonary infiltrates on imaging may have larvae visible on sputum cultures. A final diagnostic method is direct visualization via biopsy during endoscopy or bronchoscopy, which is typically recommended in cases where suspicion is high yet stool studies have been negative.¹³ Our patient’s diagnosis was made by duodenal biopsy after her stool study was negative for S stercoralis.

Deciding who to test is difficult given the nonspecific nature of the symptoms but critically important because of the potential for mortality if the disease progresses to hyperinfection. Diagnosis should be suspected in a patient who has spent time in an endemic area and presents with any combination of pulmonary, dermatologic, or GI symptoms. If suspicion for infection is high in a patient being assessed for solid organ transplant or high-dose steroids, prophylactic treatment with ivermectin should be considered. Given the difficulty in diagnosis, some have suggested using eosinophilia as a key diagnostic element, but this has poor predictive value, particularly if the patient is on corticosteroids.⁷ This patient did not manifest with significant eosinophilia throughout her hospitalization.

This case highlights the difficulties of S stercoralis diagnosis given the nonspecific and variable symptoms, limitations in testing, and potential for remote travel history to endemic regions. It further underscores the need for provider vigilance when starting patients on immunosuppression, even with steroids, given the potential to accelerate chronic infections that were previously buried deep in the mucosa into a lethal hyperinfectious state.

• The cycle of autoinfection by S stercoralis allows it to persist for decades even while asymptomatic. This means patients can present with infection years after travel to endemic regions.
• Because progression to hyperinfection syndrome carries a high mortality rate and is associated with immunosuppressants, particularly corticosteroids, screening patients from or who have spent time in endemic regions for chronic S stercoralis infection is recommended prior to beginning immunosuppression.
• Diagnosing chronic S stercoralis infection is difficult given the lack of a highly accurate, gold-standard test. Therefore, if suspicion for infection is high yet low-sensitivity stool studies have been negative, direct visualization with a biopsy is a diagnostic option.

Acknowledgment

The authors thank Dr Nicholas Moore, microbiologist at Rush University Medical Center, for his assistance in obtaining and preparing the histology images.

This icon represents the patient’s case. Each paragraph that follows represents the discussant’s thoughts.

A 56-year-old-woman with a history of HIV and locally invasive ductal carcinoma recently treated with mastectomy and adjuvant doxorubicin and cyclophosphamide, now on paclitaxel, was transferred from another hospital with worsening nausea, epigastric pain, and dyspnea. She had been admitted multiple times to both this hospital and another hospital and had extensive workup over the previous 2 months for gastrointestinal (GI) bleeding and progressive dyspnea with orthopnea and paroxysmal nocturnal dyspnea in the setting of a documented 43-lb weight loss.

The patient presents with two clusters of symptoms in the context of chemotherapy regimens following mastectomy for local breast cancer. The first cluster of nausea, epigastric pain, and weight loss indicates upper-GI dysfunction/pathology, such as gastroparesis from paclitaxel-induced autonomic neuropathy given onset of symptoms following paclitaxel chemotherapy. The pretest probability of this is fairly high; however, in light of the GI bleeding, it is prudent to evaluate for other etiologies. The absence of vomiting does not exclude gastroparesis. Other considerations would be peptic ulcer disease or, less likely, the development of a gastric malignancy. A mechanical cause of gastric outlet obstruction from her known cancer is unlikely in this patient with localized breast cancer and recently completed chemotherapy, as breast cancer is unlikely to metastasize to the stomach. Preliminary workup would consist of computed tomography (CT) scan of the abdomen and an esophagogastroduodenoscopy (EGD) to further evaluate these etiologies. The second cluster of dyspnea, paroxysmal nocturnal dyspnea, and orthopnea suggests heart failure, given recent treatment with doxorubicin, and prompts consideration of anthracycline-induced cardiomyopathy. Work-up would include obtaining a transthoracic echocardiogram (TTE) to look for a decrease in her ejection fraction.

Her past medical history was otherwise significant only for the events of the previous few months. Eight months earlier, she was diagnosed with grade 3 triple-negative (estrogen receptor, progesterone receptor, and human epidermal growth factor receptor 2) invasive ductal carcinoma and underwent mastectomy with negative sentinel lymph node biopsy. She completed four cycles of adjuvant doxorubicin and cyclophosphamide and most recently completed cycle three of paclitaxel chemotherapy.

Her HIV disease was controlled with an antiretroviral regimen of dolutegravir/rilpivirine. She had an undetectable viral load for 20 years (CD4, 239 cells/μL 2 weeks prior to transfer).

Her social history included a 1-pack-year smoking history. She denied alcohol or illicit drug use. Family history included pancreatic cancer in her father and endometrial cancer in her paternal grandmother. She was originally from Mexico but moved to Illinois 27 years earlier.

Work-up for her dyspnea was initiated 7 weeks earlier: noncontrast CT of the chest showed extensive diffuse interstitial thickening and ground-glass opacities bilaterally. Bronchoscopy showed no gross abnormalities, and bronchial washings were negative for bacteria, fungi, Pneumocystis jirovecii , acid-fast bacilli, and cancer. She also had a TTE, which showed an ejection fraction of 65% to 70% and was only significant for a pulmonary artery systolic pressure of 45 mm Hg . She was diagnosed with paclitaxel-induced pneumonitis and was discharged home with prednisone 50 mg daily, dapsone, pantoprazole, and 2 L oxygen via nasal cannula.

Two weeks later, she was admitted for coffee-ground emesis and epigastric pain. Her hemoglobin was 5.9 g/dL, for which she was transfused 3 units of packed red blood cells. EGD showed bleeding from diffuse duodenitis, which was treated with argon plasma coagulation. She was also found to have bilateral pulmonary emboli and lower-extremity deep venous thromboses. An inferior vena cava filter was placed, and she was discharged. One week later, she was readmitted with melena, and repeat EGD showed multiple duodenal ulcers with no active bleeding. Colonoscopy was normal. She was continued on prednisone 40 mg daily, as any attempts at tapering the dose resulted in hypotension.

At the time of transfer, she had presented to the outside hospital with worsening nausea and epigastric pain, increasing postprandial abdominal pain, ongoing weight loss, worsening dyspnea on exertion, paroxysmal nocturnal dyspnea, and orthopnea. She denied symptoms of GI bleeding at that time.

Her imaging is consistent with, albeit not specific for, paclitaxel-induced acute pneumonitis. Her persistent dyspnea may be due to worsening of this pneumonitis. Given her GI bleeding and continued epigastric pain, worsening peptic ulcer disease is the likely culprit. The patient has had multiple episodes of bleeding from multiple duodenal ulcers while on proton pump inhibitors, so we should consider reasons for this progression, such as use of nonsteroidal anti-inflammatory drugs (NSAIDs); Helicobacter pylori, herpes simplex virus (HSV), Epstein-Barr virus (EBV), or cytomegalovirus (CMV) infection; Crohn disease, and antral G-cell hyperfunction, as well as lymphoproliferative disease. I would start with histopathology, urease testing, and viral cultures from biopsy specimens at EGD to help eliminate these possibilities for progression of her peptic ulcer disease. Another consideration, albeit rare for progression of peptic ulcer disease and specifically duodenal ulcers, is a gastrinoma-driven acid hypersecretion state, as in the Zollinger-Ellison syndrome. If the above studies do not reveal a cause for her continued peptic ulcer disease, I would check a fasting serum gastrin level and gastric pH to further evaluate this possibility.

Upon arrival on physical exam, her temperature was 35.4° C, heart rate 112 beats per minute, blood pressure 135/96 mm Hg, respiratory rate 34 breaths per minute, and oxygen saturation 97% on room air. She was ill- appearing and in mild respiratory distress with severe muscle wasting. Cervical and supraclavicular lymphadenopathy were not detected. Heart sounds were normal without murmurs. Her jugular venous pressure was approximately 7 cm H₂O. She had no lower-extremity edema. On lung exam, diffuse rhonchi were audible bilaterally with no crackles or wheezing. There was no accessory muscle use. No clubbing was present. Her abdomen was soft and mildly tender in the epigastrium with normal bowel sounds.

Her labs revealed a white blood cell (WBC) count of 5,050/μL (neutrophils, 3,600/μL; lymphocytes, 560/μL; eosinophils, 560/μL; hemoglobin, 8.7 g/dL; mean corpuscular volume, 89.3 fL; and platelets, 402,000/μL). Her CD4 count was 235 cells/μL. Her comprehensive metabolic panel demonstrated a sodium of 127 mmol/L; potassium, 4.0 mmol/L; albumin, 2.0 g/dL; calcium, 8.6 mg/dL; creatinine, 0.41 mg/dL; aspartate aminotransferase (AST), 11 U/L; alanine aminotransferase (ALT), 17 U/L; and serum osmolarity, 258 mOs/kg. Her lipase was 30 U/L, and lactate was 0.8 mmol/L. Urine studies showed creatinine 41 mg/dL, osmolality 503 mOs/kg, and sodium 53 mmol/L.

At this point, the patient has been diagnosed with multiple pulmonary emboli and recurrent GI bleeding from duodenal ulcers with chest imaging suggestive of taxane-induced pulmonary toxicity. She now presents with worsening dyspnea and upper-GI symptoms.

Her dyspnea may represent worsening of her taxane-induced lung disease. However, she may have developed a superimposed infection, heart failure, or further pulmonary emboli. A CT pulmonary angiogram to evaluate for infection, recurrent pulmonary emboli, and worsening of her known taxane-induced pulmonary toxicity should be performed. If the CT demonstrates new areas of consolidation suggestive of a superimposed infection, a bronchoscopy would be warranted to obtain specimens for bacterial, fungal, and viral cultures.

On exam, she is in respiratory distress, almost mildly hypothermic and tachycardic with rhonchi on auscultation. This combination of findings could reflect worsening of her pulmonary disease and/or infection on the background of her cachectic state. Her epigastric tenderness, upper-GI symptoms, and anemia have continued to cause concern for persistent duodenal ulcers. Repeat EGD is warranted to evaluate for ulcers, recent bleed stigmata, and/or complications of ulcers, such as stricture formation, obstruction, or perforation. An abdominal CT scan should be done if the EGD does not show worsening of her peptic ulcer disease or obvious complications. The CT scan is more sensitive for certain complications of peptic ulcer disease, such as perforation or proximal small bowel obstruction due to strictures.

Her anemia could represent ongoing blood loss since her last EGD or an inflammatory state due to infection. Also of concern is her use of dapsone, which can lead to hemolysis with or without glucose-6-phosphate dehydrogenase deficiency (G6PD), and this should be excluded.

She has hypotonic hyponatremia and apparent euvolemia with a high urine sodium and osmolality; this suggests syndrome of inappropriate antidiuretic hormone secretion, which may be due to her ongoing pulmonary disease process.

On day 3 of her hospitalization, her abdominal pain became more diffuse and colicky, with two episodes of associated nonbloody bilious vomiting. During the next 48 hours, her abdominal pain and tenderness worsened diffusely but without rigidity or peritoneal signs. She developed mild abdominal distention. An abdominal X-ray showed moderate to large stool burden and increased bowel dilation concerning for small bowel obstruction. A nasogastric tube was placed, with initial improvement of her abdominal pain and distention. On the morning of day six of hospitalization, she had approximately 100 mL of hematemesis. She immediately became hypotensive to the 50s/20s, and roughly 400 mL of sanguineous fluid was suctioned from her nasogastric tube. She was promptly given intravenous (IV) fluids and 2 units of cross-matched packed red blood cells with normalization of her blood pressure and was transferred to the medical intensive care unit (MICU).

Later that day, she had an EGD that showed copious clots and a severely friable duodenum with duodenal narrowing. Duodenal biopsies were taken.

The duodenal ulcers have led to a complication of stricture formation and obstruction resulting in some degree of small bowel obstruction. EGD with biopsies can shed light on the etiology of these ulcers and can specifically exclude viral, fungal, protozoal, or mycobacterial infection; infiltrative diseases (lymphoma, sarcoidosis, amyloidosis); cancer; and inflammatory noninfectious diseases such as vasculitis/connective tissue disorder. Biopsy specimens should undergo light and electron microscopy (for protozoa-like Cryptosporidium); stains for fungal infections such as histoplasmosis, Candida, and Cryptococcus; and stains for mycobacterium. Immunohistochemistry and polymerase chain reaction (PCR) testing can identify CMV, HIV, HSV, and EBV within the duodenal tissue.

She remained on methylprednisolone 30 mg IV because of her known history of pneumonitis and concern for adrenal insufficiency in the setting of acute illness. Over the next 3 days, she remained normotensive with a stable hemoglobin and had no further episodes of hematemesis. She was transferred to the general medical floor.

One day later, she required an additional unit of cross-matched red blood cells because of a hemoglobin decrease to 6.4 g/dL. The next day, she developed acute-onset respiratory distress and was intubated for hypoxemic respiratory failure and readmitted to the MICU.

Her drop in hemoglobin may reflect ongoing bleeding from the duodenum or may be due to diffuse alveolar hemorrhage (DAH) complicating her pneumonitis. The deterioration in the patient’s respiratory status could represent worsening of her taxane pneumonitis (possibly complicated by DAH or acute respiratory distress syndrome), as fatalities have been reported despite steroid treatment. However, as stated earlier, it is prudent to exclude superimposed pulmonary infection or recurrent pulmonary embolism. Broad-spectrum antibiotics should be provided to cover hospital-acquired pneumonia. Transfusion-related acute lung injury (TRALI) as a cause of her respiratory distress is much less likely given onset after 24 hours from transfusion. Symptoms of TRALI almost always develop within 1 to 2 hours of starting a transfusion, with most starting within minutes. The timing of respiratory distress after 24 hours of transfusion also makes transfusion-associated circulatory overload unlikely, as this presents within 6 to 12 hours of a transfusion being completed and generally in patients receiving large transfusion volumes who have underlying cardiac or renal disease.

Her duodenal pathology revealed Strongyloides stercoralis infection (Figure 1), and she was placed on ivermectin. Steroids were continued due to concern for adrenal insufficiency in the setting of critical illness and later septic shock. Bronchoscopy was also performed, and a specimen grew S stercoralis. She developed septic shock from disseminated S stercoralis infection that required vasopressors. Her sanguineous orogastric output increased, and her abdominal distension worsened, concerning for an intra-abdominal bleed or possible duodenal perforation. As attempts were made to stabilize the patient, ultimately, she experienced cardiac arrest and died.

The patient succumbed to hyperinfection/dissemination of strongyloidiasis. Her risk factors for superinfection included chemotherapy and high-dose steroids, which led to an unchecked autoinfection.

A high index of suspicion remains the most effective overall diagnostic tool for superinfection, which carries a mortality rate of up to 85% even with treatment. Therefore, prevention is the best treatment. Asymptomatic patients with epidemiological exposure or from endemic areas should be evaluated for empiric treatment of S stercoralis prior to initiation of immunosuppressive treatment.

Strongyloides stercoralis is a helminth responsible for one of the most overlooked tropical diseases worldwide.¹ It is estimated that 370 million individuals are infected with S stercoralis globally, and prevalence in the endemic tropics and subtropics is 10% to 40%.^2,3Strongyloides stercoralis infection is characterized by typically nonspecific cutaneous, pulmonary, and GI symptoms, and chronic infection can often be asymptomatic. Once the infection is established, the entirety of the S stercoralis unique life cycle can occur inside the human host, forming a cycle of endogenous autoinfection that can keep the host chronically infected and infectious for decades (Figure 2⁴). While our patient was likely chronically infected for 27 years, cases of patients being infected for up to 75 years have been reported.⁵ Though mostly identified in societies where fecal contamination of soil and poor sanitation are common, S stercoralis should be considered among populations who have traveled to endemic areas and are immunocompromised.

Most chronic S stercoralis infections are asymptomatic, but infection can progress to the life-threatening hyperinfection phase, which has a mortality rate of approximately 85%.⁶ Hyperinfection and disseminated disease occur when there is a rapid proliferation of larvae within the pulmonary and GI tracts, but in the case of disseminated disease, may travel to the liver, brain, and kidneys.^7,8 Typically, this is caused by decreased cellular immunity, often due to preexisting conditions such as human T-cell leukemia virus type 1 (HTLV-1) or medications that allow larvae proliferation to go unchecked.^6,7 One common class of medications known to increase risk of progression to hyperinfection is corticosteroids, which are thought to both depress immunity and directly increase larvae population growth.^6,9 Our patient had been on a prolonged course of steroids for her pulmonary symptoms, with increased doses during her acute illness because of concern for adrenal insufficiency; this likely further contributed to her progression to hyperinfection syndrome. Furthermore, the patient was also immunocompromised from chemotherapy. In addition, she had HIV, which has a controversial association with S stercoralis infection. While previously an AIDS-defining illness, prevalence data indicate a significant co-infection rate between S stercoralis and HIV, but it is unlikely that HIV increases progression to hyperinfection.³

Diagnosing chronic S stercoralis infection is difficult given the lack of a widely accepted gold standard for diagnosis. Traditionally, diagnosis relied on direct visualization of larvae with stool microscopy studies. However, to obtain adequate sensitivity from this method, up to seven serial stool samples must be examined, which is impractical from patient, cost, and efficiency standpoints.¹⁰ While other stool-based techniques, such as enriching the stool sample, stool agar plate culture, or PCR-based stool analysis, improve sensitivity, all stool-based studies are limited by intermittent larvae shedding and low worm burden associated with chronic infection.¹¹ Conversely, serologic studies have higher sensitivity, but concerns exist about lower specificity due to potential cross-reactions with other helminths and the persistence of antibodies even after larvae eradication.^11,12 Patients with suspected S stercoralis infection and pulmonary infiltrates on imaging may have larvae visible on sputum cultures. A final diagnostic method is direct visualization via biopsy during endoscopy or bronchoscopy, which is typically recommended in cases where suspicion is high yet stool studies have been negative.¹³ Our patient’s diagnosis was made by duodenal biopsy after her stool study was negative for S stercoralis.

Deciding who to test is difficult given the nonspecific nature of the symptoms but critically important because of the potential for mortality if the disease progresses to hyperinfection. Diagnosis should be suspected in a patient who has spent time in an endemic area and presents with any combination of pulmonary, dermatologic, or GI symptoms. If suspicion for infection is high in a patient being assessed for solid organ transplant or high-dose steroids, prophylactic treatment with ivermectin should be considered. Given the difficulty in diagnosis, some have suggested using eosinophilia as a key diagnostic element, but this has poor predictive value, particularly if the patient is on corticosteroids.⁷ This patient did not manifest with significant eosinophilia throughout her hospitalization.

This case highlights the difficulties of S stercoralis diagnosis given the nonspecific and variable symptoms, limitations in testing, and potential for remote travel history to endemic regions. It further underscores the need for provider vigilance when starting patients on immunosuppression, even with steroids, given the potential to accelerate chronic infections that were previously buried deep in the mucosa into a lethal hyperinfectious state.

• The cycle of autoinfection by S stercoralis allows it to persist for decades even while asymptomatic. This means patients can present with infection years after travel to endemic regions.
• Because progression to hyperinfection syndrome carries a high mortality rate and is associated with immunosuppressants, particularly corticosteroids, screening patients from or who have spent time in endemic regions for chronic S stercoralis infection is recommended prior to beginning immunosuppression.
• Diagnosing chronic S stercoralis infection is difficult given the lack of a highly accurate, gold-standard test. Therefore, if suspicion for infection is high yet low-sensitivity stool studies have been negative, direct visualization with a biopsy is a diagnostic option.

Acknowledgment

The authors thank Dr Nicholas Moore, microbiologist at Rush University Medical Center, for his assistance in obtaining and preparing the histology images.

This icon represents the patient’s case. Each paragraph that follows represents the discussant’s thoughts.

A 56-year-old-woman with a history of HIV and locally invasive ductal carcinoma recently treated with mastectomy and adjuvant doxorubicin and cyclophosphamide, now on paclitaxel, was transferred from another hospital with worsening nausea, epigastric pain, and dyspnea. She had been admitted multiple times to both this hospital and another hospital and had extensive workup over the previous 2 months for gastrointestinal (GI) bleeding and progressive dyspnea with orthopnea and paroxysmal nocturnal dyspnea in the setting of a documented 43-lb weight loss.

The patient presents with two clusters of symptoms in the context of chemotherapy regimens following mastectomy for local breast cancer. The first cluster of nausea, epigastric pain, and weight loss indicates upper-GI dysfunction/pathology, such as gastroparesis from paclitaxel-induced autonomic neuropathy given onset of symptoms following paclitaxel chemotherapy. The pretest probability of this is fairly high; however, in light of the GI bleeding, it is prudent to evaluate for other etiologies. The absence of vomiting does not exclude gastroparesis. Other considerations would be peptic ulcer disease or, less likely, the development of a gastric malignancy. A mechanical cause of gastric outlet obstruction from her known cancer is unlikely in this patient with localized breast cancer and recently completed chemotherapy, as breast cancer is unlikely to metastasize to the stomach. Preliminary workup would consist of computed tomography (CT) scan of the abdomen and an esophagogastroduodenoscopy (EGD) to further evaluate these etiologies. The second cluster of dyspnea, paroxysmal nocturnal dyspnea, and orthopnea suggests heart failure, given recent treatment with doxorubicin, and prompts consideration of anthracycline-induced cardiomyopathy. Work-up would include obtaining a transthoracic echocardiogram (TTE) to look for a decrease in her ejection fraction.

Her past medical history was otherwise significant only for the events of the previous few months. Eight months earlier, she was diagnosed with grade 3 triple-negative (estrogen receptor, progesterone receptor, and human epidermal growth factor receptor 2) invasive ductal carcinoma and underwent mastectomy with negative sentinel lymph node biopsy. She completed four cycles of adjuvant doxorubicin and cyclophosphamide and most recently completed cycle three of paclitaxel chemotherapy.

Her HIV disease was controlled with an antiretroviral regimen of dolutegravir/rilpivirine. She had an undetectable viral load for 20 years (CD4, 239 cells/μL 2 weeks prior to transfer).

Her social history included a 1-pack-year smoking history. She denied alcohol or illicit drug use. Family history included pancreatic cancer in her father and endometrial cancer in her paternal grandmother. She was originally from Mexico but moved to Illinois 27 years earlier.

Work-up for her dyspnea was initiated 7 weeks earlier: noncontrast CT of the chest showed extensive diffuse interstitial thickening and ground-glass opacities bilaterally. Bronchoscopy showed no gross abnormalities, and bronchial washings were negative for bacteria, fungi, Pneumocystis jirovecii , acid-fast bacilli, and cancer. She also had a TTE, which showed an ejection fraction of 65% to 70% and was only significant for a pulmonary artery systolic pressure of 45 mm Hg . She was diagnosed with paclitaxel-induced pneumonitis and was discharged home with prednisone 50 mg daily, dapsone, pantoprazole, and 2 L oxygen via nasal cannula.

Two weeks later, she was admitted for coffee-ground emesis and epigastric pain. Her hemoglobin was 5.9 g/dL, for which she was transfused 3 units of packed red blood cells. EGD showed bleeding from diffuse duodenitis, which was treated with argon plasma coagulation. She was also found to have bilateral pulmonary emboli and lower-extremity deep venous thromboses. An inferior vena cava filter was placed, and she was discharged. One week later, she was readmitted with melena, and repeat EGD showed multiple duodenal ulcers with no active bleeding. Colonoscopy was normal. She was continued on prednisone 40 mg daily, as any attempts at tapering the dose resulted in hypotension.

At the time of transfer, she had presented to the outside hospital with worsening nausea and epigastric pain, increasing postprandial abdominal pain, ongoing weight loss, worsening dyspnea on exertion, paroxysmal nocturnal dyspnea, and orthopnea. She denied symptoms of GI bleeding at that time.

Her imaging is consistent with, albeit not specific for, paclitaxel-induced acute pneumonitis. Her persistent dyspnea may be due to worsening of this pneumonitis. Given her GI bleeding and continued epigastric pain, worsening peptic ulcer disease is the likely culprit. The patient has had multiple episodes of bleeding from multiple duodenal ulcers while on proton pump inhibitors, so we should consider reasons for this progression, such as use of nonsteroidal anti-inflammatory drugs (NSAIDs); Helicobacter pylori, herpes simplex virus (HSV), Epstein-Barr virus (EBV), or cytomegalovirus (CMV) infection; Crohn disease, and antral G-cell hyperfunction, as well as lymphoproliferative disease. I would start with histopathology, urease testing, and viral cultures from biopsy specimens at EGD to help eliminate these possibilities for progression of her peptic ulcer disease. Another consideration, albeit rare for progression of peptic ulcer disease and specifically duodenal ulcers, is a gastrinoma-driven acid hypersecretion state, as in the Zollinger-Ellison syndrome. If the above studies do not reveal a cause for her continued peptic ulcer disease, I would check a fasting serum gastrin level and gastric pH to further evaluate this possibility.

Upon arrival on physical exam, her temperature was 35.4° C, heart rate 112 beats per minute, blood pressure 135/96 mm Hg, respiratory rate 34 breaths per minute, and oxygen saturation 97% on room air. She was ill- appearing and in mild respiratory distress with severe muscle wasting. Cervical and supraclavicular lymphadenopathy were not detected. Heart sounds were normal without murmurs. Her jugular venous pressure was approximately 7 cm H₂O. She had no lower-extremity edema. On lung exam, diffuse rhonchi were audible bilaterally with no crackles or wheezing. There was no accessory muscle use. No clubbing was present. Her abdomen was soft and mildly tender in the epigastrium with normal bowel sounds.

Her labs revealed a white blood cell (WBC) count of 5,050/μL (neutrophils, 3,600/μL; lymphocytes, 560/μL; eosinophils, 560/μL; hemoglobin, 8.7 g/dL; mean corpuscular volume, 89.3 fL; and platelets, 402,000/μL). Her CD4 count was 235 cells/μL. Her comprehensive metabolic panel demonstrated a sodium of 127 mmol/L; potassium, 4.0 mmol/L; albumin, 2.0 g/dL; calcium, 8.6 mg/dL; creatinine, 0.41 mg/dL; aspartate aminotransferase (AST), 11 U/L; alanine aminotransferase (ALT), 17 U/L; and serum osmolarity, 258 mOs/kg. Her lipase was 30 U/L, and lactate was 0.8 mmol/L. Urine studies showed creatinine 41 mg/dL, osmolality 503 mOs/kg, and sodium 53 mmol/L.

At this point, the patient has been diagnosed with multiple pulmonary emboli and recurrent GI bleeding from duodenal ulcers with chest imaging suggestive of taxane-induced pulmonary toxicity. She now presents with worsening dyspnea and upper-GI symptoms.

Her dyspnea may represent worsening of her taxane-induced lung disease. However, she may have developed a superimposed infection, heart failure, or further pulmonary emboli. A CT pulmonary angiogram to evaluate for infection, recurrent pulmonary emboli, and worsening of her known taxane-induced pulmonary toxicity should be performed. If the CT demonstrates new areas of consolidation suggestive of a superimposed infection, a bronchoscopy would be warranted to obtain specimens for bacterial, fungal, and viral cultures.

On exam, she is in respiratory distress, almost mildly hypothermic and tachycardic with rhonchi on auscultation. This combination of findings could reflect worsening of her pulmonary disease and/or infection on the background of her cachectic state. Her epigastric tenderness, upper-GI symptoms, and anemia have continued to cause concern for persistent duodenal ulcers. Repeat EGD is warranted to evaluate for ulcers, recent bleed stigmata, and/or complications of ulcers, such as stricture formation, obstruction, or perforation. An abdominal CT scan should be done if the EGD does not show worsening of her peptic ulcer disease or obvious complications. The CT scan is more sensitive for certain complications of peptic ulcer disease, such as perforation or proximal small bowel obstruction due to strictures.

Her anemia could represent ongoing blood loss since her last EGD or an inflammatory state due to infection. Also of concern is her use of dapsone, which can lead to hemolysis with or without glucose-6-phosphate dehydrogenase deficiency (G6PD), and this should be excluded.

She has hypotonic hyponatremia and apparent euvolemia with a high urine sodium and osmolality; this suggests syndrome of inappropriate antidiuretic hormone secretion, which may be due to her ongoing pulmonary disease process.

On day 3 of her hospitalization, her abdominal pain became more diffuse and colicky, with two episodes of associated nonbloody bilious vomiting. During the next 48 hours, her abdominal pain and tenderness worsened diffusely but without rigidity or peritoneal signs. She developed mild abdominal distention. An abdominal X-ray showed moderate to large stool burden and increased bowel dilation concerning for small bowel obstruction. A nasogastric tube was placed, with initial improvement of her abdominal pain and distention. On the morning of day six of hospitalization, she had approximately 100 mL of hematemesis. She immediately became hypotensive to the 50s/20s, and roughly 400 mL of sanguineous fluid was suctioned from her nasogastric tube. She was promptly given intravenous (IV) fluids and 2 units of cross-matched packed red blood cells with normalization of her blood pressure and was transferred to the medical intensive care unit (MICU).

Later that day, she had an EGD that showed copious clots and a severely friable duodenum with duodenal narrowing. Duodenal biopsies were taken.

The duodenal ulcers have led to a complication of stricture formation and obstruction resulting in some degree of small bowel obstruction. EGD with biopsies can shed light on the etiology of these ulcers and can specifically exclude viral, fungal, protozoal, or mycobacterial infection; infiltrative diseases (lymphoma, sarcoidosis, amyloidosis); cancer; and inflammatory noninfectious diseases such as vasculitis/connective tissue disorder. Biopsy specimens should undergo light and electron microscopy (for protozoa-like Cryptosporidium); stains for fungal infections such as histoplasmosis, Candida, and Cryptococcus; and stains for mycobacterium. Immunohistochemistry and polymerase chain reaction (PCR) testing can identify CMV, HIV, HSV, and EBV within the duodenal tissue.

She remained on methylprednisolone 30 mg IV because of her known history of pneumonitis and concern for adrenal insufficiency in the setting of acute illness. Over the next 3 days, she remained normotensive with a stable hemoglobin and had no further episodes of hematemesis. She was transferred to the general medical floor.

One day later, she required an additional unit of cross-matched red blood cells because of a hemoglobin decrease to 6.4 g/dL. The next day, she developed acute-onset respiratory distress and was intubated for hypoxemic respiratory failure and readmitted to the MICU.

Her drop in hemoglobin may reflect ongoing bleeding from the duodenum or may be due to diffuse alveolar hemorrhage (DAH) complicating her pneumonitis. The deterioration in the patient’s respiratory status could represent worsening of her taxane pneumonitis (possibly complicated by DAH or acute respiratory distress syndrome), as fatalities have been reported despite steroid treatment. However, as stated earlier, it is prudent to exclude superimposed pulmonary infection or recurrent pulmonary embolism. Broad-spectrum antibiotics should be provided to cover hospital-acquired pneumonia. Transfusion-related acute lung injury (TRALI) as a cause of her respiratory distress is much less likely given onset after 24 hours from transfusion. Symptoms of TRALI almost always develop within 1 to 2 hours of starting a transfusion, with most starting within minutes. The timing of respiratory distress after 24 hours of transfusion also makes transfusion-associated circulatory overload unlikely, as this presents within 6 to 12 hours of a transfusion being completed and generally in patients receiving large transfusion volumes who have underlying cardiac or renal disease.

Her duodenal pathology revealed Strongyloides stercoralis infection (Figure 1), and she was placed on ivermectin. Steroids were continued due to concern for adrenal insufficiency in the setting of critical illness and later septic shock. Bronchoscopy was also performed, and a specimen grew S stercoralis. She developed septic shock from disseminated S stercoralis infection that required vasopressors. Her sanguineous orogastric output increased, and her abdominal distension worsened, concerning for an intra-abdominal bleed or possible duodenal perforation. As attempts were made to stabilize the patient, ultimately, she experienced cardiac arrest and died.

The patient succumbed to hyperinfection/dissemination of strongyloidiasis. Her risk factors for superinfection included chemotherapy and high-dose steroids, which led to an unchecked autoinfection.

A high index of suspicion remains the most effective overall diagnostic tool for superinfection, which carries a mortality rate of up to 85% even with treatment. Therefore, prevention is the best treatment. Asymptomatic patients with epidemiological exposure or from endemic areas should be evaluated for empiric treatment of S stercoralis prior to initiation of immunosuppressive treatment.

Strongyloides stercoralis is a helminth responsible for one of the most overlooked tropical diseases worldwide.¹ It is estimated that 370 million individuals are infected with S stercoralis globally, and prevalence in the endemic tropics and subtropics is 10% to 40%.^2,3Strongyloides stercoralis infection is characterized by typically nonspecific cutaneous, pulmonary, and GI symptoms, and chronic infection can often be asymptomatic. Once the infection is established, the entirety of the S stercoralis unique life cycle can occur inside the human host, forming a cycle of endogenous autoinfection that can keep the host chronically infected and infectious for decades (Figure 2⁴). While our patient was likely chronically infected for 27 years, cases of patients being infected for up to 75 years have been reported.⁵ Though mostly identified in societies where fecal contamination of soil and poor sanitation are common, S stercoralis should be considered among populations who have traveled to endemic areas and are immunocompromised.

Most chronic S stercoralis infections are asymptomatic, but infection can progress to the life-threatening hyperinfection phase, which has a mortality rate of approximately 85%.⁶ Hyperinfection and disseminated disease occur when there is a rapid proliferation of larvae within the pulmonary and GI tracts, but in the case of disseminated disease, may travel to the liver, brain, and kidneys.^7,8 Typically, this is caused by decreased cellular immunity, often due to preexisting conditions such as human T-cell leukemia virus type 1 (HTLV-1) or medications that allow larvae proliferation to go unchecked.^6,7 One common class of medications known to increase risk of progression to hyperinfection is corticosteroids, which are thought to both depress immunity and directly increase larvae population growth.^6,9 Our patient had been on a prolonged course of steroids for her pulmonary symptoms, with increased doses during her acute illness because of concern for adrenal insufficiency; this likely further contributed to her progression to hyperinfection syndrome. Furthermore, the patient was also immunocompromised from chemotherapy. In addition, she had HIV, which has a controversial association with S stercoralis infection. While previously an AIDS-defining illness, prevalence data indicate a significant co-infection rate between S stercoralis and HIV, but it is unlikely that HIV increases progression to hyperinfection.³

Diagnosing chronic S stercoralis infection is difficult given the lack of a widely accepted gold standard for diagnosis. Traditionally, diagnosis relied on direct visualization of larvae with stool microscopy studies. However, to obtain adequate sensitivity from this method, up to seven serial stool samples must be examined, which is impractical from patient, cost, and efficiency standpoints.¹⁰ While other stool-based techniques, such as enriching the stool sample, stool agar plate culture, or PCR-based stool analysis, improve sensitivity, all stool-based studies are limited by intermittent larvae shedding and low worm burden associated with chronic infection.¹¹ Conversely, serologic studies have higher sensitivity, but concerns exist about lower specificity due to potential cross-reactions with other helminths and the persistence of antibodies even after larvae eradication.^11,12 Patients with suspected S stercoralis infection and pulmonary infiltrates on imaging may have larvae visible on sputum cultures. A final diagnostic method is direct visualization via biopsy during endoscopy or bronchoscopy, which is typically recommended in cases where suspicion is high yet stool studies have been negative.¹³ Our patient’s diagnosis was made by duodenal biopsy after her stool study was negative for S stercoralis.

Deciding who to test is difficult given the nonspecific nature of the symptoms but critically important because of the potential for mortality if the disease progresses to hyperinfection. Diagnosis should be suspected in a patient who has spent time in an endemic area and presents with any combination of pulmonary, dermatologic, or GI symptoms. If suspicion for infection is high in a patient being assessed for solid organ transplant or high-dose steroids, prophylactic treatment with ivermectin should be considered. Given the difficulty in diagnosis, some have suggested using eosinophilia as a key diagnostic element, but this has poor predictive value, particularly if the patient is on corticosteroids.⁷ This patient did not manifest with significant eosinophilia throughout her hospitalization.

This case highlights the difficulties of S stercoralis diagnosis given the nonspecific and variable symptoms, limitations in testing, and potential for remote travel history to endemic regions. It further underscores the need for provider vigilance when starting patients on immunosuppression, even with steroids, given the potential to accelerate chronic infections that were previously buried deep in the mucosa into a lethal hyperinfectious state.

• The cycle of autoinfection by S stercoralis allows it to persist for decades even while asymptomatic. This means patients can present with infection years after travel to endemic regions.
• Because progression to hyperinfection syndrome carries a high mortality rate and is associated with immunosuppressants, particularly corticosteroids, screening patients from or who have spent time in endemic regions for chronic S stercoralis infection is recommended prior to beginning immunosuppression.
• Diagnosing chronic S stercoralis infection is difficult given the lack of a highly accurate, gold-standard test. Therefore, if suspicion for infection is high yet low-sensitivity stool studies have been negative, direct visualization with a biopsy is a diagnostic option.

Acknowledgment

The authors thank Dr Nicholas Moore, microbiologist at Rush University Medical Center, for his assistance in obtaining and preparing the histology images.

Inspired by the ABIM Foundation’s Choosing Wisely^® campaign, the “Things We Do for No Reason^™” (TWDFNR) series reviews practices that have become common parts of hospital care but may provide little value to our patients. Practices reviewed in the TWDFNR series do not represent clear-cut conclusions or clinical practice standards but are meant as a starting place for research and active discussions among hospitalists and patients. We invite you to be part of that discussion.

A 78-year-old female nursing home resident presents to the emergency department for evaluation of a several-hour history of confusion and restlessness. The patient is accompanied by one of her caregivers from the nursing home. Initial evaluation reveals an awake but inattentive, disoriented, and agitated woman who can answer basic questions appropriately. The caregiver denies the patient having had any antecedent concerns, such as pain with urination, abdominal pain, subjective fevers, chills, or night sweats. Vital signs include a temperature of 37.5 °C (99.5 °F), heart rate of 90 beats per minute, blood pressure of 110/60 mm Hg, respiratory rate of 14 breaths per minute, and oxygen saturation of 98% on room air. The patient has a normal lung and abdominal exam without any suprapubic or flank tenderness. There is no Foley catheter in place.

Delirium, defined by the World Health Organization’s 10th revision of the International Classification of Diseases as “an etiologically nonspecific organic cerebral syndrome characterized by concurrent disturbances of consciousness and attention, perception, thinking, memory, psychomotor behavior, emotion, and the sleep-wake schedule,” is associated with poor clinical outcomes in older patients.^1,2 Mental status changes, which can arise rapidly over the course of hours to days, often fluctuate, with most cases resolving within days of onset.³ In the United States, more than 2.6 million adults aged 65 years and older develop delirium each year, accounting for an estimated $38 to $152 billion in annual healthcare expenditures.⁴

Some clinicians believe that the evaluation for delirium should include an empiric urinary infectious workup with urinalysis and/or urine cultures, even in the absence of local genitourinary symptoms or other signs of infection. In fact, altered mental status is the most common indication for ordering a urine culture in older adult patients.⁵

Urinary tract infections (UTIs) account for almost 25% of all reported infections in older patients, with delirium occurring in up to 30% of this patient population.⁶ As one study demonstrated, given this population’s very high prevalence of asymptomatic bacteriuria (ASB), urine studies sent during a delirium work-up often yield positive findings (defined as ≥10⁵ colony-forming units [CFU]/mL [≥10⁸ CFU/L]) in older patients with no signs or symptoms attributable to UTI.⁷ The incidence of ASB increases significantly with age, with prevalence estimated to be between 6% and 10% in women older than 60 years and approximately 5% in men older than 65 years.⁵ Among older patients residing in long-term care facilities, up to 50% of female residents and up to 40% of male residents have ASB.⁸ These findings, in part, created the common perception of causation between UTI and delirium.

A recent systematic review demonstrated that there is insufficient evidence to associate UTI with acute confusion in older patients.⁹ The Centers for Disease Control and Prevention’s National Health Safety Network notes that at least one of the following criteria must be present for the diagnosis of UTI in noncatheterized patients: fever (>38 °C), suprapubic tenderness, costovertebral angle tenderness, urinary frequency, urinary urgency, or dysuria.¹⁰ Recent studies have identified that ASB—by definition, without dysuria, frequency, bladder discomfort, or fever—is an unlikely cause of delirium.^6,11

The 2019 Infectious Diseases Society of America (IDSA) practice guidelines suggest that clinicians not screen for ASB in older functionally or cognitively impaired patients with no local genitourinary symptoms or other signs of infection. The IDSA acknowledges that the potential adverse outcomes of antimicrobial therapy, including Clostridioides difficile infection, increased antimicrobial resistance, or adverse drug effects, outweigh the potential benefit of treatment given the absence of evidence that such treatment improves outcomes for this vulnerable patient population (strong recommendation, very low-quality evidence).¹² Per the IDSA guidelines, recommendations are strong when there is “moderate- or high-quality evidence that the desirable consequences outweigh the undesirable consequences for a course of action” and “may also be strong when there is high-quality evidence of harm and benefits are uncertain (ie, low or very low quality),” as in this case scenario. Studies of older institutionalized and hospitalized patients have found that ASB often results in inappropriate antimicrobial use with limited benefit.^7,13,14 In addition to noting the lack of benefit from treatment, these studies have found that these patients treated with antimicrobials have worse outcomes when compared to untreated patients with ASB. One study of hospitalized patients treated for ASB concluded that participants given antimicrobial agents experienced longer durations of hospitalization, with no benefits from treatment.¹³ Moreover, another study identified poor long-term functional recovery in patients treated for ASB.¹⁴

Overtreatment also has public health implications given that it may increase the prevalence of multidrug-resistant bacteria in long-term care facilities.¹⁵ One recent study of nursing home residents demonstrated an association between bacteriuria, increased antibiotic use, and subsequent isolation of multidrug-resistant gram-negative organisms.¹⁶ The increased prevalence of these organisms limits options for oral antibiotic therapies in the outpatient setting, potentially leading to increased healthcare utilization and further harms relating to institutionalization in this vulnerable patient population. In light of the ethical concept of nonmaleficence, recognizing the potential harms of treating ASB without clear benefit is important for clinicians to take into account when considering urinalysis in this patient population.

In addition, obtaining a urine culture in an older patient with no signs or symptoms of UTI may lead to premature closure from a diagnostic perspective, resulting in missed diagnoses during clinical evaluation. A missed alternative diagnosis could then cause additional, ongoing harm to the patient if left untreated. Subsequent harms from delayed treatment can thus compound the direct harms and added costs incurred by inappropriate testing and treatment of patients with delirium.

Since 2013, the American Geriatrics Society (AGS) has recommended against the use of antimicrobials in older patients with no urinary tract symptoms, stating that “Antimicrobial treatment studies for asymptomatic bacteriuria in older adults demonstrate no benefits and show increased adverse antimicrobial effects.”¹⁷ The IDSA practice guidelines state the following: “In older patients with functional and/or cognitive impairment with bacteriuria and delirium (acute mental status change, confusion) and without local genitourinary symptoms or other systemic signs of infection (eg, fever or hemodynamic instability), we recommend assessment for other causes and careful observation rather than antimicrobial treatment (strong recommendation, very low-quality evidence).”¹²

Older patients presenting with confusion in the setting of recognized symptoms of UTI (eg, acute dysuria, urinary urgency or frequency) warrant urinalysis and urine culture. Additionally, urinalysis and urine cultures may be warranted to assess for UTI—even in the absence of a localizing source—in older patients with signs and symptoms of delirium who also exhibit systemic signs of infection (eg, fever, leukocytosis, hemodynamic instability).¹²

Initial evaluation of an older patient with delirium should include a thorough review of their recent history and baseline mental status with a knowledgeable informant, a careful physical and neurologic examination, and laboratory studies to determine the presence of electrolyte or metabolic derangements as well as infection and organ failure.⁴ Clinicians should take into account nonmodifiable risk factors for delirium and conduct a careful review of the time course of changes in mental status and modifiable risk factors, including environment, sleep deprivation, medications, immobilization, and sensory impairments.¹⁸

To manage delirium in older patients, clinicians should identify reversible causes of the delirium and minimize modifiable exacerbating factors (eg, sensory impairment, sleep deprivation) in the immediate environment of the patient. They should also carefully review medications that may contribute to delirium, using tools such as the AGS Beers Criteria to identify high-risk medications and concerning medication combinations.¹⁹ Patients who develop local or systemic signs of infection (ie, fevers, chills, dysuria) should undergo appropriate testing, including urinalysis if there is clinical suspicion for urinary etiology.

• For older patients presenting with delirium without localized urinary symptoms or systemic signs of a serious infection, forgo routine ordering of urinalysis and urine culture.
• For older patients presenting with delirium and localized or systemic signs of infection, routine urine studies and antimicrobial therapy may be appropriate.
• For older patients presenting with delirium without localized symptoms or systemic signs of serious infection, attempt to first identify the cause of the change in mental status by obtaining history from a reliable informant, performing a thorough physical and neurologic examination, and evaluating for metabolic and electrolyte derangements.

Returning to the clinical scenario, older patients presenting with signs and symptoms of delirium should undergo further work-up to determine underlying causes for their altered mental status. The patient’s history, ideally obtained from a knowledgeable informant, should offer insight into her baseline mental status and risk factors for delirium. This should be followed by a careful physical and neurologic examination, and evaluation for electrolyte, metabolic, and other derangements. In patients without localized or systemic signs of infection, routine urine testing and treatment of bacteriuria should be avoided.

Do you think this is a low-value practice? Is this truly a “Thing We Do for No Reason^™”? Share what you do in your practice and join in the conversation online by retweeting it on Twitter (#TWDFNR) and liking it on Facebook. We invite you to pro­pose ideas for other “Things We Do for No Reason^™" topics by emailing [email protected]

Inspired by the ABIM Foundation’s Choosing Wisely^® campaign, the “Things We Do for No Reason^™” (TWDFNR) series reviews practices that have become common parts of hospital care but may provide little value to our patients. Practices reviewed in the TWDFNR series do not represent clear-cut conclusions or clinical practice standards but are meant as a starting place for research and active discussions among hospitalists and patients. We invite you to be part of that discussion.

A 78-year-old female nursing home resident presents to the emergency department for evaluation of a several-hour history of confusion and restlessness. The patient is accompanied by one of her caregivers from the nursing home. Initial evaluation reveals an awake but inattentive, disoriented, and agitated woman who can answer basic questions appropriately. The caregiver denies the patient having had any antecedent concerns, such as pain with urination, abdominal pain, subjective fevers, chills, or night sweats. Vital signs include a temperature of 37.5 °C (99.5 °F), heart rate of 90 beats per minute, blood pressure of 110/60 mm Hg, respiratory rate of 14 breaths per minute, and oxygen saturation of 98% on room air. The patient has a normal lung and abdominal exam without any suprapubic or flank tenderness. There is no Foley catheter in place.

Delirium, defined by the World Health Organization’s 10th revision of the International Classification of Diseases as “an etiologically nonspecific organic cerebral syndrome characterized by concurrent disturbances of consciousness and attention, perception, thinking, memory, psychomotor behavior, emotion, and the sleep-wake schedule,” is associated with poor clinical outcomes in older patients.^1,2 Mental status changes, which can arise rapidly over the course of hours to days, often fluctuate, with most cases resolving within days of onset.³ In the United States, more than 2.6 million adults aged 65 years and older develop delirium each year, accounting for an estimated $38 to $152 billion in annual healthcare expenditures.⁴

Some clinicians believe that the evaluation for delirium should include an empiric urinary infectious workup with urinalysis and/or urine cultures, even in the absence of local genitourinary symptoms or other signs of infection. In fact, altered mental status is the most common indication for ordering a urine culture in older adult patients.⁵

Urinary tract infections (UTIs) account for almost 25% of all reported infections in older patients, with delirium occurring in up to 30% of this patient population.⁶ As one study demonstrated, given this population’s very high prevalence of asymptomatic bacteriuria (ASB), urine studies sent during a delirium work-up often yield positive findings (defined as ≥10⁵ colony-forming units [CFU]/mL [≥10⁸ CFU/L]) in older patients with no signs or symptoms attributable to UTI.⁷ The incidence of ASB increases significantly with age, with prevalence estimated to be between 6% and 10% in women older than 60 years and approximately 5% in men older than 65 years.⁵ Among older patients residing in long-term care facilities, up to 50% of female residents and up to 40% of male residents have ASB.⁸ These findings, in part, created the common perception of causation between UTI and delirium.

A recent systematic review demonstrated that there is insufficient evidence to associate UTI with acute confusion in older patients.⁹ The Centers for Disease Control and Prevention’s National Health Safety Network notes that at least one of the following criteria must be present for the diagnosis of UTI in noncatheterized patients: fever (>38 °C), suprapubic tenderness, costovertebral angle tenderness, urinary frequency, urinary urgency, or dysuria.¹⁰ Recent studies have identified that ASB—by definition, without dysuria, frequency, bladder discomfort, or fever—is an unlikely cause of delirium.^6,11

The 2019 Infectious Diseases Society of America (IDSA) practice guidelines suggest that clinicians not screen for ASB in older functionally or cognitively impaired patients with no local genitourinary symptoms or other signs of infection. The IDSA acknowledges that the potential adverse outcomes of antimicrobial therapy, including Clostridioides difficile infection, increased antimicrobial resistance, or adverse drug effects, outweigh the potential benefit of treatment given the absence of evidence that such treatment improves outcomes for this vulnerable patient population (strong recommendation, very low-quality evidence).¹² Per the IDSA guidelines, recommendations are strong when there is “moderate- or high-quality evidence that the desirable consequences outweigh the undesirable consequences for a course of action” and “may also be strong when there is high-quality evidence of harm and benefits are uncertain (ie, low or very low quality),” as in this case scenario. Studies of older institutionalized and hospitalized patients have found that ASB often results in inappropriate antimicrobial use with limited benefit.^7,13,14 In addition to noting the lack of benefit from treatment, these studies have found that these patients treated with antimicrobials have worse outcomes when compared to untreated patients with ASB. One study of hospitalized patients treated for ASB concluded that participants given antimicrobial agents experienced longer durations of hospitalization, with no benefits from treatment.¹³ Moreover, another study identified poor long-term functional recovery in patients treated for ASB.¹⁴

Overtreatment also has public health implications given that it may increase the prevalence of multidrug-resistant bacteria in long-term care facilities.¹⁵ One recent study of nursing home residents demonstrated an association between bacteriuria, increased antibiotic use, and subsequent isolation of multidrug-resistant gram-negative organisms.¹⁶ The increased prevalence of these organisms limits options for oral antibiotic therapies in the outpatient setting, potentially leading to increased healthcare utilization and further harms relating to institutionalization in this vulnerable patient population. In light of the ethical concept of nonmaleficence, recognizing the potential harms of treating ASB without clear benefit is important for clinicians to take into account when considering urinalysis in this patient population.

In addition, obtaining a urine culture in an older patient with no signs or symptoms of UTI may lead to premature closure from a diagnostic perspective, resulting in missed diagnoses during clinical evaluation. A missed alternative diagnosis could then cause additional, ongoing harm to the patient if left untreated. Subsequent harms from delayed treatment can thus compound the direct harms and added costs incurred by inappropriate testing and treatment of patients with delirium.

Since 2013, the American Geriatrics Society (AGS) has recommended against the use of antimicrobials in older patients with no urinary tract symptoms, stating that “Antimicrobial treatment studies for asymptomatic bacteriuria in older adults demonstrate no benefits and show increased adverse antimicrobial effects.”¹⁷ The IDSA practice guidelines state the following: “In older patients with functional and/or cognitive impairment with bacteriuria and delirium (acute mental status change, confusion) and without local genitourinary symptoms or other systemic signs of infection (eg, fever or hemodynamic instability), we recommend assessment for other causes and careful observation rather than antimicrobial treatment (strong recommendation, very low-quality evidence).”¹²

Older patients presenting with confusion in the setting of recognized symptoms of UTI (eg, acute dysuria, urinary urgency or frequency) warrant urinalysis and urine culture. Additionally, urinalysis and urine cultures may be warranted to assess for UTI—even in the absence of a localizing source—in older patients with signs and symptoms of delirium who also exhibit systemic signs of infection (eg, fever, leukocytosis, hemodynamic instability).¹²

Initial evaluation of an older patient with delirium should include a thorough review of their recent history and baseline mental status with a knowledgeable informant, a careful physical and neurologic examination, and laboratory studies to determine the presence of electrolyte or metabolic derangements as well as infection and organ failure.⁴ Clinicians should take into account nonmodifiable risk factors for delirium and conduct a careful review of the time course of changes in mental status and modifiable risk factors, including environment, sleep deprivation, medications, immobilization, and sensory impairments.¹⁸

To manage delirium in older patients, clinicians should identify reversible causes of the delirium and minimize modifiable exacerbating factors (eg, sensory impairment, sleep deprivation) in the immediate environment of the patient. They should also carefully review medications that may contribute to delirium, using tools such as the AGS Beers Criteria to identify high-risk medications and concerning medication combinations.¹⁹ Patients who develop local or systemic signs of infection (ie, fevers, chills, dysuria) should undergo appropriate testing, including urinalysis if there is clinical suspicion for urinary etiology.

• For older patients presenting with delirium without localized urinary symptoms or systemic signs of a serious infection, forgo routine ordering of urinalysis and urine culture.
• For older patients presenting with delirium and localized or systemic signs of infection, routine urine studies and antimicrobial therapy may be appropriate.
• For older patients presenting with delirium without localized symptoms or systemic signs of serious infection, attempt to first identify the cause of the change in mental status by obtaining history from a reliable informant, performing a thorough physical and neurologic examination, and evaluating for metabolic and electrolyte derangements.

Returning to the clinical scenario, older patients presenting with signs and symptoms of delirium should undergo further work-up to determine underlying causes for their altered mental status. The patient’s history, ideally obtained from a knowledgeable informant, should offer insight into her baseline mental status and risk factors for delirium. This should be followed by a careful physical and neurologic examination, and evaluation for electrolyte, metabolic, and other derangements. In patients without localized or systemic signs of infection, routine urine testing and treatment of bacteriuria should be avoided.

Do you think this is a low-value practice? Is this truly a “Thing We Do for No Reason^™”? Share what you do in your practice and join in the conversation online by retweeting it on Twitter (#TWDFNR) and liking it on Facebook. We invite you to pro­pose ideas for other “Things We Do for No Reason^™" topics by emailing [email protected]

Inspired by the ABIM Foundation’s Choosing Wisely^® campaign, the “Things We Do for No Reason^™” (TWDFNR) series reviews practices that have become common parts of hospital care but may provide little value to our patients. Practices reviewed in the TWDFNR series do not represent clear-cut conclusions or clinical practice standards but are meant as a starting place for research and active discussions among hospitalists and patients. We invite you to be part of that discussion.

A 78-year-old female nursing home resident presents to the emergency department for evaluation of a several-hour history of confusion and restlessness. The patient is accompanied by one of her caregivers from the nursing home. Initial evaluation reveals an awake but inattentive, disoriented, and agitated woman who can answer basic questions appropriately. The caregiver denies the patient having had any antecedent concerns, such as pain with urination, abdominal pain, subjective fevers, chills, or night sweats. Vital signs include a temperature of 37.5 °C (99.5 °F), heart rate of 90 beats per minute, blood pressure of 110/60 mm Hg, respiratory rate of 14 breaths per minute, and oxygen saturation of 98% on room air. The patient has a normal lung and abdominal exam without any suprapubic or flank tenderness. There is no Foley catheter in place.

Delirium, defined by the World Health Organization’s 10th revision of the International Classification of Diseases as “an etiologically nonspecific organic cerebral syndrome characterized by concurrent disturbances of consciousness and attention, perception, thinking, memory, psychomotor behavior, emotion, and the sleep-wake schedule,” is associated with poor clinical outcomes in older patients.^1,2 Mental status changes, which can arise rapidly over the course of hours to days, often fluctuate, with most cases resolving within days of onset.³ In the United States, more than 2.6 million adults aged 65 years and older develop delirium each year, accounting for an estimated $38 to $152 billion in annual healthcare expenditures.⁴

Some clinicians believe that the evaluation for delirium should include an empiric urinary infectious workup with urinalysis and/or urine cultures, even in the absence of local genitourinary symptoms or other signs of infection. In fact, altered mental status is the most common indication for ordering a urine culture in older adult patients.⁵

Urinary tract infections (UTIs) account for almost 25% of all reported infections in older patients, with delirium occurring in up to 30% of this patient population.⁶ As one study demonstrated, given this population’s very high prevalence of asymptomatic bacteriuria (ASB), urine studies sent during a delirium work-up often yield positive findings (defined as ≥10⁵ colony-forming units [CFU]/mL [≥10⁸ CFU/L]) in older patients with no signs or symptoms attributable to UTI.⁷ The incidence of ASB increases significantly with age, with prevalence estimated to be between 6% and 10% in women older than 60 years and approximately 5% in men older than 65 years.⁵ Among older patients residing in long-term care facilities, up to 50% of female residents and up to 40% of male residents have ASB.⁸ These findings, in part, created the common perception of causation between UTI and delirium.

A recent systematic review demonstrated that there is insufficient evidence to associate UTI with acute confusion in older patients.⁹ The Centers for Disease Control and Prevention’s National Health Safety Network notes that at least one of the following criteria must be present for the diagnosis of UTI in noncatheterized patients: fever (>38 °C), suprapubic tenderness, costovertebral angle tenderness, urinary frequency, urinary urgency, or dysuria.¹⁰ Recent studies have identified that ASB—by definition, without dysuria, frequency, bladder discomfort, or fever—is an unlikely cause of delirium.^6,11

The 2019 Infectious Diseases Society of America (IDSA) practice guidelines suggest that clinicians not screen for ASB in older functionally or cognitively impaired patients with no local genitourinary symptoms or other signs of infection. The IDSA acknowledges that the potential adverse outcomes of antimicrobial therapy, including Clostridioides difficile infection, increased antimicrobial resistance, or adverse drug effects, outweigh the potential benefit of treatment given the absence of evidence that such treatment improves outcomes for this vulnerable patient population (strong recommendation, very low-quality evidence).¹² Per the IDSA guidelines, recommendations are strong when there is “moderate- or high-quality evidence that the desirable consequences outweigh the undesirable consequences for a course of action” and “may also be strong when there is high-quality evidence of harm and benefits are uncertain (ie, low or very low quality),” as in this case scenario. Studies of older institutionalized and hospitalized patients have found that ASB often results in inappropriate antimicrobial use with limited benefit.^7,13,14 In addition to noting the lack of benefit from treatment, these studies have found that these patients treated with antimicrobials have worse outcomes when compared to untreated patients with ASB. One study of hospitalized patients treated for ASB concluded that participants given antimicrobial agents experienced longer durations of hospitalization, with no benefits from treatment.¹³ Moreover, another study identified poor long-term functional recovery in patients treated for ASB.¹⁴

Overtreatment also has public health implications given that it may increase the prevalence of multidrug-resistant bacteria in long-term care facilities.¹⁵ One recent study of nursing home residents demonstrated an association between bacteriuria, increased antibiotic use, and subsequent isolation of multidrug-resistant gram-negative organisms.¹⁶ The increased prevalence of these organisms limits options for oral antibiotic therapies in the outpatient setting, potentially leading to increased healthcare utilization and further harms relating to institutionalization in this vulnerable patient population. In light of the ethical concept of nonmaleficence, recognizing the potential harms of treating ASB without clear benefit is important for clinicians to take into account when considering urinalysis in this patient population.

In addition, obtaining a urine culture in an older patient with no signs or symptoms of UTI may lead to premature closure from a diagnostic perspective, resulting in missed diagnoses during clinical evaluation. A missed alternative diagnosis could then cause additional, ongoing harm to the patient if left untreated. Subsequent harms from delayed treatment can thus compound the direct harms and added costs incurred by inappropriate testing and treatment of patients with delirium.

Since 2013, the American Geriatrics Society (AGS) has recommended against the use of antimicrobials in older patients with no urinary tract symptoms, stating that “Antimicrobial treatment studies for asymptomatic bacteriuria in older adults demonstrate no benefits and show increased adverse antimicrobial effects.”¹⁷ The IDSA practice guidelines state the following: “In older patients with functional and/or cognitive impairment with bacteriuria and delirium (acute mental status change, confusion) and without local genitourinary symptoms or other systemic signs of infection (eg, fever or hemodynamic instability), we recommend assessment for other causes and careful observation rather than antimicrobial treatment (strong recommendation, very low-quality evidence).”¹²

Older patients presenting with confusion in the setting of recognized symptoms of UTI (eg, acute dysuria, urinary urgency or frequency) warrant urinalysis and urine culture. Additionally, urinalysis and urine cultures may be warranted to assess for UTI—even in the absence of a localizing source—in older patients with signs and symptoms of delirium who also exhibit systemic signs of infection (eg, fever, leukocytosis, hemodynamic instability).¹²

Initial evaluation of an older patient with delirium should include a thorough review of their recent history and baseline mental status with a knowledgeable informant, a careful physical and neurologic examination, and laboratory studies to determine the presence of electrolyte or metabolic derangements as well as infection and organ failure.⁴ Clinicians should take into account nonmodifiable risk factors for delirium and conduct a careful review of the time course of changes in mental status and modifiable risk factors, including environment, sleep deprivation, medications, immobilization, and sensory impairments.¹⁸

To manage delirium in older patients, clinicians should identify reversible causes of the delirium and minimize modifiable exacerbating factors (eg, sensory impairment, sleep deprivation) in the immediate environment of the patient. They should also carefully review medications that may contribute to delirium, using tools such as the AGS Beers Criteria to identify high-risk medications and concerning medication combinations.¹⁹ Patients who develop local or systemic signs of infection (ie, fevers, chills, dysuria) should undergo appropriate testing, including urinalysis if there is clinical suspicion for urinary etiology.

• For older patients presenting with delirium without localized urinary symptoms or systemic signs of a serious infection, forgo routine ordering of urinalysis and urine culture.
• For older patients presenting with delirium and localized or systemic signs of infection, routine urine studies and antimicrobial therapy may be appropriate.
• For older patients presenting with delirium without localized symptoms or systemic signs of serious infection, attempt to first identify the cause of the change in mental status by obtaining history from a reliable informant, performing a thorough physical and neurologic examination, and evaluating for metabolic and electrolyte derangements.

Returning to the clinical scenario, older patients presenting with signs and symptoms of delirium should undergo further work-up to determine underlying causes for their altered mental status. The patient’s history, ideally obtained from a knowledgeable informant, should offer insight into her baseline mental status and risk factors for delirium. This should be followed by a careful physical and neurologic examination, and evaluation for electrolyte, metabolic, and other derangements. In patients without localized or systemic signs of infection, routine urine testing and treatment of bacteriuria should be avoided.

Do you think this is a low-value practice? Is this truly a “Thing We Do for No Reason^™”? Share what you do in your practice and join in the conversation online by retweeting it on Twitter (#TWDFNR) and liking it on Facebook. We invite you to pro­pose ideas for other “Things We Do for No Reason^™" topics by emailing [email protected]

Acute illnesses requiring hospitalization serve as a sentinel event, with many older adults requiring assistance with activities of daily living (ADLs) upon discharge.^1-3 Older adults who are frail experience even higher rates of hospital-associated disability, and rates of recovery to baseline functional status have varied.^4,5 Loss of independence in ADLs has been associated with nursing home (NH) utilization, caregiver burden, and mortality.⁶

To date, studies have characterized functional trajectories before and after hospitalization in older persons for broad medical conditions, noting persistence of disability and incomplete recovery to baseline functional status.⁷ Prior evaluations have also noted the long-term disabling impact of critical conditions such as acute myocardial infarction, stroke, and sepsis,^8,9 but a knowledge gap exists regarding the subsequent functional disability, recovery, and incident NH admission among older persons who are hospitalized for ambulatory care sensitive conditions (ACSCs). Often considered potentially preventable with optimal ambulatory care,^10,11 ACSCs represent acute, chronic, and vaccine-preventable conditions, including urinary tract infection, congestive heart failure, diabetes mellitus, and pneumonia. Investigating the aforementioned patient-centered measures post hospitalization could provide valuable supporting evidence for the continued recognition of ACSC-related hospitalizations in national quality payment programs set forth by the Centers for Medicare & Medicaid Services (CMS).¹² Demonstrating adverse outcomes after ACSC-related hospitalizations may help support interventions that target potentially preventable ACSC-related hospitalizations, such as home-based care or telehealth, with the goal of improving functional outcomes and reducing NH admission in older persons.

To address these gaps, we evaluated ACSC-related hospitalizations among participants of the Precipitating Events Project (PEP), a 19-year longitudinal study of community-living persons who were initially nondisabled in their basic functional activities. In the 6 months following an ACSC-related hospitalization, our objectives were to describe: (1) the 6-month course of postdischarge functional disability, (2) the cumulative monthly probability of functional recovery, and (3) the cumulative monthly probability of incident NH admission.

Study Population

Participants were drawn from the PEP study, an ongoing, prospective, longitudinal study of 754 community-dwelling persons aged 70 years or older.¹³ Potential participants were members of a large health plan in greater New Haven, Connecticut, and were enrolled from March 1998 through October 1999. As previously described,¹⁴ persons were oversampled if they were physically frail, as denoted by a timed score >10 seconds on the rapid gait test. Exclusion criteria included significant cognitive impairment with no available proxy, life expectancy less than 12 months, plans to leave the area, and inability to speak English. Participants were initially required to be nondisabled in four basic activities of daily living (bathing, dressing, walking across a room, and transferring from a chair). Eligibility was determined during a screening telephone interview and was confirmed during an in-home assessment. Of the eligible members, 75.2% agreed to participate in the project, and persons who declined to participate did not significantly differ in age or sex from those who were enrolled. The Yale Human Investigation Committee approved the study protocol, and all participants provided verbal informed consent.

Data Collection

From 1998 to 2017, comprehensive home-based assessments were completed by trained research nurses at baseline and at 18-month intervals over 234 months (except at 126 months), and telephone interviews were completed monthly through June 2018, to obtain information on disability over time. For participants who had significant cognitive impairment or who were unavailable, we interviewed a proxy informant using a rigorous protocol with demonstrated reliability and validity.¹⁴ All incident NH admissions, including both short- and long-term stays, were identified using the CMS Skilled Nursing Facility claims file and Long Term Care Minimum Data Set. Deaths were ascertained by review of obituaries and/or from a proxy informant, with a completion rate of 100%. A total of 688 participants (91.2%) had died after a median follow-up of 108 months, while 43 participants (5.7%) dropped out of the study after a median follow-up of 27 months. Among all participants, data were otherwise available for 99.2% of 85,531 monthly telephone interviews.

Assembly of Analytic Sample

PEP participants were considered for inclusion in the analytic sample if they had a hospitalization with an ACSC as the primary diagnosis on linked Medicare claims data. The complete list of ACSCs was defined using specifications from the Agency for Healthcare Research and Quality,¹⁵ and was assembled using the International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM) classification prior to October 1, 2015, and ICD Tenth Revision, Clinical Modification (ICD-10-CM) classification after October 1, 2015 (Appendix Table 1). Examples of ACSCs include congestive heart failure, dehydration, urinary tract infection, and angina without procedure. As performed previously,^16,17 two ACSCs (low birthweight; asthma in younger adults 18-39 years) were not included in this analysis because they were not based on full adult populations.

ACSC-related hospitalizations were included through December 2017. Participants could contribute more than one ACSC-related hospitalization over the course of the study based on the following criteria: (1) participant did not have a prior non-ACSC-related hospitalization within an 18-month interval; (2) participant did not have a prior ACSC-related hospitalization or treat-and-release emergency department (ED) visit within an 18-month interval (to ensure independence of observations if the participant was still recovering from the prior event and because some of the characteristics within Table 1 are susceptible to change in the setting of an intervening event and, hence, would not accurately reflect the status of the participant prior to ACSC-related hospitalization); (3) participant was not admitted from a NH; (4) participant did not have an in-hospital intensive care unit (ICU) stay (because persons with critical illness are a distinct population with frequent disability and prolonged recovery, as previously described¹⁸), in-hospital death, or death before first follow-up interview (because our aim was to evaluate disability and recovery after the hospitalization⁷).

Assembly of the primary analytic sample is depicted in the Appendix Figure. Of the 814 patients who were identified with ACSC-related hospitalizations, 107 had a prior non-ACSC-related hospitalization and 275 had a prior ACSC-related hospitalization or a treat-and-release ED visit within an 18-month interval. Of the remaining 432 ACSC-related hospitalizations, 181 were excluded: 114 patients were admitted from a NH, 38 had an in-hospital ICU stay, 3 died in the hospital, 11 died before their first follow-up interview, and 15 had withdrawn from the study. The primary analytic sample included the remaining 251 ACSC-related hospitalizations, contributed by 196 participants. Specifically, nine participants contributed three ACSC-related hospitalizations each, 37 participants contributed two hospitalizations each, and the remaining 150 participants contributed one hospitalization each. During the 6-month follow-up period, 40 participants contributing ACSC-related hospitalizations died after a median (interquartile range [IQR]) of 4 (2-5) months, and 1 person refused continued participation.

Comprehensive Assessments

During the comprehensive in-home assessments, data were obtained on demographic characteristics. Age was measured in years at the time of the ACSC-related hospitalization. In addition, we describe factors from the comprehensive assessment immediately prior to the ACSC-related hospitalization, grouped into two additional domains related to disability¹⁹: health-related and cognitive-psychosocial. The health-related factors included nine self-reported, physician-diagnosed chronic conditions and frailty. The cognitive-psychosocial factors included social support, cognitive impairment, and depressive symptoms.

Assessment of Disability

Complete details about the assessment of disability have been previously described.^13,14,19,20 Briefly, disability was assessed during the monthly telephone interviews, and included four basic activities (bathing, dressing, walking across a room, and transferring from a chair), five instrumental activities (shopping, housework, meal preparation, taking medications, and managing finances), and three mobility activities (walking a quarter mile, climbing a flight of stairs, and lifting or carrying 10 lb). Participants were asked, “At the present time, do you need help from another person to [complete the task]?” Disability was operationalized as the need for personal assistance or an inability to perform the task. Participants were also asked about a fourth mobility activity, “Have you driven a car during the past month?” Those who responded no were classified as being disabled in driving.¹⁹

The number of disabilities overall and for each functional domain (basic, instrumental, and mobility) was summed. Possible disability scores ranged from 0 to 13, with a score of 0 indicating complete independence in all of the items, and a score of 13 indicating complete dependence. Worse postdischarge disability was defined as a total disability score (0-13) at the first telephone interview after an ACSC-related hospitalization that was greater than the total disability score from the telephone interview immediately preceding hospitalization.

Outcome Measures

The primary outcome was the number of disabilities in all 13 basic, instrumental, and mobility activities in each of the 6 months following discharge from an ACSC-related hospitalization. To determine whether our findings were consistent across the three functional domains, we also evaluated the number of disabilities in the four basic, five instrumental, and four mobility activities separately. As secondary outcomes, we evaluated: (1) the cumulative probability of recovery within the 6-month follow-up time frame after an ACSC-related hospitalization, with “recovery” defined as return to the participant’s pre-ACSC-related hospitalization total disability score, and (2) the cumulative probability of incident NH admission within the 6 months after an ACSC-related hospitalization. Aligned with CMS and prior literature,^21,22 we defined a short-term NH stay as ≤100 days and a long-term NH stay as >100 days.

Statistical Analysis

Pre-ACSC-related hospitalization characteristics were summarized by means (SDs) and frequencies with proportions. We determined the mean number of disabilities in each of the 6 months following hospital discharge, with the prehospitalization value included as a reference point. We also determined the mean (SD) number of disabilities for the three subscales of disability (basic activities of daily living [BADLs], instrumental activities of daily living [IADLs], and mobility activities). We calculated the cumulative probability of recovery within 6 months of hospital discharge. Finally, we determined the cumulative probability of incident NH admission during the 6 months after hospital discharge.

To test the robustness of our main results, we conducted a sensitivity analysis assessing disability scores of the 150 participants that contributed only one ACSC-related hospitalization. All analyses were performed using Stata, version 16.0, statistical software (StataCorp).

Table 1 shows the characteristics of the 251 ACSC-related hospitalizations immediately prior to hospitalization. Participants’ mean (SD) age was 85.1 (6.0) years, and the mean total disability score was 5.4. The majority were female, non-Hispanic White, frail, and lived alone. As shown in Appendix Table 2, the three most common reasons for ACSC-related hospitalizations were congestive heart failure (n = 69), bacterial pneumonia (n = 53), and dehydration (n = 44).

The Figure shows the disability scores during the 6-month follow-up period for total, basic, instrumental, and mobility activities, in panels A, B, C, and D, respectively. The exact values are provided in Appendix Table 3. After hospitalization, disability scores for total, basic, instrumental, and mobility activities peaked at month 1 and tended to improve modestly over the next 5 months, but remained greater, on average, than pre-hospitalization scores. Of the 40 participants who died within the 6-month follow-up period, 36 (90%) had worse disability scores in their last month of life than in the month prior to their ACSC-related hospitalization.

Table 2 shows the cumulative probability of functional recovery after ACSC-related hospitalizations. Recovery was incomplete, with only 70% (95% CI, 64%-76%) of hospitalizations achieving a return to the pre-hospitalization total disability score within 6 months of hospitalization.

Table 3 shows the cumulative probability of incident NH admission after an ACSC-related hospitalization. Of the 251 ACSC-related hospitalizations, incident NH admission was experienced by 38% (95% CI, 32%-44%) within 1 month and 50% (95% CI, 43%-56%) within 6 months of discharge. Short-term NH stays accounted for 90 (75.6%) of the 119 incident NH admissions within the 6 months after ACSC-related hospitalizations. Sensitivity analyses yielded comparable disability scores, shown in Appendix Table 4.

In this longitudinal study of community-living older persons, we evaluated functional disability, recovery, and incident NH admission within 6 months of hospitalization for an ACSC. Our study has three major findings. First, disability scores for total, basic, instrumental, and mobility activities at months 1 to 6 of follow-up were greater on average than pre-hospitalization scores. Second, functional recovery was not achieved by 3 of 10 participants after an ACSC-related hospitalization. Third, half of them experienced an incident NH admission within 6 months of discharge from an ACSC-related hospitalization, although about three-quarters of these were short-term stays. Our findings provide evidence that older persons experience clinically meaningful adverse patient-reported outcomes after ACSC-related hospitalizations.

Prior research involving ACSCs has focused largely on rates of hospitalization as a measure of access to primary care and the associated factors predictive of ACSC-related hospitalizations,^23-26 and has not addressed subsequent patient-reported outcomes. The findings in this analysis highlight that older persons experience worsening disability immediately after an ACSC-related hospitalization, which persists for prolonged periods and often results in incomplete recovery. Prior research has assessed pre-hospitalization functional status through retrospective recall approaches,² included only older adults discharged with incident disability,³ and examined functional status after all-cause medical illness hospitalizations.⁵ Our prospective analysis extends the literature by reliably capturing pre-hospital disability scores and uniquely assessing the cohort of older persons hospitalized with ACSCs.

Our work is relevant to the continued evaluation of ACSC-related hospitalizations in national quality measurement and payment initiatives among Medicare beneficiaries. In prior evaluations of ACSC-related quality measures, stakeholders have criticized the measures for limited validity due to a lack of evidence linking each utilization outcome to other patient-centered outcomes.^10,27 Our work addresses this gap by demonstrating that ACSC-related hospitalizations are linked to persistent disability, incomplete functional recovery, and incident NH admissions. Given the large body of evidence demonstrating the priority older persons place on these patient-reported outcomes,^28,29 our work should reassure policymakers seeking to transform quality measurement programs into a more patient-oriented enterprise.

Our findings have several clinical practice, research, and policy implications. First, more-effective clinical strategies to minimize the level of care required for acute exacerbations of ACSC-related illnesses may include: (1) substituting home-based care³⁰ and telehealth interventions³¹ for traditional inpatient hospitalization, (2) making in-ED resources (ie, case management services, geriatric-focused advanced practice providers) more accessible for older persons with ACSC-related illnesses, thereby enhancing care transitions and follow-up to avoid potential current and subsequent hospitalizations, and (3) ensuring adequate ambulatory care access to all older persons, as prior work has shown variation in ACSC hospital admission rates dependent on population factors such as high-poverty neighborhoods,¹⁶ insurance status,^16,32 and race/ethnicity.³³

Clinical strategies have been narrow and not holistic for ACSCs; for example, many institutions have focused on pneumonia vaccinations to reduce hospitalizations, but our work supports the need to further evaluate the impact of preventing ACSC-related hospitalizations and their associated disabling consequences. For patients admitted to the hospital, clinical strategies, such as in-hospital or post-hospital mobility and activity programs, have been shown to be protective against hospital-associated disability.^34,35 Furthermore, hospital discharge planning could include preparing older persons for anticipated functional disabilities, associated recoveries, and NH admission after ACSC-related hospitalizations. Risk factors contributing to post-hospitalization functional disability and recovery have been identified,^19,20,36 but future work is needed to: (1) identify target populations (including those most likely to worsen) so that interventions can be offered earlier in the course of care to those who would benefit most, and (2) identify and learn from those who are resilient and have recovered, to better understand factors contributing to their success.

Our study has several strengths. First, the study is unique due to its longitudinal design, with monthly assessments of functional status. Since functional status was assessed prospectively before the ACSC-related hospitalization, we also have avoided any potential concern for recall bias that may be present if assessed after the hospitalization. Additionally, through the use of Medicare claims and the Minimum Data Set, the ascertainment of hospitalizations and NH admissions was likely complete for the studied population.

However, the study has limitations. First, functional measures were based on self-reports rather than objective measurements. Nevertheless, the self-report function is often used to guide coverage determinations in the Medicare program, as it has been shown to be associated with poor health outcomes.³⁷ Second, we are unable to comment on the rate of functional decline or NH admission when an older person was not hospitalized in relation to an ACSC. Future analyses may benefit from using a control group (eg, older adults without an ACSC hospitalization or older adults with a non-ACSC hospitalization). Third, we used strict exclusion criteria to identify a population of older adults without recent hospitalizations to determine the isolated impact of ACSC hospitalization on disability, incident NH admission, and functional recovery. Considering this potential selection bias, our findings are likely conservative estimates of the patient-centered outcomes evaluated. Fourth, participants were not asked about feeding and toileting. However, the incidence of disability in these ADLs is low among nondisabled, community-living older persons, and it is highly uncommon for disability to develop in these ADLs without concurrent disability in the ADLs within this analysis.^14,38

Finally, because our study participants were members of a single health plan in a small urban area and included nondisabled older persons living in the community, our findings may not be generalizable to geriatric patients in other settings. Nonetheless, the demographics of our cohort reflect those of older persons in New Haven County, Connecticut, which are similar to the demographics of the US population, with the exception of race and ethnicity. In addition, the generalizability of our results are strengthened by the study’s high participation rate and minimal attrition.

Within 6 months of ACSC-related hospitalizations, community-living older persons exhibited greater total disability scores than those immediately preceding hospitalization. In the same time frame, 3 of 10 older persons did not achieve functional recovery, and half experienced incident NH admission. These results provide evidence regarding the continued recognition of ACSC-related hospitalizations in federal quality measurement and payment programs and suggests the need for preventive and comprehensive interventions to meaningfully improve longitudinal outcomes.

We thank Denise Shepard, BSN, MBA, Andrea Benjamin, BSN, Barbara Foster, and Amy Shelton, MPH, for assistance with data collection; Geraldine Hawthorne, BS, for assistance with data entry and management; Peter Charpentier, MPH, for design and development of the study database and participant tracking system; and Joanne McGloin, MDiv, MBA, for leadership and advice as the Project Director. Each of these persons were paid employees of Yale School of Medicine during the conduct of this study.

Acute illnesses requiring hospitalization serve as a sentinel event, with many older adults requiring assistance with activities of daily living (ADLs) upon discharge.^1-3 Older adults who are frail experience even higher rates of hospital-associated disability, and rates of recovery to baseline functional status have varied.^4,5 Loss of independence in ADLs has been associated with nursing home (NH) utilization, caregiver burden, and mortality.⁶

To date, studies have characterized functional trajectories before and after hospitalization in older persons for broad medical conditions, noting persistence of disability and incomplete recovery to baseline functional status.⁷ Prior evaluations have also noted the long-term disabling impact of critical conditions such as acute myocardial infarction, stroke, and sepsis,^8,9 but a knowledge gap exists regarding the subsequent functional disability, recovery, and incident NH admission among older persons who are hospitalized for ambulatory care sensitive conditions (ACSCs). Often considered potentially preventable with optimal ambulatory care,^10,11 ACSCs represent acute, chronic, and vaccine-preventable conditions, including urinary tract infection, congestive heart failure, diabetes mellitus, and pneumonia. Investigating the aforementioned patient-centered measures post hospitalization could provide valuable supporting evidence for the continued recognition of ACSC-related hospitalizations in national quality payment programs set forth by the Centers for Medicare & Medicaid Services (CMS).¹² Demonstrating adverse outcomes after ACSC-related hospitalizations may help support interventions that target potentially preventable ACSC-related hospitalizations, such as home-based care or telehealth, with the goal of improving functional outcomes and reducing NH admission in older persons.

To address these gaps, we evaluated ACSC-related hospitalizations among participants of the Precipitating Events Project (PEP), a 19-year longitudinal study of community-living persons who were initially nondisabled in their basic functional activities. In the 6 months following an ACSC-related hospitalization, our objectives were to describe: (1) the 6-month course of postdischarge functional disability, (2) the cumulative monthly probability of functional recovery, and (3) the cumulative monthly probability of incident NH admission.

Study Population

Participants were drawn from the PEP study, an ongoing, prospective, longitudinal study of 754 community-dwelling persons aged 70 years or older.¹³ Potential participants were members of a large health plan in greater New Haven, Connecticut, and were enrolled from March 1998 through October 1999. As previously described,¹⁴ persons were oversampled if they were physically frail, as denoted by a timed score >10 seconds on the rapid gait test. Exclusion criteria included significant cognitive impairment with no available proxy, life expectancy less than 12 months, plans to leave the area, and inability to speak English. Participants were initially required to be nondisabled in four basic activities of daily living (bathing, dressing, walking across a room, and transferring from a chair). Eligibility was determined during a screening telephone interview and was confirmed during an in-home assessment. Of the eligible members, 75.2% agreed to participate in the project, and persons who declined to participate did not significantly differ in age or sex from those who were enrolled. The Yale Human Investigation Committee approved the study protocol, and all participants provided verbal informed consent.

Data Collection

From 1998 to 2017, comprehensive home-based assessments were completed by trained research nurses at baseline and at 18-month intervals over 234 months (except at 126 months), and telephone interviews were completed monthly through June 2018, to obtain information on disability over time. For participants who had significant cognitive impairment or who were unavailable, we interviewed a proxy informant using a rigorous protocol with demonstrated reliability and validity.¹⁴ All incident NH admissions, including both short- and long-term stays, were identified using the CMS Skilled Nursing Facility claims file and Long Term Care Minimum Data Set. Deaths were ascertained by review of obituaries and/or from a proxy informant, with a completion rate of 100%. A total of 688 participants (91.2%) had died after a median follow-up of 108 months, while 43 participants (5.7%) dropped out of the study after a median follow-up of 27 months. Among all participants, data were otherwise available for 99.2% of 85,531 monthly telephone interviews.

Assembly of Analytic Sample

PEP participants were considered for inclusion in the analytic sample if they had a hospitalization with an ACSC as the primary diagnosis on linked Medicare claims data. The complete list of ACSCs was defined using specifications from the Agency for Healthcare Research and Quality,¹⁵ and was assembled using the International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM) classification prior to October 1, 2015, and ICD Tenth Revision, Clinical Modification (ICD-10-CM) classification after October 1, 2015 (Appendix Table 1). Examples of ACSCs include congestive heart failure, dehydration, urinary tract infection, and angina without procedure. As performed previously,^16,17 two ACSCs (low birthweight; asthma in younger adults 18-39 years) were not included in this analysis because they were not based on full adult populations.

ACSC-related hospitalizations were included through December 2017. Participants could contribute more than one ACSC-related hospitalization over the course of the study based on the following criteria: (1) participant did not have a prior non-ACSC-related hospitalization within an 18-month interval; (2) participant did not have a prior ACSC-related hospitalization or treat-and-release emergency department (ED) visit within an 18-month interval (to ensure independence of observations if the participant was still recovering from the prior event and because some of the characteristics within Table 1 are susceptible to change in the setting of an intervening event and, hence, would not accurately reflect the status of the participant prior to ACSC-related hospitalization); (3) participant was not admitted from a NH; (4) participant did not have an in-hospital intensive care unit (ICU) stay (because persons with critical illness are a distinct population with frequent disability and prolonged recovery, as previously described¹⁸), in-hospital death, or death before first follow-up interview (because our aim was to evaluate disability and recovery after the hospitalization⁷).

Assembly of the primary analytic sample is depicted in the Appendix Figure. Of the 814 patients who were identified with ACSC-related hospitalizations, 107 had a prior non-ACSC-related hospitalization and 275 had a prior ACSC-related hospitalization or a treat-and-release ED visit within an 18-month interval. Of the remaining 432 ACSC-related hospitalizations, 181 were excluded: 114 patients were admitted from a NH, 38 had an in-hospital ICU stay, 3 died in the hospital, 11 died before their first follow-up interview, and 15 had withdrawn from the study. The primary analytic sample included the remaining 251 ACSC-related hospitalizations, contributed by 196 participants. Specifically, nine participants contributed three ACSC-related hospitalizations each, 37 participants contributed two hospitalizations each, and the remaining 150 participants contributed one hospitalization each. During the 6-month follow-up period, 40 participants contributing ACSC-related hospitalizations died after a median (interquartile range [IQR]) of 4 (2-5) months, and 1 person refused continued participation.

Comprehensive Assessments

During the comprehensive in-home assessments, data were obtained on demographic characteristics. Age was measured in years at the time of the ACSC-related hospitalization. In addition, we describe factors from the comprehensive assessment immediately prior to the ACSC-related hospitalization, grouped into two additional domains related to disability¹⁹: health-related and cognitive-psychosocial. The health-related factors included nine self-reported, physician-diagnosed chronic conditions and frailty. The cognitive-psychosocial factors included social support, cognitive impairment, and depressive symptoms.

Assessment of Disability

Complete details about the assessment of disability have been previously described.^13,14,19,20 Briefly, disability was assessed during the monthly telephone interviews, and included four basic activities (bathing, dressing, walking across a room, and transferring from a chair), five instrumental activities (shopping, housework, meal preparation, taking medications, and managing finances), and three mobility activities (walking a quarter mile, climbing a flight of stairs, and lifting or carrying 10 lb). Participants were asked, “At the present time, do you need help from another person to [complete the task]?” Disability was operationalized as the need for personal assistance or an inability to perform the task. Participants were also asked about a fourth mobility activity, “Have you driven a car during the past month?” Those who responded no were classified as being disabled in driving.¹⁹

The number of disabilities overall and for each functional domain (basic, instrumental, and mobility) was summed. Possible disability scores ranged from 0 to 13, with a score of 0 indicating complete independence in all of the items, and a score of 13 indicating complete dependence. Worse postdischarge disability was defined as a total disability score (0-13) at the first telephone interview after an ACSC-related hospitalization that was greater than the total disability score from the telephone interview immediately preceding hospitalization.

Outcome Measures

The primary outcome was the number of disabilities in all 13 basic, instrumental, and mobility activities in each of the 6 months following discharge from an ACSC-related hospitalization. To determine whether our findings were consistent across the three functional domains, we also evaluated the number of disabilities in the four basic, five instrumental, and four mobility activities separately. As secondary outcomes, we evaluated: (1) the cumulative probability of recovery within the 6-month follow-up time frame after an ACSC-related hospitalization, with “recovery” defined as return to the participant’s pre-ACSC-related hospitalization total disability score, and (2) the cumulative probability of incident NH admission within the 6 months after an ACSC-related hospitalization. Aligned with CMS and prior literature,^21,22 we defined a short-term NH stay as ≤100 days and a long-term NH stay as >100 days.

Statistical Analysis

Pre-ACSC-related hospitalization characteristics were summarized by means (SDs) and frequencies with proportions. We determined the mean number of disabilities in each of the 6 months following hospital discharge, with the prehospitalization value included as a reference point. We also determined the mean (SD) number of disabilities for the three subscales of disability (basic activities of daily living [BADLs], instrumental activities of daily living [IADLs], and mobility activities). We calculated the cumulative probability of recovery within 6 months of hospital discharge. Finally, we determined the cumulative probability of incident NH admission during the 6 months after hospital discharge.

To test the robustness of our main results, we conducted a sensitivity analysis assessing disability scores of the 150 participants that contributed only one ACSC-related hospitalization. All analyses were performed using Stata, version 16.0, statistical software (StataCorp).

Table 1 shows the characteristics of the 251 ACSC-related hospitalizations immediately prior to hospitalization. Participants’ mean (SD) age was 85.1 (6.0) years, and the mean total disability score was 5.4. The majority were female, non-Hispanic White, frail, and lived alone. As shown in Appendix Table 2, the three most common reasons for ACSC-related hospitalizations were congestive heart failure (n = 69), bacterial pneumonia (n = 53), and dehydration (n = 44).

The Figure shows the disability scores during the 6-month follow-up period for total, basic, instrumental, and mobility activities, in panels A, B, C, and D, respectively. The exact values are provided in Appendix Table 3. After hospitalization, disability scores for total, basic, instrumental, and mobility activities peaked at month 1 and tended to improve modestly over the next 5 months, but remained greater, on average, than pre-hospitalization scores. Of the 40 participants who died within the 6-month follow-up period, 36 (90%) had worse disability scores in their last month of life than in the month prior to their ACSC-related hospitalization.

Table 2 shows the cumulative probability of functional recovery after ACSC-related hospitalizations. Recovery was incomplete, with only 70% (95% CI, 64%-76%) of hospitalizations achieving a return to the pre-hospitalization total disability score within 6 months of hospitalization.

Table 3 shows the cumulative probability of incident NH admission after an ACSC-related hospitalization. Of the 251 ACSC-related hospitalizations, incident NH admission was experienced by 38% (95% CI, 32%-44%) within 1 month and 50% (95% CI, 43%-56%) within 6 months of discharge. Short-term NH stays accounted for 90 (75.6%) of the 119 incident NH admissions within the 6 months after ACSC-related hospitalizations. Sensitivity analyses yielded comparable disability scores, shown in Appendix Table 4.

In this longitudinal study of community-living older persons, we evaluated functional disability, recovery, and incident NH admission within 6 months of hospitalization for an ACSC. Our study has three major findings. First, disability scores for total, basic, instrumental, and mobility activities at months 1 to 6 of follow-up were greater on average than pre-hospitalization scores. Second, functional recovery was not achieved by 3 of 10 participants after an ACSC-related hospitalization. Third, half of them experienced an incident NH admission within 6 months of discharge from an ACSC-related hospitalization, although about three-quarters of these were short-term stays. Our findings provide evidence that older persons experience clinically meaningful adverse patient-reported outcomes after ACSC-related hospitalizations.

Prior research involving ACSCs has focused largely on rates of hospitalization as a measure of access to primary care and the associated factors predictive of ACSC-related hospitalizations,^23-26 and has not addressed subsequent patient-reported outcomes. The findings in this analysis highlight that older persons experience worsening disability immediately after an ACSC-related hospitalization, which persists for prolonged periods and often results in incomplete recovery. Prior research has assessed pre-hospitalization functional status through retrospective recall approaches,² included only older adults discharged with incident disability,³ and examined functional status after all-cause medical illness hospitalizations.⁵ Our prospective analysis extends the literature by reliably capturing pre-hospital disability scores and uniquely assessing the cohort of older persons hospitalized with ACSCs.

Our work is relevant to the continued evaluation of ACSC-related hospitalizations in national quality measurement and payment initiatives among Medicare beneficiaries. In prior evaluations of ACSC-related quality measures, stakeholders have criticized the measures for limited validity due to a lack of evidence linking each utilization outcome to other patient-centered outcomes.^10,27 Our work addresses this gap by demonstrating that ACSC-related hospitalizations are linked to persistent disability, incomplete functional recovery, and incident NH admissions. Given the large body of evidence demonstrating the priority older persons place on these patient-reported outcomes,^28,29 our work should reassure policymakers seeking to transform quality measurement programs into a more patient-oriented enterprise.

Our findings have several clinical practice, research, and policy implications. First, more-effective clinical strategies to minimize the level of care required for acute exacerbations of ACSC-related illnesses may include: (1) substituting home-based care³⁰ and telehealth interventions³¹ for traditional inpatient hospitalization, (2) making in-ED resources (ie, case management services, geriatric-focused advanced practice providers) more accessible for older persons with ACSC-related illnesses, thereby enhancing care transitions and follow-up to avoid potential current and subsequent hospitalizations, and (3) ensuring adequate ambulatory care access to all older persons, as prior work has shown variation in ACSC hospital admission rates dependent on population factors such as high-poverty neighborhoods,¹⁶ insurance status,^16,32 and race/ethnicity.³³

Clinical strategies have been narrow and not holistic for ACSCs; for example, many institutions have focused on pneumonia vaccinations to reduce hospitalizations, but our work supports the need to further evaluate the impact of preventing ACSC-related hospitalizations and their associated disabling consequences. For patients admitted to the hospital, clinical strategies, such as in-hospital or post-hospital mobility and activity programs, have been shown to be protective against hospital-associated disability.^34,35 Furthermore, hospital discharge planning could include preparing older persons for anticipated functional disabilities, associated recoveries, and NH admission after ACSC-related hospitalizations. Risk factors contributing to post-hospitalization functional disability and recovery have been identified,^19,20,36 but future work is needed to: (1) identify target populations (including those most likely to worsen) so that interventions can be offered earlier in the course of care to those who would benefit most, and (2) identify and learn from those who are resilient and have recovered, to better understand factors contributing to their success.

Our study has several strengths. First, the study is unique due to its longitudinal design, with monthly assessments of functional status. Since functional status was assessed prospectively before the ACSC-related hospitalization, we also have avoided any potential concern for recall bias that may be present if assessed after the hospitalization. Additionally, through the use of Medicare claims and the Minimum Data Set, the ascertainment of hospitalizations and NH admissions was likely complete for the studied population.

However, the study has limitations. First, functional measures were based on self-reports rather than objective measurements. Nevertheless, the self-report function is often used to guide coverage determinations in the Medicare program, as it has been shown to be associated with poor health outcomes.³⁷ Second, we are unable to comment on the rate of functional decline or NH admission when an older person was not hospitalized in relation to an ACSC. Future analyses may benefit from using a control group (eg, older adults without an ACSC hospitalization or older adults with a non-ACSC hospitalization). Third, we used strict exclusion criteria to identify a population of older adults without recent hospitalizations to determine the isolated impact of ACSC hospitalization on disability, incident NH admission, and functional recovery. Considering this potential selection bias, our findings are likely conservative estimates of the patient-centered outcomes evaluated. Fourth, participants were not asked about feeding and toileting. However, the incidence of disability in these ADLs is low among nondisabled, community-living older persons, and it is highly uncommon for disability to develop in these ADLs without concurrent disability in the ADLs within this analysis.^14,38

Finally, because our study participants were members of a single health plan in a small urban area and included nondisabled older persons living in the community, our findings may not be generalizable to geriatric patients in other settings. Nonetheless, the demographics of our cohort reflect those of older persons in New Haven County, Connecticut, which are similar to the demographics of the US population, with the exception of race and ethnicity. In addition, the generalizability of our results are strengthened by the study’s high participation rate and minimal attrition.

Within 6 months of ACSC-related hospitalizations, community-living older persons exhibited greater total disability scores than those immediately preceding hospitalization. In the same time frame, 3 of 10 older persons did not achieve functional recovery, and half experienced incident NH admission. These results provide evidence regarding the continued recognition of ACSC-related hospitalizations in federal quality measurement and payment programs and suggests the need for preventive and comprehensive interventions to meaningfully improve longitudinal outcomes.

We thank Denise Shepard, BSN, MBA, Andrea Benjamin, BSN, Barbara Foster, and Amy Shelton, MPH, for assistance with data collection; Geraldine Hawthorne, BS, for assistance with data entry and management; Peter Charpentier, MPH, for design and development of the study database and participant tracking system; and Joanne McGloin, MDiv, MBA, for leadership and advice as the Project Director. Each of these persons were paid employees of Yale School of Medicine during the conduct of this study.

Acute illnesses requiring hospitalization serve as a sentinel event, with many older adults requiring assistance with activities of daily living (ADLs) upon discharge.^1-3 Older adults who are frail experience even higher rates of hospital-associated disability, and rates of recovery to baseline functional status have varied.^4,5 Loss of independence in ADLs has been associated with nursing home (NH) utilization, caregiver burden, and mortality.⁶

To date, studies have characterized functional trajectories before and after hospitalization in older persons for broad medical conditions, noting persistence of disability and incomplete recovery to baseline functional status.⁷ Prior evaluations have also noted the long-term disabling impact of critical conditions such as acute myocardial infarction, stroke, and sepsis,^8,9 but a knowledge gap exists regarding the subsequent functional disability, recovery, and incident NH admission among older persons who are hospitalized for ambulatory care sensitive conditions (ACSCs). Often considered potentially preventable with optimal ambulatory care,^10,11 ACSCs represent acute, chronic, and vaccine-preventable conditions, including urinary tract infection, congestive heart failure, diabetes mellitus, and pneumonia. Investigating the aforementioned patient-centered measures post hospitalization could provide valuable supporting evidence for the continued recognition of ACSC-related hospitalizations in national quality payment programs set forth by the Centers for Medicare & Medicaid Services (CMS).¹² Demonstrating adverse outcomes after ACSC-related hospitalizations may help support interventions that target potentially preventable ACSC-related hospitalizations, such as home-based care or telehealth, with the goal of improving functional outcomes and reducing NH admission in older persons.

To address these gaps, we evaluated ACSC-related hospitalizations among participants of the Precipitating Events Project (PEP), a 19-year longitudinal study of community-living persons who were initially nondisabled in their basic functional activities. In the 6 months following an ACSC-related hospitalization, our objectives were to describe: (1) the 6-month course of postdischarge functional disability, (2) the cumulative monthly probability of functional recovery, and (3) the cumulative monthly probability of incident NH admission.

Study Population

Participants were drawn from the PEP study, an ongoing, prospective, longitudinal study of 754 community-dwelling persons aged 70 years or older.¹³ Potential participants were members of a large health plan in greater New Haven, Connecticut, and were enrolled from March 1998 through October 1999. As previously described,¹⁴ persons were oversampled if they were physically frail, as denoted by a timed score >10 seconds on the rapid gait test. Exclusion criteria included significant cognitive impairment with no available proxy, life expectancy less than 12 months, plans to leave the area, and inability to speak English. Participants were initially required to be nondisabled in four basic activities of daily living (bathing, dressing, walking across a room, and transferring from a chair). Eligibility was determined during a screening telephone interview and was confirmed during an in-home assessment. Of the eligible members, 75.2% agreed to participate in the project, and persons who declined to participate did not significantly differ in age or sex from those who were enrolled. The Yale Human Investigation Committee approved the study protocol, and all participants provided verbal informed consent.

Data Collection

From 1998 to 2017, comprehensive home-based assessments were completed by trained research nurses at baseline and at 18-month intervals over 234 months (except at 126 months), and telephone interviews were completed monthly through June 2018, to obtain information on disability over time. For participants who had significant cognitive impairment or who were unavailable, we interviewed a proxy informant using a rigorous protocol with demonstrated reliability and validity.¹⁴ All incident NH admissions, including both short- and long-term stays, were identified using the CMS Skilled Nursing Facility claims file and Long Term Care Minimum Data Set. Deaths were ascertained by review of obituaries and/or from a proxy informant, with a completion rate of 100%. A total of 688 participants (91.2%) had died after a median follow-up of 108 months, while 43 participants (5.7%) dropped out of the study after a median follow-up of 27 months. Among all participants, data were otherwise available for 99.2% of 85,531 monthly telephone interviews.

Assembly of Analytic Sample

PEP participants were considered for inclusion in the analytic sample if they had a hospitalization with an ACSC as the primary diagnosis on linked Medicare claims data. The complete list of ACSCs was defined using specifications from the Agency for Healthcare Research and Quality,¹⁵ and was assembled using the International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM) classification prior to October 1, 2015, and ICD Tenth Revision, Clinical Modification (ICD-10-CM) classification after October 1, 2015 (Appendix Table 1). Examples of ACSCs include congestive heart failure, dehydration, urinary tract infection, and angina without procedure. As performed previously,^16,17 two ACSCs (low birthweight; asthma in younger adults 18-39 years) were not included in this analysis because they were not based on full adult populations.

ACSC-related hospitalizations were included through December 2017. Participants could contribute more than one ACSC-related hospitalization over the course of the study based on the following criteria: (1) participant did not have a prior non-ACSC-related hospitalization within an 18-month interval; (2) participant did not have a prior ACSC-related hospitalization or treat-and-release emergency department (ED) visit within an 18-month interval (to ensure independence of observations if the participant was still recovering from the prior event and because some of the characteristics within Table 1 are susceptible to change in the setting of an intervening event and, hence, would not accurately reflect the status of the participant prior to ACSC-related hospitalization); (3) participant was not admitted from a NH; (4) participant did not have an in-hospital intensive care unit (ICU) stay (because persons with critical illness are a distinct population with frequent disability and prolonged recovery, as previously described¹⁸), in-hospital death, or death before first follow-up interview (because our aim was to evaluate disability and recovery after the hospitalization⁷).

Assembly of the primary analytic sample is depicted in the Appendix Figure. Of the 814 patients who were identified with ACSC-related hospitalizations, 107 had a prior non-ACSC-related hospitalization and 275 had a prior ACSC-related hospitalization or a treat-and-release ED visit within an 18-month interval. Of the remaining 432 ACSC-related hospitalizations, 181 were excluded: 114 patients were admitted from a NH, 38 had an in-hospital ICU stay, 3 died in the hospital, 11 died before their first follow-up interview, and 15 had withdrawn from the study. The primary analytic sample included the remaining 251 ACSC-related hospitalizations, contributed by 196 participants. Specifically, nine participants contributed three ACSC-related hospitalizations each, 37 participants contributed two hospitalizations each, and the remaining 150 participants contributed one hospitalization each. During the 6-month follow-up period, 40 participants contributing ACSC-related hospitalizations died after a median (interquartile range [IQR]) of 4 (2-5) months, and 1 person refused continued participation.

Comprehensive Assessments

During the comprehensive in-home assessments, data were obtained on demographic characteristics. Age was measured in years at the time of the ACSC-related hospitalization. In addition, we describe factors from the comprehensive assessment immediately prior to the ACSC-related hospitalization, grouped into two additional domains related to disability¹⁹: health-related and cognitive-psychosocial. The health-related factors included nine self-reported, physician-diagnosed chronic conditions and frailty. The cognitive-psychosocial factors included social support, cognitive impairment, and depressive symptoms.

Assessment of Disability

Complete details about the assessment of disability have been previously described.^13,14,19,20 Briefly, disability was assessed during the monthly telephone interviews, and included four basic activities (bathing, dressing, walking across a room, and transferring from a chair), five instrumental activities (shopping, housework, meal preparation, taking medications, and managing finances), and three mobility activities (walking a quarter mile, climbing a flight of stairs, and lifting or carrying 10 lb). Participants were asked, “At the present time, do you need help from another person to [complete the task]?” Disability was operationalized as the need for personal assistance or an inability to perform the task. Participants were also asked about a fourth mobility activity, “Have you driven a car during the past month?” Those who responded no were classified as being disabled in driving.¹⁹

The number of disabilities overall and for each functional domain (basic, instrumental, and mobility) was summed. Possible disability scores ranged from 0 to 13, with a score of 0 indicating complete independence in all of the items, and a score of 13 indicating complete dependence. Worse postdischarge disability was defined as a total disability score (0-13) at the first telephone interview after an ACSC-related hospitalization that was greater than the total disability score from the telephone interview immediately preceding hospitalization.

Outcome Measures

The primary outcome was the number of disabilities in all 13 basic, instrumental, and mobility activities in each of the 6 months following discharge from an ACSC-related hospitalization. To determine whether our findings were consistent across the three functional domains, we also evaluated the number of disabilities in the four basic, five instrumental, and four mobility activities separately. As secondary outcomes, we evaluated: (1) the cumulative probability of recovery within the 6-month follow-up time frame after an ACSC-related hospitalization, with “recovery” defined as return to the participant’s pre-ACSC-related hospitalization total disability score, and (2) the cumulative probability of incident NH admission within the 6 months after an ACSC-related hospitalization. Aligned with CMS and prior literature,^21,22 we defined a short-term NH stay as ≤100 days and a long-term NH stay as >100 days.

Statistical Analysis

Pre-ACSC-related hospitalization characteristics were summarized by means (SDs) and frequencies with proportions. We determined the mean number of disabilities in each of the 6 months following hospital discharge, with the prehospitalization value included as a reference point. We also determined the mean (SD) number of disabilities for the three subscales of disability (basic activities of daily living [BADLs], instrumental activities of daily living [IADLs], and mobility activities). We calculated the cumulative probability of recovery within 6 months of hospital discharge. Finally, we determined the cumulative probability of incident NH admission during the 6 months after hospital discharge.

To test the robustness of our main results, we conducted a sensitivity analysis assessing disability scores of the 150 participants that contributed only one ACSC-related hospitalization. All analyses were performed using Stata, version 16.0, statistical software (StataCorp).

Table 1 shows the characteristics of the 251 ACSC-related hospitalizations immediately prior to hospitalization. Participants’ mean (SD) age was 85.1 (6.0) years, and the mean total disability score was 5.4. The majority were female, non-Hispanic White, frail, and lived alone. As shown in Appendix Table 2, the three most common reasons for ACSC-related hospitalizations were congestive heart failure (n = 69), bacterial pneumonia (n = 53), and dehydration (n = 44).

The Figure shows the disability scores during the 6-month follow-up period for total, basic, instrumental, and mobility activities, in panels A, B, C, and D, respectively. The exact values are provided in Appendix Table 3. After hospitalization, disability scores for total, basic, instrumental, and mobility activities peaked at month 1 and tended to improve modestly over the next 5 months, but remained greater, on average, than pre-hospitalization scores. Of the 40 participants who died within the 6-month follow-up period, 36 (90%) had worse disability scores in their last month of life than in the month prior to their ACSC-related hospitalization.

Table 2 shows the cumulative probability of functional recovery after ACSC-related hospitalizations. Recovery was incomplete, with only 70% (95% CI, 64%-76%) of hospitalizations achieving a return to the pre-hospitalization total disability score within 6 months of hospitalization.

Table 3 shows the cumulative probability of incident NH admission after an ACSC-related hospitalization. Of the 251 ACSC-related hospitalizations, incident NH admission was experienced by 38% (95% CI, 32%-44%) within 1 month and 50% (95% CI, 43%-56%) within 6 months of discharge. Short-term NH stays accounted for 90 (75.6%) of the 119 incident NH admissions within the 6 months after ACSC-related hospitalizations. Sensitivity analyses yielded comparable disability scores, shown in Appendix Table 4.

In this longitudinal study of community-living older persons, we evaluated functional disability, recovery, and incident NH admission within 6 months of hospitalization for an ACSC. Our study has three major findings. First, disability scores for total, basic, instrumental, and mobility activities at months 1 to 6 of follow-up were greater on average than pre-hospitalization scores. Second, functional recovery was not achieved by 3 of 10 participants after an ACSC-related hospitalization. Third, half of them experienced an incident NH admission within 6 months of discharge from an ACSC-related hospitalization, although about three-quarters of these were short-term stays. Our findings provide evidence that older persons experience clinically meaningful adverse patient-reported outcomes after ACSC-related hospitalizations.

Prior research involving ACSCs has focused largely on rates of hospitalization as a measure of access to primary care and the associated factors predictive of ACSC-related hospitalizations,^23-26 and has not addressed subsequent patient-reported outcomes. The findings in this analysis highlight that older persons experience worsening disability immediately after an ACSC-related hospitalization, which persists for prolonged periods and often results in incomplete recovery. Prior research has assessed pre-hospitalization functional status through retrospective recall approaches,² included only older adults discharged with incident disability,³ and examined functional status after all-cause medical illness hospitalizations.⁵ Our prospective analysis extends the literature by reliably capturing pre-hospital disability scores and uniquely assessing the cohort of older persons hospitalized with ACSCs.

Our work is relevant to the continued evaluation of ACSC-related hospitalizations in national quality measurement and payment initiatives among Medicare beneficiaries. In prior evaluations of ACSC-related quality measures, stakeholders have criticized the measures for limited validity due to a lack of evidence linking each utilization outcome to other patient-centered outcomes.^10,27 Our work addresses this gap by demonstrating that ACSC-related hospitalizations are linked to persistent disability, incomplete functional recovery, and incident NH admissions. Given the large body of evidence demonstrating the priority older persons place on these patient-reported outcomes,^28,29 our work should reassure policymakers seeking to transform quality measurement programs into a more patient-oriented enterprise.

Our findings have several clinical practice, research, and policy implications. First, more-effective clinical strategies to minimize the level of care required for acute exacerbations of ACSC-related illnesses may include: (1) substituting home-based care³⁰ and telehealth interventions³¹ for traditional inpatient hospitalization, (2) making in-ED resources (ie, case management services, geriatric-focused advanced practice providers) more accessible for older persons with ACSC-related illnesses, thereby enhancing care transitions and follow-up to avoid potential current and subsequent hospitalizations, and (3) ensuring adequate ambulatory care access to all older persons, as prior work has shown variation in ACSC hospital admission rates dependent on population factors such as high-poverty neighborhoods,¹⁶ insurance status,^16,32 and race/ethnicity.³³

Clinical strategies have been narrow and not holistic for ACSCs; for example, many institutions have focused on pneumonia vaccinations to reduce hospitalizations, but our work supports the need to further evaluate the impact of preventing ACSC-related hospitalizations and their associated disabling consequences. For patients admitted to the hospital, clinical strategies, such as in-hospital or post-hospital mobility and activity programs, have been shown to be protective against hospital-associated disability.^34,35 Furthermore, hospital discharge planning could include preparing older persons for anticipated functional disabilities, associated recoveries, and NH admission after ACSC-related hospitalizations. Risk factors contributing to post-hospitalization functional disability and recovery have been identified,^19,20,36 but future work is needed to: (1) identify target populations (including those most likely to worsen) so that interventions can be offered earlier in the course of care to those who would benefit most, and (2) identify and learn from those who are resilient and have recovered, to better understand factors contributing to their success.

Our study has several strengths. First, the study is unique due to its longitudinal design, with monthly assessments of functional status. Since functional status was assessed prospectively before the ACSC-related hospitalization, we also have avoided any potential concern for recall bias that may be present if assessed after the hospitalization. Additionally, through the use of Medicare claims and the Minimum Data Set, the ascertainment of hospitalizations and NH admissions was likely complete for the studied population.

However, the study has limitations. First, functional measures were based on self-reports rather than objective measurements. Nevertheless, the self-report function is often used to guide coverage determinations in the Medicare program, as it has been shown to be associated with poor health outcomes.³⁷ Second, we are unable to comment on the rate of functional decline or NH admission when an older person was not hospitalized in relation to an ACSC. Future analyses may benefit from using a control group (eg, older adults without an ACSC hospitalization or older adults with a non-ACSC hospitalization). Third, we used strict exclusion criteria to identify a population of older adults without recent hospitalizations to determine the isolated impact of ACSC hospitalization on disability, incident NH admission, and functional recovery. Considering this potential selection bias, our findings are likely conservative estimates of the patient-centered outcomes evaluated. Fourth, participants were not asked about feeding and toileting. However, the incidence of disability in these ADLs is low among nondisabled, community-living older persons, and it is highly uncommon for disability to develop in these ADLs without concurrent disability in the ADLs within this analysis.^14,38

Finally, because our study participants were members of a single health plan in a small urban area and included nondisabled older persons living in the community, our findings may not be generalizable to geriatric patients in other settings. Nonetheless, the demographics of our cohort reflect those of older persons in New Haven County, Connecticut, which are similar to the demographics of the US population, with the exception of race and ethnicity. In addition, the generalizability of our results are strengthened by the study’s high participation rate and minimal attrition.

Within 6 months of ACSC-related hospitalizations, community-living older persons exhibited greater total disability scores than those immediately preceding hospitalization. In the same time frame, 3 of 10 older persons did not achieve functional recovery, and half experienced incident NH admission. These results provide evidence regarding the continued recognition of ACSC-related hospitalizations in federal quality measurement and payment programs and suggests the need for preventive and comprehensive interventions to meaningfully improve longitudinal outcomes.

We thank Denise Shepard, BSN, MBA, Andrea Benjamin, BSN, Barbara Foster, and Amy Shelton, MPH, for assistance with data collection; Geraldine Hawthorne, BS, for assistance with data entry and management; Peter Charpentier, MPH, for design and development of the study database and participant tracking system; and Joanne McGloin, MDiv, MBA, for leadership and advice as the Project Director. Each of these persons were paid employees of Yale School of Medicine during the conduct of this study.

We thank Dr Zorian and colleagues for their editorial¹ addressing our retrospective multicenter cohort study, “Morning Discharges and Patient Length-of-Stay in Inpatient General Internal Medicine.”² Dr Zorian and colleagues raised a question about whether morning discharges were associated with emergency department (ED) boarding times (ie, the time between the decision to admit a patient and their departure from the ED). We also received correspondence from other readers expressing interest in this metric.

We measured the association between morning discharges from general internal medicine (GIM) and ED boarding time using the same methodology and cohort as previously described in our article.² A total of 37 admissions out of 189,781 admissions (<0.1%) did not have an ED boarding time available and were excluded. The mean (SD) boarding time for the remaining cohort (n = 189,744) was 9.63 (11.67) hours. After categorizing days in the study period into quartiles based on the number of morning discharges from GIM, we did not find a strong unadjusted association with ED boarding times (Table). After multivariable adjustment with negative binomial regression models, as previously described,² there was a weak, statistically significant association between the number of morning discharges and ED boarding time (adjusted rate ratio, 0.995; 95% CI, 0.991-1.000), corresponding to 2.4 minutes less in ED boarding time for every additional morning discharge. Ultimately, we agree with Dr Zorian and colleagues that instead of focusing on discharge-before-noon, hospitals should consider patient flow and discharge quality more holistically.

We thank Dr Zorian and colleagues for their editorial¹ addressing our retrospective multicenter cohort study, “Morning Discharges and Patient Length-of-Stay in Inpatient General Internal Medicine.”² Dr Zorian and colleagues raised a question about whether morning discharges were associated with emergency department (ED) boarding times (ie, the time between the decision to admit a patient and their departure from the ED). We also received correspondence from other readers expressing interest in this metric.

We measured the association between morning discharges from general internal medicine (GIM) and ED boarding time using the same methodology and cohort as previously described in our article.² A total of 37 admissions out of 189,781 admissions (<0.1%) did not have an ED boarding time available and were excluded. The mean (SD) boarding time for the remaining cohort (n = 189,744) was 9.63 (11.67) hours. After categorizing days in the study period into quartiles based on the number of morning discharges from GIM, we did not find a strong unadjusted association with ED boarding times (Table). After multivariable adjustment with negative binomial regression models, as previously described,² there was a weak, statistically significant association between the number of morning discharges and ED boarding time (adjusted rate ratio, 0.995; 95% CI, 0.991-1.000), corresponding to 2.4 minutes less in ED boarding time for every additional morning discharge. Ultimately, we agree with Dr Zorian and colleagues that instead of focusing on discharge-before-noon, hospitals should consider patient flow and discharge quality more holistically.

We thank Dr Zorian and colleagues for their editorial¹ addressing our retrospective multicenter cohort study, “Morning Discharges and Patient Length-of-Stay in Inpatient General Internal Medicine.”² Dr Zorian and colleagues raised a question about whether morning discharges were associated with emergency department (ED) boarding times (ie, the time between the decision to admit a patient and their departure from the ED). We also received correspondence from other readers expressing interest in this metric.

We measured the association between morning discharges from general internal medicine (GIM) and ED boarding time using the same methodology and cohort as previously described in our article.² A total of 37 admissions out of 189,781 admissions (<0.1%) did not have an ED boarding time available and were excluded. The mean (SD) boarding time for the remaining cohort (n = 189,744) was 9.63 (11.67) hours. After categorizing days in the study period into quartiles based on the number of morning discharges from GIM, we did not find a strong unadjusted association with ED boarding times (Table). After multivariable adjustment with negative binomial regression models, as previously described,² there was a weak, statistically significant association between the number of morning discharges and ED boarding time (adjusted rate ratio, 0.995; 95% CI, 0.991-1.000), corresponding to 2.4 minutes less in ED boarding time for every additional morning discharge. Ultimately, we agree with Dr Zorian and colleagues that instead of focusing on discharge-before-noon, hospitals should consider patient flow and discharge quality more holistically.