Federal Practitioner

In 1997, Elizabeth McGlynn wrote, “Measuring quality is no simple task.”¹ We are reminded of this seminal Health Affairs article at a very pertinent point—as health care practice progresses, measuring the impact of performance improvement initiatives on clinical care delivery remains integral to monitoring overall effectiveness of quality. Mortality outcomes are a major focus of quality.

Inpatient mortality within the Veterans Health Administration (VHA) was measured as actual number of deaths (unadjusted mortality), and adjusted mortality was calculated using the standardized mortality ratio (SMR). SMR included actual number of deaths during hospitalization or within 1 day of hospital discharge divided by predicted number of deaths using a risk-adjusted formula and was calculated separately for acute level of care (LOC) and the intensive care unit (ICU). Using risk-adjusted SMR, if an observed/expected ratio was > 1.0, there were more inpatient deaths than expected; if < 1.0, fewer inpatient deaths occurred than predicted; and if 1.0, observed number of inpatient deaths was equivalent to expected number of deaths.²

Mortality reduction is a complex area of performance improvement. Health care facilities often focus their efforts on the biggest mortality contributors. According to Dantes and Epstein, sepsis results in about 265,000 deaths annually in the United States.³ Reinhart and colleagues demonstrated that sepsis is a worldwide issue resulting in approximately 30 million cases and 6 million deaths annually.⁴ Furthermore, Kumar and colleagues have noted that when sepsis progresses to septic shock, survival decreases by almost 8% for each hour delay in sepsis identification and treatment.⁵

Improvements in sepsis management have been multifaceted. The Surviving Sepsis Campaign guidelines created sepsis treatment bundles to guide early diagnosis/treatment of sepsis.⁶ In addition to awareness and sepsis care bundles, a plethora of informatics solutions within electronic health record (EHR) systems have demonstrated improved sepsis care.^7-16 Various approaches to early diagnosis and management of sepsis have been collectively referred to as an early warning sepsis system (EWSS).

An EWSS typically contains automated decision support tools that are integrated in the EHR and meant to assist health care professionals with clinical workflow decision-making. Automated decision support tools within the EHR have a variety of functions, such as clinical care reminders and alerts.¹⁷

Sepsis screening tools function as a form of automated decision support and may be incorporated into the EHR to support the EWSS. Although sepsis screening tools vary, they frequently include a combination of data involving vital signs, laboratory values and/or physical examination findings, such as mental status evaluation.The Modified Early Warning Signs (MEWS) + Sepsis Recognition Score (SRS) is one example of a sepsis screening tool.^7,16

At Malcom Randall Veterans Affairs Medical Center (MRVAMC) in Gainesville, Florida, we identified a quality improvement project opportunity to improve sepsis care in the emergency department (ED) and inpatient wards using the VHA EHR system, the Computerized Patient Record System (CPRS), which is supported by the Veterans Information Systems and Technology Architecture (VistA).¹⁸ A VistA/CPRS EWSS was developed using Lean Six Sigma DMAIC (define, measure, analyze, improve, and control) methodology.¹⁹ During the improve stage, informatics solutions were applied and included a combination of EHR interventions, such as template design, an order set, and clinical reminders. Clinical reminders have a wide variety of use, such as reminders for clinical tasks and as automated decision support within clinical workflows using Boolean logic.

To the best of our knowledge, there has been no published application of an EWSS within VistA/CPRS. In this study, we outline the strategic development of an EWSS in VistA/CPRS that assisted clinical staff with identification and treatment of sepsis; improved documentation of sepsis when present; and associated with improvement in unadjusted and adjusted inpatient mortality.

Methods 

According to policy activities that constitute research at MRVAMC, no institutional review board approval was required as this work met criteria for operational improvement activities exempt from ethics review.

The North Florida/South Georgia Veterans Health System (NF/SGVHS) includes MRVAMC, a large academic hospital with rotating residents/fellows and multiple specialty care services. MRVAMC comprised 144 beds on the medicine/surgery wards; 48 beds in the psychiatry unit; 18 intermediate LOC beds; and 27 ICU beds. The MRVAMC SMR was identified as an improvement opportunity during fiscal year (FY) 2017 (Table 1). Its adjusted mortality for acute LOC demonstrated an observed/expected ratio of > 1.0 suggesting more inpatient deaths were observed than expected. The number of deaths (unadjusted mortality) on acute LOC at MRVAMC was noted to be rising during the first 3 quarters of FY 2017. A deeper examination of data by Pyramid Analytics (www.pyramidanalytics.com) discovered that sepsis was the primary driver for inpatient mortality on acute LOC at MRVAMC. Our goal was to reduce inpatient sepsis-related mortality via development of an EWSS that leveraged VistA/CPRS to improve early identification and treatment of sepsis in the ED and inpatient wards.

Emergency Department

Given the importance of recognizing sepsis early, the sepsis team focused on improvement opportunities at the initial point of patient contact: ED triage. The goal was to incorporate automated VistA/CPRS decision support to assist clinicians with identifying sepsis in triage using MEWS, which was chosen to optimize immediate hospital-wide buy-in. Clinical staff were already familiar with MEWS, which was in use on the inpatient wards.

Flow through the ED and availability of resources differed from the wards. Hence, modification to MEWS on the wards was necessary to fit clinical workflow in the ED. Temperature, heart rate (HR), respiratory rate (RR), systolic blood pressure (SBP), mental status, and white blood cell count (WBC) factored into a MEWS + SRS score on the wards (Table 2). For the ED, MEWS included temperature, HR, RR and SBP, but excluded mental status and WBC. Mental status assessment was excluded due to technical infeasibility (while vital signs could be automatically calculated in real time for a MEWS score, that was not possible for mental status changes). WBC was excluded from the ED as laboratory test results would not be available in triage.

MEWS + SRS scores were calculated in VistA by using clinical reminders. Clinical reminder logic included a series of conditional statements based on various combinations of MEWS + SRS clinical data entered in the EHR. When ED triage vital signs data were entered in CPRS, clinical data were stored and processed according to clinical reminder logic in VistA and displayed to the user in CPRS. While MEWS of ≥ 5 triggered a sepsis alert on the wards, the ≥ 4 threshold was used in the ED given mental status and WBC were excluded from calculations in triage (eAppendix 1 available at doi:10.12788/fp.0194).

Once a sepsis alert was triggered in triage for MEWS ≥ 4, ED nursing staff prioritized bed location and expedited staffing with an ED attending physician for early assessment. The ED attending then performed an assessment to confirm whether sepsis was present and direct early treatment. Although every patient who triggered a sepsis alert in triage did not meet clinical findings of sepsis, patients with MEWS ≥ 4 were frequently ill and required timely intervention.

If an ED attending physician agreed with a sepsis diagnosis, the physician had access to a sepsis workup and treatment order set in CPRS (eAppendix 2 available at doi:10.12788/fp.0194). The sepsis order set incorporated recommendations from the Surviving Sepsis Campaign guidelines and included orders for 2 large-bore peripheral IV lines; aggressive fluid resuscitation (30 mL/kg) for patients with clinical findings of hypoperfusion; broad-spectrum antibiotics; and frequent ordering of laboratory tests and imaging during initial sepsis workup.⁶ Vancomycin and cefepime were selected as routine broad-spectrum antibiotics in the order set when sepsis was suspected based on local antimicrobial stewardship and safety-efficacy profiles. For example, Luther and colleagues demonstrated that cefepime has lower rates of acute kidney injury when combined with vancomycin vs vancomycin + piperacillin-tazobactam.²⁰ If a β-lactam antibiotic could not be used due to a patient’s drug allergy history, aztreonam was available as an alternative option.

The design of the order set also functioned as a communication interface with clinical pharmacists. Given the large volume of antibiotics ordered in the ED, it was difficult for pharmacists to prioritize antibiotic order verification. While stat orders convey high priority, they often lack specificity. When antibiotic orders were selected from the sepsis order set, comments were already included that stated: “STAT. First dose for sepsis protocol” (eAppendix 3 available at doi:10.12788/fp.0194). This standardized communication conveyed a sense of urgency and a collective understanding that patients with suspected sepsis required timely order verification and administration of antibiotics.

Hospital Ward

Mental status and WBC were included on the wards to monitor for possible signs of sepsis, using MEWS + SRS, which was routinely monitored by nursing every 4 to 8 hours. When MEWS + SRS was ≥ 5 points, ward nursing staff called a sepsis alert.^7,16 Early response team (ERT) members received telephone notifications of the alert. ERT staff proceeded with immediate evaluation and treatment at the bedside along with determination for most appropriate LOC. The ERT members included an ICU physician and nurse; respiratory therapist; and nursing supervisor/bed flow coordinator. During bedside evaluation, if the ERT or primary team agreed with a sepsis diagnosis, the ERT or primary team used the sepsis order set to ensure standardized procedures. Stat orders generated through the sepsis order set pathway conveyed a sense of urgency and need for immediate order verification and administration of antibiotics.

In addition to clinical care process improvement, accurate documentation also was emphasized in the EWSS. When a sepsis alert was called, a clinician from the primary team was expected to complete a standardized progress note, which communicated clinical findings, a treatment plan, and captured severity of illness (eAppendix 4 available at doi:10.12788/fp.0194). It included sections for subjective, objective, assessment, and plan. In addition, data objects were created for vital signs and common laboratory findings that retrieved important clinical data from VistA and inserted it into the CPRS note.²¹

Nursing staff on the wards were expected to communicate results with the primary team for clinical decision making when a patient had a MEWS + SRS of 3 to 4. A sepsis alert may have been called at the discretion of clinical team members but was not required if the score was < 5. Additionally, vital signs were expected to be checked by the nursing staff on the wards at least every 4 hours for closer monitoring.

Sepsis Review Meetings

Weekly meetings were scheduled to review sepsis cases to assess diagnosis, treatment, and documentation entered in the patient record. The team conducting sepsis reviews comprised the chief of staff, chief of quality management, director of patient safety, physician utilization management advisor, chief resident in quality and patient safety (CRQS), and inpatient pharmacy supervisor. In addition, ad hoc physicians and nurses from different specialty areas, such as infectious diseases, hospitalist section, ICU, and the ED participated on request for subject matter expertise when needed. At the conclusion of weekly sepsis meetings, sepsis team members provided feedback to the clinical staff for continuous improvement purposes.

Results

Before implementation of an EWSS at NF/SGVHS, a plan was devised to increase awareness and educate staff on sepsis-related mortality in late FY 2017. Awareness and education about sepsis-related mortality was organized at physician, nursing, and pharmacy leadership clinical staff meetings. Posters about early warning signs of sepsis also were displayed on the nursing units for educational purposes and to convey the importance of early recognition/treatment of sepsis. In addition, the CRQS was the quality leader for house staff and led sepsis campaign change efforts for residents/fellows. An immediate improvement in unadjusted mortality at MRVAMC was noted with initial sepsis awareness and education. From FY 2017, quarter 3 to FY 2018, quarter 1, the number of acute LOC inpatient deaths decreased from 48 to 28, a 42% reduction in unadjusted mortality at MRVAMC (Figure 1). Additionally, the acute LOC SMR improved from 1.20 during FY 2017, quarter 3 down to as low as 0.71 during FY 2018, quarter 1 (Figure 2).

The number of MRVAMC inpatient deaths increased from 28 in FY 2018, quarter 1 to 45 in FY 2018, quarter 3. While acute LOC showed improvement in unadjusted mortality after sepsis education/awareness, it was felt continuous improvement could not be sustained with education alone. An EWSS was designed and implemented within the EHR system in FY 2018. Following implementation of EWSS and reeducating staff on early recognition and treatment of sepsis, acute LOC inpatient deaths decreased from 45 in FY 2018, quarter 3 through FY 2019 where unadjusted mortality was as low as 27 during FY 2019, quarter 4. The MRVAMC acute LOC SMR was consistently < 1.0 from FY 2018, quarter 4 through FY 2019, quarter 4.

In addition to the observed decrease in acute LOC inpatient deaths and improved SMR, the number of ERT alerts and sepsis alerts on the inpatient wards were monitored from FY 2017 through FY 2019. ERT alerts listed in Table 3 were nonspecific and initiated by nursing staff on the wards where a patient’s clinical status was identified as worsening while sepsis alerts were specific ERT alerts called by the ward nursing staff due to concerns for sepsis. The inpatient wards included inpatient medicine, surgery, and psychiatry acute care and the intermediate level of care unit while outpatient clinical areas of treatment, intensive care units, stroke alerts, and STEMI alerts were excluded.

Discussion

Implementation of the EWSS was associated with improved unadjusted mortality and adjusted mortality for acute LOC at MRVAMC. Although variation exists with application of EWSS at other medical centers, there was similarity with improved sepsis outcomes reported at other health care systems after EWSS implementation.^7-16

Improved unadjusted mortality and adjusted mortality for acute LOC at MRVAMC was likely due to multiple contributing factors. First, during design and implementation of the EWSS, project work was interdisciplinary with input from physicians, nurses, and pharmacists from multiple specialties (ie, ED, ICU, and the medicine service); quality management and data analysis specialists; and clinical informatics. Second, facility commitment to improving early recognition and treatment of sepsis from leadership level down to front-line staff was evident. Weekly sepsis meetings with the NF/SGVHS chief of staff helped to sustain EWSS efforts and to identify additional improvement opportunities. Third, integrated informatics solutions within the EHR helped identify early sepsis and minimized human error as well as assisted with coordination of sepsis care across services. Fourth, the focus was on both early identification and treatment of sepsis in the ED and hospital wards. Although it cannot be deduced whether there was causation between reduced inpatient mortality and an increased number of nonspecific ERT alerts+ sepsis alerts on the inpatient wards after EWSS implementation, inpatient deaths decreased and SMR improved. Finally, the EWSS emphasized both the importance of evidence-based clinical care of sepsis and standardized documentation to appropriately capture clinical severity of illness.

Limitations

This program has limitations. The EWSS was studied at a single VHA facility. Veteran demographics and local epidemiology may limit conclusion of outcomes to an individual VHA facility located in a specific geographical region. Additional research is necessary to demonstrate reproducibility and determine whether applicable to other VHA facilities and community care settings.

SMR is a risk-adjusted formula developed by the VHA Inpatient Evaluation Center, which included numerous factors such as diagnosis, comorbid conditions, age, marital status, procedures, source of admission, specific laboratory values, medical or surgical diagnosis-related group, ICU stays, immunosuppressive status, and a COVID-19 positive indicator (added after this study). Further research is needed to evaluate sepsis-related outcomes using the EWSS during the COVID-19 pandemic.

EWSS in the literature have demonstrated various approaches to early identification and treatment of sepsis and have used different sepsis screening tools.²² Evidence suggests that the MEWS + SRS sepsis screening tool may result in false-positive screenings.^23-27 Additional research into the specificity of this sepsis screening tool is needed. Ward nursing staff were encouraged to initiate automatic sepsis alerts when MEWS + SRS was ≥ 5; however, this still depended on human factors. Because sepsis alerts are software-specific and others were incompatible with the VHA EHR, it was necessary to design our own EWSS.

Despite improvement with MRVAMC acute LOC unadjusted and adjusted mortality with our EWSS, we did not identify any actual improvement in earlier antibiotic administration times once sepsis was recognized. While accurate documentation regarding degree of sepsis improved, a MRVAMC clinical documentation improvement program was expanded in FY 2018. Therefore, it is difficult to demonstrate causation related to improved sepsis documentation with template changes alone. While sepsis alerts on the inpatient wards were variable since EWSS implementation, nonspecific ERT alerts increased. It is unclear whether some sepsis alerts were called as nonspecific ERT alerts, making it impossible to know the true number of sepsis alerts.

MRVAMC experienced an increase in nurse turnover during FY 2018 and as a teaching hospital had frequent rotating residents and fellows new to processes/protocols. These factors may have contributed to variations in unadjusted mortality. Also the decrease in inpatient mortality and improvement in SMR on acute LOC could have been the result of factors other than the EWSS and the effect of education alone may have been at least as good as that of the EWSS intervention.

Conclusions

Education along with the possible implementation of an EWSS at NF/SGVHS was associated with a decrease in the number of inpatient deaths on MRVAMC’s acute LOC wards from as high as 48 in FY 2017, quarter 3 to as low as 27 in FY 2019, quarter 4 resulting in as large of an improvement as a 44% reduction in unadjusted mortality from FY 2017 to FY 2019. In addition, MRVAMC’s acute LOC SMR improved from > 1.0 to < 1.0, demonstrating fewer inpatient mortalities than predicted from FY 2017 to FY 2019.

This multifaceted interventional strategy may be effectively applied at other VHA health care facilities that use the same EHR system. Next steps may include determining the specificity of MEWS + SRS as a sepsis screening tool; studying outcomes of MRVAMC’s EWSS during the COVID-19 era; and conducting a multicentered study on this EWSS across multiple VHA facilities.

In 1997, Elizabeth McGlynn wrote, “Measuring quality is no simple task.”¹ We are reminded of this seminal Health Affairs article at a very pertinent point—as health care practice progresses, measuring the impact of performance improvement initiatives on clinical care delivery remains integral to monitoring overall effectiveness of quality. Mortality outcomes are a major focus of quality.

Inpatient mortality within the Veterans Health Administration (VHA) was measured as actual number of deaths (unadjusted mortality), and adjusted mortality was calculated using the standardized mortality ratio (SMR). SMR included actual number of deaths during hospitalization or within 1 day of hospital discharge divided by predicted number of deaths using a risk-adjusted formula and was calculated separately for acute level of care (LOC) and the intensive care unit (ICU). Using risk-adjusted SMR, if an observed/expected ratio was > 1.0, there were more inpatient deaths than expected; if < 1.0, fewer inpatient deaths occurred than predicted; and if 1.0, observed number of inpatient deaths was equivalent to expected number of deaths.²

Mortality reduction is a complex area of performance improvement. Health care facilities often focus their efforts on the biggest mortality contributors. According to Dantes and Epstein, sepsis results in about 265,000 deaths annually in the United States.³ Reinhart and colleagues demonstrated that sepsis is a worldwide issue resulting in approximately 30 million cases and 6 million deaths annually.⁴ Furthermore, Kumar and colleagues have noted that when sepsis progresses to septic shock, survival decreases by almost 8% for each hour delay in sepsis identification and treatment.⁵

Improvements in sepsis management have been multifaceted. The Surviving Sepsis Campaign guidelines created sepsis treatment bundles to guide early diagnosis/treatment of sepsis.⁶ In addition to awareness and sepsis care bundles, a plethora of informatics solutions within electronic health record (EHR) systems have demonstrated improved sepsis care.^7-16 Various approaches to early diagnosis and management of sepsis have been collectively referred to as an early warning sepsis system (EWSS).

An EWSS typically contains automated decision support tools that are integrated in the EHR and meant to assist health care professionals with clinical workflow decision-making. Automated decision support tools within the EHR have a variety of functions, such as clinical care reminders and alerts.¹⁷

Sepsis screening tools function as a form of automated decision support and may be incorporated into the EHR to support the EWSS. Although sepsis screening tools vary, they frequently include a combination of data involving vital signs, laboratory values and/or physical examination findings, such as mental status evaluation.The Modified Early Warning Signs (MEWS) + Sepsis Recognition Score (SRS) is one example of a sepsis screening tool.^7,16

At Malcom Randall Veterans Affairs Medical Center (MRVAMC) in Gainesville, Florida, we identified a quality improvement project opportunity to improve sepsis care in the emergency department (ED) and inpatient wards using the VHA EHR system, the Computerized Patient Record System (CPRS), which is supported by the Veterans Information Systems and Technology Architecture (VistA).¹⁸ A VistA/CPRS EWSS was developed using Lean Six Sigma DMAIC (define, measure, analyze, improve, and control) methodology.¹⁹ During the improve stage, informatics solutions were applied and included a combination of EHR interventions, such as template design, an order set, and clinical reminders. Clinical reminders have a wide variety of use, such as reminders for clinical tasks and as automated decision support within clinical workflows using Boolean logic.

To the best of our knowledge, there has been no published application of an EWSS within VistA/CPRS. In this study, we outline the strategic development of an EWSS in VistA/CPRS that assisted clinical staff with identification and treatment of sepsis; improved documentation of sepsis when present; and associated with improvement in unadjusted and adjusted inpatient mortality.

Methods 

According to policy activities that constitute research at MRVAMC, no institutional review board approval was required as this work met criteria for operational improvement activities exempt from ethics review.

The North Florida/South Georgia Veterans Health System (NF/SGVHS) includes MRVAMC, a large academic hospital with rotating residents/fellows and multiple specialty care services. MRVAMC comprised 144 beds on the medicine/surgery wards; 48 beds in the psychiatry unit; 18 intermediate LOC beds; and 27 ICU beds. The MRVAMC SMR was identified as an improvement opportunity during fiscal year (FY) 2017 (Table 1). Its adjusted mortality for acute LOC demonstrated an observed/expected ratio of > 1.0 suggesting more inpatient deaths were observed than expected. The number of deaths (unadjusted mortality) on acute LOC at MRVAMC was noted to be rising during the first 3 quarters of FY 2017. A deeper examination of data by Pyramid Analytics (www.pyramidanalytics.com) discovered that sepsis was the primary driver for inpatient mortality on acute LOC at MRVAMC. Our goal was to reduce inpatient sepsis-related mortality via development of an EWSS that leveraged VistA/CPRS to improve early identification and treatment of sepsis in the ED and inpatient wards.

Emergency Department

Given the importance of recognizing sepsis early, the sepsis team focused on improvement opportunities at the initial point of patient contact: ED triage. The goal was to incorporate automated VistA/CPRS decision support to assist clinicians with identifying sepsis in triage using MEWS, which was chosen to optimize immediate hospital-wide buy-in. Clinical staff were already familiar with MEWS, which was in use on the inpatient wards.

Flow through the ED and availability of resources differed from the wards. Hence, modification to MEWS on the wards was necessary to fit clinical workflow in the ED. Temperature, heart rate (HR), respiratory rate (RR), systolic blood pressure (SBP), mental status, and white blood cell count (WBC) factored into a MEWS + SRS score on the wards (Table 2). For the ED, MEWS included temperature, HR, RR and SBP, but excluded mental status and WBC. Mental status assessment was excluded due to technical infeasibility (while vital signs could be automatically calculated in real time for a MEWS score, that was not possible for mental status changes). WBC was excluded from the ED as laboratory test results would not be available in triage.

MEWS + SRS scores were calculated in VistA by using clinical reminders. Clinical reminder logic included a series of conditional statements based on various combinations of MEWS + SRS clinical data entered in the EHR. When ED triage vital signs data were entered in CPRS, clinical data were stored and processed according to clinical reminder logic in VistA and displayed to the user in CPRS. While MEWS of ≥ 5 triggered a sepsis alert on the wards, the ≥ 4 threshold was used in the ED given mental status and WBC were excluded from calculations in triage (eAppendix 1 available at doi:10.12788/fp.0194).

Once a sepsis alert was triggered in triage for MEWS ≥ 4, ED nursing staff prioritized bed location and expedited staffing with an ED attending physician for early assessment. The ED attending then performed an assessment to confirm whether sepsis was present and direct early treatment. Although every patient who triggered a sepsis alert in triage did not meet clinical findings of sepsis, patients with MEWS ≥ 4 were frequently ill and required timely intervention.

If an ED attending physician agreed with a sepsis diagnosis, the physician had access to a sepsis workup and treatment order set in CPRS (eAppendix 2 available at doi:10.12788/fp.0194). The sepsis order set incorporated recommendations from the Surviving Sepsis Campaign guidelines and included orders for 2 large-bore peripheral IV lines; aggressive fluid resuscitation (30 mL/kg) for patients with clinical findings of hypoperfusion; broad-spectrum antibiotics; and frequent ordering of laboratory tests and imaging during initial sepsis workup.⁶ Vancomycin and cefepime were selected as routine broad-spectrum antibiotics in the order set when sepsis was suspected based on local antimicrobial stewardship and safety-efficacy profiles. For example, Luther and colleagues demonstrated that cefepime has lower rates of acute kidney injury when combined with vancomycin vs vancomycin + piperacillin-tazobactam.²⁰ If a β-lactam antibiotic could not be used due to a patient’s drug allergy history, aztreonam was available as an alternative option.

The design of the order set also functioned as a communication interface with clinical pharmacists. Given the large volume of antibiotics ordered in the ED, it was difficult for pharmacists to prioritize antibiotic order verification. While stat orders convey high priority, they often lack specificity. When antibiotic orders were selected from the sepsis order set, comments were already included that stated: “STAT. First dose for sepsis protocol” (eAppendix 3 available at doi:10.12788/fp.0194). This standardized communication conveyed a sense of urgency and a collective understanding that patients with suspected sepsis required timely order verification and administration of antibiotics.

Hospital Ward

Mental status and WBC were included on the wards to monitor for possible signs of sepsis, using MEWS + SRS, which was routinely monitored by nursing every 4 to 8 hours. When MEWS + SRS was ≥ 5 points, ward nursing staff called a sepsis alert.^7,16 Early response team (ERT) members received telephone notifications of the alert. ERT staff proceeded with immediate evaluation and treatment at the bedside along with determination for most appropriate LOC. The ERT members included an ICU physician and nurse; respiratory therapist; and nursing supervisor/bed flow coordinator. During bedside evaluation, if the ERT or primary team agreed with a sepsis diagnosis, the ERT or primary team used the sepsis order set to ensure standardized procedures. Stat orders generated through the sepsis order set pathway conveyed a sense of urgency and need for immediate order verification and administration of antibiotics.

In addition to clinical care process improvement, accurate documentation also was emphasized in the EWSS. When a sepsis alert was called, a clinician from the primary team was expected to complete a standardized progress note, which communicated clinical findings, a treatment plan, and captured severity of illness (eAppendix 4 available at doi:10.12788/fp.0194). It included sections for subjective, objective, assessment, and plan. In addition, data objects were created for vital signs and common laboratory findings that retrieved important clinical data from VistA and inserted it into the CPRS note.²¹

Nursing staff on the wards were expected to communicate results with the primary team for clinical decision making when a patient had a MEWS + SRS of 3 to 4. A sepsis alert may have been called at the discretion of clinical team members but was not required if the score was < 5. Additionally, vital signs were expected to be checked by the nursing staff on the wards at least every 4 hours for closer monitoring.

Sepsis Review Meetings

Weekly meetings were scheduled to review sepsis cases to assess diagnosis, treatment, and documentation entered in the patient record. The team conducting sepsis reviews comprised the chief of staff, chief of quality management, director of patient safety, physician utilization management advisor, chief resident in quality and patient safety (CRQS), and inpatient pharmacy supervisor. In addition, ad hoc physicians and nurses from different specialty areas, such as infectious diseases, hospitalist section, ICU, and the ED participated on request for subject matter expertise when needed. At the conclusion of weekly sepsis meetings, sepsis team members provided feedback to the clinical staff for continuous improvement purposes.

Results

Before implementation of an EWSS at NF/SGVHS, a plan was devised to increase awareness and educate staff on sepsis-related mortality in late FY 2017. Awareness and education about sepsis-related mortality was organized at physician, nursing, and pharmacy leadership clinical staff meetings. Posters about early warning signs of sepsis also were displayed on the nursing units for educational purposes and to convey the importance of early recognition/treatment of sepsis. In addition, the CRQS was the quality leader for house staff and led sepsis campaign change efforts for residents/fellows. An immediate improvement in unadjusted mortality at MRVAMC was noted with initial sepsis awareness and education. From FY 2017, quarter 3 to FY 2018, quarter 1, the number of acute LOC inpatient deaths decreased from 48 to 28, a 42% reduction in unadjusted mortality at MRVAMC (Figure 1). Additionally, the acute LOC SMR improved from 1.20 during FY 2017, quarter 3 down to as low as 0.71 during FY 2018, quarter 1 (Figure 2).

The number of MRVAMC inpatient deaths increased from 28 in FY 2018, quarter 1 to 45 in FY 2018, quarter 3. While acute LOC showed improvement in unadjusted mortality after sepsis education/awareness, it was felt continuous improvement could not be sustained with education alone. An EWSS was designed and implemented within the EHR system in FY 2018. Following implementation of EWSS and reeducating staff on early recognition and treatment of sepsis, acute LOC inpatient deaths decreased from 45 in FY 2018, quarter 3 through FY 2019 where unadjusted mortality was as low as 27 during FY 2019, quarter 4. The MRVAMC acute LOC SMR was consistently < 1.0 from FY 2018, quarter 4 through FY 2019, quarter 4.

In addition to the observed decrease in acute LOC inpatient deaths and improved SMR, the number of ERT alerts and sepsis alerts on the inpatient wards were monitored from FY 2017 through FY 2019. ERT alerts listed in Table 3 were nonspecific and initiated by nursing staff on the wards where a patient’s clinical status was identified as worsening while sepsis alerts were specific ERT alerts called by the ward nursing staff due to concerns for sepsis. The inpatient wards included inpatient medicine, surgery, and psychiatry acute care and the intermediate level of care unit while outpatient clinical areas of treatment, intensive care units, stroke alerts, and STEMI alerts were excluded.

Discussion

Implementation of the EWSS was associated with improved unadjusted mortality and adjusted mortality for acute LOC at MRVAMC. Although variation exists with application of EWSS at other medical centers, there was similarity with improved sepsis outcomes reported at other health care systems after EWSS implementation.^7-16

Improved unadjusted mortality and adjusted mortality for acute LOC at MRVAMC was likely due to multiple contributing factors. First, during design and implementation of the EWSS, project work was interdisciplinary with input from physicians, nurses, and pharmacists from multiple specialties (ie, ED, ICU, and the medicine service); quality management and data analysis specialists; and clinical informatics. Second, facility commitment to improving early recognition and treatment of sepsis from leadership level down to front-line staff was evident. Weekly sepsis meetings with the NF/SGVHS chief of staff helped to sustain EWSS efforts and to identify additional improvement opportunities. Third, integrated informatics solutions within the EHR helped identify early sepsis and minimized human error as well as assisted with coordination of sepsis care across services. Fourth, the focus was on both early identification and treatment of sepsis in the ED and hospital wards. Although it cannot be deduced whether there was causation between reduced inpatient mortality and an increased number of nonspecific ERT alerts+ sepsis alerts on the inpatient wards after EWSS implementation, inpatient deaths decreased and SMR improved. Finally, the EWSS emphasized both the importance of evidence-based clinical care of sepsis and standardized documentation to appropriately capture clinical severity of illness.

Limitations

This program has limitations. The EWSS was studied at a single VHA facility. Veteran demographics and local epidemiology may limit conclusion of outcomes to an individual VHA facility located in a specific geographical region. Additional research is necessary to demonstrate reproducibility and determine whether applicable to other VHA facilities and community care settings.

SMR is a risk-adjusted formula developed by the VHA Inpatient Evaluation Center, which included numerous factors such as diagnosis, comorbid conditions, age, marital status, procedures, source of admission, specific laboratory values, medical or surgical diagnosis-related group, ICU stays, immunosuppressive status, and a COVID-19 positive indicator (added after this study). Further research is needed to evaluate sepsis-related outcomes using the EWSS during the COVID-19 pandemic.

EWSS in the literature have demonstrated various approaches to early identification and treatment of sepsis and have used different sepsis screening tools.²² Evidence suggests that the MEWS + SRS sepsis screening tool may result in false-positive screenings.^23-27 Additional research into the specificity of this sepsis screening tool is needed. Ward nursing staff were encouraged to initiate automatic sepsis alerts when MEWS + SRS was ≥ 5; however, this still depended on human factors. Because sepsis alerts are software-specific and others were incompatible with the VHA EHR, it was necessary to design our own EWSS.

Despite improvement with MRVAMC acute LOC unadjusted and adjusted mortality with our EWSS, we did not identify any actual improvement in earlier antibiotic administration times once sepsis was recognized. While accurate documentation regarding degree of sepsis improved, a MRVAMC clinical documentation improvement program was expanded in FY 2018. Therefore, it is difficult to demonstrate causation related to improved sepsis documentation with template changes alone. While sepsis alerts on the inpatient wards were variable since EWSS implementation, nonspecific ERT alerts increased. It is unclear whether some sepsis alerts were called as nonspecific ERT alerts, making it impossible to know the true number of sepsis alerts.

MRVAMC experienced an increase in nurse turnover during FY 2018 and as a teaching hospital had frequent rotating residents and fellows new to processes/protocols. These factors may have contributed to variations in unadjusted mortality. Also the decrease in inpatient mortality and improvement in SMR on acute LOC could have been the result of factors other than the EWSS and the effect of education alone may have been at least as good as that of the EWSS intervention.

Conclusions

Education along with the possible implementation of an EWSS at NF/SGVHS was associated with a decrease in the number of inpatient deaths on MRVAMC’s acute LOC wards from as high as 48 in FY 2017, quarter 3 to as low as 27 in FY 2019, quarter 4 resulting in as large of an improvement as a 44% reduction in unadjusted mortality from FY 2017 to FY 2019. In addition, MRVAMC’s acute LOC SMR improved from > 1.0 to < 1.0, demonstrating fewer inpatient mortalities than predicted from FY 2017 to FY 2019.

This multifaceted interventional strategy may be effectively applied at other VHA health care facilities that use the same EHR system. Next steps may include determining the specificity of MEWS + SRS as a sepsis screening tool; studying outcomes of MRVAMC’s EWSS during the COVID-19 era; and conducting a multicentered study on this EWSS across multiple VHA facilities.

In 1997, Elizabeth McGlynn wrote, “Measuring quality is no simple task.”¹ We are reminded of this seminal Health Affairs article at a very pertinent point—as health care practice progresses, measuring the impact of performance improvement initiatives on clinical care delivery remains integral to monitoring overall effectiveness of quality. Mortality outcomes are a major focus of quality.

Inpatient mortality within the Veterans Health Administration (VHA) was measured as actual number of deaths (unadjusted mortality), and adjusted mortality was calculated using the standardized mortality ratio (SMR). SMR included actual number of deaths during hospitalization or within 1 day of hospital discharge divided by predicted number of deaths using a risk-adjusted formula and was calculated separately for acute level of care (LOC) and the intensive care unit (ICU). Using risk-adjusted SMR, if an observed/expected ratio was > 1.0, there were more inpatient deaths than expected; if < 1.0, fewer inpatient deaths occurred than predicted; and if 1.0, observed number of inpatient deaths was equivalent to expected number of deaths.²

Mortality reduction is a complex area of performance improvement. Health care facilities often focus their efforts on the biggest mortality contributors. According to Dantes and Epstein, sepsis results in about 265,000 deaths annually in the United States.³ Reinhart and colleagues demonstrated that sepsis is a worldwide issue resulting in approximately 30 million cases and 6 million deaths annually.⁴ Furthermore, Kumar and colleagues have noted that when sepsis progresses to septic shock, survival decreases by almost 8% for each hour delay in sepsis identification and treatment.⁵

Improvements in sepsis management have been multifaceted. The Surviving Sepsis Campaign guidelines created sepsis treatment bundles to guide early diagnosis/treatment of sepsis.⁶ In addition to awareness and sepsis care bundles, a plethora of informatics solutions within electronic health record (EHR) systems have demonstrated improved sepsis care.^7-16 Various approaches to early diagnosis and management of sepsis have been collectively referred to as an early warning sepsis system (EWSS).

An EWSS typically contains automated decision support tools that are integrated in the EHR and meant to assist health care professionals with clinical workflow decision-making. Automated decision support tools within the EHR have a variety of functions, such as clinical care reminders and alerts.¹⁷

Sepsis screening tools function as a form of automated decision support and may be incorporated into the EHR to support the EWSS. Although sepsis screening tools vary, they frequently include a combination of data involving vital signs, laboratory values and/or physical examination findings, such as mental status evaluation.The Modified Early Warning Signs (MEWS) + Sepsis Recognition Score (SRS) is one example of a sepsis screening tool.^7,16

At Malcom Randall Veterans Affairs Medical Center (MRVAMC) in Gainesville, Florida, we identified a quality improvement project opportunity to improve sepsis care in the emergency department (ED) and inpatient wards using the VHA EHR system, the Computerized Patient Record System (CPRS), which is supported by the Veterans Information Systems and Technology Architecture (VistA).¹⁸ A VistA/CPRS EWSS was developed using Lean Six Sigma DMAIC (define, measure, analyze, improve, and control) methodology.¹⁹ During the improve stage, informatics solutions were applied and included a combination of EHR interventions, such as template design, an order set, and clinical reminders. Clinical reminders have a wide variety of use, such as reminders for clinical tasks and as automated decision support within clinical workflows using Boolean logic.

To the best of our knowledge, there has been no published application of an EWSS within VistA/CPRS. In this study, we outline the strategic development of an EWSS in VistA/CPRS that assisted clinical staff with identification and treatment of sepsis; improved documentation of sepsis when present; and associated with improvement in unadjusted and adjusted inpatient mortality.

Methods 

According to policy activities that constitute research at MRVAMC, no institutional review board approval was required as this work met criteria for operational improvement activities exempt from ethics review.

The North Florida/South Georgia Veterans Health System (NF/SGVHS) includes MRVAMC, a large academic hospital with rotating residents/fellows and multiple specialty care services. MRVAMC comprised 144 beds on the medicine/surgery wards; 48 beds in the psychiatry unit; 18 intermediate LOC beds; and 27 ICU beds. The MRVAMC SMR was identified as an improvement opportunity during fiscal year (FY) 2017 (Table 1). Its adjusted mortality for acute LOC demonstrated an observed/expected ratio of > 1.0 suggesting more inpatient deaths were observed than expected. The number of deaths (unadjusted mortality) on acute LOC at MRVAMC was noted to be rising during the first 3 quarters of FY 2017. A deeper examination of data by Pyramid Analytics (www.pyramidanalytics.com) discovered that sepsis was the primary driver for inpatient mortality on acute LOC at MRVAMC. Our goal was to reduce inpatient sepsis-related mortality via development of an EWSS that leveraged VistA/CPRS to improve early identification and treatment of sepsis in the ED and inpatient wards.

Emergency Department

Given the importance of recognizing sepsis early, the sepsis team focused on improvement opportunities at the initial point of patient contact: ED triage. The goal was to incorporate automated VistA/CPRS decision support to assist clinicians with identifying sepsis in triage using MEWS, which was chosen to optimize immediate hospital-wide buy-in. Clinical staff were already familiar with MEWS, which was in use on the inpatient wards.

Flow through the ED and availability of resources differed from the wards. Hence, modification to MEWS on the wards was necessary to fit clinical workflow in the ED. Temperature, heart rate (HR), respiratory rate (RR), systolic blood pressure (SBP), mental status, and white blood cell count (WBC) factored into a MEWS + SRS score on the wards (Table 2). For the ED, MEWS included temperature, HR, RR and SBP, but excluded mental status and WBC. Mental status assessment was excluded due to technical infeasibility (while vital signs could be automatically calculated in real time for a MEWS score, that was not possible for mental status changes). WBC was excluded from the ED as laboratory test results would not be available in triage.

MEWS + SRS scores were calculated in VistA by using clinical reminders. Clinical reminder logic included a series of conditional statements based on various combinations of MEWS + SRS clinical data entered in the EHR. When ED triage vital signs data were entered in CPRS, clinical data were stored and processed according to clinical reminder logic in VistA and displayed to the user in CPRS. While MEWS of ≥ 5 triggered a sepsis alert on the wards, the ≥ 4 threshold was used in the ED given mental status and WBC were excluded from calculations in triage (eAppendix 1 available at doi:10.12788/fp.0194).

Once a sepsis alert was triggered in triage for MEWS ≥ 4, ED nursing staff prioritized bed location and expedited staffing with an ED attending physician for early assessment. The ED attending then performed an assessment to confirm whether sepsis was present and direct early treatment. Although every patient who triggered a sepsis alert in triage did not meet clinical findings of sepsis, patients with MEWS ≥ 4 were frequently ill and required timely intervention.

If an ED attending physician agreed with a sepsis diagnosis, the physician had access to a sepsis workup and treatment order set in CPRS (eAppendix 2 available at doi:10.12788/fp.0194). The sepsis order set incorporated recommendations from the Surviving Sepsis Campaign guidelines and included orders for 2 large-bore peripheral IV lines; aggressive fluid resuscitation (30 mL/kg) for patients with clinical findings of hypoperfusion; broad-spectrum antibiotics; and frequent ordering of laboratory tests and imaging during initial sepsis workup.⁶ Vancomycin and cefepime were selected as routine broad-spectrum antibiotics in the order set when sepsis was suspected based on local antimicrobial stewardship and safety-efficacy profiles. For example, Luther and colleagues demonstrated that cefepime has lower rates of acute kidney injury when combined with vancomycin vs vancomycin + piperacillin-tazobactam.²⁰ If a β-lactam antibiotic could not be used due to a patient’s drug allergy history, aztreonam was available as an alternative option.

The design of the order set also functioned as a communication interface with clinical pharmacists. Given the large volume of antibiotics ordered in the ED, it was difficult for pharmacists to prioritize antibiotic order verification. While stat orders convey high priority, they often lack specificity. When antibiotic orders were selected from the sepsis order set, comments were already included that stated: “STAT. First dose for sepsis protocol” (eAppendix 3 available at doi:10.12788/fp.0194). This standardized communication conveyed a sense of urgency and a collective understanding that patients with suspected sepsis required timely order verification and administration of antibiotics.

Hospital Ward

Mental status and WBC were included on the wards to monitor for possible signs of sepsis, using MEWS + SRS, which was routinely monitored by nursing every 4 to 8 hours. When MEWS + SRS was ≥ 5 points, ward nursing staff called a sepsis alert.^7,16 Early response team (ERT) members received telephone notifications of the alert. ERT staff proceeded with immediate evaluation and treatment at the bedside along with determination for most appropriate LOC. The ERT members included an ICU physician and nurse; respiratory therapist; and nursing supervisor/bed flow coordinator. During bedside evaluation, if the ERT or primary team agreed with a sepsis diagnosis, the ERT or primary team used the sepsis order set to ensure standardized procedures. Stat orders generated through the sepsis order set pathway conveyed a sense of urgency and need for immediate order verification and administration of antibiotics.

In addition to clinical care process improvement, accurate documentation also was emphasized in the EWSS. When a sepsis alert was called, a clinician from the primary team was expected to complete a standardized progress note, which communicated clinical findings, a treatment plan, and captured severity of illness (eAppendix 4 available at doi:10.12788/fp.0194). It included sections for subjective, objective, assessment, and plan. In addition, data objects were created for vital signs and common laboratory findings that retrieved important clinical data from VistA and inserted it into the CPRS note.²¹

Nursing staff on the wards were expected to communicate results with the primary team for clinical decision making when a patient had a MEWS + SRS of 3 to 4. A sepsis alert may have been called at the discretion of clinical team members but was not required if the score was < 5. Additionally, vital signs were expected to be checked by the nursing staff on the wards at least every 4 hours for closer monitoring.

Sepsis Review Meetings

Weekly meetings were scheduled to review sepsis cases to assess diagnosis, treatment, and documentation entered in the patient record. The team conducting sepsis reviews comprised the chief of staff, chief of quality management, director of patient safety, physician utilization management advisor, chief resident in quality and patient safety (CRQS), and inpatient pharmacy supervisor. In addition, ad hoc physicians and nurses from different specialty areas, such as infectious diseases, hospitalist section, ICU, and the ED participated on request for subject matter expertise when needed. At the conclusion of weekly sepsis meetings, sepsis team members provided feedback to the clinical staff for continuous improvement purposes.

Results

Before implementation of an EWSS at NF/SGVHS, a plan was devised to increase awareness and educate staff on sepsis-related mortality in late FY 2017. Awareness and education about sepsis-related mortality was organized at physician, nursing, and pharmacy leadership clinical staff meetings. Posters about early warning signs of sepsis also were displayed on the nursing units for educational purposes and to convey the importance of early recognition/treatment of sepsis. In addition, the CRQS was the quality leader for house staff and led sepsis campaign change efforts for residents/fellows. An immediate improvement in unadjusted mortality at MRVAMC was noted with initial sepsis awareness and education. From FY 2017, quarter 3 to FY 2018, quarter 1, the number of acute LOC inpatient deaths decreased from 48 to 28, a 42% reduction in unadjusted mortality at MRVAMC (Figure 1). Additionally, the acute LOC SMR improved from 1.20 during FY 2017, quarter 3 down to as low as 0.71 during FY 2018, quarter 1 (Figure 2).

The number of MRVAMC inpatient deaths increased from 28 in FY 2018, quarter 1 to 45 in FY 2018, quarter 3. While acute LOC showed improvement in unadjusted mortality after sepsis education/awareness, it was felt continuous improvement could not be sustained with education alone. An EWSS was designed and implemented within the EHR system in FY 2018. Following implementation of EWSS and reeducating staff on early recognition and treatment of sepsis, acute LOC inpatient deaths decreased from 45 in FY 2018, quarter 3 through FY 2019 where unadjusted mortality was as low as 27 during FY 2019, quarter 4. The MRVAMC acute LOC SMR was consistently < 1.0 from FY 2018, quarter 4 through FY 2019, quarter 4.

In addition to the observed decrease in acute LOC inpatient deaths and improved SMR, the number of ERT alerts and sepsis alerts on the inpatient wards were monitored from FY 2017 through FY 2019. ERT alerts listed in Table 3 were nonspecific and initiated by nursing staff on the wards where a patient’s clinical status was identified as worsening while sepsis alerts were specific ERT alerts called by the ward nursing staff due to concerns for sepsis. The inpatient wards included inpatient medicine, surgery, and psychiatry acute care and the intermediate level of care unit while outpatient clinical areas of treatment, intensive care units, stroke alerts, and STEMI alerts were excluded.

Discussion

Implementation of the EWSS was associated with improved unadjusted mortality and adjusted mortality for acute LOC at MRVAMC. Although variation exists with application of EWSS at other medical centers, there was similarity with improved sepsis outcomes reported at other health care systems after EWSS implementation.^7-16

Improved unadjusted mortality and adjusted mortality for acute LOC at MRVAMC was likely due to multiple contributing factors. First, during design and implementation of the EWSS, project work was interdisciplinary with input from physicians, nurses, and pharmacists from multiple specialties (ie, ED, ICU, and the medicine service); quality management and data analysis specialists; and clinical informatics. Second, facility commitment to improving early recognition and treatment of sepsis from leadership level down to front-line staff was evident. Weekly sepsis meetings with the NF/SGVHS chief of staff helped to sustain EWSS efforts and to identify additional improvement opportunities. Third, integrated informatics solutions within the EHR helped identify early sepsis and minimized human error as well as assisted with coordination of sepsis care across services. Fourth, the focus was on both early identification and treatment of sepsis in the ED and hospital wards. Although it cannot be deduced whether there was causation between reduced inpatient mortality and an increased number of nonspecific ERT alerts+ sepsis alerts on the inpatient wards after EWSS implementation, inpatient deaths decreased and SMR improved. Finally, the EWSS emphasized both the importance of evidence-based clinical care of sepsis and standardized documentation to appropriately capture clinical severity of illness.

Limitations

This program has limitations. The EWSS was studied at a single VHA facility. Veteran demographics and local epidemiology may limit conclusion of outcomes to an individual VHA facility located in a specific geographical region. Additional research is necessary to demonstrate reproducibility and determine whether applicable to other VHA facilities and community care settings.

SMR is a risk-adjusted formula developed by the VHA Inpatient Evaluation Center, which included numerous factors such as diagnosis, comorbid conditions, age, marital status, procedures, source of admission, specific laboratory values, medical or surgical diagnosis-related group, ICU stays, immunosuppressive status, and a COVID-19 positive indicator (added after this study). Further research is needed to evaluate sepsis-related outcomes using the EWSS during the COVID-19 pandemic.

EWSS in the literature have demonstrated various approaches to early identification and treatment of sepsis and have used different sepsis screening tools.²² Evidence suggests that the MEWS + SRS sepsis screening tool may result in false-positive screenings.^23-27 Additional research into the specificity of this sepsis screening tool is needed. Ward nursing staff were encouraged to initiate automatic sepsis alerts when MEWS + SRS was ≥ 5; however, this still depended on human factors. Because sepsis alerts are software-specific and others were incompatible with the VHA EHR, it was necessary to design our own EWSS.

Despite improvement with MRVAMC acute LOC unadjusted and adjusted mortality with our EWSS, we did not identify any actual improvement in earlier antibiotic administration times once sepsis was recognized. While accurate documentation regarding degree of sepsis improved, a MRVAMC clinical documentation improvement program was expanded in FY 2018. Therefore, it is difficult to demonstrate causation related to improved sepsis documentation with template changes alone. While sepsis alerts on the inpatient wards were variable since EWSS implementation, nonspecific ERT alerts increased. It is unclear whether some sepsis alerts were called as nonspecific ERT alerts, making it impossible to know the true number of sepsis alerts.

MRVAMC experienced an increase in nurse turnover during FY 2018 and as a teaching hospital had frequent rotating residents and fellows new to processes/protocols. These factors may have contributed to variations in unadjusted mortality. Also the decrease in inpatient mortality and improvement in SMR on acute LOC could have been the result of factors other than the EWSS and the effect of education alone may have been at least as good as that of the EWSS intervention.

Conclusions

Education along with the possible implementation of an EWSS at NF/SGVHS was associated with a decrease in the number of inpatient deaths on MRVAMC’s acute LOC wards from as high as 48 in FY 2017, quarter 3 to as low as 27 in FY 2019, quarter 4 resulting in as large of an improvement as a 44% reduction in unadjusted mortality from FY 2017 to FY 2019. In addition, MRVAMC’s acute LOC SMR improved from > 1.0 to < 1.0, demonstrating fewer inpatient mortalities than predicted from FY 2017 to FY 2019.

This multifaceted interventional strategy may be effectively applied at other VHA health care facilities that use the same EHR system. Next steps may include determining the specificity of MEWS + SRS as a sepsis screening tool; studying outcomes of MRVAMC’s EWSS during the COVID-19 era; and conducting a multicentered study on this EWSS across multiple VHA facilities.

The effectiveness of COVID-19 vaccines produced by Pfizer/BioNTech, Moderna, and Johnson & Johnson dropped dramatically as the Delta variant swept the United States, a study of almost 800,000 veterans found.

The study, published in the journal Science ., says the three vaccines offered about the same protection against the virus in March, when the Delta variant was first detected in the United States, but that changed 6 months later.

The Moderna two-dose vaccine went from being 89% effective in March to 58% effective in September, according to a story about the study in theLos Angeles Times.

Meanwhile, the Pfizer/BioNTech vaccine went from being 87% effective to 45% effective over the same time period.

The Johnson & Johnson vaccine showed the biggest drop -- from 86% effectiveness to 13% over those 6 months.

“In summary, although vaccination remains protective against SARS-CoV-2 infection, protection waned as the Delta variant emerged in the U.S., and this decline did not differ by age,” the study said.

The three vaccines also lost effectiveness in the ability to protect against death in veterans 65 and over after only 3 months, the Los Angeles Times reported.

Compared to unvaccinated veterans in that age group, veterans who got the Moderna vaccine and had a breakthrough case were 76% less likely to die of COVID-19 by July.

The protection was 70% for Pfizer/BioNTech vaccine recipients and 52% for J&J vaccine recipients for the same age group, compared to unvaccinated veterans, according to the newspaper.

For veterans under 65, the protectiveness against a fatal case of COVID was 84% for Pfizer/BioNTech recipients, 82% for Moderna recipients, and 73% for J&J recipients, compared to unvaccinated veterans in that age group.

The study confirms the need for booster vaccines and protective measures such as vaccine passports, vaccine mandates, masking, hand-washing, and social distancing, the researchers said.

Of the veterans studied, about 500,000 were vaccinated and 300,000 were not. Researchers noted that the study population had 6 times as many men as women. About 48% of the study group was 65 or older, 29% was 50-64, while 24% was under 50.

Researchers from the Public Health Institute in Oakland, the Veterans Affairs Medical Center in San Francisco, and the University of Texas Health Science Center conducted the study.

A version of this article first appeared on WebMD.com.

The effectiveness of COVID-19 vaccines produced by Pfizer/BioNTech, Moderna, and Johnson & Johnson dropped dramatically as the Delta variant swept the United States, a study of almost 800,000 veterans found.

The study, published in the journal Science ., says the three vaccines offered about the same protection against the virus in March, when the Delta variant was first detected in the United States, but that changed 6 months later.

The Moderna two-dose vaccine went from being 89% effective in March to 58% effective in September, according to a story about the study in theLos Angeles Times.

Meanwhile, the Pfizer/BioNTech vaccine went from being 87% effective to 45% effective over the same time period.

The Johnson & Johnson vaccine showed the biggest drop -- from 86% effectiveness to 13% over those 6 months.

“In summary, although vaccination remains protective against SARS-CoV-2 infection, protection waned as the Delta variant emerged in the U.S., and this decline did not differ by age,” the study said.

The three vaccines also lost effectiveness in the ability to protect against death in veterans 65 and over after only 3 months, the Los Angeles Times reported.

Compared to unvaccinated veterans in that age group, veterans who got the Moderna vaccine and had a breakthrough case were 76% less likely to die of COVID-19 by July.

The protection was 70% for Pfizer/BioNTech vaccine recipients and 52% for J&J vaccine recipients for the same age group, compared to unvaccinated veterans, according to the newspaper.

For veterans under 65, the protectiveness against a fatal case of COVID was 84% for Pfizer/BioNTech recipients, 82% for Moderna recipients, and 73% for J&J recipients, compared to unvaccinated veterans in that age group.

The study confirms the need for booster vaccines and protective measures such as vaccine passports, vaccine mandates, masking, hand-washing, and social distancing, the researchers said.

Of the veterans studied, about 500,000 were vaccinated and 300,000 were not. Researchers noted that the study population had 6 times as many men as women. About 48% of the study group was 65 or older, 29% was 50-64, while 24% was under 50.

Researchers from the Public Health Institute in Oakland, the Veterans Affairs Medical Center in San Francisco, and the University of Texas Health Science Center conducted the study.

A version of this article first appeared on WebMD.com.

The effectiveness of COVID-19 vaccines produced by Pfizer/BioNTech, Moderna, and Johnson & Johnson dropped dramatically as the Delta variant swept the United States, a study of almost 800,000 veterans found.

The study, published in the journal Science ., says the three vaccines offered about the same protection against the virus in March, when the Delta variant was first detected in the United States, but that changed 6 months later.

The Moderna two-dose vaccine went from being 89% effective in March to 58% effective in September, according to a story about the study in theLos Angeles Times.

Meanwhile, the Pfizer/BioNTech vaccine went from being 87% effective to 45% effective over the same time period.

The Johnson & Johnson vaccine showed the biggest drop -- from 86% effectiveness to 13% over those 6 months.

“In summary, although vaccination remains protective against SARS-CoV-2 infection, protection waned as the Delta variant emerged in the U.S., and this decline did not differ by age,” the study said.

The three vaccines also lost effectiveness in the ability to protect against death in veterans 65 and over after only 3 months, the Los Angeles Times reported.

Compared to unvaccinated veterans in that age group, veterans who got the Moderna vaccine and had a breakthrough case were 76% less likely to die of COVID-19 by July.

The protection was 70% for Pfizer/BioNTech vaccine recipients and 52% for J&J vaccine recipients for the same age group, compared to unvaccinated veterans, according to the newspaper.

For veterans under 65, the protectiveness against a fatal case of COVID was 84% for Pfizer/BioNTech recipients, 82% for Moderna recipients, and 73% for J&J recipients, compared to unvaccinated veterans in that age group.

The study confirms the need for booster vaccines and protective measures such as vaccine passports, vaccine mandates, masking, hand-washing, and social distancing, the researchers said.

Of the veterans studied, about 500,000 were vaccinated and 300,000 were not. Researchers noted that the study population had 6 times as many men as women. About 48% of the study group was 65 or older, 29% was 50-64, while 24% was under 50.

Researchers from the Public Health Institute in Oakland, the Veterans Affairs Medical Center in San Francisco, and the University of Texas Health Science Center conducted the study.

A version of this article first appeared on WebMD.com.

A new systematic review and meta-analysis finds that unvaccinated cancer patients who contracted COVID-19 last year, were more than two times more likely – than people without cancer – to develop a case of COVID-19 so severe it required hospitalization in an intensive care unit.

“Our study provides the most precise measure to date of the effect of COVID-19 in cancer patients,” wrote researchers who were led by Paolo Boffetta, MD, MPH, a specialist in population science with the Stony Brook Cancer Center in New York.

Dr. Boffetta and colleagues also found that patients with hematologic neoplasms had a higher mortality rate from COVID-19 comparable to that of all cancers combined.

Cancer patients have long been considered to be among those patients who are at high risk of developing COVID-19, and if they contract the disease, they are at high risk of having poor outcomes. Other high-risk patients include those with hypertension, diabetes, chronic kidney disease, or COPD, or the elderly. But how high the risk of developing severe COVID-19 disease is for cancer patients hasn’t yet been documented on a wide scale.

The study, which was made available as a preprint on medRxiv on Oct. 23, is based on an analysis of COVID-19 cases that were documented in 35 reviews, meta-analyses, case reports, and studies indexed in PubMed from authors in North America, Europe, and Asia.

In this study, the pooled odds ratio for mortality for all patients with any cancer was 2.32 (95% confidence interval, 1.82-2.94; 24 studies). For ICU admission, the odds ratio was 2.39 (95% CI, 1.90-3.02; I2 0.0%; 5 studies). And, for disease severity or hospitalization, it was 2.08 (95% CI, 1.60-2.72; I2 92.1%; 15 studies). The pooled mortality odds ratio for hematologic neoplasms was 2.14 (95% CI, 1.87-2.44; I2 20.8%; 8 studies).

Their findings, which have not yet been peer reviewed, confirmed the results of a similar analysis from China published as a preprint in May 2020. The analysis included 181,323 patients (23,736 cancer patients) from 26 studies reported an odds ratio of 2.54 (95% CI, 1.47-4.42). “Cancer patients with COVID-19 have an increased likelihood of death compared to non-cancer COVID-19 patients,” Venkatesulu et al. wrote. And a systematic review and meta-analysis of five studies of 2,619 patients published in October 2020 in Medicine also found a significantly higher risk of death from COVID-19 among cancer patients (odds ratio, 2.63; 95% confidence interval, 1.14-6.06; P = .023; I2 = 26.4%).

Fakih et al., writing in the journal Hematology/Oncology and Stem Cell Therapy conducted a meta-analysis early last year finding a threefold increase for admission to the intensive care unit, an almost fourfold increase for a severe SARS-CoV-2 infection, and a fivefold increase for being intubated.

The three studies show that mortality rates were higher early in the pandemic “when diagnosis and treatment for SARS-CoV-2 might have been delayed, resulting in higher death rate,” Boffetta et al. wrote, adding that their analysis showed only a twofold increase most likely because it was a year-long analysis.

“Future studies will be able to better analyze this association for the different subtypes of cancer. Furthermore, they will eventually be able to evaluate whether the difference among vaccinated population is reduced,” Boffetta et al. wrote.

The authors noted several limitations for the study, including the fact that many of the studies included in the analysis did not include sex, age, comorbidities, and therapy. Nor were the authors able to analyze specific cancers other than hematologic neoplasms.

The authors declared no conflicts of interest.

A new systematic review and meta-analysis finds that unvaccinated cancer patients who contracted COVID-19 last year, were more than two times more likely – than people without cancer – to develop a case of COVID-19 so severe it required hospitalization in an intensive care unit.

“Our study provides the most precise measure to date of the effect of COVID-19 in cancer patients,” wrote researchers who were led by Paolo Boffetta, MD, MPH, a specialist in population science with the Stony Brook Cancer Center in New York.

Dr. Boffetta and colleagues also found that patients with hematologic neoplasms had a higher mortality rate from COVID-19 comparable to that of all cancers combined.

Cancer patients have long been considered to be among those patients who are at high risk of developing COVID-19, and if they contract the disease, they are at high risk of having poor outcomes. Other high-risk patients include those with hypertension, diabetes, chronic kidney disease, or COPD, or the elderly. But how high the risk of developing severe COVID-19 disease is for cancer patients hasn’t yet been documented on a wide scale.

The study, which was made available as a preprint on medRxiv on Oct. 23, is based on an analysis of COVID-19 cases that were documented in 35 reviews, meta-analyses, case reports, and studies indexed in PubMed from authors in North America, Europe, and Asia.

In this study, the pooled odds ratio for mortality for all patients with any cancer was 2.32 (95% confidence interval, 1.82-2.94; 24 studies). For ICU admission, the odds ratio was 2.39 (95% CI, 1.90-3.02; I2 0.0%; 5 studies). And, for disease severity or hospitalization, it was 2.08 (95% CI, 1.60-2.72; I2 92.1%; 15 studies). The pooled mortality odds ratio for hematologic neoplasms was 2.14 (95% CI, 1.87-2.44; I2 20.8%; 8 studies).

Their findings, which have not yet been peer reviewed, confirmed the results of a similar analysis from China published as a preprint in May 2020. The analysis included 181,323 patients (23,736 cancer patients) from 26 studies reported an odds ratio of 2.54 (95% CI, 1.47-4.42). “Cancer patients with COVID-19 have an increased likelihood of death compared to non-cancer COVID-19 patients,” Venkatesulu et al. wrote. And a systematic review and meta-analysis of five studies of 2,619 patients published in October 2020 in Medicine also found a significantly higher risk of death from COVID-19 among cancer patients (odds ratio, 2.63; 95% confidence interval, 1.14-6.06; P = .023; I2 = 26.4%).

Fakih et al., writing in the journal Hematology/Oncology and Stem Cell Therapy conducted a meta-analysis early last year finding a threefold increase for admission to the intensive care unit, an almost fourfold increase for a severe SARS-CoV-2 infection, and a fivefold increase for being intubated.

The three studies show that mortality rates were higher early in the pandemic “when diagnosis and treatment for SARS-CoV-2 might have been delayed, resulting in higher death rate,” Boffetta et al. wrote, adding that their analysis showed only a twofold increase most likely because it was a year-long analysis.

“Future studies will be able to better analyze this association for the different subtypes of cancer. Furthermore, they will eventually be able to evaluate whether the difference among vaccinated population is reduced,” Boffetta et al. wrote.

The authors noted several limitations for the study, including the fact that many of the studies included in the analysis did not include sex, age, comorbidities, and therapy. Nor were the authors able to analyze specific cancers other than hematologic neoplasms.

The authors declared no conflicts of interest.

A new systematic review and meta-analysis finds that unvaccinated cancer patients who contracted COVID-19 last year, were more than two times more likely – than people without cancer – to develop a case of COVID-19 so severe it required hospitalization in an intensive care unit.

“Our study provides the most precise measure to date of the effect of COVID-19 in cancer patients,” wrote researchers who were led by Paolo Boffetta, MD, MPH, a specialist in population science with the Stony Brook Cancer Center in New York.

Dr. Boffetta and colleagues also found that patients with hematologic neoplasms had a higher mortality rate from COVID-19 comparable to that of all cancers combined.

Cancer patients have long been considered to be among those patients who are at high risk of developing COVID-19, and if they contract the disease, they are at high risk of having poor outcomes. Other high-risk patients include those with hypertension, diabetes, chronic kidney disease, or COPD, or the elderly. But how high the risk of developing severe COVID-19 disease is for cancer patients hasn’t yet been documented on a wide scale.

The study, which was made available as a preprint on medRxiv on Oct. 23, is based on an analysis of COVID-19 cases that were documented in 35 reviews, meta-analyses, case reports, and studies indexed in PubMed from authors in North America, Europe, and Asia.

In this study, the pooled odds ratio for mortality for all patients with any cancer was 2.32 (95% confidence interval, 1.82-2.94; 24 studies). For ICU admission, the odds ratio was 2.39 (95% CI, 1.90-3.02; I2 0.0%; 5 studies). And, for disease severity or hospitalization, it was 2.08 (95% CI, 1.60-2.72; I2 92.1%; 15 studies). The pooled mortality odds ratio for hematologic neoplasms was 2.14 (95% CI, 1.87-2.44; I2 20.8%; 8 studies).

Their findings, which have not yet been peer reviewed, confirmed the results of a similar analysis from China published as a preprint in May 2020. The analysis included 181,323 patients (23,736 cancer patients) from 26 studies reported an odds ratio of 2.54 (95% CI, 1.47-4.42). “Cancer patients with COVID-19 have an increased likelihood of death compared to non-cancer COVID-19 patients,” Venkatesulu et al. wrote. And a systematic review and meta-analysis of five studies of 2,619 patients published in October 2020 in Medicine also found a significantly higher risk of death from COVID-19 among cancer patients (odds ratio, 2.63; 95% confidence interval, 1.14-6.06; P = .023; I2 = 26.4%).

Fakih et al., writing in the journal Hematology/Oncology and Stem Cell Therapy conducted a meta-analysis early last year finding a threefold increase for admission to the intensive care unit, an almost fourfold increase for a severe SARS-CoV-2 infection, and a fivefold increase for being intubated.

The three studies show that mortality rates were higher early in the pandemic “when diagnosis and treatment for SARS-CoV-2 might have been delayed, resulting in higher death rate,” Boffetta et al. wrote, adding that their analysis showed only a twofold increase most likely because it was a year-long analysis.

“Future studies will be able to better analyze this association for the different subtypes of cancer. Furthermore, they will eventually be able to evaluate whether the difference among vaccinated population is reduced,” Boffetta et al. wrote.

The authors noted several limitations for the study, including the fact that many of the studies included in the analysis did not include sex, age, comorbidities, and therapy. Nor were the authors able to analyze specific cancers other than hematologic neoplasms.

The authors declared no conflicts of interest.

Mary Lou McMaster, MD, has spent her entire career at the National Cancer Institute (NCI) searching for the genetic underpinnings that give rise to Waldenstrom's macroglobulinemia (WM). 
After searching for decades, she has yet to uncover a "smoking gun," though a few tantalizing clues have emerged along the way. 
"Our questions are pretty basic: Why are some people more susceptible to developing WM, and why does WM sometimes cluster in families?" she explained. It turns out that the answers are not at all simple. 
Dr. McMaster described some of the clues that her team at the Clinical Genetics Branch of the NCI has unearthed in a presentation at the recent International Waldenstrom's Macroglobulinemia Foundation (IWMF) 2021 Virtual Educational Forum. 
Commenting after the presentation, Steven Treon, MD, PhD, professor of medicine, Harvard Medical School, Boston, who is collaborating with Dr. McMaster on this work, said: "From these familial studies, we can learn how familial genomics may give us insights into disease prevention and treatment." 

Identifying affected families  

Work began in 2001 to identify families in which two or more family members had been diagnosed with WM or in which there was one patient with WM and at least one other relative with a related B-cell cancer, such as chronic lymphocytic leukemia. 
For a frame of reference, they enrolled some families with only one member with WM and in which there was no known family history of the disease. 
"Overall, we have learned that familial WM is a rare disease but not nearly as rare as we first thought," Dr. McMaster said. 
For example, in a referral hospital setting, 5% of WM patients will report having a family member with the same disorder, and up to 20% of WM patients report having a family member with a related but different B-cell cancer, she noted. 
NCI researchers also discovered that environmental factors contribute to the development of WM. Notable chemical or occupational exposures include exposures to pesticides, herbicides, and fertilizers. Infections and autoimmune disease are additional factors. 
"This was not a surprise," Dr. McMaster commented regarding the role of occupational exposures. The research community has known for decades that a "lymphoma belt" cuts through the Midwest farming states. 
Focusing on genetic susceptibility, Dr. McMaster and colleagues first tried to identify a rare germline variant that can be passed down to offspring and that might confer high risk for the disease. 
"We used our high-risk families to study these types of changes, although they may be modified by other genes and environmental factors," Dr. McMaster explained. 
Much to their collective disappointment, the research team has been unable to identify any rare germline variant that could account for WM in many families. What they did find were many small changes in genes that are known to be important in B-cell development and function, but all of those would lead to only a small increase in WM risk. 
"What is holding us back is that, so far, we are not seeing the same gene affected in more than one family, so this suggests to us either that this is not the mechanism behind the development of WM in families, or we have an unfortunate situation where each family is going to have a genetic change that is private to that family and which is not found in other families," Dr. McMaster acknowledged. 

Sheer difficulty  

Given the difficulty of determining whether these small genetic changes had any detrimental functional effect in each and every family with a member who had WM, Dr. McMaster and colleagues have now turned their attention to genes that exert only a small effect on disease risk. 
"Here, we focused on specific genes that we knew were important in the function of the immune system," she explained. "We did find a few genes that may contribute to risk, but those have not yet been confirmed by us or others, and we cannot say they are causative without that confirmation," she said. 
The team has gone on to scan the highway of our genetic material so as to isolate genetic "mile markers." They then examine the area around a particular marker that they suspect contains genes that may be involved in WM. 
One study they conducted involved a cohort of 217 patients with WM in which numerous family members had WM and so was enriched with susceptibility genes. A second cohort comprised 312 WM patients in which there were few WM cases among family members. Both of these cohorts were compared with a group of healthy control persons. 
From these genome studies, "we found there are at least two regions of the genome that can contribute to WM susceptibility, the largest effect being on the short arm of chromosome 6, and the other on the long arm of chromosome 14," Dr. McMaster reported. Dr. McMaster feels that there are probably more regions on the genome that also contribute to WM, although they do not yet understand how these regions contribute to susceptibility. 
"It's more evidence that WM likely results from a combination of events rather than one single gene variant," she observed. Dr. McMaster and colleagues are now collaborating with a large consortium of WM researchers to confirm and extend their findings. Plans are underway to analyze data from approximately 1,350 WM patients and more than 20,000 control persons within the next year. 
"Our hope is that we will confirm our original findings and, because we now have a much larger sample, we will be able to discover additional regions of the genome that are contributing to susceptibility," Dr. McMaster said. 
"A single gene is not likely to account for all WM, as we've looked carefully and others have looked too," she commented. 
"So the risk for WM depends on a combination of genes and environmental exposures and possibly lifestyle factors as well, although we still estimate that approximately 25% of the heritability of WM can be attributed to these kinds of genetic changes," Dr. McMaster predicted. 
Dr. McMaster has disclosed no relevant financial relationships. Dr. Treon has served as a director, officer, partner, employee, advisor, consultant, or trustee for Janssen, Pfizer, PCYC, and BioGene.  

A version of this article first appeared on Medscape.com

Mary Lou McMaster, MD, has spent her entire career at the National Cancer Institute (NCI) searching for the genetic underpinnings that give rise to Waldenstrom's macroglobulinemia (WM). 
After searching for decades, she has yet to uncover a "smoking gun," though a few tantalizing clues have emerged along the way. 
"Our questions are pretty basic: Why are some people more susceptible to developing WM, and why does WM sometimes cluster in families?" she explained. It turns out that the answers are not at all simple. 
Dr. McMaster described some of the clues that her team at the Clinical Genetics Branch of the NCI has unearthed in a presentation at the recent International Waldenstrom's Macroglobulinemia Foundation (IWMF) 2021 Virtual Educational Forum. 
Commenting after the presentation, Steven Treon, MD, PhD, professor of medicine, Harvard Medical School, Boston, who is collaborating with Dr. McMaster on this work, said: "From these familial studies, we can learn how familial genomics may give us insights into disease prevention and treatment." 

Identifying affected families  

Work began in 2001 to identify families in which two or more family members had been diagnosed with WM or in which there was one patient with WM and at least one other relative with a related B-cell cancer, such as chronic lymphocytic leukemia. 
For a frame of reference, they enrolled some families with only one member with WM and in which there was no known family history of the disease. 
"Overall, we have learned that familial WM is a rare disease but not nearly as rare as we first thought," Dr. McMaster said. 
For example, in a referral hospital setting, 5% of WM patients will report having a family member with the same disorder, and up to 20% of WM patients report having a family member with a related but different B-cell cancer, she noted. 
NCI researchers also discovered that environmental factors contribute to the development of WM. Notable chemical or occupational exposures include exposures to pesticides, herbicides, and fertilizers. Infections and autoimmune disease are additional factors. 
"This was not a surprise," Dr. McMaster commented regarding the role of occupational exposures. The research community has known for decades that a "lymphoma belt" cuts through the Midwest farming states. 
Focusing on genetic susceptibility, Dr. McMaster and colleagues first tried to identify a rare germline variant that can be passed down to offspring and that might confer high risk for the disease. 
"We used our high-risk families to study these types of changes, although they may be modified by other genes and environmental factors," Dr. McMaster explained. 
Much to their collective disappointment, the research team has been unable to identify any rare germline variant that could account for WM in many families. What they did find were many small changes in genes that are known to be important in B-cell development and function, but all of those would lead to only a small increase in WM risk. 
"What is holding us back is that, so far, we are not seeing the same gene affected in more than one family, so this suggests to us either that this is not the mechanism behind the development of WM in families, or we have an unfortunate situation where each family is going to have a genetic change that is private to that family and which is not found in other families," Dr. McMaster acknowledged. 

Sheer difficulty  

Given the difficulty of determining whether these small genetic changes had any detrimental functional effect in each and every family with a member who had WM, Dr. McMaster and colleagues have now turned their attention to genes that exert only a small effect on disease risk. 
"Here, we focused on specific genes that we knew were important in the function of the immune system," she explained. "We did find a few genes that may contribute to risk, but those have not yet been confirmed by us or others, and we cannot say they are causative without that confirmation," she said. 
The team has gone on to scan the highway of our genetic material so as to isolate genetic "mile markers." They then examine the area around a particular marker that they suspect contains genes that may be involved in WM. 
One study they conducted involved a cohort of 217 patients with WM in which numerous family members had WM and so was enriched with susceptibility genes. A second cohort comprised 312 WM patients in which there were few WM cases among family members. Both of these cohorts were compared with a group of healthy control persons. 
From these genome studies, "we found there are at least two regions of the genome that can contribute to WM susceptibility, the largest effect being on the short arm of chromosome 6, and the other on the long arm of chromosome 14," Dr. McMaster reported. Dr. McMaster feels that there are probably more regions on the genome that also contribute to WM, although they do not yet understand how these regions contribute to susceptibility. 
"It's more evidence that WM likely results from a combination of events rather than one single gene variant," she observed. Dr. McMaster and colleagues are now collaborating with a large consortium of WM researchers to confirm and extend their findings. Plans are underway to analyze data from approximately 1,350 WM patients and more than 20,000 control persons within the next year. 
"Our hope is that we will confirm our original findings and, because we now have a much larger sample, we will be able to discover additional regions of the genome that are contributing to susceptibility," Dr. McMaster said. 
"A single gene is not likely to account for all WM, as we've looked carefully and others have looked too," she commented. 
"So the risk for WM depends on a combination of genes and environmental exposures and possibly lifestyle factors as well, although we still estimate that approximately 25% of the heritability of WM can be attributed to these kinds of genetic changes," Dr. McMaster predicted. 
Dr. McMaster has disclosed no relevant financial relationships. Dr. Treon has served as a director, officer, partner, employee, advisor, consultant, or trustee for Janssen, Pfizer, PCYC, and BioGene.  

A version of this article first appeared on Medscape.com

Mary Lou McMaster, MD, has spent her entire career at the National Cancer Institute (NCI) searching for the genetic underpinnings that give rise to Waldenstrom's macroglobulinemia (WM). 
After searching for decades, she has yet to uncover a "smoking gun," though a few tantalizing clues have emerged along the way. 
"Our questions are pretty basic: Why are some people more susceptible to developing WM, and why does WM sometimes cluster in families?" she explained. It turns out that the answers are not at all simple. 
Dr. McMaster described some of the clues that her team at the Clinical Genetics Branch of the NCI has unearthed in a presentation at the recent International Waldenstrom's Macroglobulinemia Foundation (IWMF) 2021 Virtual Educational Forum. 
Commenting after the presentation, Steven Treon, MD, PhD, professor of medicine, Harvard Medical School, Boston, who is collaborating with Dr. McMaster on this work, said: "From these familial studies, we can learn how familial genomics may give us insights into disease prevention and treatment." 

Identifying affected families  

Work began in 2001 to identify families in which two or more family members had been diagnosed with WM or in which there was one patient with WM and at least one other relative with a related B-cell cancer, such as chronic lymphocytic leukemia. 
For a frame of reference, they enrolled some families with only one member with WM and in which there was no known family history of the disease. 
"Overall, we have learned that familial WM is a rare disease but not nearly as rare as we first thought," Dr. McMaster said. 
For example, in a referral hospital setting, 5% of WM patients will report having a family member with the same disorder, and up to 20% of WM patients report having a family member with a related but different B-cell cancer, she noted. 
NCI researchers also discovered that environmental factors contribute to the development of WM. Notable chemical or occupational exposures include exposures to pesticides, herbicides, and fertilizers. Infections and autoimmune disease are additional factors. 
"This was not a surprise," Dr. McMaster commented regarding the role of occupational exposures. The research community has known for decades that a "lymphoma belt" cuts through the Midwest farming states. 
Focusing on genetic susceptibility, Dr. McMaster and colleagues first tried to identify a rare germline variant that can be passed down to offspring and that might confer high risk for the disease. 
"We used our high-risk families to study these types of changes, although they may be modified by other genes and environmental factors," Dr. McMaster explained. 
Much to their collective disappointment, the research team has been unable to identify any rare germline variant that could account for WM in many families. What they did find were many small changes in genes that are known to be important in B-cell development and function, but all of those would lead to only a small increase in WM risk. 
"What is holding us back is that, so far, we are not seeing the same gene affected in more than one family, so this suggests to us either that this is not the mechanism behind the development of WM in families, or we have an unfortunate situation where each family is going to have a genetic change that is private to that family and which is not found in other families," Dr. McMaster acknowledged. 

Sheer difficulty  

Given the difficulty of determining whether these small genetic changes had any detrimental functional effect in each and every family with a member who had WM, Dr. McMaster and colleagues have now turned their attention to genes that exert only a small effect on disease risk. 
"Here, we focused on specific genes that we knew were important in the function of the immune system," she explained. "We did find a few genes that may contribute to risk, but those have not yet been confirmed by us or others, and we cannot say they are causative without that confirmation," she said. 
The team has gone on to scan the highway of our genetic material so as to isolate genetic "mile markers." They then examine the area around a particular marker that they suspect contains genes that may be involved in WM. 
One study they conducted involved a cohort of 217 patients with WM in which numerous family members had WM and so was enriched with susceptibility genes. A second cohort comprised 312 WM patients in which there were few WM cases among family members. Both of these cohorts were compared with a group of healthy control persons. 
From these genome studies, "we found there are at least two regions of the genome that can contribute to WM susceptibility, the largest effect being on the short arm of chromosome 6, and the other on the long arm of chromosome 14," Dr. McMaster reported. Dr. McMaster feels that there are probably more regions on the genome that also contribute to WM, although they do not yet understand how these regions contribute to susceptibility. 
"It's more evidence that WM likely results from a combination of events rather than one single gene variant," she observed. Dr. McMaster and colleagues are now collaborating with a large consortium of WM researchers to confirm and extend their findings. Plans are underway to analyze data from approximately 1,350 WM patients and more than 20,000 control persons within the next year. 
"Our hope is that we will confirm our original findings and, because we now have a much larger sample, we will be able to discover additional regions of the genome that are contributing to susceptibility," Dr. McMaster said. 
"A single gene is not likely to account for all WM, as we've looked carefully and others have looked too," she commented. 
"So the risk for WM depends on a combination of genes and environmental exposures and possibly lifestyle factors as well, although we still estimate that approximately 25% of the heritability of WM can be attributed to these kinds of genetic changes," Dr. McMaster predicted. 
Dr. McMaster has disclosed no relevant financial relationships. Dr. Treon has served as a director, officer, partner, employee, advisor, consultant, or trustee for Janssen, Pfizer, PCYC, and BioGene.  

A version of this article first appeared on Medscape.com

For those of us who cannot sit in the sun and fish all day, the next best thing for preventing autoimmune diseases may be supplementation with vitamin D and fish oil-derived omega-3 fatty acids, results of a large prospective randomized trial suggest.

Ziga Plahutar

Among nearly 26,000 adults enrolled in a randomized trial designed primarily to study the effects of vitamin D and omega-3 supplementation on incident cancer and cardiovascular disease, 5 years of vitamin D supplementation was associated with a 22% reduction in risk for confirmed autoimmune diseases, and 5 years of omega-3 fatty acid supplementation was associated with an 18% reduction in confirmed and probable incident autoimmune diseases, reported Karen H. Costenbader, MD, MPH, of Brigham & Women’s Hospital in Boston.

“The clinical importance of these results is very high, given that these are nontoxic, well-tolerated supplements, and that there are no other known effective therapies to reduce the incidence of autoimmune diseases,” she said during the virtual annual meeting of the American College of Rheumatology.

“People do have to take the supplements a long time to start to see the reduction in risk, especially for vitamin D, but they make biological sense, and autoimmune diseases develop slowly over time, so taking it today isn’t going to reduce risk of developing something tomorrow,” Dr. Costenbader said in an interview.

“These supplements have other health benefits. Obviously, fish oil is anti-inflammatory, and vitamin D is good for osteoporosis prevention, especially in our patients who take glucocorticoids. People who are otherwise healthy and have a family history of autoimmune disease might also consider starting to take these supplements,” she said.

After watching her presentation, session co-moderator Gregg Silverman, MD, from the NYU Langone School of Medicine in New York, who was not involved in the study, commented “I’m going to [nutrition store] GNC to get some vitamins.”

When asked for comment, the other session moderator, Tracy Frech, MD, of Vanderbilt University, Nashville, said, “I think Dr. Costenbader’s work is very important and her presentation excellent. My current practice is replacement of vitamin D in all autoimmune disease patients with low levels and per bone health guidelines. Additionally, I discuss omega-3 supplementation with Sjögren’s [syndrome] patients as a consideration.”

Evidence base

Dr. Costenbader noted that in a 2013 observational study from France, vitamin D derived through ultraviolet (UV) light exposure was associated with a lower risk for incident Crohn’s disease but not ulcerative colitis, and in two analyses of data in 2014 from the Nurses’ Health Study, both high plasma levels of 25-OH vitamin D and geographic residence in areas of high UV exposure were associated with a decreased incidence of rheumatoid arthritis (RA).

Other observational studies have supported omega-3 fatty acids for their anti-inflammatory properties, including a 2005 Danish prospective cohort study showing a lower risk for RA in participants who reported higher levels of fatty fish intake. In a separate study conducted in 2017, healthy volunteers with higher omega-3 fatty acid/total lipid proportions in red blood cell membranes had a lower prevalence of anti-cyclic citrullinated peptide (anti-CCP) antibodies and rheumatoid factor and a lower incidence of progression to inflammatory arthritis, she said.

Ancillary study

Despite the evidence, however, there have been no prospective randomized trials to test the effects of either vitamin D or omega-3 fatty acid supplementation on the incidence of autoimmune disease over time.

To rectify this, Dr. Costenbader and colleagues piggybacked an ancillary study onto the Vitamin D and Omega-3 Trial (VITAL), which had primary outcomes of cancer and cardiovascular disease incidence.

A total of 25,871 participants were enrolled, including 12,786 men aged 50 and older, and 13,085 women aged 55 and older.

The study had a 2 x 2 factorial design, with patients randomly assigned to vitamin D 2,000 IU/day or placebo, and then further randomized to either 1 g/day omega-3 fatty acids or placebo in both the vitamin D and placebo primary randomization arms.

At baseline 16,956 participants were assayed for 25-OH vitamin D and plasma omega 3 index, the ratio of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) to total fatty acids. Participants self-reported baseline and all incident autoimmune diseases annually, with the reports confirmed by medical record review and disease criteria whenever possible.

Results

At 5 years of follow-up, confirmed incident autoimmune diseases had occurred in 123 patients in the active vitamin D group, compared with 155 in the placebo vitamin D group, translating into a hazard ratio (HR) for vitamin D of 0.78 (P = .045).

In the active omega-3 arm, 130 participants developed an autoimmune disease, compared with 148 in the placebo omega-3 arm, which translated into a nonsignificant HR of 0.85.

There was no statistical interaction between the two supplements. The investigators did observe an interaction between vitamin D and body mass index, with the effect stronger among participants with low BMI (P = .02). There also was an interaction between omega-3 fatty acids with a family history of autoimmune disease (P = .03).

In multivariate analysis adjusted for age, sex, race, and other supplement arm, vitamin D alone was associated with an HR for incident autoimmune disease of 0.68 (P = .02), omega-3 alone was associated with a nonsignificant HR of 0.74, and the combination was associated with an HR of 0.69 (P = .03).

Dr. Costenbader and colleagues acknowledged that the study was limited by the lack of a high-risk or nutritionally-deficient population, where the effects of supplementation might be larger; the restriction of the sample to older adults; and to the difficulty of confirming incident autoimmune thyroid disease from patient reports.

Cheryl Koehn, an arthritis patient advocate from Vancouver, Canada, who was not involved in the study, commented in the “chat” section of the presentation that her rheumatologist “has recommended vitamin D for years now. Says basically everyone north of Boston is vitamin D deficient. I take 1,000 IU per day. Been taking it for years.” Ms. Koehn is the founder and president of Arthritis Consumer Experts, a website that provides education to those with arthritis.

“Agreed. I tell every patient to take vitamin D supplement,” commented Fatma Dedeoglu, MD, a rheumatologist at Boston Children’s Hospital.



A version of this article first appeared on Medscape.com.

For those of us who cannot sit in the sun and fish all day, the next best thing for preventing autoimmune diseases may be supplementation with vitamin D and fish oil-derived omega-3 fatty acids, results of a large prospective randomized trial suggest.

Ziga Plahutar

Among nearly 26,000 adults enrolled in a randomized trial designed primarily to study the effects of vitamin D and omega-3 supplementation on incident cancer and cardiovascular disease, 5 years of vitamin D supplementation was associated with a 22% reduction in risk for confirmed autoimmune diseases, and 5 years of omega-3 fatty acid supplementation was associated with an 18% reduction in confirmed and probable incident autoimmune diseases, reported Karen H. Costenbader, MD, MPH, of Brigham & Women’s Hospital in Boston.

“The clinical importance of these results is very high, given that these are nontoxic, well-tolerated supplements, and that there are no other known effective therapies to reduce the incidence of autoimmune diseases,” she said during the virtual annual meeting of the American College of Rheumatology.

“People do have to take the supplements a long time to start to see the reduction in risk, especially for vitamin D, but they make biological sense, and autoimmune diseases develop slowly over time, so taking it today isn’t going to reduce risk of developing something tomorrow,” Dr. Costenbader said in an interview.

“These supplements have other health benefits. Obviously, fish oil is anti-inflammatory, and vitamin D is good for osteoporosis prevention, especially in our patients who take glucocorticoids. People who are otherwise healthy and have a family history of autoimmune disease might also consider starting to take these supplements,” she said.

After watching her presentation, session co-moderator Gregg Silverman, MD, from the NYU Langone School of Medicine in New York, who was not involved in the study, commented “I’m going to [nutrition store] GNC to get some vitamins.”

When asked for comment, the other session moderator, Tracy Frech, MD, of Vanderbilt University, Nashville, said, “I think Dr. Costenbader’s work is very important and her presentation excellent. My current practice is replacement of vitamin D in all autoimmune disease patients with low levels and per bone health guidelines. Additionally, I discuss omega-3 supplementation with Sjögren’s [syndrome] patients as a consideration.”

Evidence base

Dr. Costenbader noted that in a 2013 observational study from France, vitamin D derived through ultraviolet (UV) light exposure was associated with a lower risk for incident Crohn’s disease but not ulcerative colitis, and in two analyses of data in 2014 from the Nurses’ Health Study, both high plasma levels of 25-OH vitamin D and geographic residence in areas of high UV exposure were associated with a decreased incidence of rheumatoid arthritis (RA).

Other observational studies have supported omega-3 fatty acids for their anti-inflammatory properties, including a 2005 Danish prospective cohort study showing a lower risk for RA in participants who reported higher levels of fatty fish intake. In a separate study conducted in 2017, healthy volunteers with higher omega-3 fatty acid/total lipid proportions in red blood cell membranes had a lower prevalence of anti-cyclic citrullinated peptide (anti-CCP) antibodies and rheumatoid factor and a lower incidence of progression to inflammatory arthritis, she said.

Ancillary study

Despite the evidence, however, there have been no prospective randomized trials to test the effects of either vitamin D or omega-3 fatty acid supplementation on the incidence of autoimmune disease over time.

To rectify this, Dr. Costenbader and colleagues piggybacked an ancillary study onto the Vitamin D and Omega-3 Trial (VITAL), which had primary outcomes of cancer and cardiovascular disease incidence.

A total of 25,871 participants were enrolled, including 12,786 men aged 50 and older, and 13,085 women aged 55 and older.

The study had a 2 x 2 factorial design, with patients randomly assigned to vitamin D 2,000 IU/day or placebo, and then further randomized to either 1 g/day omega-3 fatty acids or placebo in both the vitamin D and placebo primary randomization arms.

At baseline 16,956 participants were assayed for 25-OH vitamin D and plasma omega 3 index, the ratio of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) to total fatty acids. Participants self-reported baseline and all incident autoimmune diseases annually, with the reports confirmed by medical record review and disease criteria whenever possible.

Results

At 5 years of follow-up, confirmed incident autoimmune diseases had occurred in 123 patients in the active vitamin D group, compared with 155 in the placebo vitamin D group, translating into a hazard ratio (HR) for vitamin D of 0.78 (P = .045).

In the active omega-3 arm, 130 participants developed an autoimmune disease, compared with 148 in the placebo omega-3 arm, which translated into a nonsignificant HR of 0.85.

There was no statistical interaction between the two supplements. The investigators did observe an interaction between vitamin D and body mass index, with the effect stronger among participants with low BMI (P = .02). There also was an interaction between omega-3 fatty acids with a family history of autoimmune disease (P = .03).

In multivariate analysis adjusted for age, sex, race, and other supplement arm, vitamin D alone was associated with an HR for incident autoimmune disease of 0.68 (P = .02), omega-3 alone was associated with a nonsignificant HR of 0.74, and the combination was associated with an HR of 0.69 (P = .03).

Dr. Costenbader and colleagues acknowledged that the study was limited by the lack of a high-risk or nutritionally-deficient population, where the effects of supplementation might be larger; the restriction of the sample to older adults; and to the difficulty of confirming incident autoimmune thyroid disease from patient reports.

Cheryl Koehn, an arthritis patient advocate from Vancouver, Canada, who was not involved in the study, commented in the “chat” section of the presentation that her rheumatologist “has recommended vitamin D for years now. Says basically everyone north of Boston is vitamin D deficient. I take 1,000 IU per day. Been taking it for years.” Ms. Koehn is the founder and president of Arthritis Consumer Experts, a website that provides education to those with arthritis.

“Agreed. I tell every patient to take vitamin D supplement,” commented Fatma Dedeoglu, MD, a rheumatologist at Boston Children’s Hospital.



A version of this article first appeared on Medscape.com.

For those of us who cannot sit in the sun and fish all day, the next best thing for preventing autoimmune diseases may be supplementation with vitamin D and fish oil-derived omega-3 fatty acids, results of a large prospective randomized trial suggest.

Ziga Plahutar

Among nearly 26,000 adults enrolled in a randomized trial designed primarily to study the effects of vitamin D and omega-3 supplementation on incident cancer and cardiovascular disease, 5 years of vitamin D supplementation was associated with a 22% reduction in risk for confirmed autoimmune diseases, and 5 years of omega-3 fatty acid supplementation was associated with an 18% reduction in confirmed and probable incident autoimmune diseases, reported Karen H. Costenbader, MD, MPH, of Brigham & Women’s Hospital in Boston.

“The clinical importance of these results is very high, given that these are nontoxic, well-tolerated supplements, and that there are no other known effective therapies to reduce the incidence of autoimmune diseases,” she said during the virtual annual meeting of the American College of Rheumatology.

“People do have to take the supplements a long time to start to see the reduction in risk, especially for vitamin D, but they make biological sense, and autoimmune diseases develop slowly over time, so taking it today isn’t going to reduce risk of developing something tomorrow,” Dr. Costenbader said in an interview.

“These supplements have other health benefits. Obviously, fish oil is anti-inflammatory, and vitamin D is good for osteoporosis prevention, especially in our patients who take glucocorticoids. People who are otherwise healthy and have a family history of autoimmune disease might also consider starting to take these supplements,” she said.

After watching her presentation, session co-moderator Gregg Silverman, MD, from the NYU Langone School of Medicine in New York, who was not involved in the study, commented “I’m going to [nutrition store] GNC to get some vitamins.”

When asked for comment, the other session moderator, Tracy Frech, MD, of Vanderbilt University, Nashville, said, “I think Dr. Costenbader’s work is very important and her presentation excellent. My current practice is replacement of vitamin D in all autoimmune disease patients with low levels and per bone health guidelines. Additionally, I discuss omega-3 supplementation with Sjögren’s [syndrome] patients as a consideration.”

Evidence base

Dr. Costenbader noted that in a 2013 observational study from France, vitamin D derived through ultraviolet (UV) light exposure was associated with a lower risk for incident Crohn’s disease but not ulcerative colitis, and in two analyses of data in 2014 from the Nurses’ Health Study, both high plasma levels of 25-OH vitamin D and geographic residence in areas of high UV exposure were associated with a decreased incidence of rheumatoid arthritis (RA).

Other observational studies have supported omega-3 fatty acids for their anti-inflammatory properties, including a 2005 Danish prospective cohort study showing a lower risk for RA in participants who reported higher levels of fatty fish intake. In a separate study conducted in 2017, healthy volunteers with higher omega-3 fatty acid/total lipid proportions in red blood cell membranes had a lower prevalence of anti-cyclic citrullinated peptide (anti-CCP) antibodies and rheumatoid factor and a lower incidence of progression to inflammatory arthritis, she said.

Ancillary study

Despite the evidence, however, there have been no prospective randomized trials to test the effects of either vitamin D or omega-3 fatty acid supplementation on the incidence of autoimmune disease over time.

To rectify this, Dr. Costenbader and colleagues piggybacked an ancillary study onto the Vitamin D and Omega-3 Trial (VITAL), which had primary outcomes of cancer and cardiovascular disease incidence.

A total of 25,871 participants were enrolled, including 12,786 men aged 50 and older, and 13,085 women aged 55 and older.

The study had a 2 x 2 factorial design, with patients randomly assigned to vitamin D 2,000 IU/day or placebo, and then further randomized to either 1 g/day omega-3 fatty acids or placebo in both the vitamin D and placebo primary randomization arms.

At baseline 16,956 participants were assayed for 25-OH vitamin D and plasma omega 3 index, the ratio of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) to total fatty acids. Participants self-reported baseline and all incident autoimmune diseases annually, with the reports confirmed by medical record review and disease criteria whenever possible.

Results

At 5 years of follow-up, confirmed incident autoimmune diseases had occurred in 123 patients in the active vitamin D group, compared with 155 in the placebo vitamin D group, translating into a hazard ratio (HR) for vitamin D of 0.78 (P = .045).

In the active omega-3 arm, 130 participants developed an autoimmune disease, compared with 148 in the placebo omega-3 arm, which translated into a nonsignificant HR of 0.85.

There was no statistical interaction between the two supplements. The investigators did observe an interaction between vitamin D and body mass index, with the effect stronger among participants with low BMI (P = .02). There also was an interaction between omega-3 fatty acids with a family history of autoimmune disease (P = .03).

In multivariate analysis adjusted for age, sex, race, and other supplement arm, vitamin D alone was associated with an HR for incident autoimmune disease of 0.68 (P = .02), omega-3 alone was associated with a nonsignificant HR of 0.74, and the combination was associated with an HR of 0.69 (P = .03).

Dr. Costenbader and colleagues acknowledged that the study was limited by the lack of a high-risk or nutritionally-deficient population, where the effects of supplementation might be larger; the restriction of the sample to older adults; and to the difficulty of confirming incident autoimmune thyroid disease from patient reports.

Cheryl Koehn, an arthritis patient advocate from Vancouver, Canada, who was not involved in the study, commented in the “chat” section of the presentation that her rheumatologist “has recommended vitamin D for years now. Says basically everyone north of Boston is vitamin D deficient. I take 1,000 IU per day. Been taking it for years.” Ms. Koehn is the founder and president of Arthritis Consumer Experts, a website that provides education to those with arthritis.

“Agreed. I tell every patient to take vitamin D supplement,” commented Fatma Dedeoglu, MD, a rheumatologist at Boston Children’s Hospital.



A version of this article first appeared on Medscape.com.

Artificial Intelligence (AI) was first described in 1956 and refers to machines having the ability to learn as they receive and process information, resulting in the ability to “think” like humans.¹ AI’s impact in medicine is increasing; currently, at least 29 AI medical devices and algorithms are approved by the US Food and Drug Administration (FDA) in a variety of areas, including radiograph interpretation, managing glucose levels in patients with diabetes mellitus, analyzing electrocardiograms (ECGs), and diagnosing sleep disorders among others.² Significantly, in 2020, the Centers for Medicare and Medicaid Services (CMS) announced the first reimbursement to hospitals for an AI platform, a model for early detection of strokes.³ AI is rapidly becoming an integral part of health care, and its role will only increase in the future (Table).

As knowledge in medicine is expanding exponentially, AI has great potential to assist with handling complex patient care data. The concept of exponential growth is not a natural one. As Bini described, with exponential growth the volume of knowledge amassed over the past 10 years will now occur in perhaps only 1 year.¹ Likewise, equivalent advances over the past year may take just a few months. This phenomenon is partly due to the law of accelerating returns, which states that advances feed on themselves, continually increasing the rate of further advances.⁴ The volume of medical data doubles every 2 to 5 years.⁵ Fortunately, the field of AI is growing exponentially as well and can help health care practitioners (HCPs) keep pace, allowing the continued delivery of effective health care.

In this report, we review common terminology, principles, and general applications of AI, followed by current and potential applications of AI for selected medical specialties. Finally, we discuss AI’s future in health care, along with potential risks and pitfalls.

AI Overview

AI refers to machine programs that can “learn” or think based on past experiences. This functionality contrasts with simple rules-based programming available to health care for years. An example of rules-based programming is the warfarindosing.org website developed by Barnes-Jewish Hospital at Washington University Medical Center, which guides initial warfarin dosing.^6,7 The prescriber inputs detailed patient information, including age, sex, height, weight, tobacco history, medications, laboratory results, and genotype if available. The application then calculates recommended warfarin dosing regimens to avoid over- or underanticoagulation. While the dosing algorithm may be complex, it depends entirely on preprogrammed rules. The program does not learn to reach its conclusions and recommendations from patient data.

In contrast, one of the most common subsets of AI is machine learning (ML). ML describes a program that “learns from experience and improves its performance as it learns.”¹ With ML, the computer is initially provided with a training data set—data with known outcomes or labels. Because the initial data are input from known samples, this type of AI is known as supervised learning.^8-10 As an example, we recently reported using ML to diagnose various types of cancer from pathology slides.¹¹ In one experiment, we captured images of colon adenocarcinoma and normal colon (these 2 groups represent the training data set). Unlike traditional programming, we did not define characteristics that would differentiate colon cancer from normal; rather, the machine learned these characteristics independently by assessing the labeled images provided. A second data set (the validation data set) was used to evaluate the program and fine-tune the ML training model’s parameters. Finally, the program was presented with new images of cancer and normal cases for final assessment of accuracy (test data set). Our program learned to recognize differences from the images provided and was able to differentiate normal and cancer images with > 95% accuracy.

Advances in computer processing have allowed for the development of artificial neural networks (ANNs). While there are several types of ANNs, the most common types used for image classification and segmentation are known as convolutional neural networks (CNNs).^9,12-14 The programs are designed to work similar to the human brain, specifically the visual cortex.^15,16 As data are acquired, they are processed by various layers in the program. Much like neurons in the brain, one layer decides whether to advance information to the next.^13,14 CNNs can be many layers deep, leading to the term deep learning: “computational models that are composed of multiple processing layers to learn representations of data with multiple levels of abstraction.”^1,13,17

ANNs can process larger volumes of data. This advance has led to the development of unstructured or unsupervised learning. With this type of learning, imputing defined features (ie, predetermined answers) of the training data set described above is no longer required.^1,8,10,14 The advantage of unsupervised learning is that the program can be presented raw data and extract meaningful interpretation without human input, often with less bias than may exist with supervised learning.^1,18 If shown enough data, the program can extract relevant features to make conclusions independently without predefined definitions, potentially uncovering markers not previously known. For example, several studies have used unsupervised learning to search patient data to assess readmission risks of patients with congestive heart failure.^10,19,20 AI compiled features independently and not previously defined, predicting patients at greater risk for readmission superior to traditional methods.

Health Care Applications

Image analysis has seen the most AI health care applications.^8,15 AI has shown potential in interpreting many types of medical images, including pathology slides, radiographs of various types, retina and other eye scans, and photographs of skin lesions. Many studies have demonstrated that AI can interpret these images as accurately as or even better than experienced clinicians.^9,13,22-29 Studies have suggested AI interpretation of radiographs may better distinguish patients infected with COVID-19 from other causes of pneumonia, and AI interpretation of pathology slides may detect specific genetic mutations not previously identified without additional molecular tests.^{11,14,23,24,30-32}

The second area in which AI can impact health care is improving workflow and efficiency. AI has improved surgery scheduling, saving significant revenue, and decreased patient wait times for appointments.¹ AI can screen and triage radiographs, allowing attention to be directed to critical patients. This use would be valuable in many busy clinical settings, such as the recent COVID-19 pandemic.^8,23 Similarly, AI can screen retina images to prioritize urgent conditions.²⁵ AI has improved pathologists’ efficiency when used to detect breast metastases.³³ Finally, AI may reduce medical errors, thereby ensuring patient safety.^8,9,34

A third health care benefit of AI is in public health and epidemiology. AI can assist with clinical decision-making and diagnoses in low-income countries and areas with limited health care resources and personnel.^25,29 AI can improve identification of infectious outbreaks, such as tuberculosis, malaria, dengue fever, and influenza.^29,35-40 AI has been used to predict transmission patterns of the Zika virus and the current COVID-19 pandemic.^41,42 Applications can stratify the risk of outbreaks based on multiple factors, including age, income, race, atypical geographic clusters, and seasonal factors like rainfall and temperature.^35,36,38,43 AI has been used to assess morbidity and mortality, such as predicting disease severity with malaria and identifying treatment failures in tuberculosis.²⁹

Finally, AI can dramatically impact health care due to processing large data sets or disconnected volumes of patient information—so-called big data.^44-46 An example is the widespread use of electronic health records (EHRs) such as the Computerized Patient Record System used in Veteran Affairs medical centers (VAMCs). Much of patient information exists in written text: HCP notes, laboratory and radiology reports, medication records, etc. Natural language processing (NLP) allows platforms to sort through extensive volumes of data on complex patients at rates much faster than human capability, which has great potential to assist with diagnosis and treatment decisions.⁹

Medical literature is being produced at rates that exceed our ability to digest. More than 200,000 cancer-related articles were published in 2019 alone.¹⁴ NLP capabilities of AI have the potential to rapidly sort through this extensive medical literature and relate specific verbiage in patient records guiding therapy.⁴⁶ IBM Watson, a supercomputer based on ML and NLP, demonstrates this concept with many potential applications, only some of which relate to health care.^1,9 Watson has an oncology component to assimilate multiple aspects of patient care, including clinical notes, pathology results, radiograph findings, staging, and a tumor’s genetic profile. It coordinates these inputs from the EHR and mines medical literature and research databases to recommend treatment options.^1,46 AI can assess and compile far greater patient data and therapeutic options than would be feasible by individual clinicians, thus providing customized patient care.⁴⁷ Watson has partnered with numerous medical centers, including MD Anderson Cancer Center and Memorial Sloan Kettering Cancer Center, with variable success.^44,47-49 While the full potential of Watson appears not yet realized, these AI-driven approaches will likely play an important role in leveraging the hidden value in the expanding volume of health care information.

Medical Specialty Applications

Radiology

Currently > 70% of FDA-approved AI medical devices are in the field of radiology.² Most radiology departments have used AI-friendly digital imaging for years, such as the picture archiving and communication systems used by numerous health care systems, including VAMCs.^2,15 Gray-scale images common in radiology lend themselves to standardization, although AI is not limited to black-and- white image interpretation.¹⁵

An abundance of literature describes plain radiograph interpretation using AI. One FDA-approved platform improved X-ray diagnosis of wrist fractures when used by emergency medicine clinicians.^2,50 AI has been applied to chest X-ray (CXR) interpretation of many conditions, including pneumonia, tuberculosis, malignant lung lesions, and COVID-19.^{23,25,28,44,51-53} For example, Nam and colleagues suggested AI is better at diagnosing malignant pulmonary nodules from CXRs than are trained radiologists.²⁸

In addition to plain radiographs, AI has been applied to many other imaging technologies, including ultrasounds, positron emission tomography, mammograms, computed tomography (CT), and magnetic resonance imaging (MRI).^{15,26,44,48,54-56} A large study demonstrated that ML platforms significantly reduced the time to diagnose intracranial hemorrhages on CT and identified subtle hemorrhages missed by radiologists.⁵⁵ Other studies have claimed that AI programs may be better than radiologists in detecting cancer in screening mammograms, and 3 FDA-approved devices focus on mammogram interpretation.^2,15,54,57 There is also great interest in MRI applications to detect and predict prognosis for breast cancer based on imaging findings.^21,56

Aside from providing accurate diagnoses, other studies focus on AI radiograph interpretation to assist with patient screening, triage, improving time to final diagnosis, providing a rapid “second opinion,” and even monitoring disease progression and offering insights into prognosis.^{8,21,23,52,55,56,58} These features help in busy urban centers but may play an even greater role in areas with limited access to health care or trained specialists such as radiologists.⁵²

Cardiology

Cardiology has the second highest number of FDA-approved AI applications.² Many cardiology AI platforms involve image analysis, as described in several recent reviews.^45,59,60 AI has been applied to echocardiography to measure ejection fractions, detect valvular disease, and assess heart failure from hypertrophic and restrictive cardiomyopathy and amyloidosis.^45,48,59 Applications for cardiac CT scans and CT angiography have successfully quantified both calcified and noncalcified coronary artery plaques and lumen assessments, assessed myocardial perfusion, and performed coronary artery calcium scoring.^45,59,60 Likewise, AI applications for cardiac MRI have been used to quantitate ejection fraction, large vessel flow assessment, and cardiac scar burden.^45,59

For years ECG devices have provided interpretation with limited accuracy using preprogrammed parameters.⁴⁸ However, the application of AI allows ECG interpretation on par with trained cardiologists. Numerous such AI applications exist, and 2 FDA-approved devices perform ECG interpretation.^2,61-64 One of these devices incorporates an AI-powered stethoscope to detect atrial fibrillation and heart murmurs.⁶⁵

Pathology

The advancement of whole slide imaging, wherein entire slides can be scanned and digitized at high speed and resolution, creates great potential for AI applications in pathology.^{12,24,32,33,66} A landmark study demonstrating the potential of AI for assessing whole slide imaging examined sentinel lymph node metastases in patients with breast cancer.²² Multiple algorithms in the study demonstrated that AI was equivalent or better than pathologists in detecting metastases, especially when the pathologists were time-constrained consistent with a normal working environment. Significantly, the most accurate and efficient diagnoses were achieved when the pathologist and AI interpretations were used together.^22,33

AI has shown promise in diagnosing many other entities, including cancers of the prostate (including Gleason scoring), lung, colon, breast, and skin.^{11,12,24,27,32,67} In addition, AI has shown great potential in scoring biomarkers important for prognosis and treatment, such as immunohistochemistry (IHC) labeling of Ki-67 and PD-L1.³² Pathologists can have difficulty classifying certain tumors or determining the site of origin for metastases, often having to rely on IHC with limited success. The unique features of image analysis with AI have the potential to assist in classifying difficult tumors and identifying sites of origin for metastatic disease based on morphology alone.¹¹

Oncology depends heavily on molecular pathology testing to dictate treatment options and determine prognosis. Preliminary studies suggest that AI interpretation alone has the potential to delineate whether certain molecular mutations are present in tumors from various sites.^11,14,24,32 One study combined histology and genomic results for AI interpretation that improved prognostic predictions.⁶⁸ In addition, AI analysis may have potential in predicting tumor recurrence or prognosis based on cellular features, as demonstrated for lung cancer and melanoma.^67,69,70

Ophthalmology

AI applications for ophthalmology have focused on diabetic retinopathy, age-related macular degeneration, glaucoma, retinopathy of prematurity, age-related and congenital cataracts, and retinal vein occlusion.^71-73 Diabetic retinopathy is a leading cause of blindness and has been studied by numerous platforms with good success, most having used color fundus photography.^71,72 One study showed AI could diagnose diabetic retinopathy and diabetic macular edema with specificities similar to ophthalmologists.⁷⁴ In 2018, the FDA approved the AI platform IDx-DR. This diagnostic system classifies retinal images and recommends referral for patients determined to have “more than mild diabetic retinopathy” and reexamination within a year for other patients.^8,75 Significantly, the platform recommendations do not require confirmation by a clinician.⁸

AI has been applied to other modalities in ophthalmology such as optical coherence tomography (OCT) to diagnose retinal disease and to predict appropriate management of congenital cataracts.^25,73,76 For example, an AI application using OCT has been demonstrated to match or exceed the accuracy of retinal experts in diagnosing and triaging patients with a variety of retinal pathologies, including patients needing urgent referrals.⁷⁷

Dermatology

Multiple studies demonstrate AI performs at least equal to experienced dermatologists in differentiating selected skin lesions.^78-81 For example, Esteva and colleagues demonstrated AI could differentiate keratinocyte carcinomas from benign seborrheic keratoses and malignant melanomas from benign nevi with accuracy equal to 21 board-certified dermatologists.⁷⁸

AI is applicable to various imaging procedures common to dermatology, such as dermoscopy, very high-frequency ultrasound, and reflectance confocal microscopy.⁸² Several studies have demonstrated that AI interpretation compared favorably to dermatologists evaluating dermoscopy to assess melanocytic lesions.^78-81,83

A limitation in these studies is that they differentiate only a few diagnoses.⁸² Furthermore, dermatologists have sensory input such as touch and visual examination under various conditions, something AI has yet to replicate.^15,34,84 Also, most AI devices use no or limited clinical information.⁸¹ Dermatologists can recognize rarer conditions for which AI models may have had limited or no training.³⁴ Nevertheless, a recent study assessed AI for the diagnosis of 134 separate skin disorders with promising results, including providing diagnoses with accuracy comparable to that of dermatologists and providing accurate treatment strategies.⁸⁴ As Topol points out, most skin lesions are diagnosed in the primary care setting where AI can have a greater impact when used in conjunction with the clinical impression, especially where specialists are in limited supply.^48,78

Finally, dermatology lends itself to using portable or smartphone applications (apps) wherein the user can photograph a lesion for analysis by AI algorithms to assess the need for further evaluation or make treatment recommendations.^34,84,85 Although results from currently available apps are not encouraging, they may play a greater role as the technology advances.^34,85

Oncology

Applications of AI in oncology include predicting prognosis for patients with cancer based on histologic and/or genetic information.^14,68,86 Programs can predict the risk of complications before and recurrence risks after surgery for malignancies.^44,87-89 AI can also assist in treatment planning and predict treatment failure with radiation therapy.^90,91

AI has great potential in processing the large volumes of patient data in cancer genomics. Next-generation sequencing has allowed for the identification of millions of DNA sequences in a single tumor to detect genetic anomalies.⁹² Thousands of mutations can be found in individual tumor samples, and processing this information and determining its significance can be beyond human capability.¹⁴ We know little about the effects of various mutation combinations, and most tumors have a heterogeneous molecular profile among different cell populations.^14,93 The presence or absence of various mutations can have diagnostic, prognostic, and therapeutic implications.⁹³ AI has great potential to sort through these complex data and identify actionable findings.

More than 200,000 cancer-related articles were published in 2019, and publications in the field of cancer genomics are increasing exponentially.^14,92,93 Patel and colleagues assessed the utility of IBM Watson for Genomics against results from a molecular tumor board.⁹³ Watson for Genomics identified potentially significant mutations not identified by the tumor board in 32% of patients. Most mutations were related to new clinical trials not yet added to the tumor board watch list, demonstrating the role AI will have in processing the large volume of genetic data required to deliver personalized medicine moving forward.

Gastroenterology

AI has shown promise in predicting risk or outcomes based on clinical parameters in various common gastroenterology problems, including gastric reflux, acute pancreatitis, gastrointestinal bleeding, celiac disease, and inflammatory bowel disease.^94,95 AI endoscopic analysis has demonstrated potential in assessing Barrett’s esophagus, gastric Helicobacter pylori infections, gastric atrophy, and gastric intestinal metaplasia.⁹⁵ Applications have been used to assess esophageal, gastric, and colonic malignancies, including depth of invasion based on endoscopic images.⁹⁵ Finally, studies have evaluated AI to assess small colon polyps during colonoscopy, including differentiating benign and premalignant polyps with success comparable to gastroenterologists.^94,95 AI has been shown to increase the speed and accuracy of gastroenterologists in detecting small polyps during colonoscopy.⁴⁸ In a prospective randomized study, colonoscopies performed using an AI device identified significantly more small adenomatous polyps than colonoscopies without AI.⁹⁶

Neurology

It has been suggested that AI technologies are well suited for application in neurology due to the subtle presentation of many neurologic diseases.¹⁶ Viz LVO, the first CMS-approved AI reimbursement for the diagnosis of strokes, analyzes CTs to detect early ischemic strokes and alerts the medical team, thus shortening time to treatment.^3,97 Many other AI platforms are in use or development that use CT and MRI for the early detection of strokes as well as for treatment and prognosis.^9,97

AI technologies have been applied to neurodegenerative diseases, such as Alzheimer and Parkinson diseases.^16,98 For example, several studies have evaluated patient movements in Parkinson disease for both early diagnosis and to assess response to treatment.⁹⁸ These evaluations included assessment with both external cameras as well as wearable devices and smartphone apps.

Mental Health

Due to the interactive nature of mental health care, the field has been slower to develop AI applications.¹⁸ With heavy reliance on textual information (eg, clinic notes, mood rating scales, and documentation of conversations), successful AI applications in this field will likely rely heavily on NLP.¹⁸ However, studies investigating the application of AI to mental health have also incorporated data such as brain imaging, smartphone monitoring, and social media platforms, such as Facebook and Twitter.^18,103,104

The risk of suicide is higher in veteran patients, and ML algorithms have had limited success in predicting suicide risk in both veteran and nonveteran populations.^104-106 While early models have low positive predictive values and low sensitivities, they still promise to be a useful tool in conjunction with traditional risk assessments.¹⁰⁶ Kessler and colleagues suggest that combining multiple rather than single ML algorithms might lead to greater success.^105,106

AI may assist in diagnosing other mental health disorders, including major depressive disorder, attention deficit hyperactivity disorder (ADHD), schizophrenia, posttraumatic stress disorder, and Alzheimer disease.^103,104,107 These investigations are in the early stages with limited clinical applicability. However, 2 AI applications awaiting FDA approval relate to ADHD and opioid use.² Furthermore, potential exists for AI to not only assist with prevention and diagnosis of ADHD, but also to identify optimal treatment options.^2,103

General and Personalized Medicine

Additional AI applications include diagnosing patients with suspected sepsis, measuring liver iron concentrations, predicting hospital mortality at the time of admission, and more.^2,108,109 AI can guide end-of-life decisions such as resuscitation status or whether to initiate mechanical ventilation.⁴⁸

AI-driven smartphone apps can be beneficial to both patients and clinicians. Examples include predicting nonadherence to anticoagulation therapy, monitoring heart rhythms for atrial fibrillation or signs of hyperkalemia in patients with renal failure, and improving outcomes for patients with diabetes mellitus by decreasing glycemic variability and reducing hypoglycemia.^8,48,110,111 The potential for AI applications to health care and personalized medicine are almost limitless.

Discussion

With ever-increasing expectations for all health care sectors to deliver timely, fiscally-responsible, high-quality health care, AI has the potential to have numerous impacts. AI can improve diagnostic accuracy while limiting errors and impact patient safety such as assisting with prescription delivery.^8,9,34 It can screen and triage patients, alerting clinicians to those needing more urgent evaluation.^8,23,77,97 AI also may increase a clinician’s efficiency and speed to render a diagnosis.^12,13,55,97 AI can provide a rapid second opinion, an ability especially beneficial in underserved areas with shortages of specialists.^{23,25,26,29,34} Similarly, AI may decrease the inter- and intraobserver variability common in many medical specialties.^12,27,45 AI applications can also monitor disease progression, identifying patients at greatest risk, and provide information for prognosis.^21,23,56,58 Finally, as described with applications using IBM Watson, AI can allow for an integrated approach to health care that is currently lacking.

We have described many reports suggesting AI can render diagnoses as well as or better than experienced clinicians, and speculation exists that AI will replace many roles currently performed by health care practitioners.^9,26 However, most studies demonstrate that AI’s diagnostic benefits are best realized when used to supplement a clinician’s impression.^{8,22,30,33,52,54,56,69,84} AI is not likely to replace humans in health care in the foreseeable future. The technology can be likened to the impact of CT scans developed in the 1970s in neurology. Prior to such detailed imaging, neurologists spent extensive time performing detailed physicals to render diagnoses and locate lesions before surgery. There was mistrust of this new technology and concern that CT scans would eliminate the need for neurologists.¹¹² On the contrary, neurology is alive and well, frequently being augmented by the technologies once speculated to replace it.

Commercial AI health care platforms represented a $2 billion industry in 2018 and are growing rapidly each year.^13,32 Many AI products are offered ready for implementation for various tasks, including diagnostics, patient management, and improved efficiency. Others will likely be provided as templates suitable for modification to meet the specific needs of the facility, practice, or specialty for its patient population.

AI Risks and Limitations

AI has several risks and limitations. Although there is progress in explainable AI, at times we still struggle to understand how the output provided by machine learning algorithms was created.^44,48 The many layers associated with deep learning self-determine the criteria to reach its conclusion, and these criteria can continually evolve. The parameters of deep learning are not preprogrammed, and there are too many individual data points to be extrapolated or deconvoluted for evaluation at our current level of knowledge.^26,51 These apparent lack of constraints cause concern for patient safety and suggest that greater validation and continued scrutiny of validity is required.^8,48 Efforts are underway to create explainable AI programs to make their processes more transparent, but such clarification is limited presently.^14,26,48,77

Another challenge of AI is determining the amount of training data required to function optimally. Also, if the output describes multiple variables or diagnoses, are each equally valid?¹¹³ Furthermore, many AI applications look for a specific process, such as cancer diagnoses on CXRs. However, how coexisting conditions like cardiomegaly, emphysema, pneumonia, etc, seen on CXRs will affect the diagnosis needs to be considered.^51,52 Zech and colleagues provide the example that diagnoses for pneumothorax are frequently rendered on CXRs with chest tubes in place.⁵¹ They suggest that CNNs may develop a bias toward diagnosing pneumothorax when chest tubes are present. Many current studies approach an issue in isolation, a situation not realistic in real-world clinical practice.²⁶

Most studies on AI have been retrospective, and frequently data used to train the program are preselected.^13,26 The data are typically validated on available databases rather than actual patients in the clinical setting, limiting confidence in the validity of the AI output when applied to real-world situations. Currently, fewer than 12 prospective trials had been published comparing AI with traditional clinical care.^13,114 Randomized prospective clinical trials are even fewer, with none currently reported from the United States.^13,114 The results from several studies have been shown to diminish when repeated prospectively.¹¹⁴

The FDA has created a new category known as Software as a Medical Device and has a Digital Health Innovation Action Plan to regulate AI platforms. Still, the process of AI regulation is of necessity different from traditional approval processes and is continually evolving.⁸ The FDA approval process cannot account for the fact that the program’s parameters may continually evolve or adapt.²

Guidelines for investigating and reporting AI research with its unique attributes are being developed. Examples include the TRIPOD-ML statement and others.^49,115 In September 2020, 2 publications addressed the paucity of gold-standard randomized clinical trials in clinical AI applications.^116,117 The SPIRIT-AI statement expands on the original SPIRIT statement published in 2013 to guide minimal reporting standards for AI clinical trial protocols to promote transparency of design and methodology.¹¹⁶ Similarly, the CONSORT-AI extension, stemming from the original CONSORT statement in 1996, aims to ensure quality reporting of randomized controlled trials in AI.¹¹⁷

Another risk with AI is that while an individual physician making a mistake may adversely affect 1 patient, a single mistake in an AI algorithm could potentially affect thousands of patients.⁴⁸ Also, AI programs developed for patient populations at a facility may not translate to another. Referred to as overfitting, this phenomenon relates to selection bias in training data sets.^{15,34,49,51,52} Studies have shown that programs that underrepresent certain group characteristics such as age, sex, or race may be less effective when applied to a population in which these characteristics have differing representations.^8,48,49 This problem of underrepresentation has been demonstrated in programs interpreting pathology slides, radiographs, and skin lesions.^15,32,51

Admittedly, most of these challenges are not specific to AI and existed in health care previously. Physicians make mistakes, treatments are sometimes used without adequate prospective studies, and medications are given without understanding their mechanism of action, much like AI-facilitated processes reach a conclusion that cannot be fully explained.⁴⁸

Conclusions

The view that AI will dramatically impact health care in the coming years will likely prove true. However, much work is needed, especially because of the paucity of prospective clinical trials as has been historically required in medical research. Any concern that AI will replace HCPs seems unwarranted. Early studies suggest that even AI programs that appear to exceed human interpretation perform best when working in cooperation with and oversight from clinicians. AI’s greatest potential appears to be its ability to augment care from health professionals, improving efficiency and accuracy, and should be anticipated with enthusiasm as the field moves forward at an exponential rate.

Acknowledgments

The authors thank Makenna G. Thomas for proofreading and review of the manuscript. This material is the result of work supported with resources and the use of facilities at the James A. Haley Veterans’ Hospital. This research has been approved by the James A. Haley Veteran’s Hospital Office of Communications and Media.

Artificial Intelligence (AI) was first described in 1956 and refers to machines having the ability to learn as they receive and process information, resulting in the ability to “think” like humans.¹ AI’s impact in medicine is increasing; currently, at least 29 AI medical devices and algorithms are approved by the US Food and Drug Administration (FDA) in a variety of areas, including radiograph interpretation, managing glucose levels in patients with diabetes mellitus, analyzing electrocardiograms (ECGs), and diagnosing sleep disorders among others.² Significantly, in 2020, the Centers for Medicare and Medicaid Services (CMS) announced the first reimbursement to hospitals for an AI platform, a model for early detection of strokes.³ AI is rapidly becoming an integral part of health care, and its role will only increase in the future (Table).

As knowledge in medicine is expanding exponentially, AI has great potential to assist with handling complex patient care data. The concept of exponential growth is not a natural one. As Bini described, with exponential growth the volume of knowledge amassed over the past 10 years will now occur in perhaps only 1 year.¹ Likewise, equivalent advances over the past year may take just a few months. This phenomenon is partly due to the law of accelerating returns, which states that advances feed on themselves, continually increasing the rate of further advances.⁴ The volume of medical data doubles every 2 to 5 years.⁵ Fortunately, the field of AI is growing exponentially as well and can help health care practitioners (HCPs) keep pace, allowing the continued delivery of effective health care.

In this report, we review common terminology, principles, and general applications of AI, followed by current and potential applications of AI for selected medical specialties. Finally, we discuss AI’s future in health care, along with potential risks and pitfalls.

AI Overview

AI refers to machine programs that can “learn” or think based on past experiences. This functionality contrasts with simple rules-based programming available to health care for years. An example of rules-based programming is the warfarindosing.org website developed by Barnes-Jewish Hospital at Washington University Medical Center, which guides initial warfarin dosing.^6,7 The prescriber inputs detailed patient information, including age, sex, height, weight, tobacco history, medications, laboratory results, and genotype if available. The application then calculates recommended warfarin dosing regimens to avoid over- or underanticoagulation. While the dosing algorithm may be complex, it depends entirely on preprogrammed rules. The program does not learn to reach its conclusions and recommendations from patient data.

In contrast, one of the most common subsets of AI is machine learning (ML). ML describes a program that “learns from experience and improves its performance as it learns.”¹ With ML, the computer is initially provided with a training data set—data with known outcomes or labels. Because the initial data are input from known samples, this type of AI is known as supervised learning.^8-10 As an example, we recently reported using ML to diagnose various types of cancer from pathology slides.¹¹ In one experiment, we captured images of colon adenocarcinoma and normal colon (these 2 groups represent the training data set). Unlike traditional programming, we did not define characteristics that would differentiate colon cancer from normal; rather, the machine learned these characteristics independently by assessing the labeled images provided. A second data set (the validation data set) was used to evaluate the program and fine-tune the ML training model’s parameters. Finally, the program was presented with new images of cancer and normal cases for final assessment of accuracy (test data set). Our program learned to recognize differences from the images provided and was able to differentiate normal and cancer images with > 95% accuracy.

Advances in computer processing have allowed for the development of artificial neural networks (ANNs). While there are several types of ANNs, the most common types used for image classification and segmentation are known as convolutional neural networks (CNNs).^9,12-14 The programs are designed to work similar to the human brain, specifically the visual cortex.^15,16 As data are acquired, they are processed by various layers in the program. Much like neurons in the brain, one layer decides whether to advance information to the next.^13,14 CNNs can be many layers deep, leading to the term deep learning: “computational models that are composed of multiple processing layers to learn representations of data with multiple levels of abstraction.”^1,13,17

ANNs can process larger volumes of data. This advance has led to the development of unstructured or unsupervised learning. With this type of learning, imputing defined features (ie, predetermined answers) of the training data set described above is no longer required.^1,8,10,14 The advantage of unsupervised learning is that the program can be presented raw data and extract meaningful interpretation without human input, often with less bias than may exist with supervised learning.^1,18 If shown enough data, the program can extract relevant features to make conclusions independently without predefined definitions, potentially uncovering markers not previously known. For example, several studies have used unsupervised learning to search patient data to assess readmission risks of patients with congestive heart failure.^10,19,20 AI compiled features independently and not previously defined, predicting patients at greater risk for readmission superior to traditional methods.

Health Care Applications

Image analysis has seen the most AI health care applications.^8,15 AI has shown potential in interpreting many types of medical images, including pathology slides, radiographs of various types, retina and other eye scans, and photographs of skin lesions. Many studies have demonstrated that AI can interpret these images as accurately as or even better than experienced clinicians.^9,13,22-29 Studies have suggested AI interpretation of radiographs may better distinguish patients infected with COVID-19 from other causes of pneumonia, and AI interpretation of pathology slides may detect specific genetic mutations not previously identified without additional molecular tests.^{11,14,23,24,30-32}

The second area in which AI can impact health care is improving workflow and efficiency. AI has improved surgery scheduling, saving significant revenue, and decreased patient wait times for appointments.¹ AI can screen and triage radiographs, allowing attention to be directed to critical patients. This use would be valuable in many busy clinical settings, such as the recent COVID-19 pandemic.^8,23 Similarly, AI can screen retina images to prioritize urgent conditions.²⁵ AI has improved pathologists’ efficiency when used to detect breast metastases.³³ Finally, AI may reduce medical errors, thereby ensuring patient safety.^8,9,34

A third health care benefit of AI is in public health and epidemiology. AI can assist with clinical decision-making and diagnoses in low-income countries and areas with limited health care resources and personnel.^25,29 AI can improve identification of infectious outbreaks, such as tuberculosis, malaria, dengue fever, and influenza.^29,35-40 AI has been used to predict transmission patterns of the Zika virus and the current COVID-19 pandemic.^41,42 Applications can stratify the risk of outbreaks based on multiple factors, including age, income, race, atypical geographic clusters, and seasonal factors like rainfall and temperature.^35,36,38,43 AI has been used to assess morbidity and mortality, such as predicting disease severity with malaria and identifying treatment failures in tuberculosis.²⁹

Finally, AI can dramatically impact health care due to processing large data sets or disconnected volumes of patient information—so-called big data.^44-46 An example is the widespread use of electronic health records (EHRs) such as the Computerized Patient Record System used in Veteran Affairs medical centers (VAMCs). Much of patient information exists in written text: HCP notes, laboratory and radiology reports, medication records, etc. Natural language processing (NLP) allows platforms to sort through extensive volumes of data on complex patients at rates much faster than human capability, which has great potential to assist with diagnosis and treatment decisions.⁹

Medical literature is being produced at rates that exceed our ability to digest. More than 200,000 cancer-related articles were published in 2019 alone.¹⁴ NLP capabilities of AI have the potential to rapidly sort through this extensive medical literature and relate specific verbiage in patient records guiding therapy.⁴⁶ IBM Watson, a supercomputer based on ML and NLP, demonstrates this concept with many potential applications, only some of which relate to health care.^1,9 Watson has an oncology component to assimilate multiple aspects of patient care, including clinical notes, pathology results, radiograph findings, staging, and a tumor’s genetic profile. It coordinates these inputs from the EHR and mines medical literature and research databases to recommend treatment options.^1,46 AI can assess and compile far greater patient data and therapeutic options than would be feasible by individual clinicians, thus providing customized patient care.⁴⁷ Watson has partnered with numerous medical centers, including MD Anderson Cancer Center and Memorial Sloan Kettering Cancer Center, with variable success.^44,47-49 While the full potential of Watson appears not yet realized, these AI-driven approaches will likely play an important role in leveraging the hidden value in the expanding volume of health care information.

Medical Specialty Applications

Radiology

Currently > 70% of FDA-approved AI medical devices are in the field of radiology.² Most radiology departments have used AI-friendly digital imaging for years, such as the picture archiving and communication systems used by numerous health care systems, including VAMCs.^2,15 Gray-scale images common in radiology lend themselves to standardization, although AI is not limited to black-and- white image interpretation.¹⁵

An abundance of literature describes plain radiograph interpretation using AI. One FDA-approved platform improved X-ray diagnosis of wrist fractures when used by emergency medicine clinicians.^2,50 AI has been applied to chest X-ray (CXR) interpretation of many conditions, including pneumonia, tuberculosis, malignant lung lesions, and COVID-19.^{23,25,28,44,51-53} For example, Nam and colleagues suggested AI is better at diagnosing malignant pulmonary nodules from CXRs than are trained radiologists.²⁸

In addition to plain radiographs, AI has been applied to many other imaging technologies, including ultrasounds, positron emission tomography, mammograms, computed tomography (CT), and magnetic resonance imaging (MRI).^{15,26,44,48,54-56} A large study demonstrated that ML platforms significantly reduced the time to diagnose intracranial hemorrhages on CT and identified subtle hemorrhages missed by radiologists.⁵⁵ Other studies have claimed that AI programs may be better than radiologists in detecting cancer in screening mammograms, and 3 FDA-approved devices focus on mammogram interpretation.^2,15,54,57 There is also great interest in MRI applications to detect and predict prognosis for breast cancer based on imaging findings.^21,56

Aside from providing accurate diagnoses, other studies focus on AI radiograph interpretation to assist with patient screening, triage, improving time to final diagnosis, providing a rapid “second opinion,” and even monitoring disease progression and offering insights into prognosis.^{8,21,23,52,55,56,58} These features help in busy urban centers but may play an even greater role in areas with limited access to health care or trained specialists such as radiologists.⁵²