Federal Practitioner

Opioid overdose is a major public health challenge, with recent reports estimating 41 deaths per day in the United States from prescription opioid overdose.^1,2 Prescribing naloxone has increasingly been advocated to reduce the risk of opioid overdose for patients identified as high risk. Naloxone distribution has been shown to decrease the incidence of opioid overdoses in the general population.^3,4 The Centers for Disease Control and Prevention (CDC) Guideline for Prescribing Opioids for Chronic Pain recommends considering naloxone prescription for patients with a history of overdose or substance use disorder, opioid dosages ≥ 50 morphine equivalent daily dose (MEDD), and concurrent use of benzodiazepines.⁵

Although the CDC guidelines are intended for primary care clinicians in outpatient settings, naloxone prescribing is also relevant in the postsurgical setting.⁵ Many surgical patients are at risk for opioid overdose and data from the Veterans Health Administration (VHA) has shown that risk of opioid overdose is 11-fold higher in the 30 days following discharge from a surgical admission, when compared with the subsequent calendar year.^6,7 This likely occurs due to new prescriptions or escalated doses of opioids following surgery. Overdose risk may be particularly relevant to orthopedic surgery as postoperative opioids are commonly prescribed.⁸ Patients undergoing total knee arthroplasty (TKA) may represent a vulnerable population to overdose as it is one of the most commonly performed surgeries for the treatment of chronic pain, and is frequently performed in older adults with medical comorbidities.^9,10

Identifying patients at high risk for opioid overdose is important for targeted naloxone dispensing.⁵ A risk index for overdose or serious opioid-induced respiratory depression (RIOSORD) tool has been developed and validated in veteran and other populations to identify such patients.¹¹ The RIOSORD tool classifies patients by risk level (1-10) and predicts probability of overdose or serious opioid-induced respiratory depression (OSORD). A patient’s level of risk is based on a weighted combination of the 15 independent risk factors most highly associated with OSORD, including comorbid conditions, prescription drug use, and health care utilization.¹² Using the RIOSORD tool, the VHA Opioid Education and Naloxone Distribution (OEND) program is a risk mitigation initiative that aims to decrease opioid-related overdose morbidity and mortality. This is achieved via opioid overdose education for prevention, recognition, and response and includes outpatient naloxone prescription.^13,14

Despite the comprehensive OEND program, there exists very little data to guide postsurgical naloxone prescribing. The prevalence of known risk factors for overdose in surgical patients remains unknown, as does the prevalence of perioperative naloxone distribution. Understanding overdose risk factors and naloxone prescribing patterns in surgical patients may identify potential targets for OEND efforts. This study retrospectively estimated RIOSORD scores for TKA patients between 2013 to 2016 and described naloxone distribution based on RIOSORD scores and risk factors.

Methods

We identified patients who had undergone primary TKA at VHA hospitals using Current Procedural Terminology (CPT), International Classification of Diseases, Ninth Revision (ICD-9) procedure codes, and data extracted from the VHA Corporate Data Warehouse (CDW) of electronic health records (EHRs). Our study was granted approval with exemption from informed consent by the Durham Veteran Affairs Healthcare System Institutional Review Board.

This retrospective cohort study included all veterans who underwent elective primary TKA from January 1, 2013 through December 31, 2016. We excluded patients who died before discharge.

Outcomes

Our primary outcome was being dispensed an outpatient naloxone prescription following TKA. Naloxone dispensing was identified by examining CDW outpatient pharmacy records with a final dispense date from 1 year before surgery through 7 days after discharge following TKA. To exclude naloxone administration that may have been given in a clinic, prescription data included only records with an outpatient prescription copay. Naloxone dispensing in the year before surgery was chosen to estimate likely preoperative possession of naloxone which could be available in the postoperative period. Naloxone dispensing until 7 days after discharge was chosen to identify any new dispensing that would be available in the postoperative period. These outcomes were examined over the study time frame on an annual basis.

Patient Factors

Demographic variables included age, sex, and race/ethnicity. Independent risk factors for overdose from RIOSORD were identified for each patient.¹⁵ These risk factors included comorbidities (opioid use disorder, schizophrenia, bipolar disorder, liver disease, chronic kidney disease, sleep apnea, or lung disease) and prescription drug use (use of opioids, benzodiazepines, long-acting opioids, ≥ 50 MEDD or ≥ 100 MEDD). ICD-9 and ICD-10 diagnosis codes were used to identify comorbidities. Risk classes on day of surgery were identified using a RIOSORD algorithm code. Consistent with the display of RIOSORD risk classes on the VHA Academic Detailing Service OEND risk report, patients were grouped into 3 groups based on their RIOSORD score: classes 1 to 4 (low risk), 5 to 7 (moderate risk), and 8 to 10 (high risk).

Descriptive statistics were used to summarize data on patient demographics, RIOSORD risk factors, overdose events, and naloxone dispensing over time.

Results

The study cohort included 38,011 veterans who underwent primary TKA in the VHA between January 1, 2013 and December 30, 2016. In this cohort, the mean age was 65 years, 93% were male, and 77% were White patients (Table 1). The most common comorbidities were lung disease in 9170 (24.1%) patients, sleep apnea in 6630 (17.4%) patients, chronic kidney disease in 4036 (10.6%) patients, liver disease in 2822 (7.4%) patients, and bipolar disorder in 1748 (4.6%) patients.

In 2013, 63.1% of patients presenting for surgery were actively prescribed opioids. By 2016, this decreased to 50.5%. Benzodiazepine use decreased from 13.2 to 8.8% and long-acting opioid use decreased from 8.5 to 5.8% over the same period. Patients taking ≥ 50 MEDD decreased from 8.0 to 5.3% and patients taking ≥ 100 MEDD decreased from 3.3 to 2.2%. The prevalence of moderate-risk patients decreased from 2.5 to 1.6% and high-risk patients decreased from 0.8 to 0.6% (Figure 1). Cumulatively, the prevalence of presenting with either moderate or high risk of overdose decreased from 3.3 to 2.2% between 2013 to 2016.

Naloxone Dispensing

In 2013, naloxone was not dispensed to any patients at moderate or high risk for overdose between 365 days prior to surgery until 7 days after discharge (Table 2 and Figure 2). Low-risk group naloxone dispensing increased to 2 (0.0%) in 2014, to 13 (0.1%), in 2015, and to 86 (0.9%) in 2016. Moderate-risk group naloxone dispensing remained at 0 (0.0%) in 2014, but increased to 8 (3.5%) in 2015, and to 18 (10.9%) in 2016. High-risk group naloxone dispensing remained at 0 (0.0%) in 2014, but increased to 5 (5.8%) in 2015, and to 8 (12.7%) in 2016 (Figure 3).

Discussion

Our data demonstrate that patients presenting for TKA between 2013 and 2016 routinely had individual risk factors for overdose related to either prescription drug use or comorbidities. We also show that, although the number of patients at moderate and high risk for opioid overdose is decreasing, 2.2% of TKA patients remain at moderate or high risk for opioid overdose based on a weighted combination of these individual risk factors using RIOSORD. As demand for primary TKA is projected to grow to 3.5 million procedures by 2030, using prevalence from 2016, we estimate that 76,560 patients may present for TKA across the US with moderate or high risk for opioid overdose.⁹ Following discharge, this risk may be even higher as this estimate does not yet account for postoperative opioid use. We demonstrate that through a VHA OEND initiative, naloxone distribution increased and appeared to be targeted to those most at risk using a simple validated tool like RIOSORD.

Presence of an individual risk factor for overdose was present in as many as 63.1% of patients presenting for TKA, as was seen in 2013 with preoperative opioid use. The 3 highest scoring prescription use–related risk factors in RIOSORD are use of opioids ≥ 100 MEDD (16 points), ≥ 50 MEDD (9 points), and long-acting formulations (9 points). All 3 decreased in prevalence over the study period but by 2016 were still seen in 2.2% for ≥ 100 MEDD, 5.3% for ≥ 50 MEDD, and 5.8% for long-acting opioids. This decrease was not surprising given implementation of a VHA-wide opioid safety initiative and the OEND program, but this could also be related to changes in patient selection for surgery in the context of increased awareness of the opioid epidemic. Despite the trend toward safer opioid prescribing, by 2016 over half of patients (50.5%) who presented for TKA were already taking opioids, with 10.6% (543 of 5127) on doses ≥ 50 MEDD.

We observed a decrease in RIOSORD risk each year, consistent with decreasing prescription-related risk factors over time. This was most obvious in the moderate-risk group. It is unclear why a similar decrease was not as obvious in the high-risk group, but this in part may be due to the already low numbers of patients in the high-risk group. This may also represent the high-risk group being somewhat resistant to the initiatives that shifted moderate-risk patients to the low-risk group. There were proportionately more patients in the moderate- and high-risk groups in the original RIOSORD population than in our surgical population, which may be attributed to the fewer comorbidities seen in our surgical population, as well as the higher opioid-prescribing patterns seen prior to the VA OEND initiative.¹²

Naloxone prescribing was rare prior to the OEND initiative and increased from 2013 to 2016. Increases were most marked in those in moderate- and high-risk groups, although naloxone prescribing also increased among the low-risk group. Integration of RIOSORD stratification into the OEND initiative likely played a role in targeting increased access to naloxone among those at highest risk of overdose. Naloxone dispensing increased for every group, although a significant proportion of moderate- and high-risk patients, 89.1% and 87.3%, respectively, were still not dispensed naloxone by 2016. Moreover, our estimates of perioperative naloxone access were likely an overestimate by including patients dispensed naloxone up to 1 year before surgery until 7 days after surgery. The aim was to include patients who may not have been prescribed naloxone postoperatively because of an existing naloxone prescription at home. Perioperative naloxone access estimates would have been even lower if a narrower window had been used to approximate perioperative access. This identifies an important gap between those who may benefit from naloxone dispensing and those who received naloxone. This in part may be because OEND has not been implemented as routinely in surgical settings as other settings (eg, primary care). OEND efforts may more effectively increase naloxone prescribing among surgical patients if these efforts were targeted at surgical and anesthesia departments. Given that the Comprehensive Addiction and Recovery Act of 2016 requires an assessment of patient risk prior to opioid prescribing and VHA efforts to increase utilization of tools like the Stratification Tool for Opioid Risk Mitigation (STORM), which estimates patient risk when initiating an opioid prescription and includes naloxone as one of many risk mitigation strategies, we anticipate that rates of naloxone prescribing will increase over time.

Limitations

Our study captures a large number of patients across VHA hospitals of varying size nationwide, including a mix of those with and without academic medical center affiliations. This veteran population may not represent the US commercially insured population (CIP). Zedler and colleagues highlighted the differences in prevalence of individual risk factors: notably, the CIP had a substantially higher proportion of females and younger patients.¹¹ VHA had a greater prevalence of common chronic conditions associated with older age. The frequency of opioid dependence was similar among CIP and VHA. However, substance abuse and nonopioid substance dependence diagnoses were 4-fold more frequent among VHA controls as CIP controls. Prescribing of all opioids, except morphine and methadone, was substantially greater in CIP than in VHA.¹¹ Despite a difference in individual risk factors, a CIP-specific RIOSORD has been validated and can be used outside of the VHA to obviate the limitations of the VHA-specific RIOSORD.¹¹

Other limitations include our estimation of naloxone access. We do not know whether naloxone was administered or have a reliable estimate of overdose incidence in this postoperative TKA population. Also, it is important to note that RIOSORD was not developed for surgical patients. The use of RIOSORD in a postoperative population likely underestimates risk of opioid overdose due to the frequent prescriptions of new opioids or escalation of existing MEDD to the postoperative patient. Our study was also retrospective in nature and reliant on accurate coding of patient risk factors. It is possible that comorbidities were not accurately identified by EHR and therefore subject to inconsistency.

Conclusions

Veterans presenting for TKA routinely have risk factors for opioid overdose. We observed a trend toward decreasing overdose risk which coincided with the Opioid Safety and OEND initiatives within the VHA. We also observed an increase in naloxone prescription for moderate- and high-risk patients undergoing TKA, although most of these patients still did not receive naloxone as of 2016. More research is needed to refine and validate the RIOSORD score for surgical populations. Expanding initiatives such as OEND to include surgical patients presents an opportunity to improve access to naloxone for postoperative patients that may help reduce opioid overdose in this population.

Opioid overdose is a major public health challenge, with recent reports estimating 41 deaths per day in the United States from prescription opioid overdose.^1,2 Prescribing naloxone has increasingly been advocated to reduce the risk of opioid overdose for patients identified as high risk. Naloxone distribution has been shown to decrease the incidence of opioid overdoses in the general population.^3,4 The Centers for Disease Control and Prevention (CDC) Guideline for Prescribing Opioids for Chronic Pain recommends considering naloxone prescription for patients with a history of overdose or substance use disorder, opioid dosages ≥ 50 morphine equivalent daily dose (MEDD), and concurrent use of benzodiazepines.⁵

Although the CDC guidelines are intended for primary care clinicians in outpatient settings, naloxone prescribing is also relevant in the postsurgical setting.⁵ Many surgical patients are at risk for opioid overdose and data from the Veterans Health Administration (VHA) has shown that risk of opioid overdose is 11-fold higher in the 30 days following discharge from a surgical admission, when compared with the subsequent calendar year.^6,7 This likely occurs due to new prescriptions or escalated doses of opioids following surgery. Overdose risk may be particularly relevant to orthopedic surgery as postoperative opioids are commonly prescribed.⁸ Patients undergoing total knee arthroplasty (TKA) may represent a vulnerable population to overdose as it is one of the most commonly performed surgeries for the treatment of chronic pain, and is frequently performed in older adults with medical comorbidities.^9,10

Identifying patients at high risk for opioid overdose is important for targeted naloxone dispensing.⁵ A risk index for overdose or serious opioid-induced respiratory depression (RIOSORD) tool has been developed and validated in veteran and other populations to identify such patients.¹¹ The RIOSORD tool classifies patients by risk level (1-10) and predicts probability of overdose or serious opioid-induced respiratory depression (OSORD). A patient’s level of risk is based on a weighted combination of the 15 independent risk factors most highly associated with OSORD, including comorbid conditions, prescription drug use, and health care utilization.¹² Using the RIOSORD tool, the VHA Opioid Education and Naloxone Distribution (OEND) program is a risk mitigation initiative that aims to decrease opioid-related overdose morbidity and mortality. This is achieved via opioid overdose education for prevention, recognition, and response and includes outpatient naloxone prescription.^13,14

Despite the comprehensive OEND program, there exists very little data to guide postsurgical naloxone prescribing. The prevalence of known risk factors for overdose in surgical patients remains unknown, as does the prevalence of perioperative naloxone distribution. Understanding overdose risk factors and naloxone prescribing patterns in surgical patients may identify potential targets for OEND efforts. This study retrospectively estimated RIOSORD scores for TKA patients between 2013 to 2016 and described naloxone distribution based on RIOSORD scores and risk factors.

Methods

We identified patients who had undergone primary TKA at VHA hospitals using Current Procedural Terminology (CPT), International Classification of Diseases, Ninth Revision (ICD-9) procedure codes, and data extracted from the VHA Corporate Data Warehouse (CDW) of electronic health records (EHRs). Our study was granted approval with exemption from informed consent by the Durham Veteran Affairs Healthcare System Institutional Review Board.

This retrospective cohort study included all veterans who underwent elective primary TKA from January 1, 2013 through December 31, 2016. We excluded patients who died before discharge.

Outcomes

Our primary outcome was being dispensed an outpatient naloxone prescription following TKA. Naloxone dispensing was identified by examining CDW outpatient pharmacy records with a final dispense date from 1 year before surgery through 7 days after discharge following TKA. To exclude naloxone administration that may have been given in a clinic, prescription data included only records with an outpatient prescription copay. Naloxone dispensing in the year before surgery was chosen to estimate likely preoperative possession of naloxone which could be available in the postoperative period. Naloxone dispensing until 7 days after discharge was chosen to identify any new dispensing that would be available in the postoperative period. These outcomes were examined over the study time frame on an annual basis.

Patient Factors

Demographic variables included age, sex, and race/ethnicity. Independent risk factors for overdose from RIOSORD were identified for each patient.¹⁵ These risk factors included comorbidities (opioid use disorder, schizophrenia, bipolar disorder, liver disease, chronic kidney disease, sleep apnea, or lung disease) and prescription drug use (use of opioids, benzodiazepines, long-acting opioids, ≥ 50 MEDD or ≥ 100 MEDD). ICD-9 and ICD-10 diagnosis codes were used to identify comorbidities. Risk classes on day of surgery were identified using a RIOSORD algorithm code. Consistent with the display of RIOSORD risk classes on the VHA Academic Detailing Service OEND risk report, patients were grouped into 3 groups based on their RIOSORD score: classes 1 to 4 (low risk), 5 to 7 (moderate risk), and 8 to 10 (high risk).

Descriptive statistics were used to summarize data on patient demographics, RIOSORD risk factors, overdose events, and naloxone dispensing over time.

Results

The study cohort included 38,011 veterans who underwent primary TKA in the VHA between January 1, 2013 and December 30, 2016. In this cohort, the mean age was 65 years, 93% were male, and 77% were White patients (Table 1). The most common comorbidities were lung disease in 9170 (24.1%) patients, sleep apnea in 6630 (17.4%) patients, chronic kidney disease in 4036 (10.6%) patients, liver disease in 2822 (7.4%) patients, and bipolar disorder in 1748 (4.6%) patients.

In 2013, 63.1% of patients presenting for surgery were actively prescribed opioids. By 2016, this decreased to 50.5%. Benzodiazepine use decreased from 13.2 to 8.8% and long-acting opioid use decreased from 8.5 to 5.8% over the same period. Patients taking ≥ 50 MEDD decreased from 8.0 to 5.3% and patients taking ≥ 100 MEDD decreased from 3.3 to 2.2%. The prevalence of moderate-risk patients decreased from 2.5 to 1.6% and high-risk patients decreased from 0.8 to 0.6% (Figure 1). Cumulatively, the prevalence of presenting with either moderate or high risk of overdose decreased from 3.3 to 2.2% between 2013 to 2016.

Naloxone Dispensing

In 2013, naloxone was not dispensed to any patients at moderate or high risk for overdose between 365 days prior to surgery until 7 days after discharge (Table 2 and Figure 2). Low-risk group naloxone dispensing increased to 2 (0.0%) in 2014, to 13 (0.1%), in 2015, and to 86 (0.9%) in 2016. Moderate-risk group naloxone dispensing remained at 0 (0.0%) in 2014, but increased to 8 (3.5%) in 2015, and to 18 (10.9%) in 2016. High-risk group naloxone dispensing remained at 0 (0.0%) in 2014, but increased to 5 (5.8%) in 2015, and to 8 (12.7%) in 2016 (Figure 3).

Discussion

Our data demonstrate that patients presenting for TKA between 2013 and 2016 routinely had individual risk factors for overdose related to either prescription drug use or comorbidities. We also show that, although the number of patients at moderate and high risk for opioid overdose is decreasing, 2.2% of TKA patients remain at moderate or high risk for opioid overdose based on a weighted combination of these individual risk factors using RIOSORD. As demand for primary TKA is projected to grow to 3.5 million procedures by 2030, using prevalence from 2016, we estimate that 76,560 patients may present for TKA across the US with moderate or high risk for opioid overdose.⁹ Following discharge, this risk may be even higher as this estimate does not yet account for postoperative opioid use. We demonstrate that through a VHA OEND initiative, naloxone distribution increased and appeared to be targeted to those most at risk using a simple validated tool like RIOSORD.

Presence of an individual risk factor for overdose was present in as many as 63.1% of patients presenting for TKA, as was seen in 2013 with preoperative opioid use. The 3 highest scoring prescription use–related risk factors in RIOSORD are use of opioids ≥ 100 MEDD (16 points), ≥ 50 MEDD (9 points), and long-acting formulations (9 points). All 3 decreased in prevalence over the study period but by 2016 were still seen in 2.2% for ≥ 100 MEDD, 5.3% for ≥ 50 MEDD, and 5.8% for long-acting opioids. This decrease was not surprising given implementation of a VHA-wide opioid safety initiative and the OEND program, but this could also be related to changes in patient selection for surgery in the context of increased awareness of the opioid epidemic. Despite the trend toward safer opioid prescribing, by 2016 over half of patients (50.5%) who presented for TKA were already taking opioids, with 10.6% (543 of 5127) on doses ≥ 50 MEDD.

We observed a decrease in RIOSORD risk each year, consistent with decreasing prescription-related risk factors over time. This was most obvious in the moderate-risk group. It is unclear why a similar decrease was not as obvious in the high-risk group, but this in part may be due to the already low numbers of patients in the high-risk group. This may also represent the high-risk group being somewhat resistant to the initiatives that shifted moderate-risk patients to the low-risk group. There were proportionately more patients in the moderate- and high-risk groups in the original RIOSORD population than in our surgical population, which may be attributed to the fewer comorbidities seen in our surgical population, as well as the higher opioid-prescribing patterns seen prior to the VA OEND initiative.¹²

Naloxone prescribing was rare prior to the OEND initiative and increased from 2013 to 2016. Increases were most marked in those in moderate- and high-risk groups, although naloxone prescribing also increased among the low-risk group. Integration of RIOSORD stratification into the OEND initiative likely played a role in targeting increased access to naloxone among those at highest risk of overdose. Naloxone dispensing increased for every group, although a significant proportion of moderate- and high-risk patients, 89.1% and 87.3%, respectively, were still not dispensed naloxone by 2016. Moreover, our estimates of perioperative naloxone access were likely an overestimate by including patients dispensed naloxone up to 1 year before surgery until 7 days after surgery. The aim was to include patients who may not have been prescribed naloxone postoperatively because of an existing naloxone prescription at home. Perioperative naloxone access estimates would have been even lower if a narrower window had been used to approximate perioperative access. This identifies an important gap between those who may benefit from naloxone dispensing and those who received naloxone. This in part may be because OEND has not been implemented as routinely in surgical settings as other settings (eg, primary care). OEND efforts may more effectively increase naloxone prescribing among surgical patients if these efforts were targeted at surgical and anesthesia departments. Given that the Comprehensive Addiction and Recovery Act of 2016 requires an assessment of patient risk prior to opioid prescribing and VHA efforts to increase utilization of tools like the Stratification Tool for Opioid Risk Mitigation (STORM), which estimates patient risk when initiating an opioid prescription and includes naloxone as one of many risk mitigation strategies, we anticipate that rates of naloxone prescribing will increase over time.

Limitations

Our study captures a large number of patients across VHA hospitals of varying size nationwide, including a mix of those with and without academic medical center affiliations. This veteran population may not represent the US commercially insured population (CIP). Zedler and colleagues highlighted the differences in prevalence of individual risk factors: notably, the CIP had a substantially higher proportion of females and younger patients.¹¹ VHA had a greater prevalence of common chronic conditions associated with older age. The frequency of opioid dependence was similar among CIP and VHA. However, substance abuse and nonopioid substance dependence diagnoses were 4-fold more frequent among VHA controls as CIP controls. Prescribing of all opioids, except morphine and methadone, was substantially greater in CIP than in VHA.¹¹ Despite a difference in individual risk factors, a CIP-specific RIOSORD has been validated and can be used outside of the VHA to obviate the limitations of the VHA-specific RIOSORD.¹¹

Other limitations include our estimation of naloxone access. We do not know whether naloxone was administered or have a reliable estimate of overdose incidence in this postoperative TKA population. Also, it is important to note that RIOSORD was not developed for surgical patients. The use of RIOSORD in a postoperative population likely underestimates risk of opioid overdose due to the frequent prescriptions of new opioids or escalation of existing MEDD to the postoperative patient. Our study was also retrospective in nature and reliant on accurate coding of patient risk factors. It is possible that comorbidities were not accurately identified by EHR and therefore subject to inconsistency.

Conclusions

Veterans presenting for TKA routinely have risk factors for opioid overdose. We observed a trend toward decreasing overdose risk which coincided with the Opioid Safety and OEND initiatives within the VHA. We also observed an increase in naloxone prescription for moderate- and high-risk patients undergoing TKA, although most of these patients still did not receive naloxone as of 2016. More research is needed to refine and validate the RIOSORD score for surgical populations. Expanding initiatives such as OEND to include surgical patients presents an opportunity to improve access to naloxone for postoperative patients that may help reduce opioid overdose in this population.

Opioid overdose is a major public health challenge, with recent reports estimating 41 deaths per day in the United States from prescription opioid overdose.^1,2 Prescribing naloxone has increasingly been advocated to reduce the risk of opioid overdose for patients identified as high risk. Naloxone distribution has been shown to decrease the incidence of opioid overdoses in the general population.^3,4 The Centers for Disease Control and Prevention (CDC) Guideline for Prescribing Opioids for Chronic Pain recommends considering naloxone prescription for patients with a history of overdose or substance use disorder, opioid dosages ≥ 50 morphine equivalent daily dose (MEDD), and concurrent use of benzodiazepines.⁵

Although the CDC guidelines are intended for primary care clinicians in outpatient settings, naloxone prescribing is also relevant in the postsurgical setting.⁵ Many surgical patients are at risk for opioid overdose and data from the Veterans Health Administration (VHA) has shown that risk of opioid overdose is 11-fold higher in the 30 days following discharge from a surgical admission, when compared with the subsequent calendar year.^6,7 This likely occurs due to new prescriptions or escalated doses of opioids following surgery. Overdose risk may be particularly relevant to orthopedic surgery as postoperative opioids are commonly prescribed.⁸ Patients undergoing total knee arthroplasty (TKA) may represent a vulnerable population to overdose as it is one of the most commonly performed surgeries for the treatment of chronic pain, and is frequently performed in older adults with medical comorbidities.^9,10

Identifying patients at high risk for opioid overdose is important for targeted naloxone dispensing.⁵ A risk index for overdose or serious opioid-induced respiratory depression (RIOSORD) tool has been developed and validated in veteran and other populations to identify such patients.¹¹ The RIOSORD tool classifies patients by risk level (1-10) and predicts probability of overdose or serious opioid-induced respiratory depression (OSORD). A patient’s level of risk is based on a weighted combination of the 15 independent risk factors most highly associated with OSORD, including comorbid conditions, prescription drug use, and health care utilization.¹² Using the RIOSORD tool, the VHA Opioid Education and Naloxone Distribution (OEND) program is a risk mitigation initiative that aims to decrease opioid-related overdose morbidity and mortality. This is achieved via opioid overdose education for prevention, recognition, and response and includes outpatient naloxone prescription.^13,14

Despite the comprehensive OEND program, there exists very little data to guide postsurgical naloxone prescribing. The prevalence of known risk factors for overdose in surgical patients remains unknown, as does the prevalence of perioperative naloxone distribution. Understanding overdose risk factors and naloxone prescribing patterns in surgical patients may identify potential targets for OEND efforts. This study retrospectively estimated RIOSORD scores for TKA patients between 2013 to 2016 and described naloxone distribution based on RIOSORD scores and risk factors.

Methods

We identified patients who had undergone primary TKA at VHA hospitals using Current Procedural Terminology (CPT), International Classification of Diseases, Ninth Revision (ICD-9) procedure codes, and data extracted from the VHA Corporate Data Warehouse (CDW) of electronic health records (EHRs). Our study was granted approval with exemption from informed consent by the Durham Veteran Affairs Healthcare System Institutional Review Board.

This retrospective cohort study included all veterans who underwent elective primary TKA from January 1, 2013 through December 31, 2016. We excluded patients who died before discharge.

Outcomes

Our primary outcome was being dispensed an outpatient naloxone prescription following TKA. Naloxone dispensing was identified by examining CDW outpatient pharmacy records with a final dispense date from 1 year before surgery through 7 days after discharge following TKA. To exclude naloxone administration that may have been given in a clinic, prescription data included only records with an outpatient prescription copay. Naloxone dispensing in the year before surgery was chosen to estimate likely preoperative possession of naloxone which could be available in the postoperative period. Naloxone dispensing until 7 days after discharge was chosen to identify any new dispensing that would be available in the postoperative period. These outcomes were examined over the study time frame on an annual basis.

Patient Factors

Demographic variables included age, sex, and race/ethnicity. Independent risk factors for overdose from RIOSORD were identified for each patient.¹⁵ These risk factors included comorbidities (opioid use disorder, schizophrenia, bipolar disorder, liver disease, chronic kidney disease, sleep apnea, or lung disease) and prescription drug use (use of opioids, benzodiazepines, long-acting opioids, ≥ 50 MEDD or ≥ 100 MEDD). ICD-9 and ICD-10 diagnosis codes were used to identify comorbidities. Risk classes on day of surgery were identified using a RIOSORD algorithm code. Consistent with the display of RIOSORD risk classes on the VHA Academic Detailing Service OEND risk report, patients were grouped into 3 groups based on their RIOSORD score: classes 1 to 4 (low risk), 5 to 7 (moderate risk), and 8 to 10 (high risk).

Descriptive statistics were used to summarize data on patient demographics, RIOSORD risk factors, overdose events, and naloxone dispensing over time.

Results

The study cohort included 38,011 veterans who underwent primary TKA in the VHA between January 1, 2013 and December 30, 2016. In this cohort, the mean age was 65 years, 93% were male, and 77% were White patients (Table 1). The most common comorbidities were lung disease in 9170 (24.1%) patients, sleep apnea in 6630 (17.4%) patients, chronic kidney disease in 4036 (10.6%) patients, liver disease in 2822 (7.4%) patients, and bipolar disorder in 1748 (4.6%) patients.

In 2013, 63.1% of patients presenting for surgery were actively prescribed opioids. By 2016, this decreased to 50.5%. Benzodiazepine use decreased from 13.2 to 8.8% and long-acting opioid use decreased from 8.5 to 5.8% over the same period. Patients taking ≥ 50 MEDD decreased from 8.0 to 5.3% and patients taking ≥ 100 MEDD decreased from 3.3 to 2.2%. The prevalence of moderate-risk patients decreased from 2.5 to 1.6% and high-risk patients decreased from 0.8 to 0.6% (Figure 1). Cumulatively, the prevalence of presenting with either moderate or high risk of overdose decreased from 3.3 to 2.2% between 2013 to 2016.

Naloxone Dispensing

In 2013, naloxone was not dispensed to any patients at moderate or high risk for overdose between 365 days prior to surgery until 7 days after discharge (Table 2 and Figure 2). Low-risk group naloxone dispensing increased to 2 (0.0%) in 2014, to 13 (0.1%), in 2015, and to 86 (0.9%) in 2016. Moderate-risk group naloxone dispensing remained at 0 (0.0%) in 2014, but increased to 8 (3.5%) in 2015, and to 18 (10.9%) in 2016. High-risk group naloxone dispensing remained at 0 (0.0%) in 2014, but increased to 5 (5.8%) in 2015, and to 8 (12.7%) in 2016 (Figure 3).

Discussion

Our data demonstrate that patients presenting for TKA between 2013 and 2016 routinely had individual risk factors for overdose related to either prescription drug use or comorbidities. We also show that, although the number of patients at moderate and high risk for opioid overdose is decreasing, 2.2% of TKA patients remain at moderate or high risk for opioid overdose based on a weighted combination of these individual risk factors using RIOSORD. As demand for primary TKA is projected to grow to 3.5 million procedures by 2030, using prevalence from 2016, we estimate that 76,560 patients may present for TKA across the US with moderate or high risk for opioid overdose.⁹ Following discharge, this risk may be even higher as this estimate does not yet account for postoperative opioid use. We demonstrate that through a VHA OEND initiative, naloxone distribution increased and appeared to be targeted to those most at risk using a simple validated tool like RIOSORD.

Presence of an individual risk factor for overdose was present in as many as 63.1% of patients presenting for TKA, as was seen in 2013 with preoperative opioid use. The 3 highest scoring prescription use–related risk factors in RIOSORD are use of opioids ≥ 100 MEDD (16 points), ≥ 50 MEDD (9 points), and long-acting formulations (9 points). All 3 decreased in prevalence over the study period but by 2016 were still seen in 2.2% for ≥ 100 MEDD, 5.3% for ≥ 50 MEDD, and 5.8% for long-acting opioids. This decrease was not surprising given implementation of a VHA-wide opioid safety initiative and the OEND program, but this could also be related to changes in patient selection for surgery in the context of increased awareness of the opioid epidemic. Despite the trend toward safer opioid prescribing, by 2016 over half of patients (50.5%) who presented for TKA were already taking opioids, with 10.6% (543 of 5127) on doses ≥ 50 MEDD.

We observed a decrease in RIOSORD risk each year, consistent with decreasing prescription-related risk factors over time. This was most obvious in the moderate-risk group. It is unclear why a similar decrease was not as obvious in the high-risk group, but this in part may be due to the already low numbers of patients in the high-risk group. This may also represent the high-risk group being somewhat resistant to the initiatives that shifted moderate-risk patients to the low-risk group. There were proportionately more patients in the moderate- and high-risk groups in the original RIOSORD population than in our surgical population, which may be attributed to the fewer comorbidities seen in our surgical population, as well as the higher opioid-prescribing patterns seen prior to the VA OEND initiative.¹²

Naloxone prescribing was rare prior to the OEND initiative and increased from 2013 to 2016. Increases were most marked in those in moderate- and high-risk groups, although naloxone prescribing also increased among the low-risk group. Integration of RIOSORD stratification into the OEND initiative likely played a role in targeting increased access to naloxone among those at highest risk of overdose. Naloxone dispensing increased for every group, although a significant proportion of moderate- and high-risk patients, 89.1% and 87.3%, respectively, were still not dispensed naloxone by 2016. Moreover, our estimates of perioperative naloxone access were likely an overestimate by including patients dispensed naloxone up to 1 year before surgery until 7 days after surgery. The aim was to include patients who may not have been prescribed naloxone postoperatively because of an existing naloxone prescription at home. Perioperative naloxone access estimates would have been even lower if a narrower window had been used to approximate perioperative access. This identifies an important gap between those who may benefit from naloxone dispensing and those who received naloxone. This in part may be because OEND has not been implemented as routinely in surgical settings as other settings (eg, primary care). OEND efforts may more effectively increase naloxone prescribing among surgical patients if these efforts were targeted at surgical and anesthesia departments. Given that the Comprehensive Addiction and Recovery Act of 2016 requires an assessment of patient risk prior to opioid prescribing and VHA efforts to increase utilization of tools like the Stratification Tool for Opioid Risk Mitigation (STORM), which estimates patient risk when initiating an opioid prescription and includes naloxone as one of many risk mitigation strategies, we anticipate that rates of naloxone prescribing will increase over time.

Limitations

Our study captures a large number of patients across VHA hospitals of varying size nationwide, including a mix of those with and without academic medical center affiliations. This veteran population may not represent the US commercially insured population (CIP). Zedler and colleagues highlighted the differences in prevalence of individual risk factors: notably, the CIP had a substantially higher proportion of females and younger patients.¹¹ VHA had a greater prevalence of common chronic conditions associated with older age. The frequency of opioid dependence was similar among CIP and VHA. However, substance abuse and nonopioid substance dependence diagnoses were 4-fold more frequent among VHA controls as CIP controls. Prescribing of all opioids, except morphine and methadone, was substantially greater in CIP than in VHA.¹¹ Despite a difference in individual risk factors, a CIP-specific RIOSORD has been validated and can be used outside of the VHA to obviate the limitations of the VHA-specific RIOSORD.¹¹

Other limitations include our estimation of naloxone access. We do not know whether naloxone was administered or have a reliable estimate of overdose incidence in this postoperative TKA population. Also, it is important to note that RIOSORD was not developed for surgical patients. The use of RIOSORD in a postoperative population likely underestimates risk of opioid overdose due to the frequent prescriptions of new opioids or escalation of existing MEDD to the postoperative patient. Our study was also retrospective in nature and reliant on accurate coding of patient risk factors. It is possible that comorbidities were not accurately identified by EHR and therefore subject to inconsistency.

Conclusions

Veterans presenting for TKA routinely have risk factors for opioid overdose. We observed a trend toward decreasing overdose risk which coincided with the Opioid Safety and OEND initiatives within the VHA. We also observed an increase in naloxone prescription for moderate- and high-risk patients undergoing TKA, although most of these patients still did not receive naloxone as of 2016. More research is needed to refine and validate the RIOSORD score for surgical populations. Expanding initiatives such as OEND to include surgical patients presents an opportunity to improve access to naloxone for postoperative patients that may help reduce opioid overdose in this population.

There are approximately 2 million female veterans in the United States, representing about 10% of the veteran population.¹ In 2015, 456,000 female veterans used the US Department of Veterans Affairs (VA) health care services. The VA predicts an increase in utilization over the next 20 years.²

Female veterans are more likely to have musculoskeletal disorder multimorbidity compared with male veterans and have higher rates of depressive and bipolar disorders, anxiety, and posttraumatic stress disorder (PTSD).^3,4 Compared with male veterans, female veterans are younger, more likely to be unmarried and to have served during the wars in Iraq and Afghanistan.³ Fifty-five percent of women veterans vs 41% of men veterans have a service-connected disability, and a greater percentage of women veterans have a service connection rating > 50%.⁵ The top service-connected disabilities for women veterans are PTSD, major depressive disorder, migraines, and lumbosacral or cervical strain.² In addition, one-third of women veterans using VA health care report experiencing military sexual trauma (MST).⁶ Military service may impact the health of female veterans both physically and mentally. Providing treatments and programs to improve their health and their health care experience are current VA priorities.

The VA is changing the way health care is conceptualized and delivered by implementing a holistic model of care known as Whole Health, which seeks to empower and equip patients to take charge of their health, blending conventional medicine with self-care and complementary and integrative health (CIH) approaches, such as massage therapy, yoga, acupuncture, and meditation.⁷ CIH therapies can help improve physical and mental health with little to no adverse effects.^8-10

As part of the Whole Health initiative at the VA Ann Arbor Healthcare System (VAAAHS) in Michigan, the massage program was expanded in 2017 to offer relaxation massages to female veterans attending the women’s health clinic, which provides gynecologic care. Patients visiting a gynecology clinic often experience anxiety and pain related to invasive procedures and examinations. This is especially true for female veterans who experienced MST.¹¹

VAAAHS has 1 staff massage therapist (MT). To expand the program to the women’s health clinic, volunteer licensed MTs were recruited and trained in specific procedures by the staff MT.

Several studies have demonstrated the effect of therapeutic massage on pain and anxiety in predominantly male veteran study populations, including veterans needing postsurgical and palliative care as well as those experiencing chronic pain and knee osteoarthritis.^12-16 Little is known about the effects of massage therapy on female veterans. The purpose of this pilot study was to examine the effects of massage therapy among female veterans participating in the women’s health massage program.

Methods

The setting for this pre-post intervention study was VAAAHS. Veterans were called in advance by clinic staff and scheduled for 60-minute appointments either before or after their clinic appointment, depending on availability. MTs were instructed to provide relaxation massage using Swedish massage techniques with moderate pressure, avoiding deep pressure techniques. Swedish massage was selected to compare with previous veteran studies and because these techniques were approved for delivery by volunteer MTs. Massages were given in a private space on a massage table and were limited to the back, neck, head/face, and extremities.

The volunteer MTs gave the participants a survey to provide comments and to rate baseline pain and other symptoms prior to and following the massage. The MT left the room to provide privacy while completing the survey. The staff included the symptom data in the massage note as clinical outcomes and entered them into the electronic health record. Massages were given from October 1, 2017 to June 30, 2018. Data including symptom scores, demographics, the presence of chronic pain, mental health diagnoses, patient comments, and opioid use were abstracted from the electronic health record by 2 members of the study team and entered into an Excel database. This study was approved by the VAAAHS Institutional Review Board.

Study Measures

Pain intensity, pain unpleasantness (the affective component of pain), anxiety, shortness of breath, relaxation, and inner peace were rated pre- and postmassage on a 0 to 10 scale. Shortness of breath was included due to the relationship between breathing and anxiety. Inner peace was assessed to measure the calming effects of massage therapy. Beck and colleagues found the concept of inner peace was an important outcome of massage therapy.¹⁷ The scale anchors for pain intensity were “no pain” and “severe pain”; and “not at all unpleasant” and “as unpleasant as it can be” for pain unpleasantness. For anxiety, the anchors were “no anxiety” and “as anxious as I can be.” Anchors for relaxation and inner peace were reversed so that a 0 indicated low relaxation and inner peace while a 10 indicated the highest state of relaxation and inner peace.

Chronic pain was defined as pain existing for > 3 months. A history of chronic pain was determined from a review and synthesis of primary care and specialty care recorded diagnoses, patient concerns, and service-connected disabilities. The diagnoses included lumbosacral or cervical strain, chronic low back, joint (knee, shoulder, hip, ankle), neck, or pelvic pain, fibromyalgia, headache, migraine, osteoarthritis, and myofascial pain syndrome. The presence of mental health conditions, including depression, anxiety, bipolar disorders, and PTSD, were similarly determined by a review of mental health clinical notes. Sex was determined from the gynecology note.

Statistical Analysis

Means and medians were calculated for short-term changes in symptom scores. Due to skewness in the short-term changes, significance was tested using a nonparametric sign test. Significance was adjusted using the Bonferroni correction to protect the overall type I error level at 5% from multiple testing. We also assessed for differences in symptom changes in 4 subgroups, using an unadjusted general linear model: those with (1) chronic pain vs without; (2) an anxiety diagnosis vs without; (3) depression vs without; and (4) a PTSD diagnosis vs without. Data were analyzed using SPSS 25 and SAS 9.4.

Results

Results are based on the first massage received by 96 unique individuals (Table 1). Fifty-one (53%) patients were aged 21 to 40 years, and 45 (47%) were aged ≥ 41 years. Most participants (80%) had had a previous massage. Seven (7%) participants were currently on prescription opioids; 76 (79%) participants had a history of one or more chronic pain diagnoses (eg, back pain, migraine headaches, fibromyalgia) and 78 (81%) had a history of a mental health diagnosis (eg, depression, anxiety, PTSD). Massage sessions ranged from 30 to 60 minutes; most patients received massage therapy for 50 minutes.

Prior to massage, mean scores were 3.9 pain intensity, 3.7 pain unpleasantness, 3.8 anxiety, 1.0 shortness of breath, 4.0 relaxation, and 4.2 inner peace. Short-term changes in symptom scores are shown in Table 2. The mean score for pain intensity decreased by 1.9 points, pain unpleasantness by 2.0 points, anxiety by 2.4 points. The greatest change occurred for relaxation, which increased by 4.3 points. All changes in symptoms were statistically significant (P < .001). For subgroup comparisons, there were no significant differences in symptom scores for patients with a diagnosis of anxiety vs without and depression vs without (Table 3). However, anxiety in patients diagnosed with PTSD decreased by 3.3 points compared with 2.0 in patients without PTSD (P = .005). For patients with chronic pain, inner peace increased 3.9 points compared with an increase of 2.0 in patients without chronic pain (P = .002).

Discussion

Massage therapy may be a useful treatment for female veterans experiencing chronic pain, anxiety disorders, depression, or situational anxiety related to gynecologic procedures. After receiving a relaxation massage, female veterans reported decreased pain intensity, pain unpleasantness, and anxiety while reporting increased relaxation and feelings of inner peace. The effects of massage were consistent for all the symptoms or characteristics assessed, suggesting that massage may act on the body in multiple ways.

These changes parallel those seen in a palliative care population primarily composed of male veterans.¹⁴ However, the female veterans in this cohort experienced greater changes in relaxation and feelings of inner peace, which may be partly due to relief of tension related to an upcoming stressful appointment. The large mean decrease in anxiety level among female veterans with PTSD is notable as well as the larger increase in inner peace in those with chronic pain.

Many patients expressed their gratitude for the massage and interest in having access to more massage therapy. Female patients who have experienced sexual trauma or other trauma may especially benefit from massage prior to painful, invasive gynecologic procedures. Anecdotally, 2 nurse chaperones in the clinic mentioned separately to the massage program supervisor that the massages helped some very anxious women better tolerate an invasive procedure that would have been otherwise extremely difficult.

Limitations

This study has several limitations including lack of a control group, a self-selected population, the lack of objective biochemical measurements, and possible respondent bias to please the MTs. Eighty percent had previously experienced massage therapy and may have been biased toward the effects of massage before receiving the intervention. The first report of the effects of massage therapy in an exclusively female veteran population is a major strength of this study.

Further research including randomized controlled trials is needed, especially in populations with coexisting chronic pain and mental health disorders, as is exploring the acceptability of massage therapy for female veterans with MST. Finding viable alternatives to medications has become even more important as the nation addresses the challenge of the opioid crisis.^45,46

Conclusions

Female veterans are increasingly seeking VA health care. Although further research is needed, results from this pilot study suggest massage therapy may be an effective, inexpensive, and safe treatment for pain and/or anxiety in female veterans. Massage may be especially beneficial for female veterans who experience both chronic pain and mental health conditions. Providing female veterans with access to massage therapy may encourage better self-care and utilization of other Whole Health services, leading to overall improved health and well-being. VA Whole Health programs should consider targeting female veterans for massage therapy services.

Acknowledgments

The authors express our gratitude to the Women Veteran Program Manager, Cheryl Allen, RN; Massage Therapists Denise McGee and Kimberly Morro; Dara Ganoczy, MPH, for help with statistical analysis; and Mark Hausman, MD, for leadership support.

There are approximately 2 million female veterans in the United States, representing about 10% of the veteran population.¹ In 2015, 456,000 female veterans used the US Department of Veterans Affairs (VA) health care services. The VA predicts an increase in utilization over the next 20 years.²

Female veterans are more likely to have musculoskeletal disorder multimorbidity compared with male veterans and have higher rates of depressive and bipolar disorders, anxiety, and posttraumatic stress disorder (PTSD).^3,4 Compared with male veterans, female veterans are younger, more likely to be unmarried and to have served during the wars in Iraq and Afghanistan.³ Fifty-five percent of women veterans vs 41% of men veterans have a service-connected disability, and a greater percentage of women veterans have a service connection rating > 50%.⁵ The top service-connected disabilities for women veterans are PTSD, major depressive disorder, migraines, and lumbosacral or cervical strain.² In addition, one-third of women veterans using VA health care report experiencing military sexual trauma (MST).⁶ Military service may impact the health of female veterans both physically and mentally. Providing treatments and programs to improve their health and their health care experience are current VA priorities.

The VA is changing the way health care is conceptualized and delivered by implementing a holistic model of care known as Whole Health, which seeks to empower and equip patients to take charge of their health, blending conventional medicine with self-care and complementary and integrative health (CIH) approaches, such as massage therapy, yoga, acupuncture, and meditation.⁷ CIH therapies can help improve physical and mental health with little to no adverse effects.^8-10

As part of the Whole Health initiative at the VA Ann Arbor Healthcare System (VAAAHS) in Michigan, the massage program was expanded in 2017 to offer relaxation massages to female veterans attending the women’s health clinic, which provides gynecologic care. Patients visiting a gynecology clinic often experience anxiety and pain related to invasive procedures and examinations. This is especially true for female veterans who experienced MST.¹¹

VAAAHS has 1 staff massage therapist (MT). To expand the program to the women’s health clinic, volunteer licensed MTs were recruited and trained in specific procedures by the staff MT.

Several studies have demonstrated the effect of therapeutic massage on pain and anxiety in predominantly male veteran study populations, including veterans needing postsurgical and palliative care as well as those experiencing chronic pain and knee osteoarthritis.^12-16 Little is known about the effects of massage therapy on female veterans. The purpose of this pilot study was to examine the effects of massage therapy among female veterans participating in the women’s health massage program.

Methods

The setting for this pre-post intervention study was VAAAHS. Veterans were called in advance by clinic staff and scheduled for 60-minute appointments either before or after their clinic appointment, depending on availability. MTs were instructed to provide relaxation massage using Swedish massage techniques with moderate pressure, avoiding deep pressure techniques. Swedish massage was selected to compare with previous veteran studies and because these techniques were approved for delivery by volunteer MTs. Massages were given in a private space on a massage table and were limited to the back, neck, head/face, and extremities.

The volunteer MTs gave the participants a survey to provide comments and to rate baseline pain and other symptoms prior to and following the massage. The MT left the room to provide privacy while completing the survey. The staff included the symptom data in the massage note as clinical outcomes and entered them into the electronic health record. Massages were given from October 1, 2017 to June 30, 2018. Data including symptom scores, demographics, the presence of chronic pain, mental health diagnoses, patient comments, and opioid use were abstracted from the electronic health record by 2 members of the study team and entered into an Excel database. This study was approved by the VAAAHS Institutional Review Board.

Study Measures

Pain intensity, pain unpleasantness (the affective component of pain), anxiety, shortness of breath, relaxation, and inner peace were rated pre- and postmassage on a 0 to 10 scale. Shortness of breath was included due to the relationship between breathing and anxiety. Inner peace was assessed to measure the calming effects of massage therapy. Beck and colleagues found the concept of inner peace was an important outcome of massage therapy.¹⁷ The scale anchors for pain intensity were “no pain” and “severe pain”; and “not at all unpleasant” and “as unpleasant as it can be” for pain unpleasantness. For anxiety, the anchors were “no anxiety” and “as anxious as I can be.” Anchors for relaxation and inner peace were reversed so that a 0 indicated low relaxation and inner peace while a 10 indicated the highest state of relaxation and inner peace.

Chronic pain was defined as pain existing for > 3 months. A history of chronic pain was determined from a review and synthesis of primary care and specialty care recorded diagnoses, patient concerns, and service-connected disabilities. The diagnoses included lumbosacral or cervical strain, chronic low back, joint (knee, shoulder, hip, ankle), neck, or pelvic pain, fibromyalgia, headache, migraine, osteoarthritis, and myofascial pain syndrome. The presence of mental health conditions, including depression, anxiety, bipolar disorders, and PTSD, were similarly determined by a review of mental health clinical notes. Sex was determined from the gynecology note.

Statistical Analysis

Means and medians were calculated for short-term changes in symptom scores. Due to skewness in the short-term changes, significance was tested using a nonparametric sign test. Significance was adjusted using the Bonferroni correction to protect the overall type I error level at 5% from multiple testing. We also assessed for differences in symptom changes in 4 subgroups, using an unadjusted general linear model: those with (1) chronic pain vs without; (2) an anxiety diagnosis vs without; (3) depression vs without; and (4) a PTSD diagnosis vs without. Data were analyzed using SPSS 25 and SAS 9.4.

Results

Results are based on the first massage received by 96 unique individuals (Table 1). Fifty-one (53%) patients were aged 21 to 40 years, and 45 (47%) were aged ≥ 41 years. Most participants (80%) had had a previous massage. Seven (7%) participants were currently on prescription opioids; 76 (79%) participants had a history of one or more chronic pain diagnoses (eg, back pain, migraine headaches, fibromyalgia) and 78 (81%) had a history of a mental health diagnosis (eg, depression, anxiety, PTSD). Massage sessions ranged from 30 to 60 minutes; most patients received massage therapy for 50 minutes.

Prior to massage, mean scores were 3.9 pain intensity, 3.7 pain unpleasantness, 3.8 anxiety, 1.0 shortness of breath, 4.0 relaxation, and 4.2 inner peace. Short-term changes in symptom scores are shown in Table 2. The mean score for pain intensity decreased by 1.9 points, pain unpleasantness by 2.0 points, anxiety by 2.4 points. The greatest change occurred for relaxation, which increased by 4.3 points. All changes in symptoms were statistically significant (P < .001). For subgroup comparisons, there were no significant differences in symptom scores for patients with a diagnosis of anxiety vs without and depression vs without (Table 3). However, anxiety in patients diagnosed with PTSD decreased by 3.3 points compared with 2.0 in patients without PTSD (P = .005). For patients with chronic pain, inner peace increased 3.9 points compared with an increase of 2.0 in patients without chronic pain (P = .002).

Discussion

Massage therapy may be a useful treatment for female veterans experiencing chronic pain, anxiety disorders, depression, or situational anxiety related to gynecologic procedures. After receiving a relaxation massage, female veterans reported decreased pain intensity, pain unpleasantness, and anxiety while reporting increased relaxation and feelings of inner peace. The effects of massage were consistent for all the symptoms or characteristics assessed, suggesting that massage may act on the body in multiple ways.

These changes parallel those seen in a palliative care population primarily composed of male veterans.¹⁴ However, the female veterans in this cohort experienced greater changes in relaxation and feelings of inner peace, which may be partly due to relief of tension related to an upcoming stressful appointment. The large mean decrease in anxiety level among female veterans with PTSD is notable as well as the larger increase in inner peace in those with chronic pain.

Many patients expressed their gratitude for the massage and interest in having access to more massage therapy. Female patients who have experienced sexual trauma or other trauma may especially benefit from massage prior to painful, invasive gynecologic procedures. Anecdotally, 2 nurse chaperones in the clinic mentioned separately to the massage program supervisor that the massages helped some very anxious women better tolerate an invasive procedure that would have been otherwise extremely difficult.

Limitations

This study has several limitations including lack of a control group, a self-selected population, the lack of objective biochemical measurements, and possible respondent bias to please the MTs. Eighty percent had previously experienced massage therapy and may have been biased toward the effects of massage before receiving the intervention. The first report of the effects of massage therapy in an exclusively female veteran population is a major strength of this study.

Further research including randomized controlled trials is needed, especially in populations with coexisting chronic pain and mental health disorders, as is exploring the acceptability of massage therapy for female veterans with MST. Finding viable alternatives to medications has become even more important as the nation addresses the challenge of the opioid crisis.^45,46

Conclusions

Female veterans are increasingly seeking VA health care. Although further research is needed, results from this pilot study suggest massage therapy may be an effective, inexpensive, and safe treatment for pain and/or anxiety in female veterans. Massage may be especially beneficial for female veterans who experience both chronic pain and mental health conditions. Providing female veterans with access to massage therapy may encourage better self-care and utilization of other Whole Health services, leading to overall improved health and well-being. VA Whole Health programs should consider targeting female veterans for massage therapy services.

Acknowledgments

The authors express our gratitude to the Women Veteran Program Manager, Cheryl Allen, RN; Massage Therapists Denise McGee and Kimberly Morro; Dara Ganoczy, MPH, for help with statistical analysis; and Mark Hausman, MD, for leadership support.

There are approximately 2 million female veterans in the United States, representing about 10% of the veteran population.¹ In 2015, 456,000 female veterans used the US Department of Veterans Affairs (VA) health care services. The VA predicts an increase in utilization over the next 20 years.²

Female veterans are more likely to have musculoskeletal disorder multimorbidity compared with male veterans and have higher rates of depressive and bipolar disorders, anxiety, and posttraumatic stress disorder (PTSD).^3,4 Compared with male veterans, female veterans are younger, more likely to be unmarried and to have served during the wars in Iraq and Afghanistan.³ Fifty-five percent of women veterans vs 41% of men veterans have a service-connected disability, and a greater percentage of women veterans have a service connection rating > 50%.⁵ The top service-connected disabilities for women veterans are PTSD, major depressive disorder, migraines, and lumbosacral or cervical strain.² In addition, one-third of women veterans using VA health care report experiencing military sexual trauma (MST).⁶ Military service may impact the health of female veterans both physically and mentally. Providing treatments and programs to improve their health and their health care experience are current VA priorities.

The VA is changing the way health care is conceptualized and delivered by implementing a holistic model of care known as Whole Health, which seeks to empower and equip patients to take charge of their health, blending conventional medicine with self-care and complementary and integrative health (CIH) approaches, such as massage therapy, yoga, acupuncture, and meditation.⁷ CIH therapies can help improve physical and mental health with little to no adverse effects.^8-10

As part of the Whole Health initiative at the VA Ann Arbor Healthcare System (VAAAHS) in Michigan, the massage program was expanded in 2017 to offer relaxation massages to female veterans attending the women’s health clinic, which provides gynecologic care. Patients visiting a gynecology clinic often experience anxiety and pain related to invasive procedures and examinations. This is especially true for female veterans who experienced MST.¹¹

VAAAHS has 1 staff massage therapist (MT). To expand the program to the women’s health clinic, volunteer licensed MTs were recruited and trained in specific procedures by the staff MT.

Several studies have demonstrated the effect of therapeutic massage on pain and anxiety in predominantly male veteran study populations, including veterans needing postsurgical and palliative care as well as those experiencing chronic pain and knee osteoarthritis.^12-16 Little is known about the effects of massage therapy on female veterans. The purpose of this pilot study was to examine the effects of massage therapy among female veterans participating in the women’s health massage program.

Methods

The setting for this pre-post intervention study was VAAAHS. Veterans were called in advance by clinic staff and scheduled for 60-minute appointments either before or after their clinic appointment, depending on availability. MTs were instructed to provide relaxation massage using Swedish massage techniques with moderate pressure, avoiding deep pressure techniques. Swedish massage was selected to compare with previous veteran studies and because these techniques were approved for delivery by volunteer MTs. Massages were given in a private space on a massage table and were limited to the back, neck, head/face, and extremities.

The volunteer MTs gave the participants a survey to provide comments and to rate baseline pain and other symptoms prior to and following the massage. The MT left the room to provide privacy while completing the survey. The staff included the symptom data in the massage note as clinical outcomes and entered them into the electronic health record. Massages were given from October 1, 2017 to June 30, 2018. Data including symptom scores, demographics, the presence of chronic pain, mental health diagnoses, patient comments, and opioid use were abstracted from the electronic health record by 2 members of the study team and entered into an Excel database. This study was approved by the VAAAHS Institutional Review Board.

Study Measures

Pain intensity, pain unpleasantness (the affective component of pain), anxiety, shortness of breath, relaxation, and inner peace were rated pre- and postmassage on a 0 to 10 scale. Shortness of breath was included due to the relationship between breathing and anxiety. Inner peace was assessed to measure the calming effects of massage therapy. Beck and colleagues found the concept of inner peace was an important outcome of massage therapy.¹⁷ The scale anchors for pain intensity were “no pain” and “severe pain”; and “not at all unpleasant” and “as unpleasant as it can be” for pain unpleasantness. For anxiety, the anchors were “no anxiety” and “as anxious as I can be.” Anchors for relaxation and inner peace were reversed so that a 0 indicated low relaxation and inner peace while a 10 indicated the highest state of relaxation and inner peace.

Chronic pain was defined as pain existing for > 3 months. A history of chronic pain was determined from a review and synthesis of primary care and specialty care recorded diagnoses, patient concerns, and service-connected disabilities. The diagnoses included lumbosacral or cervical strain, chronic low back, joint (knee, shoulder, hip, ankle), neck, or pelvic pain, fibromyalgia, headache, migraine, osteoarthritis, and myofascial pain syndrome. The presence of mental health conditions, including depression, anxiety, bipolar disorders, and PTSD, were similarly determined by a review of mental health clinical notes. Sex was determined from the gynecology note.

Statistical Analysis

Means and medians were calculated for short-term changes in symptom scores. Due to skewness in the short-term changes, significance was tested using a nonparametric sign test. Significance was adjusted using the Bonferroni correction to protect the overall type I error level at 5% from multiple testing. We also assessed for differences in symptom changes in 4 subgroups, using an unadjusted general linear model: those with (1) chronic pain vs without; (2) an anxiety diagnosis vs without; (3) depression vs without; and (4) a PTSD diagnosis vs without. Data were analyzed using SPSS 25 and SAS 9.4.

Results

Results are based on the first massage received by 96 unique individuals (Table 1). Fifty-one (53%) patients were aged 21 to 40 years, and 45 (47%) were aged ≥ 41 years. Most participants (80%) had had a previous massage. Seven (7%) participants were currently on prescription opioids; 76 (79%) participants had a history of one or more chronic pain diagnoses (eg, back pain, migraine headaches, fibromyalgia) and 78 (81%) had a history of a mental health diagnosis (eg, depression, anxiety, PTSD). Massage sessions ranged from 30 to 60 minutes; most patients received massage therapy for 50 minutes.

Prior to massage, mean scores were 3.9 pain intensity, 3.7 pain unpleasantness, 3.8 anxiety, 1.0 shortness of breath, 4.0 relaxation, and 4.2 inner peace. Short-term changes in symptom scores are shown in Table 2. The mean score for pain intensity decreased by 1.9 points, pain unpleasantness by 2.0 points, anxiety by 2.4 points. The greatest change occurred for relaxation, which increased by 4.3 points. All changes in symptoms were statistically significant (P < .001). For subgroup comparisons, there were no significant differences in symptom scores for patients with a diagnosis of anxiety vs without and depression vs without (Table 3). However, anxiety in patients diagnosed with PTSD decreased by 3.3 points compared with 2.0 in patients without PTSD (P = .005). For patients with chronic pain, inner peace increased 3.9 points compared with an increase of 2.0 in patients without chronic pain (P = .002).

Discussion

Massage therapy may be a useful treatment for female veterans experiencing chronic pain, anxiety disorders, depression, or situational anxiety related to gynecologic procedures. After receiving a relaxation massage, female veterans reported decreased pain intensity, pain unpleasantness, and anxiety while reporting increased relaxation and feelings of inner peace. The effects of massage were consistent for all the symptoms or characteristics assessed, suggesting that massage may act on the body in multiple ways.

These changes parallel those seen in a palliative care population primarily composed of male veterans.¹⁴ However, the female veterans in this cohort experienced greater changes in relaxation and feelings of inner peace, which may be partly due to relief of tension related to an upcoming stressful appointment. The large mean decrease in anxiety level among female veterans with PTSD is notable as well as the larger increase in inner peace in those with chronic pain.

Many patients expressed their gratitude for the massage and interest in having access to more massage therapy. Female patients who have experienced sexual trauma or other trauma may especially benefit from massage prior to painful, invasive gynecologic procedures. Anecdotally, 2 nurse chaperones in the clinic mentioned separately to the massage program supervisor that the massages helped some very anxious women better tolerate an invasive procedure that would have been otherwise extremely difficult.

Limitations

This study has several limitations including lack of a control group, a self-selected population, the lack of objective biochemical measurements, and possible respondent bias to please the MTs. Eighty percent had previously experienced massage therapy and may have been biased toward the effects of massage before receiving the intervention. The first report of the effects of massage therapy in an exclusively female veteran population is a major strength of this study.

Further research including randomized controlled trials is needed, especially in populations with coexisting chronic pain and mental health disorders, as is exploring the acceptability of massage therapy for female veterans with MST. Finding viable alternatives to medications has become even more important as the nation addresses the challenge of the opioid crisis.^45,46

Conclusions

Female veterans are increasingly seeking VA health care. Although further research is needed, results from this pilot study suggest massage therapy may be an effective, inexpensive, and safe treatment for pain and/or anxiety in female veterans. Massage may be especially beneficial for female veterans who experience both chronic pain and mental health conditions. Providing female veterans with access to massage therapy may encourage better self-care and utilization of other Whole Health services, leading to overall improved health and well-being. VA Whole Health programs should consider targeting female veterans for massage therapy services.

Acknowledgments

The authors express our gratitude to the Women Veteran Program Manager, Cheryl Allen, RN; Massage Therapists Denise McGee and Kimberly Morro; Dara Ganoczy, MPH, for help with statistical analysis; and Mark Hausman, MD, for leadership support.

Updated June 17, 2022

The US Department of Veterans Affairs (VA) provides health care for > 9 million military veterans, nearly half of all former service members.¹ Over the past 15 years, there has been a steady and substantial increase in the frequency of disability awards for veterans with post-9/11 military service. Recent data from the Bureau of Labor Statistics indicate that 41% of veterans who served after 9/11 receive service-connected disability benefits compared with 28% of veterans overall.² More than 5 million veterans receive VA service-related disability benefits.^2,3 More than half of the VA $243 billion budget for fiscal year (FY) 2021 ($135.5 billion) was allocated to the Veterans Benefits Administration (VBA), of which $115.7 billion (85%) was allocated specifically for service-related compensation claims payments.⁴

The VA predicted that VBA will have completed 1.4 million ratings for disability claims in 2021.⁵ A substantial percentage of these claims will be for mental disorders, specifically posttraumatic stress disorder (PTSD). VA officials testifying before Congress in 2017 noted that the number of PTSD claims had nearly tripled in the previous 10 years.⁶ As far back as 2013, McNally and Frueh analyzed “the skyrocketing of disability claims,” particularly for PTSD, among veterans who served in Iraq and Afghanistan.⁷

This large increase has placed an unprecedented burden on the VBA to expand its capacity to conduct initial PTSD disability evaluations that by regulations are completed by psychologists or psychiatrists. This need has led the VBA to make significant changes in the compensation and pension (C&P) process, including a reduced role for Veterans Health Administration (VHA) examiners and increased reliance on non-VA (contract) examiners through the Contract Medical Disability Examination (MDE) program. In 2019, the MDE budget was $1.23 billion; in 2020, it was increased to $1.79 billion, and for 2021, it was $2.23 billion, reflecting the increasing investment of resources in non-VA examiners, ostensibly to both increase capacity and save costs.⁵

Anecdotally, concerns have been raised regarding inadequate training of contract examiners as well as inadequate reports by these examiners. A 2018 Government Accountability Office (GAO) report concluded that VA lacked the data to determine whether contract examiners were meeting standards for quality, timeliness, and accuracy.⁸ The GAO report noted that VA required 92% of contractor reports contained no obvious errors, a relatively low target; however, in the first half of 2017, only 1 contractor group met that target. The report noted further that “VBA does not verify if examiners have completed training nor does it collect information to assess training effectiveness in preparing examiners.”⁸ A subsequent analysis of contract examinations completed by the VA Office of the Inspector General (OIG) in 2019 concluded that the MDE program was “hampered in their ability to provide oversight because of limitations with VBA’s electronic examination management systems, the lack of reliable data, and inadequate staffing of the program.”⁹

These reports have focused almost exclusively on simple performance metrics, such as timeliness of examination completion. However, the 2018 GAO report referenced isolated “focused reviews” of complaints about the quality of examinations by contract examiners and gave as an example an isolated “review of one contracted examiner who had high rates of diagnosing severe posttraumatic stress disorder.”⁸ After review indicated the examiner’s reports were of poor quality, the VBA discontinued the examiner’s contract.

Unfortunately, despite such anecdotal reports and isolated actions, to date there are no published reports examining and comparing the quality of PTSD examination reports completed by VHA and contract examiners or the subsequent disability determinations made by the VBA as a result of these evaluations. In a November 2020 letter to the VA Secretary, 11 US Senators expressed “grave concerns” regarding the VA decision to privatize C&P programs noting, among other concerns, that there were “no clinical quality measurement for, or evaluation of, contractor examinations.”¹⁰ The letter cited anecdotal evidence of contract examiners not reviewing veteran’s medical records and diagnosing conditions “without supporting evidence.”¹⁰

The purpose of the present evaluation was to provide a systematic comparison of the content and quality of initial PTSD disability examinations conducted by VHA and non-VA contract examiners. In addition, this study compared the disability rating decisions resulting from VHA and contract examinations.

Methods

A random sample of 100 Initial PTSD Disability Benefits Questionnaires (DBQs)—structured forms completed by all examiners—were obtained from a list supplied by the VA Office of Performance Analysis and Integrity. All examinations were from the Veterans Integrated Service Network (VISN) 1, encompassing the New England region and were conducted in 2019 and 2020. Two of the 100 cases were excluded for technical reasons, resulting in 98 examination reports. However, the final pool yielded 62 contract examinations and only 36 VHA examinations. To make the sample sizes more comparable, an additional 15 examinations were randomly selected from the local examination database (also VISN 1) to complement the original examination pool.

Once DBQs were retrieved, all identifying information was deleted, and cases were analyzed using assigned record numbers. All coding was completed by the 2 principal investigators, both VA psychologists with extensive training and experience in C&P evaluation and treatment of veterans with PTSD. Due to inherent structural differences between the forms used for VA and contract examinations, raters could not be masked/blinded to the source of the report.

A number of measures were taken to reduce bias and enhance objectivity of rating. First, objectively coded variables (eg, age and sex of veteran, period of service, trauma type, diagnoses rendered by the examiner, impairment category endorsed, number and type of symptoms) were transcribed directly from the DBQ as recorded by the examiner. Second, to rate report quality, an initial categorical rating scale was developed based on predetermined elements of examination quality that were considered essential. After refinement and preliminary analysis of interrater reliabilities, 3 quality-related indices were identified: (1) level of detail in description of key content areas (history before service, service trauma, after service social and vocational history, mental health history, substance use); (2) synthesis of history and findings in explaining opinion rendered; and (3) clarity of opinion regarding causation required “at least as likely as not” degree of confidence. The first 2 quality ratings were based on a 3-point scale (poor, fair, good), and the third variable was coded as yes or no. (eAppendix available at doi:10.12788/fp.0225). Interrater reliabilities calculated based on a subsample of 18 cases, randomly selected and rated by both raters, yielded Cohen κ in the acceptable range (.61, .72, and .89 for detail, synthesis, and clarity, respectively). Finally, for information regarding VBA decision making, rating decision documents contained in the Veterans Benefit Management System database were reviewed to determine whether the veteran was granted service connection for PTSD or another mental disorder based on the examination report in question and, if so, the disability rating percentage awarded. These were recorded independently after all other coding had been completed.

Results

Comparison of VHA and contract examinations revealed no significant differences between groups on relevant sociodemographic and other measures (Table). Missing data were not obtained from other records or sources, and for this study, reflect only what is recorded in the examination reports except for age, which was calculated using veteran’s date of birth and the date of examination.

To examine differences between VHA and contract examinations, the groups were first compared on a set of predetermined objectively coded variables taken directly from the DBQ. The frequency of PTSD diagnoses by VHA (57%) and contract (71%) examiners was not significantly different nor were rates of non-PTSD diagnoses by VHA (51%) and contract (73%) examiners. There also was no difference in the mean number of PTSD symptoms endorsed across PTSD diagnostic criteria B, C, D, and E (maximum of 20) recorded by VHA (9.4) and contract (10.9) examiners.

Contract examiners recorded a significantly greater mean number of “other symptoms” on a checklist of 31 possible symptoms as compared to VHA examiners: 7.3 vs 5.8, respectively (t[104] = 2.27, P < .05). An initial analysis of overall social/vocational impairment ratings coded by examiners did not reveal significant differences between examiner groups. However, when the 2 most severe impairment categories were combined to create a pooled “severe” category, 31% of contract examiners rated veterans as severely impaired compared with only 12% of VHA examiners (χ² = 5.79, 1 df, P < .05) (Figure 1).

Discussion

The present study provides the first reported systematic comparison of VA disability examinations for PTSD completed by examiners employed by the VHA and those hired as contract examiners through the MDE program. Although the frequency of PTSD diagnoses by contract examiners was higher than that of VHA examiners (71% vs 57%, respectively), the difference was not statistically significant. However, contract examiners recorded significantly more symptoms for examinees and rated them as severely impaired more frequently than did their VHA counterparts. In keeping with rating guidelines used by the VBA, these differences in examination content resulted in higher disability ratings for veterans seen by contract examiners.

Along with these elevated reports in symptom and severity ratings, contract examiners were less likely to provide adequate detail in the narrative sections of their reports and less frequently provided a satisfactory explanation and synthesis of relevant history and findings in support of their conclusions. Although not reflected in the statistical analysis, case-by-case review revealed some startlingly inadequate examination reports by contract examiners, several of which contained no review of records, no report or discussion of relevant background, and no discussion or analysis of social and vocational function to inform and support their opinion about level of impairment. None of the VHA examination reports reviewed lacked information to that degree.

Such deficiencies in detail and synthesis run counter to accepted guidelines for the adequate assessment of psychological injury in general and in VA disability claims specifically.^11,12 For example, Watson and colleagues proposed that a minimum of 3 hours was required to conduct an initial PTSD examination, with more complex cases possibly taking longer.¹¹ There is no information available about how long contract examiners take to complete their examinations and how that compares with the time taken by VA examiners. The VBA failure to monitor whether or not examiners follow accepted guidelines for PTSD examination has not previously been evaluated. Historically, a large number of clinicians, researchers, and policy critics have raised concerns about the potential for exaggeration or malingering among VA PTSD disability claimants and have urged the need to adequately assess for unreliable reporting and presentation.^13,14 However, the possibility of systematic examiner deficiency and/or bias increasing the frequency of false or inflated claims being approved has received little empirical attention.

Limitations

A methodological limitation of the present study was that, due to structural differences in the DBQ forms used for VHA and contract examinations, the reports could not be de-identified as to examiner type and thus raters could not be masked/blinded. To mitigate bias, a predetermined, piloted, and refined coding and rating plan for report quality metrics was adhered to strictly, and interrater reliabilities were acceptable. Future study is suggested in which all report content is standardized for coding using the same format, which at present would require a complete rewriting of the entire report; this problem could be resolved by having the VBA adopt a more coherent system in which all reports, regardless of examiner type, use a single, standardized template. Further study using larger data sets and expanding to other VA regions also is needed.

Conclusions

The present study suggests that poor examination and report quality—by contract examiners and to a lesser degree VHA examiners—are not uncommon. The findings confirm and extend previous anecdotal reports of deficiencies in PTSD examinations performed by contract examiners and provide empirical support for concerns raised of global deficiencies in the VBA oversight of the MDE program. Such deficiencies have significant implications for the quality and integrity of the VA disability determination process for veterans claiming PTSD related to military service.

The current findings support and strengthen the call for development and management of a structured and enforced training and quality assurance/improvement program for VA PTSD disability examinations. Such training and oversight will be critical to improve the quality and integrity of these examinations, reduce error and waste in VBA’s Compensation and Pension process, and in doing so optimize VA financial resources to best serve veterans’ benefits and health care needs.

Updated June 17, 2022

The US Department of Veterans Affairs (VA) provides health care for > 9 million military veterans, nearly half of all former service members.¹ Over the past 15 years, there has been a steady and substantial increase in the frequency of disability awards for veterans with post-9/11 military service. Recent data from the Bureau of Labor Statistics indicate that 41% of veterans who served after 9/11 receive service-connected disability benefits compared with 28% of veterans overall.² More than 5 million veterans receive VA service-related disability benefits.^2,3 More than half of the VA $243 billion budget for fiscal year (FY) 2021 ($135.5 billion) was allocated to the Veterans Benefits Administration (VBA), of which $115.7 billion (85%) was allocated specifically for service-related compensation claims payments.⁴

The VA predicted that VBA will have completed 1.4 million ratings for disability claims in 2021.⁵ A substantial percentage of these claims will be for mental disorders, specifically posttraumatic stress disorder (PTSD). VA officials testifying before Congress in 2017 noted that the number of PTSD claims had nearly tripled in the previous 10 years.⁶ As far back as 2013, McNally and Frueh analyzed “the skyrocketing of disability claims,” particularly for PTSD, among veterans who served in Iraq and Afghanistan.⁷

This large increase has placed an unprecedented burden on the VBA to expand its capacity to conduct initial PTSD disability evaluations that by regulations are completed by psychologists or psychiatrists. This need has led the VBA to make significant changes in the compensation and pension (C&P) process, including a reduced role for Veterans Health Administration (VHA) examiners and increased reliance on non-VA (contract) examiners through the Contract Medical Disability Examination (MDE) program. In 2019, the MDE budget was $1.23 billion; in 2020, it was increased to $1.79 billion, and for 2021, it was $2.23 billion, reflecting the increasing investment of resources in non-VA examiners, ostensibly to both increase capacity and save costs.⁵

Anecdotally, concerns have been raised regarding inadequate training of contract examiners as well as inadequate reports by these examiners. A 2018 Government Accountability Office (GAO) report concluded that VA lacked the data to determine whether contract examiners were meeting standards for quality, timeliness, and accuracy.⁸ The GAO report noted that VA required 92% of contractor reports contained no obvious errors, a relatively low target; however, in the first half of 2017, only 1 contractor group met that target. The report noted further that “VBA does not verify if examiners have completed training nor does it collect information to assess training effectiveness in preparing examiners.”⁸ A subsequent analysis of contract examinations completed by the VA Office of the Inspector General (OIG) in 2019 concluded that the MDE program was “hampered in their ability to provide oversight because of limitations with VBA’s electronic examination management systems, the lack of reliable data, and inadequate staffing of the program.”⁹

These reports have focused almost exclusively on simple performance metrics, such as timeliness of examination completion. However, the 2018 GAO report referenced isolated “focused reviews” of complaints about the quality of examinations by contract examiners and gave as an example an isolated “review of one contracted examiner who had high rates of diagnosing severe posttraumatic stress disorder.”⁸ After review indicated the examiner’s reports were of poor quality, the VBA discontinued the examiner’s contract.

Unfortunately, despite such anecdotal reports and isolated actions, to date there are no published reports examining and comparing the quality of PTSD examination reports completed by VHA and contract examiners or the subsequent disability determinations made by the VBA as a result of these evaluations. In a November 2020 letter to the VA Secretary, 11 US Senators expressed “grave concerns” regarding the VA decision to privatize C&P programs noting, among other concerns, that there were “no clinical quality measurement for, or evaluation of, contractor examinations.”¹⁰ The letter cited anecdotal evidence of contract examiners not reviewing veteran’s medical records and diagnosing conditions “without supporting evidence.”¹⁰

The purpose of the present evaluation was to provide a systematic comparison of the content and quality of initial PTSD disability examinations conducted by VHA and non-VA contract examiners. In addition, this study compared the disability rating decisions resulting from VHA and contract examinations.

Methods

A random sample of 100 Initial PTSD Disability Benefits Questionnaires (DBQs)—structured forms completed by all examiners—were obtained from a list supplied by the VA Office of Performance Analysis and Integrity. All examinations were from the Veterans Integrated Service Network (VISN) 1, encompassing the New England region and were conducted in 2019 and 2020. Two of the 100 cases were excluded for technical reasons, resulting in 98 examination reports. However, the final pool yielded 62 contract examinations and only 36 VHA examinations. To make the sample sizes more comparable, an additional 15 examinations were randomly selected from the local examination database (also VISN 1) to complement the original examination pool.

Once DBQs were retrieved, all identifying information was deleted, and cases were analyzed using assigned record numbers. All coding was completed by the 2 principal investigators, both VA psychologists with extensive training and experience in C&P evaluation and treatment of veterans with PTSD. Due to inherent structural differences between the forms used for VA and contract examinations, raters could not be masked/blinded to the source of the report.

A number of measures were taken to reduce bias and enhance objectivity of rating. First, objectively coded variables (eg, age and sex of veteran, period of service, trauma type, diagnoses rendered by the examiner, impairment category endorsed, number and type of symptoms) were transcribed directly from the DBQ as recorded by the examiner. Second, to rate report quality, an initial categorical rating scale was developed based on predetermined elements of examination quality that were considered essential. After refinement and preliminary analysis of interrater reliabilities, 3 quality-related indices were identified: (1) level of detail in description of key content areas (history before service, service trauma, after service social and vocational history, mental health history, substance use); (2) synthesis of history and findings in explaining opinion rendered; and (3) clarity of opinion regarding causation required “at least as likely as not” degree of confidence. The first 2 quality ratings were based on a 3-point scale (poor, fair, good), and the third variable was coded as yes or no. (eAppendix available at doi:10.12788/fp.0225). Interrater reliabilities calculated based on a subsample of 18 cases, randomly selected and rated by both raters, yielded Cohen κ in the acceptable range (.61, .72, and .89 for detail, synthesis, and clarity, respectively). Finally, for information regarding VBA decision making, rating decision documents contained in the Veterans Benefit Management System database were reviewed to determine whether the veteran was granted service connection for PTSD or another mental disorder based on the examination report in question and, if so, the disability rating percentage awarded. These were recorded independently after all other coding had been completed.

Results

Comparison of VHA and contract examinations revealed no significant differences between groups on relevant sociodemographic and other measures (Table). Missing data were not obtained from other records or sources, and for this study, reflect only what is recorded in the examination reports except for age, which was calculated using veteran’s date of birth and the date of examination.

To examine differences between VHA and contract examinations, the groups were first compared on a set of predetermined objectively coded variables taken directly from the DBQ. The frequency of PTSD diagnoses by VHA (57%) and contract (71%) examiners was not significantly different nor were rates of non-PTSD diagnoses by VHA (51%) and contract (73%) examiners. There also was no difference in the mean number of PTSD symptoms endorsed across PTSD diagnostic criteria B, C, D, and E (maximum of 20) recorded by VHA (9.4) and contract (10.9) examiners.

Contract examiners recorded a significantly greater mean number of “other symptoms” on a checklist of 31 possible symptoms as compared to VHA examiners: 7.3 vs 5.8, respectively (t[104] = 2.27, P < .05). An initial analysis of overall social/vocational impairment ratings coded by examiners did not reveal significant differences between examiner groups. However, when the 2 most severe impairment categories were combined to create a pooled “severe” category, 31% of contract examiners rated veterans as severely impaired compared with only 12% of VHA examiners (χ² = 5.79, 1 df, P < .05) (Figure 1).

Discussion

The present study provides the first reported systematic comparison of VA disability examinations for PTSD completed by examiners employed by the VHA and those hired as contract examiners through the MDE program. Although the frequency of PTSD diagnoses by contract examiners was higher than that of VHA examiners (71% vs 57%, respectively), the difference was not statistically significant. However, contract examiners recorded significantly more symptoms for examinees and rated them as severely impaired more frequently than did their VHA counterparts. In keeping with rating guidelines used by the VBA, these differences in examination content resulted in higher disability ratings for veterans seen by contract examiners.

Along with these elevated reports in symptom and severity ratings, contract examiners were less likely to provide adequate detail in the narrative sections of their reports and less frequently provided a satisfactory explanation and synthesis of relevant history and findings in support of their conclusions. Although not reflected in the statistical analysis, case-by-case review revealed some startlingly inadequate examination reports by contract examiners, several of which contained no review of records, no report or discussion of relevant background, and no discussion or analysis of social and vocational function to inform and support their opinion about level of impairment. None of the VHA examination reports reviewed lacked information to that degree.

Such deficiencies in detail and synthesis run counter to accepted guidelines for the adequate assessment of psychological injury in general and in VA disability claims specifically.^11,12 For example, Watson and colleagues proposed that a minimum of 3 hours was required to conduct an initial PTSD examination, with more complex cases possibly taking longer.¹¹ There is no information available about how long contract examiners take to complete their examinations and how that compares with the time taken by VA examiners. The VBA failure to monitor whether or not examiners follow accepted guidelines for PTSD examination has not previously been evaluated. Historically, a large number of clinicians, researchers, and policy critics have raised concerns about the potential for exaggeration or malingering among VA PTSD disability claimants and have urged the need to adequately assess for unreliable reporting and presentation.^13,14 However, the possibility of systematic examiner deficiency and/or bias increasing the frequency of false or inflated claims being approved has received little empirical attention.

Limitations

A methodological limitation of the present study was that, due to structural differences in the DBQ forms used for VHA and contract examinations, the reports could not be de-identified as to examiner type and thus raters could not be masked/blinded. To mitigate bias, a predetermined, piloted, and refined coding and rating plan for report quality metrics was adhered to strictly, and interrater reliabilities were acceptable. Future study is suggested in which all report content is standardized for coding using the same format, which at present would require a complete rewriting of the entire report; this problem could be resolved by having the VBA adopt a more coherent system in which all reports, regardless of examiner type, use a single, standardized template. Further study using larger data sets and expanding to other VA regions also is needed.

Conclusions

The present study suggests that poor examination and report quality—by contract examiners and to a lesser degree VHA examiners—are not uncommon. The findings confirm and extend previous anecdotal reports of deficiencies in PTSD examinations performed by contract examiners and provide empirical support for concerns raised of global deficiencies in the VBA oversight of the MDE program. Such deficiencies have significant implications for the quality and integrity of the VA disability determination process for veterans claiming PTSD related to military service.

The current findings support and strengthen the call for development and management of a structured and enforced training and quality assurance/improvement program for VA PTSD disability examinations. Such training and oversight will be critical to improve the quality and integrity of these examinations, reduce error and waste in VBA’s Compensation and Pension process, and in doing so optimize VA financial resources to best serve veterans’ benefits and health care needs.

Updated June 17, 2022

The US Department of Veterans Affairs (VA) provides health care for > 9 million military veterans, nearly half of all former service members.¹ Over the past 15 years, there has been a steady and substantial increase in the frequency of disability awards for veterans with post-9/11 military service. Recent data from the Bureau of Labor Statistics indicate that 41% of veterans who served after 9/11 receive service-connected disability benefits compared with 28% of veterans overall.² More than 5 million veterans receive VA service-related disability benefits.^2,3 More than half of the VA $243 billion budget for fiscal year (FY) 2021 ($135.5 billion) was allocated to the Veterans Benefits Administration (VBA), of which $115.7 billion (85%) was allocated specifically for service-related compensation claims payments.⁴

The VA predicted that VBA will have completed 1.4 million ratings for disability claims in 2021.⁵ A substantial percentage of these claims will be for mental disorders, specifically posttraumatic stress disorder (PTSD). VA officials testifying before Congress in 2017 noted that the number of PTSD claims had nearly tripled in the previous 10 years.⁶ As far back as 2013, McNally and Frueh analyzed “the skyrocketing of disability claims,” particularly for PTSD, among veterans who served in Iraq and Afghanistan.⁷

This large increase has placed an unprecedented burden on the VBA to expand its capacity to conduct initial PTSD disability evaluations that by regulations are completed by psychologists or psychiatrists. This need has led the VBA to make significant changes in the compensation and pension (C&P) process, including a reduced role for Veterans Health Administration (VHA) examiners and increased reliance on non-VA (contract) examiners through the Contract Medical Disability Examination (MDE) program. In 2019, the MDE budget was $1.23 billion; in 2020, it was increased to $1.79 billion, and for 2021, it was $2.23 billion, reflecting the increasing investment of resources in non-VA examiners, ostensibly to both increase capacity and save costs.⁵

Anecdotally, concerns have been raised regarding inadequate training of contract examiners as well as inadequate reports by these examiners. A 2018 Government Accountability Office (GAO) report concluded that VA lacked the data to determine whether contract examiners were meeting standards for quality, timeliness, and accuracy.⁸ The GAO report noted that VA required 92% of contractor reports contained no obvious errors, a relatively low target; however, in the first half of 2017, only 1 contractor group met that target. The report noted further that “VBA does not verify if examiners have completed training nor does it collect information to assess training effectiveness in preparing examiners.”⁸ A subsequent analysis of contract examinations completed by the VA Office of the Inspector General (OIG) in 2019 concluded that the MDE program was “hampered in their ability to provide oversight because of limitations with VBA’s electronic examination management systems, the lack of reliable data, and inadequate staffing of the program.”⁹

These reports have focused almost exclusively on simple performance metrics, such as timeliness of examination completion. However, the 2018 GAO report referenced isolated “focused reviews” of complaints about the quality of examinations by contract examiners and gave as an example an isolated “review of one contracted examiner who had high rates of diagnosing severe posttraumatic stress disorder.”⁸ After review indicated the examiner’s reports were of poor quality, the VBA discontinued the examiner’s contract.

Unfortunately, despite such anecdotal reports and isolated actions, to date there are no published reports examining and comparing the quality of PTSD examination reports completed by VHA and contract examiners or the subsequent disability determinations made by the VBA as a result of these evaluations. In a November 2020 letter to the VA Secretary, 11 US Senators expressed “grave concerns” regarding the VA decision to privatize C&P programs noting, among other concerns, that there were “no clinical quality measurement for, or evaluation of, contractor examinations.”¹⁰ The letter cited anecdotal evidence of contract examiners not reviewing veteran’s medical records and diagnosing conditions “without supporting evidence.”¹⁰

The purpose of the present evaluation was to provide a systematic comparison of the content and quality of initial PTSD disability examinations conducted by VHA and non-VA contract examiners. In addition, this study compared the disability rating decisions resulting from VHA and contract examinations.

Methods

A random sample of 100 Initial PTSD Disability Benefits Questionnaires (DBQs)—structured forms completed by all examiners—were obtained from a list supplied by the VA Office of Performance Analysis and Integrity. All examinations were from the Veterans Integrated Service Network (VISN) 1, encompassing the New England region and were conducted in 2019 and 2020. Two of the 100 cases were excluded for technical reasons, resulting in 98 examination reports. However, the final pool yielded 62 contract examinations and only 36 VHA examinations. To make the sample sizes more comparable, an additional 15 examinations were randomly selected from the local examination database (also VISN 1) to complement the original examination pool.

Once DBQs were retrieved, all identifying information was deleted, and cases were analyzed using assigned record numbers. All coding was completed by the 2 principal investigators, both VA psychologists with extensive training and experience in C&P evaluation and treatment of veterans with PTSD. Due to inherent structural differences between the forms used for VA and contract examinations, raters could not be masked/blinded to the source of the report.

A number of measures were taken to reduce bias and enhance objectivity of rating. First, objectively coded variables (eg, age and sex of veteran, period of service, trauma type, diagnoses rendered by the examiner, impairment category endorsed, number and type of symptoms) were transcribed directly from the DBQ as recorded by the examiner. Second, to rate report quality, an initial categorical rating scale was developed based on predetermined elements of examination quality that were considered essential. After refinement and preliminary analysis of interrater reliabilities, 3 quality-related indices were identified: (1) level of detail in description of key content areas (history before service, service trauma, after service social and vocational history, mental health history, substance use); (2) synthesis of history and findings in explaining opinion rendered; and (3) clarity of opinion regarding causation required “at least as likely as not” degree of confidence. The first 2 quality ratings were based on a 3-point scale (poor, fair, good), and the third variable was coded as yes or no. (eAppendix available at doi:10.12788/fp.0225). Interrater reliabilities calculated based on a subsample of 18 cases, randomly selected and rated by both raters, yielded Cohen κ in the acceptable range (.61, .72, and .89 for detail, synthesis, and clarity, respectively). Finally, for information regarding VBA decision making, rating decision documents contained in the Veterans Benefit Management System database were reviewed to determine whether the veteran was granted service connection for PTSD or another mental disorder based on the examination report in question and, if so, the disability rating percentage awarded. These were recorded independently after all other coding had been completed.

Results

Comparison of VHA and contract examinations revealed no significant differences between groups on relevant sociodemographic and other measures (Table). Missing data were not obtained from other records or sources, and for this study, reflect only what is recorded in the examination reports except for age, which was calculated using veteran’s date of birth and the date of examination.

To examine differences between VHA and contract examinations, the groups were first compared on a set of predetermined objectively coded variables taken directly from the DBQ. The frequency of PTSD diagnoses by VHA (57%) and contract (71%) examiners was not significantly different nor were rates of non-PTSD diagnoses by VHA (51%) and contract (73%) examiners. There also was no difference in the mean number of PTSD symptoms endorsed across PTSD diagnostic criteria B, C, D, and E (maximum of 20) recorded by VHA (9.4) and contract (10.9) examiners.

Contract examiners recorded a significantly greater mean number of “other symptoms” on a checklist of 31 possible symptoms as compared to VHA examiners: 7.3 vs 5.8, respectively (t[104] = 2.27, P < .05). An initial analysis of overall social/vocational impairment ratings coded by examiners did not reveal significant differences between examiner groups. However, when the 2 most severe impairment categories were combined to create a pooled “severe” category, 31% of contract examiners rated veterans as severely impaired compared with only 12% of VHA examiners (χ² = 5.79, 1 df, P < .05) (Figure 1).

Discussion

The present study provides the first reported systematic comparison of VA disability examinations for PTSD completed by examiners employed by the VHA and those hired as contract examiners through the MDE program. Although the frequency of PTSD diagnoses by contract examiners was higher than that of VHA examiners (71% vs 57%, respectively), the difference was not statistically significant. However, contract examiners recorded significantly more symptoms for examinees and rated them as severely impaired more frequently than did their VHA counterparts. In keeping with rating guidelines used by the VBA, these differences in examination content resulted in higher disability ratings for veterans seen by contract examiners.

Along with these elevated reports in symptom and severity ratings, contract examiners were less likely to provide adequate detail in the narrative sections of their reports and less frequently provided a satisfactory explanation and synthesis of relevant history and findings in support of their conclusions. Although not reflected in the statistical analysis, case-by-case review revealed some startlingly inadequate examination reports by contract examiners, several of which contained no review of records, no report or discussion of relevant background, and no discussion or analysis of social and vocational function to inform and support their opinion about level of impairment. None of the VHA examination reports reviewed lacked information to that degree.

Such deficiencies in detail and synthesis run counter to accepted guidelines for the adequate assessment of psychological injury in general and in VA disability claims specifically.^11,12 For example, Watson and colleagues proposed that a minimum of 3 hours was required to conduct an initial PTSD examination, with more complex cases possibly taking longer.¹¹ There is no information available about how long contract examiners take to complete their examinations and how that compares with the time taken by VA examiners. The VBA failure to monitor whether or not examiners follow accepted guidelines for PTSD examination has not previously been evaluated. Historically, a large number of clinicians, researchers, and policy critics have raised concerns about the potential for exaggeration or malingering among VA PTSD disability claimants and have urged the need to adequately assess for unreliable reporting and presentation.^13,14 However, the possibility of systematic examiner deficiency and/or bias increasing the frequency of false or inflated claims being approved has received little empirical attention.

Limitations

A methodological limitation of the present study was that, due to structural differences in the DBQ forms used for VHA and contract examinations, the reports could not be de-identified as to examiner type and thus raters could not be masked/blinded. To mitigate bias, a predetermined, piloted, and refined coding and rating plan for report quality metrics was adhered to strictly, and interrater reliabilities were acceptable. Future study is suggested in which all report content is standardized for coding using the same format, which at present would require a complete rewriting of the entire report; this problem could be resolved by having the VBA adopt a more coherent system in which all reports, regardless of examiner type, use a single, standardized template. Further study using larger data sets and expanding to other VA regions also is needed.

Conclusions

The present study suggests that poor examination and report quality—by contract examiners and to a lesser degree VHA examiners—are not uncommon. The findings confirm and extend previous anecdotal reports of deficiencies in PTSD examinations performed by contract examiners and provide empirical support for concerns raised of global deficiencies in the VBA oversight of the MDE program. Such deficiencies have significant implications for the quality and integrity of the VA disability determination process for veterans claiming PTSD related to military service.

The current findings support and strengthen the call for development and management of a structured and enforced training and quality assurance/improvement program for VA PTSD disability examinations. Such training and oversight will be critical to improve the quality and integrity of these examinations, reduce error and waste in VBA’s Compensation and Pension process, and in doing so optimize VA financial resources to best serve veterans’ benefits and health care needs.

Medical case reports are as old as the healing profession itself.¹ These ancient medical stories have a modern definition: “A case report is a narrative that describes, for medical, scientific or educational purposes, a medical problem experienced by one or more patients.”² Case report experts describe the 3-fold purposes of this type of research: as a mainstay of education; a harbinger of emerging illnesses; and an appraiser of new interventions. Case-based education has long been a pillar of health professions education: Nurses, doctors, and allied health professionals are taught and learn through reading and discussing with their teachers and each other about cases of their own patients and of those in the literature.³ Case reports also have helped identify and raise awareness of new diseases and rare conditions, such as HIV.⁴ Finally, case reports have alerted regulatory agencies and the medical community about medication adverse effects, such as birth defects from thalidomide.⁵

Case reports also have been criticized on both scientific and ethical grounds. Critics argue that many case reports often lack the rigor and consistency of other types of research.⁶ Three recent trends in medical publication have strengthened the validity of these criticisms: the increase in the popularity of case reports; the corresponding increase in submissions to journals, including Federal Practitioner; and the rise of predatory publishers.^7,8

The ethical scrutiny of case reports discussed in this column focuses on the tension between providing readers with adequate, accurate information to fulfil the goals of case reports while also protecting patient confidentiality. The latter issue during most of the history of medicine was not considered by health care professionals when the prevailing paternalism supported a professional-oriented approach to health care. The rise of bioethics in the 1960s and 1970s began the shift toward patient autonomy in medical decision making and patient rights to control their protected health information that rendered case reports ethically problematic.

To address both changes in ethical standards and scientific limitations, a committee of clinicians, researchers, and journal editors formed the Case Report (CARE) group.^2,8 The group undertook an effort to improve the quality of case reports. From 2011 to 2012, they developed the CARE guidelines for clinical case reporting. The guidance took the form of a Statement and Checklist presented at the 2013 International Congress on Peer Review and Biomedical Publication. Since their presentation, multiple prestigious medical journals in many countries have implemented these recommendations.

As part of an overall effort to raise the ethical caliber of our own journal, Federal Practitioner will begin to implement the CARE guidelines for case reports for all future submissions. Use of the CARE recommendations will help prospective authors enhance the scientific value and ethical caliber of case reports submitted to the journal as well as assist the Federal Practitioner editorial team, editorial board, and peer reviewers to evaluate submissions more judiciously.

An essential part of the CARE guidelines is that the patient who is the subject of the case report provide informed consent for the publication of their personal narrative. The CARE group considers this an “ethical duty” of authors and editors alike. In “exceptional circumstances” such as if the patient is a minor or permanently incapacitated, a guardian or relative may grant consent. In the rare event that even with exhaustive attempts, if informed consent cannot be obtained from a patient or their representative, then the authors of the case report must submit a statement to this effect.⁴ Some journals may require that the authors obtain the approval of an institutional review board or the permission of an ethics or other institutional committee or a privacy officer.²

Requesting the patient’s consent is an extension of the shared decision making that is now a best practice in clinical care into the arena of research, making the patient or their representative a partner in the work. Ethicists have recommended inviting patients or relatives to read a draft of the case report and agree to its publication or request specific modifications to the manuscript. The CARE group rightly points out that with the rise of open notes in medical documentation, patients increasingly have access to their charts in near or real time.² Gone are the days of Sir William Osler when only doctors read medical journals and all of these technical developments as well as standards of research and social changes in the practitioner-patient relationship make it imperative that writers and editors join together to make case reports more transparent, accurate, and consistent.⁷

An additional step to protect patient privacy is the requirement that authors either de-identify potentially identifiable health information, such as age, birth, death, admission, and discharge dates, or in some instances obtain separate consent for the release of that protected data.⁸ These restrictions constitute a challenge to case report authors who in some instances may consider these same facts critical to the integrity of the case presentation that have made some scholars doubt their continued viability. After all, the contribution of the case to the medical literature often lies in its very particularity. Conversely, no matter how frustrated we might become during writing a case report, we would not want to see our own protected health information or that of our family on a website or in print without our knowledge or approval. Indeed, the International Committee of Medical Journal Editors states that “If identifying characteristics are de-identified, authors should provide assurance, and editors should so note, that such changes do not distort scientific meaning.”⁹

However, the exponential growth of the internet, the spread of social media, and the ubiquity of a plethora of electronic devices, which prior generations of writers and readers could not even imagine, make these limitations necessary to protect patient privacy and the public’s trust in health care professionals. The CARE guidelines can help authors of case reports hone the art of anonymizing the protected health information of subjects of case reports, such as ethnicity and occupation, while accurately conveying the clinical specifics of the case that make it valuable to students and colleagues.

We at Federal Practitioner recognize there is a real tension between truth-telling in case report publication and respect for patient confidentiality that will never be perfectly achieved, but is one that is important for medical knowledge, making it worthy of the continuous efforts of authors and editors to negotiate.

Medical case reports are as old as the healing profession itself.¹ These ancient medical stories have a modern definition: “A case report is a narrative that describes, for medical, scientific or educational purposes, a medical problem experienced by one or more patients.”² Case report experts describe the 3-fold purposes of this type of research: as a mainstay of education; a harbinger of emerging illnesses; and an appraiser of new interventions. Case-based education has long been a pillar of health professions education: Nurses, doctors, and allied health professionals are taught and learn through reading and discussing with their teachers and each other about cases of their own patients and of those in the literature.³ Case reports also have helped identify and raise awareness of new diseases and rare conditions, such as HIV.⁴ Finally, case reports have alerted regulatory agencies and the medical community about medication adverse effects, such as birth defects from thalidomide.⁵

Case reports also have been criticized on both scientific and ethical grounds. Critics argue that many case reports often lack the rigor and consistency of other types of research.⁶ Three recent trends in medical publication have strengthened the validity of these criticisms: the increase in the popularity of case reports; the corresponding increase in submissions to journals, including Federal Practitioner; and the rise of predatory publishers.^7,8

The ethical scrutiny of case reports discussed in this column focuses on the tension between providing readers with adequate, accurate information to fulfil the goals of case reports while also protecting patient confidentiality. The latter issue during most of the history of medicine was not considered by health care professionals when the prevailing paternalism supported a professional-oriented approach to health care. The rise of bioethics in the 1960s and 1970s began the shift toward patient autonomy in medical decision making and patient rights to control their protected health information that rendered case reports ethically problematic.

To address both changes in ethical standards and scientific limitations, a committee of clinicians, researchers, and journal editors formed the Case Report (CARE) group.^2,8 The group undertook an effort to improve the quality of case reports. From 2011 to 2012, they developed the CARE guidelines for clinical case reporting. The guidance took the form of a Statement and Checklist presented at the 2013 International Congress on Peer Review and Biomedical Publication. Since their presentation, multiple prestigious medical journals in many countries have implemented these recommendations.

As part of an overall effort to raise the ethical caliber of our own journal, Federal Practitioner will begin to implement the CARE guidelines for case reports for all future submissions. Use of the CARE recommendations will help prospective authors enhance the scientific value and ethical caliber of case reports submitted to the journal as well as assist the Federal Practitioner editorial team, editorial board, and peer reviewers to evaluate submissions more judiciously.

An essential part of the CARE guidelines is that the patient who is the subject of the case report provide informed consent for the publication of their personal narrative. The CARE group considers this an “ethical duty” of authors and editors alike. In “exceptional circumstances” such as if the patient is a minor or permanently incapacitated, a guardian or relative may grant consent. In the rare event that even with exhaustive attempts, if informed consent cannot be obtained from a patient or their representative, then the authors of the case report must submit a statement to this effect.⁴ Some journals may require that the authors obtain the approval of an institutional review board or the permission of an ethics or other institutional committee or a privacy officer.²

Requesting the patient’s consent is an extension of the shared decision making that is now a best practice in clinical care into the arena of research, making the patient or their representative a partner in the work. Ethicists have recommended inviting patients or relatives to read a draft of the case report and agree to its publication or request specific modifications to the manuscript. The CARE group rightly points out that with the rise of open notes in medical documentation, patients increasingly have access to their charts in near or real time.² Gone are the days of Sir William Osler when only doctors read medical journals and all of these technical developments as well as standards of research and social changes in the practitioner-patient relationship make it imperative that writers and editors join together to make case reports more transparent, accurate, and consistent.⁷

An additional step to protect patient privacy is the requirement that authors either de-identify potentially identifiable health information, such as age, birth, death, admission, and discharge dates, or in some instances obtain separate consent for the release of that protected data.⁸ These restrictions constitute a challenge to case report authors who in some instances may consider these same facts critical to the integrity of the case presentation that have made some scholars doubt their continued viability. After all, the contribution of the case to the medical literature often lies in its very particularity. Conversely, no matter how frustrated we might become during writing a case report, we would not want to see our own protected health information or that of our family on a website or in print without our knowledge or approval. Indeed, the International Committee of Medical Journal Editors states that “If identifying characteristics are de-identified, authors should provide assurance, and editors should so note, that such changes do not distort scientific meaning.”⁹

However, the exponential growth of the internet, the spread of social media, and the ubiquity of a plethora of electronic devices, which prior generations of writers and readers could not even imagine, make these limitations necessary to protect patient privacy and the public’s trust in health care professionals. The CARE guidelines can help authors of case reports hone the art of anonymizing the protected health information of subjects of case reports, such as ethnicity and occupation, while accurately conveying the clinical specifics of the case that make it valuable to students and colleagues.

We at Federal Practitioner recognize there is a real tension between truth-telling in case report publication and respect for patient confidentiality that will never be perfectly achieved, but is one that is important for medical knowledge, making it worthy of the continuous efforts of authors and editors to negotiate.

Medical case reports are as old as the healing profession itself.¹ These ancient medical stories have a modern definition: “A case report is a narrative that describes, for medical, scientific or educational purposes, a medical problem experienced by one or more patients.”² Case report experts describe the 3-fold purposes of this type of research: as a mainstay of education; a harbinger of emerging illnesses; and an appraiser of new interventions. Case-based education has long been a pillar of health professions education: Nurses, doctors, and allied health professionals are taught and learn through reading and discussing with their teachers and each other about cases of their own patients and of those in the literature.³ Case reports also have helped identify and raise awareness of new diseases and rare conditions, such as HIV.⁴ Finally, case reports have alerted regulatory agencies and the medical community about medication adverse effects, such as birth defects from thalidomide.⁵

Case reports also have been criticized on both scientific and ethical grounds. Critics argue that many case reports often lack the rigor and consistency of other types of research.⁶ Three recent trends in medical publication have strengthened the validity of these criticisms: the increase in the popularity of case reports; the corresponding increase in submissions to journals, including Federal Practitioner; and the rise of predatory publishers.^7,8

The ethical scrutiny of case reports discussed in this column focuses on the tension between providing readers with adequate, accurate information to fulfil the goals of case reports while also protecting patient confidentiality. The latter issue during most of the history of medicine was not considered by health care professionals when the prevailing paternalism supported a professional-oriented approach to health care. The rise of bioethics in the 1960s and 1970s began the shift toward patient autonomy in medical decision making and patient rights to control their protected health information that rendered case reports ethically problematic.

To address both changes in ethical standards and scientific limitations, a committee of clinicians, researchers, and journal editors formed the Case Report (CARE) group.^2,8 The group undertook an effort to improve the quality of case reports. From 2011 to 2012, they developed the CARE guidelines for clinical case reporting. The guidance took the form of a Statement and Checklist presented at the 2013 International Congress on Peer Review and Biomedical Publication. Since their presentation, multiple prestigious medical journals in many countries have implemented these recommendations.

As part of an overall effort to raise the ethical caliber of our own journal, Federal Practitioner will begin to implement the CARE guidelines for case reports for all future submissions. Use of the CARE recommendations will help prospective authors enhance the scientific value and ethical caliber of case reports submitted to the journal as well as assist the Federal Practitioner editorial team, editorial board, and peer reviewers to evaluate submissions more judiciously.

An essential part of the CARE guidelines is that the patient who is the subject of the case report provide informed consent for the publication of their personal narrative. The CARE group considers this an “ethical duty” of authors and editors alike. In “exceptional circumstances” such as if the patient is a minor or permanently incapacitated, a guardian or relative may grant consent. In the rare event that even with exhaustive attempts, if informed consent cannot be obtained from a patient or their representative, then the authors of the case report must submit a statement to this effect.⁴ Some journals may require that the authors obtain the approval of an institutional review board or the permission of an ethics or other institutional committee or a privacy officer.²

Requesting the patient’s consent is an extension of the shared decision making that is now a best practice in clinical care into the arena of research, making the patient or their representative a partner in the work. Ethicists have recommended inviting patients or relatives to read a draft of the case report and agree to its publication or request specific modifications to the manuscript. The CARE group rightly points out that with the rise of open notes in medical documentation, patients increasingly have access to their charts in near or real time.² Gone are the days of Sir William Osler when only doctors read medical journals and all of these technical developments as well as standards of research and social changes in the practitioner-patient relationship make it imperative that writers and editors join together to make case reports more transparent, accurate, and consistent.⁷

An additional step to protect patient privacy is the requirement that authors either de-identify potentially identifiable health information, such as age, birth, death, admission, and discharge dates, or in some instances obtain separate consent for the release of that protected data.⁸ These restrictions constitute a challenge to case report authors who in some instances may consider these same facts critical to the integrity of the case presentation that have made some scholars doubt their continued viability. After all, the contribution of the case to the medical literature often lies in its very particularity. Conversely, no matter how frustrated we might become during writing a case report, we would not want to see our own protected health information or that of our family on a website or in print without our knowledge or approval. Indeed, the International Committee of Medical Journal Editors states that “If identifying characteristics are de-identified, authors should provide assurance, and editors should so note, that such changes do not distort scientific meaning.”⁹

However, the exponential growth of the internet, the spread of social media, and the ubiquity of a plethora of electronic devices, which prior generations of writers and readers could not even imagine, make these limitations necessary to protect patient privacy and the public’s trust in health care professionals. The CARE guidelines can help authors of case reports hone the art of anonymizing the protected health information of subjects of case reports, such as ethnicity and occupation, while accurately conveying the clinical specifics of the case that make it valuable to students and colleagues.

We at Federal Practitioner recognize there is a real tension between truth-telling in case report publication and respect for patient confidentiality that will never be perfectly achieved, but is one that is important for medical knowledge, making it worthy of the continuous efforts of authors and editors to negotiate.

Automated urine cultures (UCs) following urinalysis (UA) are often used in emergency departments (EDs) to identify urinary tract infections (UTIs). The fast-paced environment of the ED makes this method of proactive collection and facilitation of UC favorable. However, results are often reported as no organism growth or the growth of clinically insignificant organisms, leading to the overdetection and overtreatment of asymptomatic bacteriuria (ASB).^1-3 An estimated 30 to 60% of patients with ASB receive unwarranted antibiotic treatment, which is associated with an increased risk of developing Clostridioides difficile infection and contributes to the development of antimicrobial resistance.^4-10 The costs associated with UC are an important consideration given the use of resources, the time and effort required to collect and process large numbers of negative cultures, and further efforts devoted to the follow-up of ED culture results.

Changes in traditional testing involving testing of both a UA and UC to reflex testing where urine specimens undergo culture only if they meet certain criteria have been described.^11-14 This change in traditional testing aims to reduce the number of potentially unnecessary cultures performed without compromising clinical care. Leukocyte quantity in the UA has been shown to be a reliable predictor of true infection.^11,15 Fok and colleagues demonstrated that reflex urine testing in ambulatory male urology patients in which cultures were done on only urine specimens with > 5 white blood cells per high-power field (WBC/HPF) would have missed only 7% of positive UCs, while avoiding 69% of cultures.¹¹

At the Edward Hines, Jr Veterans Affairs Hospital (Hines VA), inappropriate UC ordering and treatment for ASB has been identified as an area needing improvement. An evaluation was conducted at the facility to determine the population of inpatient veterans with a positive UC who were appropriately managed. Of the 113 study patients with a positive UC included in this review, 77 (68%) had a diagnosis of ASB, with > 80% of patients with ASB (and no other suspected infections) receiving antimicrobial therapy.⁸ A subsequent evaluation was conducted at the Hines VA ED to evaluate UTI treatment and follow-up. Of the 173 ED patients included, 23% received antibiotic therapy for an ASB and 60% had a UA and UC collected but did not report symptoms.⁹ Finally, a review by the Hines VA laboratory showed that in May 2017, of 359 UCs sent from various locations of the hospital, 38% were obtained in the setting of a negative UA.

A multidisciplinary group with representation from primary care, infectious diseases, pharmacy, nursing, laboratory, and informatics was created with a goal to improve the workup and management of UTIs. In addition to periodic education for the clinicians regarding appropriate use and interpretation of UA and UC along with judicious use of antimicrobials especially in the setting of ASB, a UA to reflex culture process change was implemented. This allowed for automatic cancellation of a UC in the setting of a negative UA, which was designed to help facilitate appropriate UC ordering.

Methods

The primary objective of this study was to compare the frequency of inappropriate UC use and inappropriate antibiotic prescribing pre- and postimplementation of this UA to reflex culture process change. An inappropriate UC was defined as a UC ordered despite a negative UA in asymptomatic patients. Inappropriate antibiotic prescribing was defined as treatment of patients with ASB. The secondary objective evaluated postintervention data to assess the frequency of outpatient, ED, and hospital visits for UTI-related symptoms in the group of patients that had a UC cancelled as a result of the new process change (within a 7-day period of the initial UA) to determine whether patients with true infections were missed due to the process change.

Study Design and Setting

This pre-post quality improvement (QI) study analyzed the UC-ordering practices for UTIs sent from the ED at the Hines VA. This VA is a 483-bed tertiary care hospital in Chicago, Illinois, and serves > 57,000 veterans and about 23,000 ED visits annually. This study was approved by the Edward Hines, Jr VA Institutional Review Board as a quality assurance/QI proposal prior to data collection.

Patient Selection

All patients who received a UA with or without a UC sent from the ED between October 17, 2017 and January 17, 2018 were identified by the microbiology laboratory and a list was generated. Postintervention data were compared with data from a previous analysis performed at the Hines VA in 2015 (baseline data), which found that UCs were collected frequently despite negative UA, and often resulted in the prescribing of unnecessary antibiotics.⁹

When comparing postintervention data with preintervention data for the primary study objective, the same exclusion criteria from the 2015 study were applied to the present study, which excluded ED patients who were admitted for inpatient care, concurrent antibiotic therapy for a non-UTI indication, duplicate cultures, and use of chronic bladder management devices. All patients identified as receiving a UA during the specified postintervention study period were included for evaluation of the secondary study objective.

Interventions

After physician education, an ED process change was implemented on October 3, 2017. This process change involved the creation of new order sets in the EHR that allowed clinicians to order a UA only, a UA with culture that would be cancelled by laboratory personnel if the UA did not result in > 5 WBC/HPF, and a UA with culture designated as do not cancel, where the UC was processed regardless of the UA results. The scenarios in which the latter option was considered appropriate were listed on the ordering screen and included pregnancy, a genitourinary procedure with necessary preoperative culture, and neutropenia.

Measurements

Postimplementation, all UAs were reviewed and grouped as follows: (1) positive UA with subsequent UC; (2) negative UA, culture cancelled; (3) only UA ordered (no culture); or (4) do not cancel UC ordered. Of the UAs that were analyzed, the following data were collected: demographics, comorbidities, concurrent medications for benign prostatic hyperplasia (BPH) and/or overactive bladder (OAB), documented allergies/adverse drug reactions to antibiotics, date of ED visit, documented UTI signs/symptoms (defined as frequency, urgency, dysuria, fever, suprapubic pain, or altered mental status in patients unable to verbalize urinary symptoms), UC results and susceptibilities, number of UCs repeated within 7 days after initial UA, requirement of antibiotic for UTI within 7 days of initial UA, antibiotic prescribed, duration of antibiotic therapy, and outpatient visits, ED visits, or need for hospital admission within 7 days of the initial UA for UTI-related symptoms. Other relevant UA and UC data that could not be obtained from the EHR were collected by generating a report using the Veterans Information Systems and Technology Architecture (VistA).

Analysis

Statistical analysis was performed using SAS v9.4. Independent t tests and Fisher exact tests were used to describe difference pre- and postintervention. Statistical significance was considered for P < .05. Based on results from the previous study conducted at this facility in addition to a literature review, it was determined that 92 patients in each group (pre- and postintervention) would be necessary to detect a 15% increase in percentage of patients appropriately treated for a UTI.

Results

There were 684 UAs evaluated from ED visits, 429 preintervention and 255 postintervention. The 255 patients were evaluated for the secondary objective of the study. Of the 255 patients with UAs identified postintervention, 150 were excluded based on the predefined exclusion criteria, and the remaining 105 were compared with the 173 patients from the preintervention group and were included in the analysis for the primary objective (Figure 1).

Patients in the postintervention group were younger than those in the preintervention group (P < .02): otherwise the groups were similar (Table 1). Inappropriate antibiotics for ASB decreased from 10.2% preintervention to 1.9% postintervention (odds ratio, 0.17; P = .01) (Table 2). UC processing despite a negative UA significantly decreased from 100% preintervention to 38.6% postintervention (P < .001) (Table 3). In patients with a negative UA, antibiotic prescribing decreased by 25.3% postintervention, but this difference was not statistically significant.

Discussion

A simple process change at the Hines VA resulted in benefits related to antimicrobial stewardship without conferring adverse outcomes on patient safety. Both UC processing despite a negative UA and inappropriate antibiotic prescribing for ASB were reduced significantly postintervention. This process change was piloted in the ED where UCs are often included as part of the initial diagnostic testing in patients who may not report UTI-related symptoms but for whom a UC is often bundled with other infectious workup, depending on the patient presentation.

Reflex testing of urine specimens has been described in the literature, both in an exploratory nature where impact of a reflex UC cancellation protocol based on certain UA criteria is measured by percent reduction of UCs processed as well as results of such interventions implemented into clinical practice.^11-13 A retrospective study performed at the University of North Carolina Medical Center evaluated patients who presented to the ED during a 6-month period and had both an automated UA and UC collected. UC processing was restricted to UA that was positive for nitrites, leukocyte esterase, bacteria, or > 10 WBC/HPF. Use of this reflex culture cancellation protocol could have eliminated 604 of the 1546 (39.1%) cultures processed. However, 11 of the 314 (3.5%) positive cultures could have been missed.¹³ This same protocol was externally validated at another large academic ED setting, where similar results were found.¹⁴

Limitations

This single-center, pre-post QI study was not without limitations. Manual chart reviews were required, and accuracy of information was dependent on clinician documentation and assessment of UTI-related symptoms. The population studied was predominately older males; thus, results may not be applicable to females or young adults. Additionally, recognition of a negative UA and subsequent cancellation of the UC was dependent on laboratory personnel. As noted in the patient group with a positive UA, some of these UAs were negative and may have been overlooked; therefore, subsequent UCs were inappropriately processed. However, this occurred infrequently and confirmed the low probability of true UTI in the setting of a negative UA. Follow-up for UTI-related symptoms may not have been captured if a patient had presented to an outside facility. Last, definitions of a positive UA differed slightly between the pre- and postintervention groups. The preintervention study defined a positive UA as a WBC count > 5 WBC/HPF and positive leukocyte esterase, whereas the present study defined a positive UA with a WBC count > 5. This may have resulted in an overestimation of positive UA in the postintervention group.

Conclusions

Better selective use of UC testing may improve stewardship resources and reduce costs impacting both ED and clinical laboratories. Furthermore, benefits can include a reduction in the use of time and resources required to collect samples for culture, use of test supplies, the time and effort required to process the large number of negative cultures, and resources devoted to the follow-up of these ED culture results. The described UA to reflex culture process change demonstrated a significant reduction in the processing of inappropriate UC and unnecessary antibiotics for ASB. There were no missed UTIs or other adverse patient outcomes noted. This process change has been implemented in all departments at the Hines VA and additional data will be collected to ensure consistent outcomes.

Automated urine cultures (UCs) following urinalysis (UA) are often used in emergency departments (EDs) to identify urinary tract infections (UTIs). The fast-paced environment of the ED makes this method of proactive collection and facilitation of UC favorable. However, results are often reported as no organism growth or the growth of clinically insignificant organisms, leading to the overdetection and overtreatment of asymptomatic bacteriuria (ASB).^1-3 An estimated 30 to 60% of patients with ASB receive unwarranted antibiotic treatment, which is associated with an increased risk of developing Clostridioides difficile infection and contributes to the development of antimicrobial resistance.^4-10 The costs associated with UC are an important consideration given the use of resources, the time and effort required to collect and process large numbers of negative cultures, and further efforts devoted to the follow-up of ED culture results.

Changes in traditional testing involving testing of both a UA and UC to reflex testing where urine specimens undergo culture only if they meet certain criteria have been described.^11-14 This change in traditional testing aims to reduce the number of potentially unnecessary cultures performed without compromising clinical care. Leukocyte quantity in the UA has been shown to be a reliable predictor of true infection.^11,15 Fok and colleagues demonstrated that reflex urine testing in ambulatory male urology patients in which cultures were done on only urine specimens with > 5 white blood cells per high-power field (WBC/HPF) would have missed only 7% of positive UCs, while avoiding 69% of cultures.¹¹

At the Edward Hines, Jr Veterans Affairs Hospital (Hines VA), inappropriate UC ordering and treatment for ASB has been identified as an area needing improvement. An evaluation was conducted at the facility to determine the population of inpatient veterans with a positive UC who were appropriately managed. Of the 113 study patients with a positive UC included in this review, 77 (68%) had a diagnosis of ASB, with > 80% of patients with ASB (and no other suspected infections) receiving antimicrobial therapy.⁸ A subsequent evaluation was conducted at the Hines VA ED to evaluate UTI treatment and follow-up. Of the 173 ED patients included, 23% received antibiotic therapy for an ASB and 60% had a UA and UC collected but did not report symptoms.⁹ Finally, a review by the Hines VA laboratory showed that in May 2017, of 359 UCs sent from various locations of the hospital, 38% were obtained in the setting of a negative UA.

A multidisciplinary group with representation from primary care, infectious diseases, pharmacy, nursing, laboratory, and informatics was created with a goal to improve the workup and management of UTIs. In addition to periodic education for the clinicians regarding appropriate use and interpretation of UA and UC along with judicious use of antimicrobials especially in the setting of ASB, a UA to reflex culture process change was implemented. This allowed for automatic cancellation of a UC in the setting of a negative UA, which was designed to help facilitate appropriate UC ordering.