Original Research

Many veterans of the Gulf War are experiencing deployment-related chronic illness, known as Gulf War illness (GWI). Symptoms of GWI include cognitive impairments (mood and memory), chronic fatigue, musculoskeletal pain, gastrointestinal (GI) disorders, respiratory problems, and skin rashes.^1-4 Three survey studies of the physical and mental health of a large cohort of Gulf War and Gulf era veterans, conducted by the US Department of Veterans Affairs (VA) Office of Public Health, established the increased prevalence of GWI in the decades that followed the end of the conflict.^5-7 Thus, GWI has become the signature adverse health-related outcome of the Gulf War. Quality improvement (QI) within the Veterans Health Administration (VHA) is needed in the diagnosis and treatment of GWI.

Background

GWI was first termed chronic multisymptom illness (CMI) by the Centers for Disease Control and Prevention (CDC). According to the CDC-10 case definition, CMI in veterans of the 1990-1991 Gulf War is defined as having ≥ 1 symptoms lasting ≥ 6 months in at least 2 of 3 categories: fatigue, depressed mood and altered cognition, and musculoskeletal pain.³ The Kansas case definition of GWI is more specific and is defined as having moderate-to-severe symptoms that are unexplained by any other diagnosis, in at least 3 of 6 categories: fatigue/sleep, somatic pain, neurologic/cognition/mood, GI, respiratory, and skin.⁴ Although chronic unexplained symptoms have occurred after other modern conflicts, the prevalence of GWI among Gulf War veterans has proven higher than those of prior conflicts.⁸

The Persian Gulf War Veterans Act of 1998 and the Veterans Programs Enhancement Act of 1998 mandated studies by the Institute of Medicine (IOM) on the biologic and chemical exposures that may have contributed to illness in the Kuwaiti theater of operations.⁹ However, elucidating the etiology and underlying pathophysiology of GWI has been a major research challenge. In the absence of objective diagnostic measures, an understanding of the fundamental pathophysiology, evidence-based treatments, a single case definition, and definitive guidelines for health care providers (HCPs) for the diagnosis and management of GWI has not been produced. As a result, veterans with GWI have struggled for nearly 3 decades to find a consistent diagnosis of and an effective treatment for their condition.

According to a report by the Government Accountability Office (GAO), the VA approved only 17% of claims for compensation for veterans with GWI from 2011 to 2015, about one-third the level of approval for all other claimed disabilities.¹⁰ Although the VA applied GAO recommendations to improve the compensation process, many veterans consider that their illness is treated as psychosomatic in clinical practice, despite emerging evidence of GWI-associated biomarkers.¹¹ Others think they have been forgotten due to their short 1-year period of service in the Gulf War.¹² To realign research, guidelines, clinical care, and the health care experience of veterans with GWI, focused QI within VHA is urgently needed.

Veterans of Operations Enduring Freedom, Iraqi Freedom, and New Dawn (OEF/OIF/OND) are experiencing similar CMI symptoms. A study of US Army Reserve OEF/OIF veterans found that > 60% met the CDC-10 case definition for GWI 1-year postdeployment.¹³ Thus, CMI is emerging as a serious health problem for post-9/11 veterans. The evidence of postdeployment CMI among veterans of recent conflicts underscores the need to increase efforts at a national level, beginning with the VHA. This report includes a summary of Gulf War veterans’ experiences at the Minneapolis VA Health Care System (MVAHCS) and a proposal for QI of MVAHCS processes focused on HCP education and clinical care.

Methods

To determine areas of GWI health care that needed QI at the MVAHCS, veterans with GWI were contacted for a telephone survey. These veterans had participated in the Gulf War Illness Inflammation Reduction Trial (ClinicalTrials.gov. Identifier: NCT02506192). Therefore, all met the Kansas case definition for GWI.⁴ The aim of the survey was to characterize veterans’ experiences seeking health care for chronic postdeployment symptoms.

Sixty Gulf War veterans were contacted by telephone and invited to participate in a 15-minute survey about their experience seeking diagnosis and treatment for GWI. They were informed that the survey was voluntary and confidential, that it was not part of the research trial in which they had been enrolled, and that their participation would not affect compensation received from VA. Verbal consent was requested, and 30 veterans agreed to participate in the survey.

The survey included questions about the course of illness, disability and service connection status, HCPs seen, and suggestions for improvement in their care (Table 1). 

Results

Of the 30 veterans who participated in the survey, most were male with only 2 female veterans. This proportion of female veterans (7%) is similar to the overall percentage of female veterans (6.7%) of the first Gulf War.² Ages ranged from 46 to 66 years with a mean age of 53. Mean duration of illness, defined as time elapsed since perceived onset of chronic systemic symptoms during or after deployment, was 22.8 years, with a range of 4 to 27 years. Most respondents reported symptom onset within a few years after the end of the conflict, while a few reported the onset within weeks of arriving in the Kuwaiti theater of operations. A little more than half the respondents considered themselves disabled due to their symptoms, while one-third reported losing the ability to work due to symptoms. Respondents described needing to reduce hours, retire early, or stop working altogether because of their symptoms.

Respondents attributed several common chronic symptoms to deployment in the Gulf Wars (Table 2). 

Most veterans surveyed were service connected for individual chronic symptoms. Some were service connected for systemic conditions such as fibromyalgia (FM), chronic fatigue syndrome (CFS), and irritable bowel syndrome (IBS) (5 veterans were connected for each condition). Three of the 30 veterans had been diagnosed with GWI—2 by past VA physicians and 1 by a physician at a GWI research center in another state. Of those 3, only 1 was service connected for the condition. Three respondents were not service connected at all.

The most common VA HCPs seen were in primary care and neurology followed by psychiatry and psychology. Of non-VA HCPs, most respondents saw primary care providers (PCPs) followed by chiropractors (Table 4).

Before taking the Gulf War survey, a broad subjective question was posed. Respondents were asked whether VA HCPs were “supportive as you sought care for chronic postdeployment symptoms.” A majority of veterans reported that their VA HCPs were supportive. Reasons veterans gave for VA HCPs lack of support included feeling that HCPs did not believe them or trust their reported symptoms; did not care about their symptoms; refused to attribute their symptoms to Gulf War deployment; attributed symptoms to mental health issues; focused on doing things a certain way; or did not have the tools or information necessary to help.

Most non-VA HCPs were supportive. Reasons community HCPs were not supportive included “not looking at the whole picture,” not knowing veteran issues, not feeling comfortable with GWI, or not having much they could do.

Veterans were then asked whether they felt their HCPs were knowledgeable about GWI, and 13 respondents reported that their HCP was knowledgeable. Reasons respondents felt VA HCPs were not knowledgeable included denying that GWI exists, attributing symptoms to other conditions, not being aware of or familiar with GWI, needing education from the veteran, avoiding discussion about GWI or not caring to learn, or not knowing the latest research evidence to talk about GWI with authority. Compared with VA HCPs, veterans found community HCPs about half as likely to be knowledgeable about GWI. Many reported that community HCPs had not heard of GWI or had no knowledge about it.

Respondents also were asked what types of treatments they tried in order to typify the care received. The most common responses were pain medications, symptom-specific treatments, or “just putting up with it” (no treatment). Many patients were also self-medicating, trying lifestyle changes, or seeking alternative therapies.

Finally, respondents were asked on a scale of 0 (very unsatisfied) to 5 (very satisfied), how satisfied they were with their overall care at the VA. The majority were satisfied with their overall care, with two-thirds very satisfied (5 of 5) or pretty satisfied (4 of 5). Only 3 (10%) were unsatisfied or very unsatisfied. Respondents had the following comments about their care: “They treat me like I am important;” “I am very thankful even though they cannot figure it out;” “They are doing the best they can with no answers and not enough help;” “[I know] it is still a work in progress.” A number of respondents were satisfied with some HCPs or care for some but not all of their symptoms. Reasons respondents were less satisfied included desiring answers, feeling they were not respected, or feeling that their concerns were not addressed.

When asked for suggestions for improvement of GWI care, the most common response was providing up-to-date HCP education (Table 5).

Discussion

The veterans participating in this QI survey had similar demographics, symptomology, and exposures as did those in other studies.^1-7 Therefore, improvements based on their responses are likely applicable to the health care of veterans experiencing GWI-associated symptoms at other VA health care systems as well.

Veterans with GWI can lose significant functional capacity and productivity due to their symptoms. The symptoms are chronic and have afflicted many Gulf War veterans for nearly 3 decades. Furthermore, the prevalence of GWI in Gulf War veterans continues to increase.^5-7 These facts testify to the enormous health-related quality-of-life impact of GWI.

Veterans who meet the Kansas case definition for GWI were not diagnosed or service connected in a uniform manner. Only 3 of the 30 veterans in this study were given a unifying diagnosis that connected their chronic illness to Gulf War deployment. Under current guidelines, Gulf War veterans are able to receive compensation for chronic symptoms in 3 ways: (1) compensation for chronic unexplained symptoms existing for ≥ 6 months that appeared during active duty in the Southwest Asia theater or by December 31, 2021, and are ≥ 10% disabling; (2) the 1995 Persian Gulf War Veterans’ Act recognizes 3 multisymptom illnesses for which veterans can be service connected: FM, CFS, and functional GI disorders, including IBS; and (3) expansion to include any CMI of unknown etiology is underway. A uniform diagnostic protocol based on biomarkers and updated understanding of disease pathology would be helpful.

Respondents shared experiences that demonstrated perceived gaps in HCP support or knowledge. Overall, more respondents found their HCPs supportive. Many of the reasons respondents found HCPs unsupportive related to acknowledgment of symptoms. Also, more respondents found that both VA and non-VA HCPs lacked knowledge about GWI symptoms. These findings further highlight the need for HCP education within the VA and in community-based care.

The treatments tried by respondents also highlight potential areas for improvement. Most of the treatments were for pain; therefore, more involvement with pain clinics and specialists could be helpful. Symptom-specific medications also are appropriate, although only one-third of patients reported use. While medications are not necessarily markers of quality care, the fact that many patients self-medicate or go without treatment suggests that access to care could be improved. In 2014, the VA and the US Department of Defense (DoD) released the “VA/DoD Clinical Practice Guideline for the Management of Chronic Multisymptom Illness,” which recommended treatments for the global disease and specific symptoms.¹⁵

Since then, GWI research points to inflammatory and metabolic disease mechanisms.^11-14,16 As the underlying pathophysiology is further elucidated, practice guidelines will need to be updated to include anti-inflammatory and antioxidant treatments used in practice for GWI and similar chronic systemic illnesses (eg, CFS, FM, and IBS).^17-19

Randomized control trials are needed to determine the efficacy of such medications for the treatment of GWI. As new results emerge, disseminating and updating evidence-based guidelines in a coordinated manner will be required for veterans to receive appropriate treatment. Veterans also seek alternative or nonpharmaceutical interventions, such as physical therapy and diet changes. Improving access to integrative medicine, physical therapy, nutritionists, and other practitioners also could optimize veterans’ health and function.

HCP Education

The Gulf War veteran respondents who participated in the survey noted HCP education, research progress, and veteran inclusion as areas for improvement. Respondents requested dissemination of information on diagnosis and treatment of GWI for HCPs and updates on research and other actions. They suggested ways research could be more effective (such as subgrouping by exposure, which researchers have been doing) and could extend to veterans experiencing CMI from other conflicts as well.²⁰ Respondents also recommended team approaches or centers of excellence in order to receive more comprehensive care.

An asset of VHA is the culture of QI and education. The VA Employee Education System previously produced “Caring for Gulf War I Veterans,” a systemwide training module.²¹ In 2014, updated clinical practice guidelines for GWI were provided by the VA and the DoD, including evidence for each recommendation. In 2016, the VA in collaboration with the IOM produced a report summarizing conclusions and recommendations regarding associations between health concerns and Gulf War deployment.²² A concise guide for HCPs caring for veterans with GWI, updated in 2018, is available.²³ Updated treatment guidelines, based on evolving understanding of GWI pathophysiology, and continuing efforts to disseminate information will be essential.

Respondents most often presented to primary care, both within and outside of MVAHCS. Therefore, VA and community PCPs who see veterans should be equipped to recognize and diagnose GWI as well as be familiar with basic disease management and specialists whom they could refer their patients. Neurology was the second most common specialty seen by respondents. The most prominent symptoms of GWI are related to nervous system function in addition to evidence of underlying neuroinflammation.²⁰ Veterans may present to a neurologist with a variety of concerns, such as cognitive issues, sleep problems, migraines and headaches, and pain. Neurologists could best manage treatments targeting common neurologic GWI symptoms and neuroinflammation, especially as new treatments are discovered.

The next 2 most common specialty services seen were psychiatry and psychology (7 responses for each). Five respondents reported mental health issues as part of their chronic postdeployment symptoms. Population-based studies have indicated that rates of PTSD in Gulf War veterans is 3% to 6%, much lower than the prevalence of GWI.^8,20 The 2010 IOM study concluded that GWI symptoms cannot be ascribed to any known psychiatric disorder. Unfortunately, several surveyed veterans made it clear that they had been denied care due to HCPs attributing their symptoms solely to mental health issues. Therefore, psychiatrists and psychologists must be educated about GWI, mental health issues occurring in Gulf War veterans, and physiologic symptoms of GWI that may mimic or coincide with mental health issues. These HCPs also would be important to include in an interdisciplinary clinic for veterans with GWI.

Finally, respondents sought care from numerous other specialties, including gastroenterology, physical therapy, pulmonology, dermatology, and surgical subspecialties, such as orthopedics and otolaryngology. This wide range of specialists seen emphasizes the need for medical education, beginning in medical school. If provided education on GWI, these specialists would be able to treat veterans with GWI, know to look for updates on GWI management, or know to look for other common symptoms, such as chronic sinusitis in otolaryngology or recurring rashes in dermatology. We also recommend identifying HCPs in these specialties who could be part of an interdisciplinary clinic or be referrals for symptom management.

Protocol Implementation

HCP education and clinical care protocol implementation should be the initial focus of improving GWI management. A team of stakeholders within the different areas of MVAHCS, including education, HCPs, and administrative staff, will need to be developed. Reaching out to VA HCPs who have seen veterans with GWI will be an essential first step to equip them with updated education about the diagnosis and management of CMI. Providing integrated widespread education to current HCPs who are likely to encounter veterans with deployment-related CMI from the Gulf War, OND/OEF/OIF, or other deployments also will be necessary. Finally, educating medical trainees, including residents and medical students, will ensure continuous care for future veterans, post-9/11 veterans.

GWI presentations at medical grand rounds or at other medical community educational events could provide educational outlets. These events create face-to-face opportunities to discuss GWI/CMI education with HCPs, giving them the opportunity to offer feedback about their experiences and create relationships with other HCPs who have seen patients with GWI/CMI. At an educational event, a short postevent feedback form that indicates whether HCPs would like more information or get involved in a clinic for veterans with CMI could be included. This information would help identify key HCPs and areas within the local VA needing further improvements, such as creating a clinic for veterans with GWI.

Since 1946, the VA has worked with academic institutions to provide state-of-the-art health care to US veterans and train new HCPs to meet the health care needs of the nation. Every year, > 40,000 residents and 20,000 medical students receive medical training at VA facilities, making VA the largest single provider of medical education in the country. Therefore, providing detailed GWI/CMI education to medical students and residents as a standard part of the VA Talent Management System would be of value for all VA professionals.

GWI Clinics

Access to comprehensive care can be accomplished by organizing a clinic for veterans with GWI. The most likely effective location would be in primary care. PCPs who have seen veterans with GWI and/or expressed interest in learning more about GWI will be the initial point of contact. As the primary care service has connections to ancillary services, such as pharmacists, dieticians, psychologists, and social workers, organizing 1 day each week to see patients with GWI would improve care.

As the need for specialty care arises, the team also would need to identify specialists willing to receive referrals from HCPs of veterans with GWI. These specialists could be identified via feedback forms from educational events, surveys after an online educational training, or through relationships among VA physicians. As the clinic becomes established, it may be effective to have certain commonly seen specialists available in person, most likely neurology, psychiatry, gastroenterology, pulmonology, and dermatology. Also, relationships with a pain clinic, sleep medicine, and integrative medicine services should be established.

Measures of improvement in the veteran health care experience could include veterans’ perceptions of the supportiveness and knowledge of physicians about GWI as well as overall satisfaction. A follow-up survey on these measures of veterans involved in a GWI clinic and those not involved would be a way to determine whether these clinics better meet veterans’ needs and what additional QI is needed.

Conclusion

A significant number of Gulf War veterans experience chronic postdeployment symptoms that need to be better addressed. Physicians need to be equipped to recognize and manage GWI and similar postdeployment CMI among veterans of OEF/OIF/OND. We recommend creating an educational initiative about GWI among VA physicians and trainees, connecting physicians who see veterans with GWI, and establishing an interdisciplinary clinic with a referral system as the next steps to improve care for veterans. An additional goal would be to reach out to veteran networks to update them on GWI research, education, and available health care, as veterans are the essential stakeholders in the QI process.

Many veterans of the Gulf War are experiencing deployment-related chronic illness, known as Gulf War illness (GWI). Symptoms of GWI include cognitive impairments (mood and memory), chronic fatigue, musculoskeletal pain, gastrointestinal (GI) disorders, respiratory problems, and skin rashes.^1-4 Three survey studies of the physical and mental health of a large cohort of Gulf War and Gulf era veterans, conducted by the US Department of Veterans Affairs (VA) Office of Public Health, established the increased prevalence of GWI in the decades that followed the end of the conflict.^5-7 Thus, GWI has become the signature adverse health-related outcome of the Gulf War. Quality improvement (QI) within the Veterans Health Administration (VHA) is needed in the diagnosis and treatment of GWI.

Background

GWI was first termed chronic multisymptom illness (CMI) by the Centers for Disease Control and Prevention (CDC). According to the CDC-10 case definition, CMI in veterans of the 1990-1991 Gulf War is defined as having ≥ 1 symptoms lasting ≥ 6 months in at least 2 of 3 categories: fatigue, depressed mood and altered cognition, and musculoskeletal pain.³ The Kansas case definition of GWI is more specific and is defined as having moderate-to-severe symptoms that are unexplained by any other diagnosis, in at least 3 of 6 categories: fatigue/sleep, somatic pain, neurologic/cognition/mood, GI, respiratory, and skin.⁴ Although chronic unexplained symptoms have occurred after other modern conflicts, the prevalence of GWI among Gulf War veterans has proven higher than those of prior conflicts.⁸

The Persian Gulf War Veterans Act of 1998 and the Veterans Programs Enhancement Act of 1998 mandated studies by the Institute of Medicine (IOM) on the biologic and chemical exposures that may have contributed to illness in the Kuwaiti theater of operations.⁹ However, elucidating the etiology and underlying pathophysiology of GWI has been a major research challenge. In the absence of objective diagnostic measures, an understanding of the fundamental pathophysiology, evidence-based treatments, a single case definition, and definitive guidelines for health care providers (HCPs) for the diagnosis and management of GWI has not been produced. As a result, veterans with GWI have struggled for nearly 3 decades to find a consistent diagnosis of and an effective treatment for their condition.

According to a report by the Government Accountability Office (GAO), the VA approved only 17% of claims for compensation for veterans with GWI from 2011 to 2015, about one-third the level of approval for all other claimed disabilities.¹⁰ Although the VA applied GAO recommendations to improve the compensation process, many veterans consider that their illness is treated as psychosomatic in clinical practice, despite emerging evidence of GWI-associated biomarkers.¹¹ Others think they have been forgotten due to their short 1-year period of service in the Gulf War.¹² To realign research, guidelines, clinical care, and the health care experience of veterans with GWI, focused QI within VHA is urgently needed.

Veterans of Operations Enduring Freedom, Iraqi Freedom, and New Dawn (OEF/OIF/OND) are experiencing similar CMI symptoms. A study of US Army Reserve OEF/OIF veterans found that > 60% met the CDC-10 case definition for GWI 1-year postdeployment.¹³ Thus, CMI is emerging as a serious health problem for post-9/11 veterans. The evidence of postdeployment CMI among veterans of recent conflicts underscores the need to increase efforts at a national level, beginning with the VHA. This report includes a summary of Gulf War veterans’ experiences at the Minneapolis VA Health Care System (MVAHCS) and a proposal for QI of MVAHCS processes focused on HCP education and clinical care.

Methods

To determine areas of GWI health care that needed QI at the MVAHCS, veterans with GWI were contacted for a telephone survey. These veterans had participated in the Gulf War Illness Inflammation Reduction Trial (ClinicalTrials.gov. Identifier: NCT02506192). Therefore, all met the Kansas case definition for GWI.⁴ The aim of the survey was to characterize veterans’ experiences seeking health care for chronic postdeployment symptoms.

Sixty Gulf War veterans were contacted by telephone and invited to participate in a 15-minute survey about their experience seeking diagnosis and treatment for GWI. They were informed that the survey was voluntary and confidential, that it was not part of the research trial in which they had been enrolled, and that their participation would not affect compensation received from VA. Verbal consent was requested, and 30 veterans agreed to participate in the survey.

The survey included questions about the course of illness, disability and service connection status, HCPs seen, and suggestions for improvement in their care (Table 1). 

Results

Of the 30 veterans who participated in the survey, most were male with only 2 female veterans. This proportion of female veterans (7%) is similar to the overall percentage of female veterans (6.7%) of the first Gulf War.² Ages ranged from 46 to 66 years with a mean age of 53. Mean duration of illness, defined as time elapsed since perceived onset of chronic systemic symptoms during or after deployment, was 22.8 years, with a range of 4 to 27 years. Most respondents reported symptom onset within a few years after the end of the conflict, while a few reported the onset within weeks of arriving in the Kuwaiti theater of operations. A little more than half the respondents considered themselves disabled due to their symptoms, while one-third reported losing the ability to work due to symptoms. Respondents described needing to reduce hours, retire early, or stop working altogether because of their symptoms.

Respondents attributed several common chronic symptoms to deployment in the Gulf Wars (Table 2). 

Most veterans surveyed were service connected for individual chronic symptoms. Some were service connected for systemic conditions such as fibromyalgia (FM), chronic fatigue syndrome (CFS), and irritable bowel syndrome (IBS) (5 veterans were connected for each condition). Three of the 30 veterans had been diagnosed with GWI—2 by past VA physicians and 1 by a physician at a GWI research center in another state. Of those 3, only 1 was service connected for the condition. Three respondents were not service connected at all.

The most common VA HCPs seen were in primary care and neurology followed by psychiatry and psychology. Of non-VA HCPs, most respondents saw primary care providers (PCPs) followed by chiropractors (Table 4).

Before taking the Gulf War survey, a broad subjective question was posed. Respondents were asked whether VA HCPs were “supportive as you sought care for chronic postdeployment symptoms.” A majority of veterans reported that their VA HCPs were supportive. Reasons veterans gave for VA HCPs lack of support included feeling that HCPs did not believe them or trust their reported symptoms; did not care about their symptoms; refused to attribute their symptoms to Gulf War deployment; attributed symptoms to mental health issues; focused on doing things a certain way; or did not have the tools or information necessary to help.

Most non-VA HCPs were supportive. Reasons community HCPs were not supportive included “not looking at the whole picture,” not knowing veteran issues, not feeling comfortable with GWI, or not having much they could do.

Veterans were then asked whether they felt their HCPs were knowledgeable about GWI, and 13 respondents reported that their HCP was knowledgeable. Reasons respondents felt VA HCPs were not knowledgeable included denying that GWI exists, attributing symptoms to other conditions, not being aware of or familiar with GWI, needing education from the veteran, avoiding discussion about GWI or not caring to learn, or not knowing the latest research evidence to talk about GWI with authority. Compared with VA HCPs, veterans found community HCPs about half as likely to be knowledgeable about GWI. Many reported that community HCPs had not heard of GWI or had no knowledge about it.

Respondents also were asked what types of treatments they tried in order to typify the care received. The most common responses were pain medications, symptom-specific treatments, or “just putting up with it” (no treatment). Many patients were also self-medicating, trying lifestyle changes, or seeking alternative therapies.

Finally, respondents were asked on a scale of 0 (very unsatisfied) to 5 (very satisfied), how satisfied they were with their overall care at the VA. The majority were satisfied with their overall care, with two-thirds very satisfied (5 of 5) or pretty satisfied (4 of 5). Only 3 (10%) were unsatisfied or very unsatisfied. Respondents had the following comments about their care: “They treat me like I am important;” “I am very thankful even though they cannot figure it out;” “They are doing the best they can with no answers and not enough help;” “[I know] it is still a work in progress.” A number of respondents were satisfied with some HCPs or care for some but not all of their symptoms. Reasons respondents were less satisfied included desiring answers, feeling they were not respected, or feeling that their concerns were not addressed.

When asked for suggestions for improvement of GWI care, the most common response was providing up-to-date HCP education (Table 5).

Discussion

The veterans participating in this QI survey had similar demographics, symptomology, and exposures as did those in other studies.^1-7 Therefore, improvements based on their responses are likely applicable to the health care of veterans experiencing GWI-associated symptoms at other VA health care systems as well.

Veterans with GWI can lose significant functional capacity and productivity due to their symptoms. The symptoms are chronic and have afflicted many Gulf War veterans for nearly 3 decades. Furthermore, the prevalence of GWI in Gulf War veterans continues to increase.^5-7 These facts testify to the enormous health-related quality-of-life impact of GWI.

Veterans who meet the Kansas case definition for GWI were not diagnosed or service connected in a uniform manner. Only 3 of the 30 veterans in this study were given a unifying diagnosis that connected their chronic illness to Gulf War deployment. Under current guidelines, Gulf War veterans are able to receive compensation for chronic symptoms in 3 ways: (1) compensation for chronic unexplained symptoms existing for ≥ 6 months that appeared during active duty in the Southwest Asia theater or by December 31, 2021, and are ≥ 10% disabling; (2) the 1995 Persian Gulf War Veterans’ Act recognizes 3 multisymptom illnesses for which veterans can be service connected: FM, CFS, and functional GI disorders, including IBS; and (3) expansion to include any CMI of unknown etiology is underway. A uniform diagnostic protocol based on biomarkers and updated understanding of disease pathology would be helpful.

Respondents shared experiences that demonstrated perceived gaps in HCP support or knowledge. Overall, more respondents found their HCPs supportive. Many of the reasons respondents found HCPs unsupportive related to acknowledgment of symptoms. Also, more respondents found that both VA and non-VA HCPs lacked knowledge about GWI symptoms. These findings further highlight the need for HCP education within the VA and in community-based care.

The treatments tried by respondents also highlight potential areas for improvement. Most of the treatments were for pain; therefore, more involvement with pain clinics and specialists could be helpful. Symptom-specific medications also are appropriate, although only one-third of patients reported use. While medications are not necessarily markers of quality care, the fact that many patients self-medicate or go without treatment suggests that access to care could be improved. In 2014, the VA and the US Department of Defense (DoD) released the “VA/DoD Clinical Practice Guideline for the Management of Chronic Multisymptom Illness,” which recommended treatments for the global disease and specific symptoms.¹⁵

Since then, GWI research points to inflammatory and metabolic disease mechanisms.^11-14,16 As the underlying pathophysiology is further elucidated, practice guidelines will need to be updated to include anti-inflammatory and antioxidant treatments used in practice for GWI and similar chronic systemic illnesses (eg, CFS, FM, and IBS).^17-19

Randomized control trials are needed to determine the efficacy of such medications for the treatment of GWI. As new results emerge, disseminating and updating evidence-based guidelines in a coordinated manner will be required for veterans to receive appropriate treatment. Veterans also seek alternative or nonpharmaceutical interventions, such as physical therapy and diet changes. Improving access to integrative medicine, physical therapy, nutritionists, and other practitioners also could optimize veterans’ health and function.

HCP Education

The Gulf War veteran respondents who participated in the survey noted HCP education, research progress, and veteran inclusion as areas for improvement. Respondents requested dissemination of information on diagnosis and treatment of GWI for HCPs and updates on research and other actions. They suggested ways research could be more effective (such as subgrouping by exposure, which researchers have been doing) and could extend to veterans experiencing CMI from other conflicts as well.²⁰ Respondents also recommended team approaches or centers of excellence in order to receive more comprehensive care.

An asset of VHA is the culture of QI and education. The VA Employee Education System previously produced “Caring for Gulf War I Veterans,” a systemwide training module.²¹ In 2014, updated clinical practice guidelines for GWI were provided by the VA and the DoD, including evidence for each recommendation. In 2016, the VA in collaboration with the IOM produced a report summarizing conclusions and recommendations regarding associations between health concerns and Gulf War deployment.²² A concise guide for HCPs caring for veterans with GWI, updated in 2018, is available.²³ Updated treatment guidelines, based on evolving understanding of GWI pathophysiology, and continuing efforts to disseminate information will be essential.

Respondents most often presented to primary care, both within and outside of MVAHCS. Therefore, VA and community PCPs who see veterans should be equipped to recognize and diagnose GWI as well as be familiar with basic disease management and specialists whom they could refer their patients. Neurology was the second most common specialty seen by respondents. The most prominent symptoms of GWI are related to nervous system function in addition to evidence of underlying neuroinflammation.²⁰ Veterans may present to a neurologist with a variety of concerns, such as cognitive issues, sleep problems, migraines and headaches, and pain. Neurologists could best manage treatments targeting common neurologic GWI symptoms and neuroinflammation, especially as new treatments are discovered.

The next 2 most common specialty services seen were psychiatry and psychology (7 responses for each). Five respondents reported mental health issues as part of their chronic postdeployment symptoms. Population-based studies have indicated that rates of PTSD in Gulf War veterans is 3% to 6%, much lower than the prevalence of GWI.^8,20 The 2010 IOM study concluded that GWI symptoms cannot be ascribed to any known psychiatric disorder. Unfortunately, several surveyed veterans made it clear that they had been denied care due to HCPs attributing their symptoms solely to mental health issues. Therefore, psychiatrists and psychologists must be educated about GWI, mental health issues occurring in Gulf War veterans, and physiologic symptoms of GWI that may mimic or coincide with mental health issues. These HCPs also would be important to include in an interdisciplinary clinic for veterans with GWI.

Finally, respondents sought care from numerous other specialties, including gastroenterology, physical therapy, pulmonology, dermatology, and surgical subspecialties, such as orthopedics and otolaryngology. This wide range of specialists seen emphasizes the need for medical education, beginning in medical school. If provided education on GWI, these specialists would be able to treat veterans with GWI, know to look for updates on GWI management, or know to look for other common symptoms, such as chronic sinusitis in otolaryngology or recurring rashes in dermatology. We also recommend identifying HCPs in these specialties who could be part of an interdisciplinary clinic or be referrals for symptom management.

Protocol Implementation

HCP education and clinical care protocol implementation should be the initial focus of improving GWI management. A team of stakeholders within the different areas of MVAHCS, including education, HCPs, and administrative staff, will need to be developed. Reaching out to VA HCPs who have seen veterans with GWI will be an essential first step to equip them with updated education about the diagnosis and management of CMI. Providing integrated widespread education to current HCPs who are likely to encounter veterans with deployment-related CMI from the Gulf War, OND/OEF/OIF, or other deployments also will be necessary. Finally, educating medical trainees, including residents and medical students, will ensure continuous care for future veterans, post-9/11 veterans.

GWI presentations at medical grand rounds or at other medical community educational events could provide educational outlets. These events create face-to-face opportunities to discuss GWI/CMI education with HCPs, giving them the opportunity to offer feedback about their experiences and create relationships with other HCPs who have seen patients with GWI/CMI. At an educational event, a short postevent feedback form that indicates whether HCPs would like more information or get involved in a clinic for veterans with CMI could be included. This information would help identify key HCPs and areas within the local VA needing further improvements, such as creating a clinic for veterans with GWI.

Since 1946, the VA has worked with academic institutions to provide state-of-the-art health care to US veterans and train new HCPs to meet the health care needs of the nation. Every year, > 40,000 residents and 20,000 medical students receive medical training at VA facilities, making VA the largest single provider of medical education in the country. Therefore, providing detailed GWI/CMI education to medical students and residents as a standard part of the VA Talent Management System would be of value for all VA professionals.

GWI Clinics

Access to comprehensive care can be accomplished by organizing a clinic for veterans with GWI. The most likely effective location would be in primary care. PCPs who have seen veterans with GWI and/or expressed interest in learning more about GWI will be the initial point of contact. As the primary care service has connections to ancillary services, such as pharmacists, dieticians, psychologists, and social workers, organizing 1 day each week to see patients with GWI would improve care.

As the need for specialty care arises, the team also would need to identify specialists willing to receive referrals from HCPs of veterans with GWI. These specialists could be identified via feedback forms from educational events, surveys after an online educational training, or through relationships among VA physicians. As the clinic becomes established, it may be effective to have certain commonly seen specialists available in person, most likely neurology, psychiatry, gastroenterology, pulmonology, and dermatology. Also, relationships with a pain clinic, sleep medicine, and integrative medicine services should be established.

Measures of improvement in the veteran health care experience could include veterans’ perceptions of the supportiveness and knowledge of physicians about GWI as well as overall satisfaction. A follow-up survey on these measures of veterans involved in a GWI clinic and those not involved would be a way to determine whether these clinics better meet veterans’ needs and what additional QI is needed.

Conclusion

A significant number of Gulf War veterans experience chronic postdeployment symptoms that need to be better addressed. Physicians need to be equipped to recognize and manage GWI and similar postdeployment CMI among veterans of OEF/OIF/OND. We recommend creating an educational initiative about GWI among VA physicians and trainees, connecting physicians who see veterans with GWI, and establishing an interdisciplinary clinic with a referral system as the next steps to improve care for veterans. An additional goal would be to reach out to veteran networks to update them on GWI research, education, and available health care, as veterans are the essential stakeholders in the QI process.

Many veterans of the Gulf War are experiencing deployment-related chronic illness, known as Gulf War illness (GWI). Symptoms of GWI include cognitive impairments (mood and memory), chronic fatigue, musculoskeletal pain, gastrointestinal (GI) disorders, respiratory problems, and skin rashes.^1-4 Three survey studies of the physical and mental health of a large cohort of Gulf War and Gulf era veterans, conducted by the US Department of Veterans Affairs (VA) Office of Public Health, established the increased prevalence of GWI in the decades that followed the end of the conflict.^5-7 Thus, GWI has become the signature adverse health-related outcome of the Gulf War. Quality improvement (QI) within the Veterans Health Administration (VHA) is needed in the diagnosis and treatment of GWI.

Background

GWI was first termed chronic multisymptom illness (CMI) by the Centers for Disease Control and Prevention (CDC). According to the CDC-10 case definition, CMI in veterans of the 1990-1991 Gulf War is defined as having ≥ 1 symptoms lasting ≥ 6 months in at least 2 of 3 categories: fatigue, depressed mood and altered cognition, and musculoskeletal pain.³ The Kansas case definition of GWI is more specific and is defined as having moderate-to-severe symptoms that are unexplained by any other diagnosis, in at least 3 of 6 categories: fatigue/sleep, somatic pain, neurologic/cognition/mood, GI, respiratory, and skin.⁴ Although chronic unexplained symptoms have occurred after other modern conflicts, the prevalence of GWI among Gulf War veterans has proven higher than those of prior conflicts.⁸

The Persian Gulf War Veterans Act of 1998 and the Veterans Programs Enhancement Act of 1998 mandated studies by the Institute of Medicine (IOM) on the biologic and chemical exposures that may have contributed to illness in the Kuwaiti theater of operations.⁹ However, elucidating the etiology and underlying pathophysiology of GWI has been a major research challenge. In the absence of objective diagnostic measures, an understanding of the fundamental pathophysiology, evidence-based treatments, a single case definition, and definitive guidelines for health care providers (HCPs) for the diagnosis and management of GWI has not been produced. As a result, veterans with GWI have struggled for nearly 3 decades to find a consistent diagnosis of and an effective treatment for their condition.

According to a report by the Government Accountability Office (GAO), the VA approved only 17% of claims for compensation for veterans with GWI from 2011 to 2015, about one-third the level of approval for all other claimed disabilities.¹⁰ Although the VA applied GAO recommendations to improve the compensation process, many veterans consider that their illness is treated as psychosomatic in clinical practice, despite emerging evidence of GWI-associated biomarkers.¹¹ Others think they have been forgotten due to their short 1-year period of service in the Gulf War.¹² To realign research, guidelines, clinical care, and the health care experience of veterans with GWI, focused QI within VHA is urgently needed.

Veterans of Operations Enduring Freedom, Iraqi Freedom, and New Dawn (OEF/OIF/OND) are experiencing similar CMI symptoms. A study of US Army Reserve OEF/OIF veterans found that > 60% met the CDC-10 case definition for GWI 1-year postdeployment.¹³ Thus, CMI is emerging as a serious health problem for post-9/11 veterans. The evidence of postdeployment CMI among veterans of recent conflicts underscores the need to increase efforts at a national level, beginning with the VHA. This report includes a summary of Gulf War veterans’ experiences at the Minneapolis VA Health Care System (MVAHCS) and a proposal for QI of MVAHCS processes focused on HCP education and clinical care.

Methods

To determine areas of GWI health care that needed QI at the MVAHCS, veterans with GWI were contacted for a telephone survey. These veterans had participated in the Gulf War Illness Inflammation Reduction Trial (ClinicalTrials.gov. Identifier: NCT02506192). Therefore, all met the Kansas case definition for GWI.⁴ The aim of the survey was to characterize veterans’ experiences seeking health care for chronic postdeployment symptoms.

Sixty Gulf War veterans were contacted by telephone and invited to participate in a 15-minute survey about their experience seeking diagnosis and treatment for GWI. They were informed that the survey was voluntary and confidential, that it was not part of the research trial in which they had been enrolled, and that their participation would not affect compensation received from VA. Verbal consent was requested, and 30 veterans agreed to participate in the survey.

The survey included questions about the course of illness, disability and service connection status, HCPs seen, and suggestions for improvement in their care (Table 1). 

Results

Of the 30 veterans who participated in the survey, most were male with only 2 female veterans. This proportion of female veterans (7%) is similar to the overall percentage of female veterans (6.7%) of the first Gulf War.² Ages ranged from 46 to 66 years with a mean age of 53. Mean duration of illness, defined as time elapsed since perceived onset of chronic systemic symptoms during or after deployment, was 22.8 years, with a range of 4 to 27 years. Most respondents reported symptom onset within a few years after the end of the conflict, while a few reported the onset within weeks of arriving in the Kuwaiti theater of operations. A little more than half the respondents considered themselves disabled due to their symptoms, while one-third reported losing the ability to work due to symptoms. Respondents described needing to reduce hours, retire early, or stop working altogether because of their symptoms.

Respondents attributed several common chronic symptoms to deployment in the Gulf Wars (Table 2). 

Most veterans surveyed were service connected for individual chronic symptoms. Some were service connected for systemic conditions such as fibromyalgia (FM), chronic fatigue syndrome (CFS), and irritable bowel syndrome (IBS) (5 veterans were connected for each condition). Three of the 30 veterans had been diagnosed with GWI—2 by past VA physicians and 1 by a physician at a GWI research center in another state. Of those 3, only 1 was service connected for the condition. Three respondents were not service connected at all.

The most common VA HCPs seen were in primary care and neurology followed by psychiatry and psychology. Of non-VA HCPs, most respondents saw primary care providers (PCPs) followed by chiropractors (Table 4).

Before taking the Gulf War survey, a broad subjective question was posed. Respondents were asked whether VA HCPs were “supportive as you sought care for chronic postdeployment symptoms.” A majority of veterans reported that their VA HCPs were supportive. Reasons veterans gave for VA HCPs lack of support included feeling that HCPs did not believe them or trust their reported symptoms; did not care about their symptoms; refused to attribute their symptoms to Gulf War deployment; attributed symptoms to mental health issues; focused on doing things a certain way; or did not have the tools or information necessary to help.

Most non-VA HCPs were supportive. Reasons community HCPs were not supportive included “not looking at the whole picture,” not knowing veteran issues, not feeling comfortable with GWI, or not having much they could do.

Veterans were then asked whether they felt their HCPs were knowledgeable about GWI, and 13 respondents reported that their HCP was knowledgeable. Reasons respondents felt VA HCPs were not knowledgeable included denying that GWI exists, attributing symptoms to other conditions, not being aware of or familiar with GWI, needing education from the veteran, avoiding discussion about GWI or not caring to learn, or not knowing the latest research evidence to talk about GWI with authority. Compared with VA HCPs, veterans found community HCPs about half as likely to be knowledgeable about GWI. Many reported that community HCPs had not heard of GWI or had no knowledge about it.

Respondents also were asked what types of treatments they tried in order to typify the care received. The most common responses were pain medications, symptom-specific treatments, or “just putting up with it” (no treatment). Many patients were also self-medicating, trying lifestyle changes, or seeking alternative therapies.

Finally, respondents were asked on a scale of 0 (very unsatisfied) to 5 (very satisfied), how satisfied they were with their overall care at the VA. The majority were satisfied with their overall care, with two-thirds very satisfied (5 of 5) or pretty satisfied (4 of 5). Only 3 (10%) were unsatisfied or very unsatisfied. Respondents had the following comments about their care: “They treat me like I am important;” “I am very thankful even though they cannot figure it out;” “They are doing the best they can with no answers and not enough help;” “[I know] it is still a work in progress.” A number of respondents were satisfied with some HCPs or care for some but not all of their symptoms. Reasons respondents were less satisfied included desiring answers, feeling they were not respected, or feeling that their concerns were not addressed.

When asked for suggestions for improvement of GWI care, the most common response was providing up-to-date HCP education (Table 5).

Discussion

The veterans participating in this QI survey had similar demographics, symptomology, and exposures as did those in other studies.^1-7 Therefore, improvements based on their responses are likely applicable to the health care of veterans experiencing GWI-associated symptoms at other VA health care systems as well.

Veterans with GWI can lose significant functional capacity and productivity due to their symptoms. The symptoms are chronic and have afflicted many Gulf War veterans for nearly 3 decades. Furthermore, the prevalence of GWI in Gulf War veterans continues to increase.^5-7 These facts testify to the enormous health-related quality-of-life impact of GWI.

Veterans who meet the Kansas case definition for GWI were not diagnosed or service connected in a uniform manner. Only 3 of the 30 veterans in this study were given a unifying diagnosis that connected their chronic illness to Gulf War deployment. Under current guidelines, Gulf War veterans are able to receive compensation for chronic symptoms in 3 ways: (1) compensation for chronic unexplained symptoms existing for ≥ 6 months that appeared during active duty in the Southwest Asia theater or by December 31, 2021, and are ≥ 10% disabling; (2) the 1995 Persian Gulf War Veterans’ Act recognizes 3 multisymptom illnesses for which veterans can be service connected: FM, CFS, and functional GI disorders, including IBS; and (3) expansion to include any CMI of unknown etiology is underway. A uniform diagnostic protocol based on biomarkers and updated understanding of disease pathology would be helpful.

Respondents shared experiences that demonstrated perceived gaps in HCP support or knowledge. Overall, more respondents found their HCPs supportive. Many of the reasons respondents found HCPs unsupportive related to acknowledgment of symptoms. Also, more respondents found that both VA and non-VA HCPs lacked knowledge about GWI symptoms. These findings further highlight the need for HCP education within the VA and in community-based care.

The treatments tried by respondents also highlight potential areas for improvement. Most of the treatments were for pain; therefore, more involvement with pain clinics and specialists could be helpful. Symptom-specific medications also are appropriate, although only one-third of patients reported use. While medications are not necessarily markers of quality care, the fact that many patients self-medicate or go without treatment suggests that access to care could be improved. In 2014, the VA and the US Department of Defense (DoD) released the “VA/DoD Clinical Practice Guideline for the Management of Chronic Multisymptom Illness,” which recommended treatments for the global disease and specific symptoms.¹⁵

Since then, GWI research points to inflammatory and metabolic disease mechanisms.^11-14,16 As the underlying pathophysiology is further elucidated, practice guidelines will need to be updated to include anti-inflammatory and antioxidant treatments used in practice for GWI and similar chronic systemic illnesses (eg, CFS, FM, and IBS).^17-19

Randomized control trials are needed to determine the efficacy of such medications for the treatment of GWI. As new results emerge, disseminating and updating evidence-based guidelines in a coordinated manner will be required for veterans to receive appropriate treatment. Veterans also seek alternative or nonpharmaceutical interventions, such as physical therapy and diet changes. Improving access to integrative medicine, physical therapy, nutritionists, and other practitioners also could optimize veterans’ health and function.

HCP Education

The Gulf War veteran respondents who participated in the survey noted HCP education, research progress, and veteran inclusion as areas for improvement. Respondents requested dissemination of information on diagnosis and treatment of GWI for HCPs and updates on research and other actions. They suggested ways research could be more effective (such as subgrouping by exposure, which researchers have been doing) and could extend to veterans experiencing CMI from other conflicts as well.²⁰ Respondents also recommended team approaches or centers of excellence in order to receive more comprehensive care.

An asset of VHA is the culture of QI and education. The VA Employee Education System previously produced “Caring for Gulf War I Veterans,” a systemwide training module.²¹ In 2014, updated clinical practice guidelines for GWI were provided by the VA and the DoD, including evidence for each recommendation. In 2016, the VA in collaboration with the IOM produced a report summarizing conclusions and recommendations regarding associations between health concerns and Gulf War deployment.²² A concise guide for HCPs caring for veterans with GWI, updated in 2018, is available.²³ Updated treatment guidelines, based on evolving understanding of GWI pathophysiology, and continuing efforts to disseminate information will be essential.

Respondents most often presented to primary care, both within and outside of MVAHCS. Therefore, VA and community PCPs who see veterans should be equipped to recognize and diagnose GWI as well as be familiar with basic disease management and specialists whom they could refer their patients. Neurology was the second most common specialty seen by respondents. The most prominent symptoms of GWI are related to nervous system function in addition to evidence of underlying neuroinflammation.²⁰ Veterans may present to a neurologist with a variety of concerns, such as cognitive issues, sleep problems, migraines and headaches, and pain. Neurologists could best manage treatments targeting common neurologic GWI symptoms and neuroinflammation, especially as new treatments are discovered.

The next 2 most common specialty services seen were psychiatry and psychology (7 responses for each). Five respondents reported mental health issues as part of their chronic postdeployment symptoms. Population-based studies have indicated that rates of PTSD in Gulf War veterans is 3% to 6%, much lower than the prevalence of GWI.^8,20 The 2010 IOM study concluded that GWI symptoms cannot be ascribed to any known psychiatric disorder. Unfortunately, several surveyed veterans made it clear that they had been denied care due to HCPs attributing their symptoms solely to mental health issues. Therefore, psychiatrists and psychologists must be educated about GWI, mental health issues occurring in Gulf War veterans, and physiologic symptoms of GWI that may mimic or coincide with mental health issues. These HCPs also would be important to include in an interdisciplinary clinic for veterans with GWI.

Finally, respondents sought care from numerous other specialties, including gastroenterology, physical therapy, pulmonology, dermatology, and surgical subspecialties, such as orthopedics and otolaryngology. This wide range of specialists seen emphasizes the need for medical education, beginning in medical school. If provided education on GWI, these specialists would be able to treat veterans with GWI, know to look for updates on GWI management, or know to look for other common symptoms, such as chronic sinusitis in otolaryngology or recurring rashes in dermatology. We also recommend identifying HCPs in these specialties who could be part of an interdisciplinary clinic or be referrals for symptom management.

Protocol Implementation

HCP education and clinical care protocol implementation should be the initial focus of improving GWI management. A team of stakeholders within the different areas of MVAHCS, including education, HCPs, and administrative staff, will need to be developed. Reaching out to VA HCPs who have seen veterans with GWI will be an essential first step to equip them with updated education about the diagnosis and management of CMI. Providing integrated widespread education to current HCPs who are likely to encounter veterans with deployment-related CMI from the Gulf War, OND/OEF/OIF, or other deployments also will be necessary. Finally, educating medical trainees, including residents and medical students, will ensure continuous care for future veterans, post-9/11 veterans.

GWI presentations at medical grand rounds or at other medical community educational events could provide educational outlets. These events create face-to-face opportunities to discuss GWI/CMI education with HCPs, giving them the opportunity to offer feedback about their experiences and create relationships with other HCPs who have seen patients with GWI/CMI. At an educational event, a short postevent feedback form that indicates whether HCPs would like more information or get involved in a clinic for veterans with CMI could be included. This information would help identify key HCPs and areas within the local VA needing further improvements, such as creating a clinic for veterans with GWI.

Since 1946, the VA has worked with academic institutions to provide state-of-the-art health care to US veterans and train new HCPs to meet the health care needs of the nation. Every year, > 40,000 residents and 20,000 medical students receive medical training at VA facilities, making VA the largest single provider of medical education in the country. Therefore, providing detailed GWI/CMI education to medical students and residents as a standard part of the VA Talent Management System would be of value for all VA professionals.

GWI Clinics

Access to comprehensive care can be accomplished by organizing a clinic for veterans with GWI. The most likely effective location would be in primary care. PCPs who have seen veterans with GWI and/or expressed interest in learning more about GWI will be the initial point of contact. As the primary care service has connections to ancillary services, such as pharmacists, dieticians, psychologists, and social workers, organizing 1 day each week to see patients with GWI would improve care.

As the need for specialty care arises, the team also would need to identify specialists willing to receive referrals from HCPs of veterans with GWI. These specialists could be identified via feedback forms from educational events, surveys after an online educational training, or through relationships among VA physicians. As the clinic becomes established, it may be effective to have certain commonly seen specialists available in person, most likely neurology, psychiatry, gastroenterology, pulmonology, and dermatology. Also, relationships with a pain clinic, sleep medicine, and integrative medicine services should be established.

Measures of improvement in the veteran health care experience could include veterans’ perceptions of the supportiveness and knowledge of physicians about GWI as well as overall satisfaction. A follow-up survey on these measures of veterans involved in a GWI clinic and those not involved would be a way to determine whether these clinics better meet veterans’ needs and what additional QI is needed.

Conclusion

A significant number of Gulf War veterans experience chronic postdeployment symptoms that need to be better addressed. Physicians need to be equipped to recognize and manage GWI and similar postdeployment CMI among veterans of OEF/OIF/OND. We recommend creating an educational initiative about GWI among VA physicians and trainees, connecting physicians who see veterans with GWI, and establishing an interdisciplinary clinic with a referral system as the next steps to improve care for veterans. An additional goal would be to reach out to veteran networks to update them on GWI research, education, and available health care, as veterans are the essential stakeholders in the QI process.

Nonablative laser therapy is emerging as an effective noninvasive treatment option for basal cell carcinoma (BCC) with reduced adverse effects and good cosmetic outcomes compared to surgery. Vascular lasers, such as the pulsed dye laser (PDL), are thought to work by selectively targeting the tumor’s vascular network while preserving normal surrounding tissue.^1,2 Although high energy and multiple passes might be required, adjunctive use of dynamic cooling reduces the risk for nonselective thermal injury vs ablative lasers, which destroy the tumor itself through vaporization of tissue water.²

With no established laser management guidelines for the treatment of BCC, earlier studies using a 595-nm PDL varied highly in their protocol.^3-8 Pulsed dye laser parameters ranged from a spot size of 7 to 10 mm, fluence of 7.5 to 15 J/cm², and pulse duration of 0.5 to 3 milliseconds. Follow-up ranged from 12 days to 25 months after the final laser treatment. The number of lesions in prior studies ranged from 7 to 100 BCCs, with the clinical clearance rate ranging from 71.4% to 75% for facial BCC and 78.6% to 95% for nonfacial BCC.^3-8 Studies with histologic confirmation had a clearance rate of 66.6% for facial BCC and 25% to 92.3% for nonfacial BCC.^3-5,7,8 Most studies examined BCCs on the trunk and extremities with few investigating facial BCC,^3-8 which is especially important given that the head and neck are the most common and cosmetically sensitive anatomic locations.^9-13

Noninvasive imaging devices, such as reflectance confocal microscopy (RCM) and optical coherence tomography (OCT) can assist with the diagnosis and treatment monitoring of BCC. These devices enable in vivo visualization of tissue in both cross-sectional and en face views and therefore can reduce the need for diagnostic biopsy. Reflectance confocal microscopy enables near-histologic visualization of the epidermis and superficial dermis with a resolution of 0.5 to 1 μm.¹⁴ Optical coherence tomography uses an infrared broadband light source that allows users to view skin architecture as deep as 1.5 to 2 mm with a resolution of 5 μm.¹⁵

When used synergistically, both devices can enhance the efficacy of nonablative laser treatment. With its increased depth and wider field of view, OCT is an optimal tool for repetitive evaluation of the same site over time and for following biopsy-confirmed tumors undergoing management.¹⁶ In addition to delineating tumor margins before treatment, imaging improves the detection of residual skin cancers, despite clearance on clinical and dermoscopic examination. Noninvasive imaging and nonsurgical management with laser therapy allow the physician to leave the skin intact and avoid scar tissue that might otherwise make it more difficult to detect and manage recurrence. The ability of OCT and RCM to monitor the efficacy of nonsurgical therapies for skin cancer has been demonstrated with imiquimod, photodynamic therapy, vismodegib, and ablative laser therapy.^17-20

With limited data on nonablative laser management of BCC, several gaps in the literature exist. First, in previously published studies the number of treatments was either determined to be an arbitrary set number or based on clinical clearance, which has the potential to miss residual tumor. Second, many follow-ups were limited to shortly after the final treatment, which limits the accuracy of the clearance rate, given that inflammation and scars can hide residual tumor.^21-23 Third, because many studies excised the treated area, long-term follow-up for recurrence was obscured. Last, only a few studies involved facial BCC, which is the most common and cosmetically concerning anatomic location.¹³

Our study attempted to address these gaps by evaluating the use of noninvasive imaging to guide management of primarily facial BCC. The objective was to perform a retrospective chart review on a subgroup of patients with BCC who were treated with combined nonablative PDL and fractional laser treatment with an extended follow-up period.

Methods

Study Design
We performed a retrospective chart review of 68 patients with 93 BCCs who had been treated with nonablative laser therapy as an alternative to surgery at the Mount Sinai Faculty Practice Associates between February 2011 and December 2018. Patients were followed throughout this period for assessment of clinical and subclinical recurrence. The Icahn School of Medicine at Mount Sinai Program for the Protection of Human Subjects provided institutional review board approval.

Patients
Inclusion criteria included the following: (1) BCC diagnosed by biopsy (see eTable 1 for subtypes) and (2) treated with a nonablative laser due to patient preference and eligibility by the principal investigator (PI). As a retrospective study, lesions were included irrespective of tumor subtype or size. Although the risk for perineural invasion (PNI) is extremely low with BCC (<0.2%), none of the cases demonstrated PNI on diagnostic biopsy and none exhibited clinical evidence of PNI, such as paresthesia, pain, facial paralysis, or diplopia.²⁴

Eligibility determined by the PI included limited clinical ulceration or bleeding, or both, and a safe distance from the eye when wearing an external eye shield (ie, outside the orbital rim). Patients who had Mohs micrographic surgery (MMS) or excision (or both) with recurrence at the treatment site were included. Detailed and thorough clinical and dermoscopic skin examination was critical in early detection of these cancers, allowing for treatment of less advanced tumors. The PI’s diagnostic approach utilized the published diagnostic color wheel algorithm,²⁵ which encompasses both clinical and dermoscopic colors and patterns for early diagnosis (ie, ulceration, pink-white to white shiny areas, absence of pigmented network, leaflike structures, large blue-gray ovoid nests or globular structures, spoke wheel structures, a crystalline pattern, a singular vascular pattern of arborizing vessels), combined with OCT or RCM, when necessary.²⁶ All lesions were imaged with OCT prior to laser treatment to confirm residual tumor following biopsy.

Although postsurgical patients were included, lesions receiving concurrent or prior nonsurgical therapy, such as a topical immunomodulator or oral hedgehog inhibitor (eg, vismodegib), were excluded.

Treatment Protocol
All patients received thorough information about the treatment, treatment alternatives, and potential adverse effects and complications. Lesions were selected based on clinical and dermoscopic findings and were biopsy confirmed. Clinical and dermoscopic photographs were taken at every visit. A camera was used for clinical photographs and a dermatoscope was attached for all contact polarized dermoscopic images. All lesions were imaged with OCT prior to laser therapy to delineate tumor margins and to confirm residual disease following biopsy to preclude biopsy-mediated regression.

Laser treatment consisted of a 595-nm PDL followed by fractional laser treatment with the 1927-nm setting. The range of PDL settings was similar to published studies of PDL for BCC (spot size, 7–10 mm; fluence, 6–15 J/cm²; pulse duration, 0.45–3 milliseconds).^3-8 The fractional laser also was used at settings similar to earlier studies for actinic keratosis (fluence, 5–20 mJ; treatment density, 40%–70%).²⁷ Laser treatment was performed by 1 of 5 medically trained providers who were fellows supervised by the PI.

All tumors received 1 to 7 treatments (average, 2.89) at 1- to 2-month intervals. Treatment end point (complete clearance) was judged on the absence of skin cancer clinically, dermoscopically on OCT, or histologically by biopsy, or a combination of these modalities. Recurrence was defined as a new histologically confirmed BCC occurring in an area that was previously documented as clear. Patients returned for follow-up 1 to 2 months after the final treatment to monitor tumor clearance and subsequently every 6 to 12 months for tumor recurrence. Posttreatment care included application of a thick emollient, such as a petrolatum-based product, until the area completely healed.

Data Collection
Clinical photographs, dermoscopic photographs, OCT scans, RCM scans, and biopsy reports were reviewed for each patient, as applicable. All patients were given an unidentifiable number; no protected health information was recorded. Data recorded for each patient included age, tumor subtype and location, tumor size, classification of the tumor as primary or a recurrence, number of treatments, treatment duration, lesion clearance, and length of follow-up.

Results

Patient and Lesion Characteristics
Sixty-eight patients with 93 BCCs (77 facial; 16 nonfacial) were included. The median age of patients was 70 years (range, 31–91 years). All 93 BCCs demonstrated residual tumor on OCT after diagnostic biopsy. Four BCCs had been treated earlier with MMS and were biopsy-proven recurrences. Most BCCs were of the nodular subtype; however, sclerosing, superficial, pigmented, morpheaform, and infiltrative subtypes also were included (eTable 1). Eight BCCs were obtained at outside institutions with no subtype provided. Facial BCCs had a mean (SD) clinical and dermoscopic diameter of 6.75 (4.71) mm (range, 2–24 mm). Patients were followed for 2.53 months to 6.03 years (mean follow-up, 2.43 years) and assessed for clinical and subclinical recurrence.

Tumor Clearance
Most lesions were effectively treated, with 89 of 93 BCCs (95.70%) demonstrating complete tumor clearance. Complete tumor clearance following laser therapy was reported in 74 of 77 facial BCCs (96.10%) and 15 of 16 nonfacial BCCs (93.75%)(eTable 2). Successfully treated BCCs underwent an average of 2.88 laser treatments over a mean duration of 3.54 months (range, 1 week to 1.92 years). Four incomplete responders underwent an average of 3.25 laser treatments over a mean duration of 3.44 months (range, 1.13–6.87 months). Of the 4 lesions that did not clear, 2 were nodular, 1 was pigmented, and 1 was sclerosing.

• Lesions receiving 3 or fewer treatments had a clearance rate of 96.05% (73/76) for all BCCs, 96.72% (59/61) for facial BCCs, and 93.33% (14/15) for nonfacial BCCs.

• Lesions receiving more than 3 laser treatments had a clearance rate of 94.12% (16/17) for all BCCs, 93.75% (15/16) for facial BCCs, and 100% (1/1) for nonfacial BCCs.

The relationship between facial BCC tumor diameter and number of treatments required for clearance had a positive correlation coefficient (Pearson r=0.319), indicating that larger BCCs required more laser treatments (eTable 3).

Comment

High Tumor Clearance Rates With OCT
This study yielded a clearance rate of 95.70% for all BCCs, 96.10% for facial BCCs, and 93.75% for nonfacial BCCs. This rate is higher than the clinical or histologic clearance rate (or both) of earlier studies on facial and nonfacial BCCs, which ranged from 25% to 95%.^8-11 In this study, we were able to utilize OCT and histology to confirm clearance. Optical coherence tomography, which has been shown to have a high sensitivity ranging from 86% to 95.7%, is therefore optimally used in treatment monitoring.^19,26,28 Optical coherence tomography has a broader specificity range of 75.3% to 98% and was not utilized for diagnostic purposes in this study. Combining OCT with a color wheel dermoscopic approach was helpful in confirming treatment efficacy of nonsurgical therapies and is significantly more accurate than clinical analysis alone (P<.01).^19,26,28

We suspect that the higher clearance rates observed in our study were due to the OCT-guided treatment protocol. Optical coherence tomography was used for margination while providing a modality for tailored treatment through visualization of residual tumor on clinically and dermoscopically clear follow-ups, given that several studies found residual tumor at the lateral edge of the tumor margin on histopathologic analysis.⁵ Utilizing noninvasive imaging technology to delineate tumor margins before treatment can improve efficacy and limit unnecessary treatment to the surrounding normal skin (eFigure 2).²⁹

After grouping lesions by number of laser treatments, the clearance rate remained similar among facial BCCs with 3 or fewer treatments (59/61 [96.72%]), but there was a slightly decreased clearance rate for facial BCCs with more than 3 treatments (15/16 [93.75%]), which may be explained by the need for more laser treatments for larger BCCs (eTable 3). The relationship between facial BCC size and number of laser treatments was found to correlate positively (Pearson r=0.319). The largest lesion (24 mm) was successfully treated with 5 treatments (Figure). The number of nonfacial lesions was limited in this study and was not statistically significant.

• The third was a BCC on the nose of a nonadherent patient, which may have contributed to the lack of clearance. We defined nonadherent patients as those who did not follow-up within the appropriate periods and who therefore ran the risk for tumor growth in between treatments.

The nonfacial BCC that did not clear had histologic features of focal sclerosing BCC, a more aggressive subtype of basal cell skin cancer.

Tumor Recurrence
Only 4 of 89 BCCs (4.49%) recurred, with a 5.41% (4/74) recurrence rate among facial BCCs. All recurrences lacked clinical and dermoscopic evidence of BCC but were found on follow-up OCT scan and confirmed with RCM. All recurrences were found 1.5 to 3.9 years posttreatment.

Recurrent tumors following MMS required, on average, more laser treatments than primary tumors to achieve successful tumor clearance, which we attribute to scar tissue from prior therapy obscuring recurrence, resulting in delayed diagnosis, and to inflammation and fibrosis masking residual tumors (eFigure 1). An added benefit of laser treatment is that all 4 recurrent tumors demonstrated improved cosmetic appearance of the original MMS scar.

The benefit of using OCT scans to check for recurrences is that OCT can find residual skin cancers despite the area looking clinically clear, which is especially important during clinical evaluation of a healed postsurgical scar for recurrence because OCT imaging allows us to look as deep as 2 mm under the skin. Nonsurgical treatments also enable us to leave skin intact and avoid creating scar tissue, which makes it easier to detect and manage recurrence.

Limitations
There were several important limitations of this retrospective study:

• Patients were treated by 1 of 5 medically trained fellows. Although the fellows worked under the supervision of the PI, variation in their work from one to another might have led to different end points.

• All patients who appeared clinically clear were offered biopsy to confirm clearance on histology. Some patients agreed to biopsy, but many did not because they were pleased with the cosmetic outcome, which is similar to other studies exhibiting only clinical clearance rates without providing histologic clearance following nonsurgical therapy.⁶ We believe that imaging with OCT circumvents this problem and offers more accurate confirmation than clinical or dermoscopic correlation alone, or the combination of the 2 modalities.

• Lack of treatment standardization and short length of follow-up can result in underestimation of the recurrence rate. In particular, most patients were followed up with OCT in less than 6 months. These are unavoidable features in a retrospective study and we are currently addressing this problem in a new prospective study.

Extended Follow-up
Although this study is not a prospective design, it does provide recurrence data over extended follow-up for the nonablative laser management of BCCs (eTables 4 and 5). Studies have demonstrated that MMS has a 5-year cure rate as high as 99% for BCC.³² Given the limited follow-up period of prior nonablative laser management studies, recurrences might not have been fully evaluated. Our study had a 4.49% recurrence rate for all BCCs and a 5.41% recurrence rate for facial BCCs but was not detectable by clinical examination combined with dermoscopic findings alone. All recurrences required the utilization of OCT or RCM or a combination of these modalities to be diagnosed. In 1 patient with recurrence, we were able to see residual tumor on both OCT and RCM without any inflammation obscuring the scan, given that 3 years had passed. Although 2 months is an optimal follow-up time for OCT, we have not found an optimal follow-up time for RCM, which is another reason why OCT might be preferable to other imaging modalities, such as RCM and high-definition OCT, that have higher resolution but provide less depth on imaging. Although only 40 of 89 patients (4.49%) had follow-up ranging from 3 years to greater than 5 years, long-term follow-up to date has been limited in prior studies.

We believe the high clearance rates and limited recurrence are secondary to the utilization of noninvasive imaging, as the majority of these recurrences would not have been diagnosed based on clinical and/or dermoscopic information alone. Additionally, the 4 biopsy-proven post-MMS recurrence patients that were treated in this study also may not have been diagnosed this early without the use of additional noninvasive imaging. In our opinion, although laser management can be used without noninvasive imaging guidance—dermoscopy, OCT, and/or RCM—this technology is critical not only for early detection but also for proper management of patients.

Conclusion

This study showed a 95.70% clearance rate for all BCCs and a 96.10% clearance rate for facial BCCs. Although we had a zero clinical recurrence rate, 4.49% of all BCCs and 5.41% of facial BCCs had recurred on subsequent monitoring with noninvasive imaging. Given the large size of the study and extended follow-up, we found nonablative laser management to be a reliable treatment alternative with improved cosmetic outcome (Figure) and minimal short-term adverse effects compared to surgery.

Tailored care for the individual patient is based on a variety of options and patient preference, including ease of compliance, number of follow-up visits, invasive vs noninvasive diagnosis and monitoring, and downtime for healing. The use of noninvasive imaging also allowed us to find a more standardized treatment regimen using this nonablative laser combination. We found that 3 or fewer and more than 3 treatments had similar efficacy in tumor clearance. We recommend a standard laser protocol of 3 treatments every 4 to 6 weeks with follow-up 2 months after the final treatment to assess for clearance with OCT.

Larger BCCs might require additional treatments; therefore, we caution against laser therapy without concomitant use of OCT imaging to visualize residual tumor. Utilizing other noninvasive modalities, such as dermoscopy, in combination with thorough skin examination also is critical in the early detection of skin cancers to improve the efficacy of this less-aggressive, nonablative, and cosmetically optimal treatment protocol.

Acknowledgement—We would like to acknowledge Dimitrios Karponis, BSc, from the Impirial College London, England, for his assistance with a portion of the statistical analysis.

Nonablative laser therapy is emerging as an effective noninvasive treatment option for basal cell carcinoma (BCC) with reduced adverse effects and good cosmetic outcomes compared to surgery. Vascular lasers, such as the pulsed dye laser (PDL), are thought to work by selectively targeting the tumor’s vascular network while preserving normal surrounding tissue.^1,2 Although high energy and multiple passes might be required, adjunctive use of dynamic cooling reduces the risk for nonselective thermal injury vs ablative lasers, which destroy the tumor itself through vaporization of tissue water.²

With no established laser management guidelines for the treatment of BCC, earlier studies using a 595-nm PDL varied highly in their protocol.^3-8 Pulsed dye laser parameters ranged from a spot size of 7 to 10 mm, fluence of 7.5 to 15 J/cm², and pulse duration of 0.5 to 3 milliseconds. Follow-up ranged from 12 days to 25 months after the final laser treatment. The number of lesions in prior studies ranged from 7 to 100 BCCs, with the clinical clearance rate ranging from 71.4% to 75% for facial BCC and 78.6% to 95% for nonfacial BCC.^3-8 Studies with histologic confirmation had a clearance rate of 66.6% for facial BCC and 25% to 92.3% for nonfacial BCC.^3-5,7,8 Most studies examined BCCs on the trunk and extremities with few investigating facial BCC,^3-8 which is especially important given that the head and neck are the most common and cosmetically sensitive anatomic locations.^9-13

Noninvasive imaging devices, such as reflectance confocal microscopy (RCM) and optical coherence tomography (OCT) can assist with the diagnosis and treatment monitoring of BCC. These devices enable in vivo visualization of tissue in both cross-sectional and en face views and therefore can reduce the need for diagnostic biopsy. Reflectance confocal microscopy enables near-histologic visualization of the epidermis and superficial dermis with a resolution of 0.5 to 1 μm.¹⁴ Optical coherence tomography uses an infrared broadband light source that allows users to view skin architecture as deep as 1.5 to 2 mm with a resolution of 5 μm.¹⁵

When used synergistically, both devices can enhance the efficacy of nonablative laser treatment. With its increased depth and wider field of view, OCT is an optimal tool for repetitive evaluation of the same site over time and for following biopsy-confirmed tumors undergoing management.¹⁶ In addition to delineating tumor margins before treatment, imaging improves the detection of residual skin cancers, despite clearance on clinical and dermoscopic examination. Noninvasive imaging and nonsurgical management with laser therapy allow the physician to leave the skin intact and avoid scar tissue that might otherwise make it more difficult to detect and manage recurrence. The ability of OCT and RCM to monitor the efficacy of nonsurgical therapies for skin cancer has been demonstrated with imiquimod, photodynamic therapy, vismodegib, and ablative laser therapy.^17-20

With limited data on nonablative laser management of BCC, several gaps in the literature exist. First, in previously published studies the number of treatments was either determined to be an arbitrary set number or based on clinical clearance, which has the potential to miss residual tumor. Second, many follow-ups were limited to shortly after the final treatment, which limits the accuracy of the clearance rate, given that inflammation and scars can hide residual tumor.^21-23 Third, because many studies excised the treated area, long-term follow-up for recurrence was obscured. Last, only a few studies involved facial BCC, which is the most common and cosmetically concerning anatomic location.¹³

Our study attempted to address these gaps by evaluating the use of noninvasive imaging to guide management of primarily facial BCC. The objective was to perform a retrospective chart review on a subgroup of patients with BCC who were treated with combined nonablative PDL and fractional laser treatment with an extended follow-up period.

Methods

Study Design
We performed a retrospective chart review of 68 patients with 93 BCCs who had been treated with nonablative laser therapy as an alternative to surgery at the Mount Sinai Faculty Practice Associates between February 2011 and December 2018. Patients were followed throughout this period for assessment of clinical and subclinical recurrence. The Icahn School of Medicine at Mount Sinai Program for the Protection of Human Subjects provided institutional review board approval.

Patients
Inclusion criteria included the following: (1) BCC diagnosed by biopsy (see eTable 1 for subtypes) and (2) treated with a nonablative laser due to patient preference and eligibility by the principal investigator (PI). As a retrospective study, lesions were included irrespective of tumor subtype or size. Although the risk for perineural invasion (PNI) is extremely low with BCC (<0.2%), none of the cases demonstrated PNI on diagnostic biopsy and none exhibited clinical evidence of PNI, such as paresthesia, pain, facial paralysis, or diplopia.²⁴

Eligibility determined by the PI included limited clinical ulceration or bleeding, or both, and a safe distance from the eye when wearing an external eye shield (ie, outside the orbital rim). Patients who had Mohs micrographic surgery (MMS) or excision (or both) with recurrence at the treatment site were included. Detailed and thorough clinical and dermoscopic skin examination was critical in early detection of these cancers, allowing for treatment of less advanced tumors. The PI’s diagnostic approach utilized the published diagnostic color wheel algorithm,²⁵ which encompasses both clinical and dermoscopic colors and patterns for early diagnosis (ie, ulceration, pink-white to white shiny areas, absence of pigmented network, leaflike structures, large blue-gray ovoid nests or globular structures, spoke wheel structures, a crystalline pattern, a singular vascular pattern of arborizing vessels), combined with OCT or RCM, when necessary.²⁶ All lesions were imaged with OCT prior to laser treatment to confirm residual tumor following biopsy.

Although postsurgical patients were included, lesions receiving concurrent or prior nonsurgical therapy, such as a topical immunomodulator or oral hedgehog inhibitor (eg, vismodegib), were excluded.

Treatment Protocol
All patients received thorough information about the treatment, treatment alternatives, and potential adverse effects and complications. Lesions were selected based on clinical and dermoscopic findings and were biopsy confirmed. Clinical and dermoscopic photographs were taken at every visit. A camera was used for clinical photographs and a dermatoscope was attached for all contact polarized dermoscopic images. All lesions were imaged with OCT prior to laser therapy to delineate tumor margins and to confirm residual disease following biopsy to preclude biopsy-mediated regression.

Laser treatment consisted of a 595-nm PDL followed by fractional laser treatment with the 1927-nm setting. The range of PDL settings was similar to published studies of PDL for BCC (spot size, 7–10 mm; fluence, 6–15 J/cm²; pulse duration, 0.45–3 milliseconds).^3-8 The fractional laser also was used at settings similar to earlier studies for actinic keratosis (fluence, 5–20 mJ; treatment density, 40%–70%).²⁷ Laser treatment was performed by 1 of 5 medically trained providers who were fellows supervised by the PI.

All tumors received 1 to 7 treatments (average, 2.89) at 1- to 2-month intervals. Treatment end point (complete clearance) was judged on the absence of skin cancer clinically, dermoscopically on OCT, or histologically by biopsy, or a combination of these modalities. Recurrence was defined as a new histologically confirmed BCC occurring in an area that was previously documented as clear. Patients returned for follow-up 1 to 2 months after the final treatment to monitor tumor clearance and subsequently every 6 to 12 months for tumor recurrence. Posttreatment care included application of a thick emollient, such as a petrolatum-based product, until the area completely healed.

Data Collection
Clinical photographs, dermoscopic photographs, OCT scans, RCM scans, and biopsy reports were reviewed for each patient, as applicable. All patients were given an unidentifiable number; no protected health information was recorded. Data recorded for each patient included age, tumor subtype and location, tumor size, classification of the tumor as primary or a recurrence, number of treatments, treatment duration, lesion clearance, and length of follow-up.

Results

Patient and Lesion Characteristics
Sixty-eight patients with 93 BCCs (77 facial; 16 nonfacial) were included. The median age of patients was 70 years (range, 31–91 years). All 93 BCCs demonstrated residual tumor on OCT after diagnostic biopsy. Four BCCs had been treated earlier with MMS and were biopsy-proven recurrences. Most BCCs were of the nodular subtype; however, sclerosing, superficial, pigmented, morpheaform, and infiltrative subtypes also were included (eTable 1). Eight BCCs were obtained at outside institutions with no subtype provided. Facial BCCs had a mean (SD) clinical and dermoscopic diameter of 6.75 (4.71) mm (range, 2–24 mm). Patients were followed for 2.53 months to 6.03 years (mean follow-up, 2.43 years) and assessed for clinical and subclinical recurrence.

Tumor Clearance
Most lesions were effectively treated, with 89 of 93 BCCs (95.70%) demonstrating complete tumor clearance. Complete tumor clearance following laser therapy was reported in 74 of 77 facial BCCs (96.10%) and 15 of 16 nonfacial BCCs (93.75%)(eTable 2). Successfully treated BCCs underwent an average of 2.88 laser treatments over a mean duration of 3.54 months (range, 1 week to 1.92 years). Four incomplete responders underwent an average of 3.25 laser treatments over a mean duration of 3.44 months (range, 1.13–6.87 months). Of the 4 lesions that did not clear, 2 were nodular, 1 was pigmented, and 1 was sclerosing.

• Lesions receiving 3 or fewer treatments had a clearance rate of 96.05% (73/76) for all BCCs, 96.72% (59/61) for facial BCCs, and 93.33% (14/15) for nonfacial BCCs.

• Lesions receiving more than 3 laser treatments had a clearance rate of 94.12% (16/17) for all BCCs, 93.75% (15/16) for facial BCCs, and 100% (1/1) for nonfacial BCCs.

The relationship between facial BCC tumor diameter and number of treatments required for clearance had a positive correlation coefficient (Pearson r=0.319), indicating that larger BCCs required more laser treatments (eTable 3).

Comment

High Tumor Clearance Rates With OCT
This study yielded a clearance rate of 95.70% for all BCCs, 96.10% for facial BCCs, and 93.75% for nonfacial BCCs. This rate is higher than the clinical or histologic clearance rate (or both) of earlier studies on facial and nonfacial BCCs, which ranged from 25% to 95%.^8-11 In this study, we were able to utilize OCT and histology to confirm clearance. Optical coherence tomography, which has been shown to have a high sensitivity ranging from 86% to 95.7%, is therefore optimally used in treatment monitoring.^19,26,28 Optical coherence tomography has a broader specificity range of 75.3% to 98% and was not utilized for diagnostic purposes in this study. Combining OCT with a color wheel dermoscopic approach was helpful in confirming treatment efficacy of nonsurgical therapies and is significantly more accurate than clinical analysis alone (P<.01).^19,26,28

We suspect that the higher clearance rates observed in our study were due to the OCT-guided treatment protocol. Optical coherence tomography was used for margination while providing a modality for tailored treatment through visualization of residual tumor on clinically and dermoscopically clear follow-ups, given that several studies found residual tumor at the lateral edge of the tumor margin on histopathologic analysis.⁵ Utilizing noninvasive imaging technology to delineate tumor margins before treatment can improve efficacy and limit unnecessary treatment to the surrounding normal skin (eFigure 2).²⁹

After grouping lesions by number of laser treatments, the clearance rate remained similar among facial BCCs with 3 or fewer treatments (59/61 [96.72%]), but there was a slightly decreased clearance rate for facial BCCs with more than 3 treatments (15/16 [93.75%]), which may be explained by the need for more laser treatments for larger BCCs (eTable 3). The relationship between facial BCC size and number of laser treatments was found to correlate positively (Pearson r=0.319). The largest lesion (24 mm) was successfully treated with 5 treatments (Figure). The number of nonfacial lesions was limited in this study and was not statistically significant.

• The third was a BCC on the nose of a nonadherent patient, which may have contributed to the lack of clearance. We defined nonadherent patients as those who did not follow-up within the appropriate periods and who therefore ran the risk for tumor growth in between treatments.

The nonfacial BCC that did not clear had histologic features of focal sclerosing BCC, a more aggressive subtype of basal cell skin cancer.

Tumor Recurrence
Only 4 of 89 BCCs (4.49%) recurred, with a 5.41% (4/74) recurrence rate among facial BCCs. All recurrences lacked clinical and dermoscopic evidence of BCC but were found on follow-up OCT scan and confirmed with RCM. All recurrences were found 1.5 to 3.9 years posttreatment.

Recurrent tumors following MMS required, on average, more laser treatments than primary tumors to achieve successful tumor clearance, which we attribute to scar tissue from prior therapy obscuring recurrence, resulting in delayed diagnosis, and to inflammation and fibrosis masking residual tumors (eFigure 1). An added benefit of laser treatment is that all 4 recurrent tumors demonstrated improved cosmetic appearance of the original MMS scar.

The benefit of using OCT scans to check for recurrences is that OCT can find residual skin cancers despite the area looking clinically clear, which is especially important during clinical evaluation of a healed postsurgical scar for recurrence because OCT imaging allows us to look as deep as 2 mm under the skin. Nonsurgical treatments also enable us to leave skin intact and avoid creating scar tissue, which makes it easier to detect and manage recurrence.

Limitations
There were several important limitations of this retrospective study:

• Patients were treated by 1 of 5 medically trained fellows. Although the fellows worked under the supervision of the PI, variation in their work from one to another might have led to different end points.

• All patients who appeared clinically clear were offered biopsy to confirm clearance on histology. Some patients agreed to biopsy, but many did not because they were pleased with the cosmetic outcome, which is similar to other studies exhibiting only clinical clearance rates without providing histologic clearance following nonsurgical therapy.⁶ We believe that imaging with OCT circumvents this problem and offers more accurate confirmation than clinical or dermoscopic correlation alone, or the combination of the 2 modalities.

• Lack of treatment standardization and short length of follow-up can result in underestimation of the recurrence rate. In particular, most patients were followed up with OCT in less than 6 months. These are unavoidable features in a retrospective study and we are currently addressing this problem in a new prospective study.

Extended Follow-up
Although this study is not a prospective design, it does provide recurrence data over extended follow-up for the nonablative laser management of BCCs (eTables 4 and 5). Studies have demonstrated that MMS has a 5-year cure rate as high as 99% for BCC.³² Given the limited follow-up period of prior nonablative laser management studies, recurrences might not have been fully evaluated. Our study had a 4.49% recurrence rate for all BCCs and a 5.41% recurrence rate for facial BCCs but was not detectable by clinical examination combined with dermoscopic findings alone. All recurrences required the utilization of OCT or RCM or a combination of these modalities to be diagnosed. In 1 patient with recurrence, we were able to see residual tumor on both OCT and RCM without any inflammation obscuring the scan, given that 3 years had passed. Although 2 months is an optimal follow-up time for OCT, we have not found an optimal follow-up time for RCM, which is another reason why OCT might be preferable to other imaging modalities, such as RCM and high-definition OCT, that have higher resolution but provide less depth on imaging. Although only 40 of 89 patients (4.49%) had follow-up ranging from 3 years to greater than 5 years, long-term follow-up to date has been limited in prior studies.

We believe the high clearance rates and limited recurrence are secondary to the utilization of noninvasive imaging, as the majority of these recurrences would not have been diagnosed based on clinical and/or dermoscopic information alone. Additionally, the 4 biopsy-proven post-MMS recurrence patients that were treated in this study also may not have been diagnosed this early without the use of additional noninvasive imaging. In our opinion, although laser management can be used without noninvasive imaging guidance—dermoscopy, OCT, and/or RCM—this technology is critical not only for early detection but also for proper management of patients.

Conclusion

This study showed a 95.70% clearance rate for all BCCs and a 96.10% clearance rate for facial BCCs. Although we had a zero clinical recurrence rate, 4.49% of all BCCs and 5.41% of facial BCCs had recurred on subsequent monitoring with noninvasive imaging. Given the large size of the study and extended follow-up, we found nonablative laser management to be a reliable treatment alternative with improved cosmetic outcome (Figure) and minimal short-term adverse effects compared to surgery.

Tailored care for the individual patient is based on a variety of options and patient preference, including ease of compliance, number of follow-up visits, invasive vs noninvasive diagnosis and monitoring, and downtime for healing. The use of noninvasive imaging also allowed us to find a more standardized treatment regimen using this nonablative laser combination. We found that 3 or fewer and more than 3 treatments had similar efficacy in tumor clearance. We recommend a standard laser protocol of 3 treatments every 4 to 6 weeks with follow-up 2 months after the final treatment to assess for clearance with OCT.

Larger BCCs might require additional treatments; therefore, we caution against laser therapy without concomitant use of OCT imaging to visualize residual tumor. Utilizing other noninvasive modalities, such as dermoscopy, in combination with thorough skin examination also is critical in the early detection of skin cancers to improve the efficacy of this less-aggressive, nonablative, and cosmetically optimal treatment protocol.

Acknowledgement—We would like to acknowledge Dimitrios Karponis, BSc, from the Impirial College London, England, for his assistance with a portion of the statistical analysis.

Nonablative laser therapy is emerging as an effective noninvasive treatment option for basal cell carcinoma (BCC) with reduced adverse effects and good cosmetic outcomes compared to surgery. Vascular lasers, such as the pulsed dye laser (PDL), are thought to work by selectively targeting the tumor’s vascular network while preserving normal surrounding tissue.^1,2 Although high energy and multiple passes might be required, adjunctive use of dynamic cooling reduces the risk for nonselective thermal injury vs ablative lasers, which destroy the tumor itself through vaporization of tissue water.²

With no established laser management guidelines for the treatment of BCC, earlier studies using a 595-nm PDL varied highly in their protocol.^3-8 Pulsed dye laser parameters ranged from a spot size of 7 to 10 mm, fluence of 7.5 to 15 J/cm², and pulse duration of 0.5 to 3 milliseconds. Follow-up ranged from 12 days to 25 months after the final laser treatment. The number of lesions in prior studies ranged from 7 to 100 BCCs, with the clinical clearance rate ranging from 71.4% to 75% for facial BCC and 78.6% to 95% for nonfacial BCC.^3-8 Studies with histologic confirmation had a clearance rate of 66.6% for facial BCC and 25% to 92.3% for nonfacial BCC.^3-5,7,8 Most studies examined BCCs on the trunk and extremities with few investigating facial BCC,^3-8 which is especially important given that the head and neck are the most common and cosmetically sensitive anatomic locations.^9-13

Noninvasive imaging devices, such as reflectance confocal microscopy (RCM) and optical coherence tomography (OCT) can assist with the diagnosis and treatment monitoring of BCC. These devices enable in vivo visualization of tissue in both cross-sectional and en face views and therefore can reduce the need for diagnostic biopsy. Reflectance confocal microscopy enables near-histologic visualization of the epidermis and superficial dermis with a resolution of 0.5 to 1 μm.¹⁴ Optical coherence tomography uses an infrared broadband light source that allows users to view skin architecture as deep as 1.5 to 2 mm with a resolution of 5 μm.¹⁵

When used synergistically, both devices can enhance the efficacy of nonablative laser treatment. With its increased depth and wider field of view, OCT is an optimal tool for repetitive evaluation of the same site over time and for following biopsy-confirmed tumors undergoing management.¹⁶ In addition to delineating tumor margins before treatment, imaging improves the detection of residual skin cancers, despite clearance on clinical and dermoscopic examination. Noninvasive imaging and nonsurgical management with laser therapy allow the physician to leave the skin intact and avoid scar tissue that might otherwise make it more difficult to detect and manage recurrence. The ability of OCT and RCM to monitor the efficacy of nonsurgical therapies for skin cancer has been demonstrated with imiquimod, photodynamic therapy, vismodegib, and ablative laser therapy.^17-20

With limited data on nonablative laser management of BCC, several gaps in the literature exist. First, in previously published studies the number of treatments was either determined to be an arbitrary set number or based on clinical clearance, which has the potential to miss residual tumor. Second, many follow-ups were limited to shortly after the final treatment, which limits the accuracy of the clearance rate, given that inflammation and scars can hide residual tumor.^21-23 Third, because many studies excised the treated area, long-term follow-up for recurrence was obscured. Last, only a few studies involved facial BCC, which is the most common and cosmetically concerning anatomic location.¹³

Our study attempted to address these gaps by evaluating the use of noninvasive imaging to guide management of primarily facial BCC. The objective was to perform a retrospective chart review on a subgroup of patients with BCC who were treated with combined nonablative PDL and fractional laser treatment with an extended follow-up period.

Methods

Study Design
We performed a retrospective chart review of 68 patients with 93 BCCs who had been treated with nonablative laser therapy as an alternative to surgery at the Mount Sinai Faculty Practice Associates between February 2011 and December 2018. Patients were followed throughout this period for assessment of clinical and subclinical recurrence. The Icahn School of Medicine at Mount Sinai Program for the Protection of Human Subjects provided institutional review board approval.

Patients
Inclusion criteria included the following: (1) BCC diagnosed by biopsy (see eTable 1 for subtypes) and (2) treated with a nonablative laser due to patient preference and eligibility by the principal investigator (PI). As a retrospective study, lesions were included irrespective of tumor subtype or size. Although the risk for perineural invasion (PNI) is extremely low with BCC (<0.2%), none of the cases demonstrated PNI on diagnostic biopsy and none exhibited clinical evidence of PNI, such as paresthesia, pain, facial paralysis, or diplopia.²⁴

Eligibility determined by the PI included limited clinical ulceration or bleeding, or both, and a safe distance from the eye when wearing an external eye shield (ie, outside the orbital rim). Patients who had Mohs micrographic surgery (MMS) or excision (or both) with recurrence at the treatment site were included. Detailed and thorough clinical and dermoscopic skin examination was critical in early detection of these cancers, allowing for treatment of less advanced tumors. The PI’s diagnostic approach utilized the published diagnostic color wheel algorithm,²⁵ which encompasses both clinical and dermoscopic colors and patterns for early diagnosis (ie, ulceration, pink-white to white shiny areas, absence of pigmented network, leaflike structures, large blue-gray ovoid nests or globular structures, spoke wheel structures, a crystalline pattern, a singular vascular pattern of arborizing vessels), combined with OCT or RCM, when necessary.²⁶ All lesions were imaged with OCT prior to laser treatment to confirm residual tumor following biopsy.

Although postsurgical patients were included, lesions receiving concurrent or prior nonsurgical therapy, such as a topical immunomodulator or oral hedgehog inhibitor (eg, vismodegib), were excluded.

Treatment Protocol
All patients received thorough information about the treatment, treatment alternatives, and potential adverse effects and complications. Lesions were selected based on clinical and dermoscopic findings and were biopsy confirmed. Clinical and dermoscopic photographs were taken at every visit. A camera was used for clinical photographs and a dermatoscope was attached for all contact polarized dermoscopic images. All lesions were imaged with OCT prior to laser therapy to delineate tumor margins and to confirm residual disease following biopsy to preclude biopsy-mediated regression.

Laser treatment consisted of a 595-nm PDL followed by fractional laser treatment with the 1927-nm setting. The range of PDL settings was similar to published studies of PDL for BCC (spot size, 7–10 mm; fluence, 6–15 J/cm²; pulse duration, 0.45–3 milliseconds).^3-8 The fractional laser also was used at settings similar to earlier studies for actinic keratosis (fluence, 5–20 mJ; treatment density, 40%–70%).²⁷ Laser treatment was performed by 1 of 5 medically trained providers who were fellows supervised by the PI.

All tumors received 1 to 7 treatments (average, 2.89) at 1- to 2-month intervals. Treatment end point (complete clearance) was judged on the absence of skin cancer clinically, dermoscopically on OCT, or histologically by biopsy, or a combination of these modalities. Recurrence was defined as a new histologically confirmed BCC occurring in an area that was previously documented as clear. Patients returned for follow-up 1 to 2 months after the final treatment to monitor tumor clearance and subsequently every 6 to 12 months for tumor recurrence. Posttreatment care included application of a thick emollient, such as a petrolatum-based product, until the area completely healed.

Data Collection
Clinical photographs, dermoscopic photographs, OCT scans, RCM scans, and biopsy reports were reviewed for each patient, as applicable. All patients were given an unidentifiable number; no protected health information was recorded. Data recorded for each patient included age, tumor subtype and location, tumor size, classification of the tumor as primary or a recurrence, number of treatments, treatment duration, lesion clearance, and length of follow-up.

Results

Patient and Lesion Characteristics
Sixty-eight patients with 93 BCCs (77 facial; 16 nonfacial) were included. The median age of patients was 70 years (range, 31–91 years). All 93 BCCs demonstrated residual tumor on OCT after diagnostic biopsy. Four BCCs had been treated earlier with MMS and were biopsy-proven recurrences. Most BCCs were of the nodular subtype; however, sclerosing, superficial, pigmented, morpheaform, and infiltrative subtypes also were included (eTable 1). Eight BCCs were obtained at outside institutions with no subtype provided. Facial BCCs had a mean (SD) clinical and dermoscopic diameter of 6.75 (4.71) mm (range, 2–24 mm). Patients were followed for 2.53 months to 6.03 years (mean follow-up, 2.43 years) and assessed for clinical and subclinical recurrence.

Tumor Clearance
Most lesions were effectively treated, with 89 of 93 BCCs (95.70%) demonstrating complete tumor clearance. Complete tumor clearance following laser therapy was reported in 74 of 77 facial BCCs (96.10%) and 15 of 16 nonfacial BCCs (93.75%)(eTable 2). Successfully treated BCCs underwent an average of 2.88 laser treatments over a mean duration of 3.54 months (range, 1 week to 1.92 years). Four incomplete responders underwent an average of 3.25 laser treatments over a mean duration of 3.44 months (range, 1.13–6.87 months). Of the 4 lesions that did not clear, 2 were nodular, 1 was pigmented, and 1 was sclerosing.

• Lesions receiving 3 or fewer treatments had a clearance rate of 96.05% (73/76) for all BCCs, 96.72% (59/61) for facial BCCs, and 93.33% (14/15) for nonfacial BCCs.

• Lesions receiving more than 3 laser treatments had a clearance rate of 94.12% (16/17) for all BCCs, 93.75% (15/16) for facial BCCs, and 100% (1/1) for nonfacial BCCs.

The relationship between facial BCC tumor diameter and number of treatments required for clearance had a positive correlation coefficient (Pearson r=0.319), indicating that larger BCCs required more laser treatments (eTable 3).

Comment

High Tumor Clearance Rates With OCT
This study yielded a clearance rate of 95.70% for all BCCs, 96.10% for facial BCCs, and 93.75% for nonfacial BCCs. This rate is higher than the clinical or histologic clearance rate (or both) of earlier studies on facial and nonfacial BCCs, which ranged from 25% to 95%.^8-11 In this study, we were able to utilize OCT and histology to confirm clearance. Optical coherence tomography, which has been shown to have a high sensitivity ranging from 86% to 95.7%, is therefore optimally used in treatment monitoring.^19,26,28 Optical coherence tomography has a broader specificity range of 75.3% to 98% and was not utilized for diagnostic purposes in this study. Combining OCT with a color wheel dermoscopic approach was helpful in confirming treatment efficacy of nonsurgical therapies and is significantly more accurate than clinical analysis alone (P<.01).^19,26,28

We suspect that the higher clearance rates observed in our study were due to the OCT-guided treatment protocol. Optical coherence tomography was used for margination while providing a modality for tailored treatment through visualization of residual tumor on clinically and dermoscopically clear follow-ups, given that several studies found residual tumor at the lateral edge of the tumor margin on histopathologic analysis.⁵ Utilizing noninvasive imaging technology to delineate tumor margins before treatment can improve efficacy and limit unnecessary treatment to the surrounding normal skin (eFigure 2).²⁹

After grouping lesions by number of laser treatments, the clearance rate remained similar among facial BCCs with 3 or fewer treatments (59/61 [96.72%]), but there was a slightly decreased clearance rate for facial BCCs with more than 3 treatments (15/16 [93.75%]), which may be explained by the need for more laser treatments for larger BCCs (eTable 3). The relationship between facial BCC size and number of laser treatments was found to correlate positively (Pearson r=0.319). The largest lesion (24 mm) was successfully treated with 5 treatments (Figure). The number of nonfacial lesions was limited in this study and was not statistically significant.

• The third was a BCC on the nose of a nonadherent patient, which may have contributed to the lack of clearance. We defined nonadherent patients as those who did not follow-up within the appropriate periods and who therefore ran the risk for tumor growth in between treatments.

The nonfacial BCC that did not clear had histologic features of focal sclerosing BCC, a more aggressive subtype of basal cell skin cancer.

Tumor Recurrence
Only 4 of 89 BCCs (4.49%) recurred, with a 5.41% (4/74) recurrence rate among facial BCCs. All recurrences lacked clinical and dermoscopic evidence of BCC but were found on follow-up OCT scan and confirmed with RCM. All recurrences were found 1.5 to 3.9 years posttreatment.

Recurrent tumors following MMS required, on average, more laser treatments than primary tumors to achieve successful tumor clearance, which we attribute to scar tissue from prior therapy obscuring recurrence, resulting in delayed diagnosis, and to inflammation and fibrosis masking residual tumors (eFigure 1). An added benefit of laser treatment is that all 4 recurrent tumors demonstrated improved cosmetic appearance of the original MMS scar.

The benefit of using OCT scans to check for recurrences is that OCT can find residual skin cancers despite the area looking clinically clear, which is especially important during clinical evaluation of a healed postsurgical scar for recurrence because OCT imaging allows us to look as deep as 2 mm under the skin. Nonsurgical treatments also enable us to leave skin intact and avoid creating scar tissue, which makes it easier to detect and manage recurrence.

Limitations
There were several important limitations of this retrospective study:

• Patients were treated by 1 of 5 medically trained fellows. Although the fellows worked under the supervision of the PI, variation in their work from one to another might have led to different end points.

• All patients who appeared clinically clear were offered biopsy to confirm clearance on histology. Some patients agreed to biopsy, but many did not because they were pleased with the cosmetic outcome, which is similar to other studies exhibiting only clinical clearance rates without providing histologic clearance following nonsurgical therapy.⁶ We believe that imaging with OCT circumvents this problem and offers more accurate confirmation than clinical or dermoscopic correlation alone, or the combination of the 2 modalities.

• Lack of treatment standardization and short length of follow-up can result in underestimation of the recurrence rate. In particular, most patients were followed up with OCT in less than 6 months. These are unavoidable features in a retrospective study and we are currently addressing this problem in a new prospective study.

Extended Follow-up
Although this study is not a prospective design, it does provide recurrence data over extended follow-up for the nonablative laser management of BCCs (eTables 4 and 5). Studies have demonstrated that MMS has a 5-year cure rate as high as 99% for BCC.³² Given the limited follow-up period of prior nonablative laser management studies, recurrences might not have been fully evaluated. Our study had a 4.49% recurrence rate for all BCCs and a 5.41% recurrence rate for facial BCCs but was not detectable by clinical examination combined with dermoscopic findings alone. All recurrences required the utilization of OCT or RCM or a combination of these modalities to be diagnosed. In 1 patient with recurrence, we were able to see residual tumor on both OCT and RCM without any inflammation obscuring the scan, given that 3 years had passed. Although 2 months is an optimal follow-up time for OCT, we have not found an optimal follow-up time for RCM, which is another reason why OCT might be preferable to other imaging modalities, such as RCM and high-definition OCT, that have higher resolution but provide less depth on imaging. Although only 40 of 89 patients (4.49%) had follow-up ranging from 3 years to greater than 5 years, long-term follow-up to date has been limited in prior studies.

We believe the high clearance rates and limited recurrence are secondary to the utilization of noninvasive imaging, as the majority of these recurrences would not have been diagnosed based on clinical and/or dermoscopic information alone. Additionally, the 4 biopsy-proven post-MMS recurrence patients that were treated in this study also may not have been diagnosed this early without the use of additional noninvasive imaging. In our opinion, although laser management can be used without noninvasive imaging guidance—dermoscopy, OCT, and/or RCM—this technology is critical not only for early detection but also for proper management of patients.

Conclusion

This study showed a 95.70% clearance rate for all BCCs and a 96.10% clearance rate for facial BCCs. Although we had a zero clinical recurrence rate, 4.49% of all BCCs and 5.41% of facial BCCs had recurred on subsequent monitoring with noninvasive imaging. Given the large size of the study and extended follow-up, we found nonablative laser management to be a reliable treatment alternative with improved cosmetic outcome (Figure) and minimal short-term adverse effects compared to surgery.

Tailored care for the individual patient is based on a variety of options and patient preference, including ease of compliance, number of follow-up visits, invasive vs noninvasive diagnosis and monitoring, and downtime for healing. The use of noninvasive imaging also allowed us to find a more standardized treatment regimen using this nonablative laser combination. We found that 3 or fewer and more than 3 treatments had similar efficacy in tumor clearance. We recommend a standard laser protocol of 3 treatments every 4 to 6 weeks with follow-up 2 months after the final treatment to assess for clearance with OCT.

Larger BCCs might require additional treatments; therefore, we caution against laser therapy without concomitant use of OCT imaging to visualize residual tumor. Utilizing other noninvasive modalities, such as dermoscopy, in combination with thorough skin examination also is critical in the early detection of skin cancers to improve the efficacy of this less-aggressive, nonablative, and cosmetically optimal treatment protocol.

Acknowledgement—We would like to acknowledge Dimitrios Karponis, BSc, from the Impirial College London, England, for his assistance with a portion of the statistical analysis.

The incidence of nonmelanoma skin cancer (NMSC) is steadily increasing, and it accounts for more annual cancer diagnoses than all other malignancies combined.^1,2 For NMSCs of the head and neck, Mohs micrographic surgery (MMS) has become a preferred technique because of its high cure rates, intraprocedural margin control, and improved tissue preservation in cosmetically sensitive areas.³ The nose and ears are especially sensitive anatomic locations given their prominent positions and relative lack of skin reservoir and laxity compared to other areas of the head and neck. For the nose and ears, both patients and referring providers may question who is best suited to surgically remove a malignancy and repair the defect with positive functional and cosmetic results, as a large portion of the defects following tumor extirpation will require a flap or graft for repair.

The notion of plastic surgery is strongly associated with supreme cosmesis for many patients and providers, as the specialty trains in several surgical and nonsurgical elective techniques to preserve and improve appearance. Consequently, patients commonly ask dermatologists if they should be referred to a plastic surgeon for skin cancer removal in cosmetically sensitive areas, especially areas that may require more complex surgical repairs. However, recent Medicare data indicate that dermatologists perform the vast majority of reconstructive skin surgeries, with more than 15 times the number of intermediate and complex closures and more than 4 times the number of flaps and grafts as the next closest specialty.⁴ Earlier studies using Medicare data revealed similar findings, with dermatologic surgeons performing more reconstructions of head and neck skin than both plastic surgeons and otorhinolaryngologists.⁵ However, these studies did not address the characteristics of the tumor, defects, or repairs performed by the specialties for comparison.

We sought to compare the quantity and characteristics of flaps or grafts performed for skin cancer on the nose or ears by fellowship-trained Mohs surgeons and plastic surgeons at 1 academic institution.

Methods

We performed a retrospective chart review of all skin cancer surgeries requiring a flap or graft on the nose or ears at Baylor Scott & White Health (Temple, Texas) from October 1, 2016, to October 1, 2017. This study was approved by the Baylor Scott & White Health institutional review board.

Data Collection
The analysis included full-time, fellowship-trained Mohs surgeons and all full-time plastic surgeons who accepted skin cancer surgery patient referrals as part of their practice and performed all procedures within our hospital system. We reviewed individual provider schedules for both outpatient consultation and operating room notes to capture each procedure performed. To ensure we captured all procedures for both Mohs and plastic surgeons, we used billing codes for any flap or graft repair done on the nose or ears to cross-reference and confirm the cases found by chart review. The total number of flaps or grafts on the nose or ears were collected. Data also were collected regarding the anatomic location of the skin cancer, final defect size prior to the repair, skin tumor type, repair type (flap or graft), and flap (transposition vs advancement) or graft (full thickness vs partial thickness) type. All surgical data were collected from operative notes. Demographic data, including age, race, and sex, also were collected. We also collected data on the specialty of the physicians who referred patients for surgical management of biopsy-proven skin malignancy.

Statistical Analysis
Sample characteristics were described using descriptive statistics. Frequencies and percentages were used to describe categorical variables. Medians and ranges were used to describe continuous variables due to nonsymmetrically distributed data. χ² tests (or Fisher exact tests when low cell counts were present) for categorical variables and Wilcoxon signed rank tests for continuous variables were used to test for associations in bivariate comparisons between MMS and plastic surgery.

Results

A total of 7 physicians (1 fellowship-trained Mohs surgeon and 6 plastic surgeons) at our institution met the inclusion criteria. The Mohs surgeon performed a significantly higher number of flaps and grafts (n=276) than the plastic surgeons (n=17 combined; average per plastic surgeon, 2.83) on the nose or ears in a 12-month period (P<.05)(Table). The median final defect size was not significantly different between MMS (1.5 cm) and plastic surgery (1.8 cm)(P=.306). Flap repairs were more common in patients undergoing MMS (80%) vs plastic surgery (53%)(P=.022)(Figure). For flap repair, advancement flaps were used more commonly (MMS, 53%; plastic surgery, 35%) than transposition flaps (MMS, 27%; plastic surgery, 12%) by both specialties.

Patient age was similar between MMS (median, 74 years) and plastic surgery (median, 73 years) patients (P=.382), but a greater percentage of women were treated by plastic surgeons (53%) compared with Mohs surgeons (33%). The predominant skin tumor type for both specialties was basal cell carcinoma (MMS, 85%; plastic surgery, 76%). Dermatology was the largest referring specialty to both MMS (98%) and plastic surgery (53%). Family medicine referrals comprised a much larger percentage of cases for plastic surgery (24%) compared to MMS (1%).

Comment

This study supports and adds to recent studies and data regarding the utilization of MMS for the treatment of NMSCs. Although the percentage of all skin cancer surgery is increasing for dermatology, little has been reported on more complex repairs. This study highlights the volume and complexity of skin surgery performed by Mohs surgeons compared to our colleagues in plastic surgery.

Defect Size
The defect sizes prior to repair were not statistically different between the 2 types of surgeries, though the median size was slightly larger for plastic surgery (1.8 cm) compared to MMS (1.5 cm). These non–statistically significant differences may be explained by potentially larger tumors requiring repair by plastic surgeons in an operating room. Plastic surgeons, however, may be more likely to take a larger margin of clinically unaffected tissue as part of the initial layer. Plastic surgeons also may be less likely to curette the lesion prior to excision to obtain more clear tumor margins, possibly leading to more stages and a subsequently larger defect. Knowing the clinical sizes of these NMSCs prior to biopsy would have been beneficial to our study, but these data often were not available from the referring providers.

Repair Type
Most patients who underwent MMS had surgical defects repaired with a flap vs a graft, and a much higher percentage of patients who had undergone MMS vs surgical excision with plastic surgery had their defects repaired with flaps. Using a visual analog scale score and Hollander Wound Evaluation Scale, Jacobs et al⁶ found flaps to be cosmetically superior to grafts following tumor extirpation on the nose. The more frequent use of grafts by plastic surgeons could be at least partially explained by larger defect size or by a few outlier larger lesions among an otherwise small sample size. Larger studies may be needed to see if a true discrepancy in repair preferences exists between the specialties.

Referring Specialty
Primary care physician referral comprised a much larger percentage of cases sent for treatment with plastic surgery (24%) compared to MMS (1%). This statistic may represent a practice gap in the perception of MMS and its benefits among our primary care colleagues, particularly among female patients, as a much higher percentage of women were treated with plastic surgery. Important potential benefits of MMS, particularly tissue conservation, cure rates for skin cancer, and the volume of repairs performed by Mohs surgeons, may need to be emphasized.

Scope of Practice
Our colleagues in plastic surgery are extremely gifted and perform numerous repairs outside the scope of most Mohs surgeons. They are vital to multidisciplinary approaches to patients with skin cancer. Although Mohs surgeons focus on treating skin cancers that arise in a narrower range of anatomic locations, the breadth and variety of surgical procedures performed by plastic surgeons is more diverse. Skin cancer surgery may account for a smaller portion of procedures in a plastic surgery practice.

Limitations
There are several limitations to this study. We did not compare cosmesis or wound healing in patients treated by MMS or plastic surgery. The sample size, particularly with plastic surgery, was small and did not allow for a larger, more powerful comparison of data between the 2 specialties. Finally, our study only represents 1 institution over the course of 1 year.

Conclusion

To provide the best care possible, it is imperative for referring physicians to possess an accurate understanding of the volume of cases and the types of repairs that treating specialties perform on a regular basis for NMSCs. This knowledge is particularly important when there is a treatment overlap among specialties. Our data show Mohs surgeons are performing more complex repairs and reconstructions on even the most cosmetically sensitive areas; therefore, primary care physicians and other specialists may be more likely to involve dermatology in the care of skin cancer.

The incidence of nonmelanoma skin cancer (NMSC) is steadily increasing, and it accounts for more annual cancer diagnoses than all other malignancies combined.^1,2 For NMSCs of the head and neck, Mohs micrographic surgery (MMS) has become a preferred technique because of its high cure rates, intraprocedural margin control, and improved tissue preservation in cosmetically sensitive areas.³ The nose and ears are especially sensitive anatomic locations given their prominent positions and relative lack of skin reservoir and laxity compared to other areas of the head and neck. For the nose and ears, both patients and referring providers may question who is best suited to surgically remove a malignancy and repair the defect with positive functional and cosmetic results, as a large portion of the defects following tumor extirpation will require a flap or graft for repair.

The notion of plastic surgery is strongly associated with supreme cosmesis for many patients and providers, as the specialty trains in several surgical and nonsurgical elective techniques to preserve and improve appearance. Consequently, patients commonly ask dermatologists if they should be referred to a plastic surgeon for skin cancer removal in cosmetically sensitive areas, especially areas that may require more complex surgical repairs. However, recent Medicare data indicate that dermatologists perform the vast majority of reconstructive skin surgeries, with more than 15 times the number of intermediate and complex closures and more than 4 times the number of flaps and grafts as the next closest specialty.⁴ Earlier studies using Medicare data revealed similar findings, with dermatologic surgeons performing more reconstructions of head and neck skin than both plastic surgeons and otorhinolaryngologists.⁵ However, these studies did not address the characteristics of the tumor, defects, or repairs performed by the specialties for comparison.

We sought to compare the quantity and characteristics of flaps or grafts performed for skin cancer on the nose or ears by fellowship-trained Mohs surgeons and plastic surgeons at 1 academic institution.

Methods

We performed a retrospective chart review of all skin cancer surgeries requiring a flap or graft on the nose or ears at Baylor Scott & White Health (Temple, Texas) from October 1, 2016, to October 1, 2017. This study was approved by the Baylor Scott & White Health institutional review board.

Data Collection
The analysis included full-time, fellowship-trained Mohs surgeons and all full-time plastic surgeons who accepted skin cancer surgery patient referrals as part of their practice and performed all procedures within our hospital system. We reviewed individual provider schedules for both outpatient consultation and operating room notes to capture each procedure performed. To ensure we captured all procedures for both Mohs and plastic surgeons, we used billing codes for any flap or graft repair done on the nose or ears to cross-reference and confirm the cases found by chart review. The total number of flaps or grafts on the nose or ears were collected. Data also were collected regarding the anatomic location of the skin cancer, final defect size prior to the repair, skin tumor type, repair type (flap or graft), and flap (transposition vs advancement) or graft (full thickness vs partial thickness) type. All surgical data were collected from operative notes. Demographic data, including age, race, and sex, also were collected. We also collected data on the specialty of the physicians who referred patients for surgical management of biopsy-proven skin malignancy.

Statistical Analysis
Sample characteristics were described using descriptive statistics. Frequencies and percentages were used to describe categorical variables. Medians and ranges were used to describe continuous variables due to nonsymmetrically distributed data. χ² tests (or Fisher exact tests when low cell counts were present) for categorical variables and Wilcoxon signed rank tests for continuous variables were used to test for associations in bivariate comparisons between MMS and plastic surgery.

Results

A total of 7 physicians (1 fellowship-trained Mohs surgeon and 6 plastic surgeons) at our institution met the inclusion criteria. The Mohs surgeon performed a significantly higher number of flaps and grafts (n=276) than the plastic surgeons (n=17 combined; average per plastic surgeon, 2.83) on the nose or ears in a 12-month period (P<.05)(Table). The median final defect size was not significantly different between MMS (1.5 cm) and plastic surgery (1.8 cm)(P=.306). Flap repairs were more common in patients undergoing MMS (80%) vs plastic surgery (53%)(P=.022)(Figure). For flap repair, advancement flaps were used more commonly (MMS, 53%; plastic surgery, 35%) than transposition flaps (MMS, 27%; plastic surgery, 12%) by both specialties.

Patient age was similar between MMS (median, 74 years) and plastic surgery (median, 73 years) patients (P=.382), but a greater percentage of women were treated by plastic surgeons (53%) compared with Mohs surgeons (33%). The predominant skin tumor type for both specialties was basal cell carcinoma (MMS, 85%; plastic surgery, 76%). Dermatology was the largest referring specialty to both MMS (98%) and plastic surgery (53%). Family medicine referrals comprised a much larger percentage of cases for plastic surgery (24%) compared to MMS (1%).

Comment

This study supports and adds to recent studies and data regarding the utilization of MMS for the treatment of NMSCs. Although the percentage of all skin cancer surgery is increasing for dermatology, little has been reported on more complex repairs. This study highlights the volume and complexity of skin surgery performed by Mohs surgeons compared to our colleagues in plastic surgery.

Defect Size
The defect sizes prior to repair were not statistically different between the 2 types of surgeries, though the median size was slightly larger for plastic surgery (1.8 cm) compared to MMS (1.5 cm). These non–statistically significant differences may be explained by potentially larger tumors requiring repair by plastic surgeons in an operating room. Plastic surgeons, however, may be more likely to take a larger margin of clinically unaffected tissue as part of the initial layer. Plastic surgeons also may be less likely to curette the lesion prior to excision to obtain more clear tumor margins, possibly leading to more stages and a subsequently larger defect. Knowing the clinical sizes of these NMSCs prior to biopsy would have been beneficial to our study, but these data often were not available from the referring providers.

Repair Type
Most patients who underwent MMS had surgical defects repaired with a flap vs a graft, and a much higher percentage of patients who had undergone MMS vs surgical excision with plastic surgery had their defects repaired with flaps. Using a visual analog scale score and Hollander Wound Evaluation Scale, Jacobs et al⁶ found flaps to be cosmetically superior to grafts following tumor extirpation on the nose. The more frequent use of grafts by plastic surgeons could be at least partially explained by larger defect size or by a few outlier larger lesions among an otherwise small sample size. Larger studies may be needed to see if a true discrepancy in repair preferences exists between the specialties.

Referring Specialty
Primary care physician referral comprised a much larger percentage of cases sent for treatment with plastic surgery (24%) compared to MMS (1%). This statistic may represent a practice gap in the perception of MMS and its benefits among our primary care colleagues, particularly among female patients, as a much higher percentage of women were treated with plastic surgery. Important potential benefits of MMS, particularly tissue conservation, cure rates for skin cancer, and the volume of repairs performed by Mohs surgeons, may need to be emphasized.

Scope of Practice
Our colleagues in plastic surgery are extremely gifted and perform numerous repairs outside the scope of most Mohs surgeons. They are vital to multidisciplinary approaches to patients with skin cancer. Although Mohs surgeons focus on treating skin cancers that arise in a narrower range of anatomic locations, the breadth and variety of surgical procedures performed by plastic surgeons is more diverse. Skin cancer surgery may account for a smaller portion of procedures in a plastic surgery practice.

Limitations
There are several limitations to this study. We did not compare cosmesis or wound healing in patients treated by MMS or plastic surgery. The sample size, particularly with plastic surgery, was small and did not allow for a larger, more powerful comparison of data between the 2 specialties. Finally, our study only represents 1 institution over the course of 1 year.

Conclusion

To provide the best care possible, it is imperative for referring physicians to possess an accurate understanding of the volume of cases and the types of repairs that treating specialties perform on a regular basis for NMSCs. This knowledge is particularly important when there is a treatment overlap among specialties. Our data show Mohs surgeons are performing more complex repairs and reconstructions on even the most cosmetically sensitive areas; therefore, primary care physicians and other specialists may be more likely to involve dermatology in the care of skin cancer.

The incidence of nonmelanoma skin cancer (NMSC) is steadily increasing, and it accounts for more annual cancer diagnoses than all other malignancies combined.^1,2 For NMSCs of the head and neck, Mohs micrographic surgery (MMS) has become a preferred technique because of its high cure rates, intraprocedural margin control, and improved tissue preservation in cosmetically sensitive areas.³ The nose and ears are especially sensitive anatomic locations given their prominent positions and relative lack of skin reservoir and laxity compared to other areas of the head and neck. For the nose and ears, both patients and referring providers may question who is best suited to surgically remove a malignancy and repair the defect with positive functional and cosmetic results, as a large portion of the defects following tumor extirpation will require a flap or graft for repair.

The notion of plastic surgery is strongly associated with supreme cosmesis for many patients and providers, as the specialty trains in several surgical and nonsurgical elective techniques to preserve and improve appearance. Consequently, patients commonly ask dermatologists if they should be referred to a plastic surgeon for skin cancer removal in cosmetically sensitive areas, especially areas that may require more complex surgical repairs. However, recent Medicare data indicate that dermatologists perform the vast majority of reconstructive skin surgeries, with more than 15 times the number of intermediate and complex closures and more than 4 times the number of flaps and grafts as the next closest specialty.⁴ Earlier studies using Medicare data revealed similar findings, with dermatologic surgeons performing more reconstructions of head and neck skin than both plastic surgeons and otorhinolaryngologists.⁵ However, these studies did not address the characteristics of the tumor, defects, or repairs performed by the specialties for comparison.

We sought to compare the quantity and characteristics of flaps or grafts performed for skin cancer on the nose or ears by fellowship-trained Mohs surgeons and plastic surgeons at 1 academic institution.

Methods

We performed a retrospective chart review of all skin cancer surgeries requiring a flap or graft on the nose or ears at Baylor Scott & White Health (Temple, Texas) from October 1, 2016, to October 1, 2017. This study was approved by the Baylor Scott & White Health institutional review board.

Data Collection
The analysis included full-time, fellowship-trained Mohs surgeons and all full-time plastic surgeons who accepted skin cancer surgery patient referrals as part of their practice and performed all procedures within our hospital system. We reviewed individual provider schedules for both outpatient consultation and operating room notes to capture each procedure performed. To ensure we captured all procedures for both Mohs and plastic surgeons, we used billing codes for any flap or graft repair done on the nose or ears to cross-reference and confirm the cases found by chart review. The total number of flaps or grafts on the nose or ears were collected. Data also were collected regarding the anatomic location of the skin cancer, final defect size prior to the repair, skin tumor type, repair type (flap or graft), and flap (transposition vs advancement) or graft (full thickness vs partial thickness) type. All surgical data were collected from operative notes. Demographic data, including age, race, and sex, also were collected. We also collected data on the specialty of the physicians who referred patients for surgical management of biopsy-proven skin malignancy.

Statistical Analysis
Sample characteristics were described using descriptive statistics. Frequencies and percentages were used to describe categorical variables. Medians and ranges were used to describe continuous variables due to nonsymmetrically distributed data. χ² tests (or Fisher exact tests when low cell counts were present) for categorical variables and Wilcoxon signed rank tests for continuous variables were used to test for associations in bivariate comparisons between MMS and plastic surgery.

Results

A total of 7 physicians (1 fellowship-trained Mohs surgeon and 6 plastic surgeons) at our institution met the inclusion criteria. The Mohs surgeon performed a significantly higher number of flaps and grafts (n=276) than the plastic surgeons (n=17 combined; average per plastic surgeon, 2.83) on the nose or ears in a 12-month period (P<.05)(Table). The median final defect size was not significantly different between MMS (1.5 cm) and plastic surgery (1.8 cm)(P=.306). Flap repairs were more common in patients undergoing MMS (80%) vs plastic surgery (53%)(P=.022)(Figure). For flap repair, advancement flaps were used more commonly (MMS, 53%; plastic surgery, 35%) than transposition flaps (MMS, 27%; plastic surgery, 12%) by both specialties.

Patient age was similar between MMS (median, 74 years) and plastic surgery (median, 73 years) patients (P=.382), but a greater percentage of women were treated by plastic surgeons (53%) compared with Mohs surgeons (33%). The predominant skin tumor type for both specialties was basal cell carcinoma (MMS, 85%; plastic surgery, 76%). Dermatology was the largest referring specialty to both MMS (98%) and plastic surgery (53%). Family medicine referrals comprised a much larger percentage of cases for plastic surgery (24%) compared to MMS (1%).

Comment

This study supports and adds to recent studies and data regarding the utilization of MMS for the treatment of NMSCs. Although the percentage of all skin cancer surgery is increasing for dermatology, little has been reported on more complex repairs. This study highlights the volume and complexity of skin surgery performed by Mohs surgeons compared to our colleagues in plastic surgery.

Defect Size
The defect sizes prior to repair were not statistically different between the 2 types of surgeries, though the median size was slightly larger for plastic surgery (1.8 cm) compared to MMS (1.5 cm). These non–statistically significant differences may be explained by potentially larger tumors requiring repair by plastic surgeons in an operating room. Plastic surgeons, however, may be more likely to take a larger margin of clinically unaffected tissue as part of the initial layer. Plastic surgeons also may be less likely to curette the lesion prior to excision to obtain more clear tumor margins, possibly leading to more stages and a subsequently larger defect. Knowing the clinical sizes of these NMSCs prior to biopsy would have been beneficial to our study, but these data often were not available from the referring providers.

Repair Type
Most patients who underwent MMS had surgical defects repaired with a flap vs a graft, and a much higher percentage of patients who had undergone MMS vs surgical excision with plastic surgery had their defects repaired with flaps. Using a visual analog scale score and Hollander Wound Evaluation Scale, Jacobs et al⁶ found flaps to be cosmetically superior to grafts following tumor extirpation on the nose. The more frequent use of grafts by plastic surgeons could be at least partially explained by larger defect size or by a few outlier larger lesions among an otherwise small sample size. Larger studies may be needed to see if a true discrepancy in repair preferences exists between the specialties.

Referring Specialty
Primary care physician referral comprised a much larger percentage of cases sent for treatment with plastic surgery (24%) compared to MMS (1%). This statistic may represent a practice gap in the perception of MMS and its benefits among our primary care colleagues, particularly among female patients, as a much higher percentage of women were treated with plastic surgery. Important potential benefits of MMS, particularly tissue conservation, cure rates for skin cancer, and the volume of repairs performed by Mohs surgeons, may need to be emphasized.

Scope of Practice
Our colleagues in plastic surgery are extremely gifted and perform numerous repairs outside the scope of most Mohs surgeons. They are vital to multidisciplinary approaches to patients with skin cancer. Although Mohs surgeons focus on treating skin cancers that arise in a narrower range of anatomic locations, the breadth and variety of surgical procedures performed by plastic surgeons is more diverse. Skin cancer surgery may account for a smaller portion of procedures in a plastic surgery practice.

Limitations
There are several limitations to this study. We did not compare cosmesis or wound healing in patients treated by MMS or plastic surgery. The sample size, particularly with plastic surgery, was small and did not allow for a larger, more powerful comparison of data between the 2 specialties. Finally, our study only represents 1 institution over the course of 1 year.

Conclusion

To provide the best care possible, it is imperative for referring physicians to possess an accurate understanding of the volume of cases and the types of repairs that treating specialties perform on a regular basis for NMSCs. This knowledge is particularly important when there is a treatment overlap among specialties. Our data show Mohs surgeons are performing more complex repairs and reconstructions on even the most cosmetically sensitive areas; therefore, primary care physicians and other specialists may be more likely to involve dermatology in the care of skin cancer.

The “human microbiota” describes all microorganisms within the human body, including bacteria, viruses, and eukaryotes. The related term “microbiome” refers to the complete catalog of these microbes and their genes.¹ There is a growing awareness that the human microbiota plays an important role in maintaining mental health, and that a disruption in its composition can contribute to manifestations of psychiatric disorders. A growing body of evidence has also linked mental health outcomes to the gut microbiome, suggesting that the gut microbiota can modulate the gut-brain axis.²

Numerous neurotransmitters, including dopamine, serotonin, gamma-aminobutyric acid, and acetylcholine, are produced in the gastrointestinal (GI) tract, and our diet is vital in sustaining and replenishing them. At the same time, our brain regulates our GI tract by secretion of hormones such as oxytocin, leptin, ghrelin, neuropeptide Y, corticotrophin-releasing factor, and a plethora of others. Dysregulation of this microbiome can lead to both physical and mental illnesses. Symptoms of psychiatric disorders, such as depression, psychosis, anxiety, and autism, can be a consequence of this dysregulation.²

Our diet can also modify the gut micro­organisms and therefore many of its metabolic pathways. More attention has been given to pre- and probiotics and their effects on DNA by epigenetic changes. One can quickly start to appreciate how this intricate crosstalk can lead to a variety of pathologic and psychiatric problems that have an adverse effect on autoimmune, inflammatory, metabolic, cognitive, and behavioral processes.^2,3

Thus far, links have mostly been reported in animal models, and human studies are limited.⁴ Researchers are just beginning to elucidate how the microbiota affect gut-brain signaling in humans. Such mechanisms may include alterations in microbial composition, immune activation, vagus nerve signaling, alterations in tryptophan metabolism, production of specific microbial neuroactive metabolites, and bacterial cell wall sugars.⁵ The microbiota-gut-brain axis plays a part in regulating/programming the hypothalamic-pituitary-adrenal (HPA) axis throughout the life span.³ The interactions between the gut microbiome, the immune system, and the CNS are regulated through pathways that involve endocrine functions (HPA axis), the immune system, and metabolic factors.^3,4 Recent research focusing on the gut microbiome has also given rise to international projects such as the Human Microbiome Project (Human Microbiome Project Consortium, 2012).³

Several studies have looked into psychiatry and inflammatory/immune pathways. Here we review 3 recent studies that have focused on the gut-brain axis (Table^6-8).

1. Rudzki L, Pawlak D, Pawlak K, et al. Immune suppression of IgG response against dairy proteins in major depression. BMC Psychiatry. 2017;17(1):268.

The aim of this study was to evaluate immunoglobulin G (IgG) response against 40 food products in patients with depression vs those in a control group, along with changes in inflammatory markers, psychological stress, and dietary variables.⁶

Study design

• N = 63, IgG levels against 44 food products, cortisol levels, tumor necrosis factor (TNF)-alpha, interleukin 6 (IL-6), and IL-1 beta levels were recorded. The psychological parameters of 34 participants with depression and 29 controls were compared using the Hamilton Depression Rating scale, (HAM-D-17), Perceived Stress scale, and Symptom Checklist scale. The study was conducted in Poland.

Continue to: Outcomes

Outcomes

• Patients who were depressed had lower IgG levels against dairy products compared to controls when there was high dairy consumption. However, there was no overall difference between patients and controls in mean IgG concentration against food products. 
• Patients who were depressed had higher levels of cortisol. Levels of cortisol had a positive correlation with HAM-D-17 score. Patients with depression had lower levels of TNF-alpha.

Conclusion

• Patients with depression had lower levels of IgG against dairy protein. Patients with depression had high cortisol levels but decreased levels of TNF-alpha, which could explain an immune suppression of IgG in these patients. There were no differences in IL-6 or IL-1beta levels.

Hypercortisolemia is present in approximately 60% of patients with depression. Elevated cortisol levels have a negative effect on lymphocyte function. B-lymphocytes (CD 10+ and CD 19+) are sensitive to glucocorticoids. Studies in mice have demonstrated that elevated glucocorticoid levels are associated with a 50% decrease in serum B-lymphocytes, and this can be explained by downregulation of c-myc protein, which plays a role in cell proliferation and cell survival. Glucocorticoids also decrease levels of protein kinases that are vital for the cell cycle to continue, and they upregulate p27 and p21, which are cell cycle inhibitors. Therefore, if high cortisol suppresses B-lymphocyte production, this can explain how patients with depression have low IgG levels, since B-lymphocytes differentiate into plasma cells that will produce antibodies.⁶

Depression can trigger an inflammatory response by increasing levels of inflammatory cytokines, acute phase reactants, and oxidative molecules. The inflammatory response can lead to intestinal wall disruption, and therefore bacteria can migrate across the GI barrier, along with food antigens, which could then lead to food antigen hypersensitivity.⁶

The significance of diet

Many studies have looked into specific types of diets, such as the Mediterranean diet, the ketogenic diet, and the addition of supplements such as probiotics, omega-3 fatty acids, zinc, and multivitamins.⁷ The Mediterranean diet is high in fiber, nuts, legumes, and fish.⁷ The ketogenic diet includes a controlled amount of fat, but is low in protein and carbohydrates.⁷ The main point is that a balanced diet can have a positive effect on mental health.⁷ The Mediterranean diet has shown to decrease the incidence of cardiovascular disease and lower the risk of depression.⁷ In animal studies, the ketogenic diet has improved anxiety, depression, and autism.⁷ Diet clearly affects gut microbiota and, as a consequence, the body’s level of inflammation.⁷ 

Continue to: The following review...

The following review highlighted the significance of diet on gut microbiome and mental health.⁷

2. Mörkl S, Wagner-Skacel J, Lahousen T, et al. The role of nutrition and the gut- brain axis in psychiatry: a review of the literature. Neuropsychobiology. 2018;17: 1-9.

Study design

• These researchers provided a narrative review of the significance of a healthy diet and nutritional supplements on the gut microbiome and the treatment of patients with psychiatric illness. 

Outcomes

• This review suggested dietary coaching as a nonpharmacologic treatment for patients with psychiatric illness. 

Conclusion

• The utilization of nutritional advice, along with medication management, therapy, and physical activity, can provide a holistic approach to the biopsychosocial treatment of patients with psychiatric illness.

This review also emphasized the poor dietary trends of Westernized countries, which include calorie-dense, genetically altered, processed meals. As Mörkl et al⁷ noted, we are overfed but undernourished. Mörkl et al⁷ reviewed studies that involve dietary coaching as part of the treatment plan of patients with mental illness. In one of these studies, patients who received nutritional advice and coaching over 6 weeks had a 40% to 50% decrease in depressive symptoms. These effects persisted for 2 more years. Mörkl et al⁷ also reviewed an Italian study that found that providing nutritional advice in patients with affective disorders and psychosis helped improve symptom severity and sleep.⁷

Continue to: Mörkl et al...

Mörkl et al⁷ also reviewed dietary supplements. Some studies have linked use of omega-3 fatty acids with improvement in affective disorders, Alzheimer’s disease, and posttraumatic stress disorder, as well as cardiovascular conditions. Omega-3 fatty acids may exert beneficial effects by enhancing brain-derived neurotrophic factor and neurogenesis as well as by decreasing inflammation.⁷

Zinc supplementation can also improve depression, as it has been linked to cytokine variation and hippocampal neuronal growth. Vitamin B₉ deficiency and vitamin D deficiency also have been associated with depression. Mörkl et al⁷ emphasized that a balanced diet that incorporates a variety of nutrients is more beneficial than supplementation of any individual vitamin alone.

Researchers have long emphasized the importance of a healthy balanced diet when treating patients with medical conditions such as cardiovascular or cerebrovascular diseases. Based on the studies Mörkl et al⁷ reviewed, the same emphasis should be communicated to our patients who suffer from psychiatric conditions.

The gut and anxiety

The gut microbiome has also been an area of research when studying generalized anxiety disorder (GAD).⁸

3. Jiang HY, Zhang X, Yu ZH, et al. Altered gut microbiota profile in patients with generalized anxiety disorder. J Psychiatr Res. 2018;104:130-136.

The aim of the study was to determine if there were changes in the composition of the gut microbiome in patients with GAD compared with healthy controls.⁸

Continue to: Study design

Study design

• A cross-sectional study of 76 patients in Zhejiang, China. Forty patients with GAD in the active state and 36 healthy controls were compared in terms of composition of GI microbacterial flora.
• Researchers also examined a subgroup of 12 patients who were treatment-naïve and 17 controls. Stool samples were collected from the 12 patients who were treatment-naïve before initiating medication.
• Researchers also conducted a prospective study in a subgroup of 9 patients with GAD in both the active state and remissive state. Two stool samples were collected from each patient—one during the active state of GAD and one during the remissive state—for a total of 18 samples. Stool samples analyzed with the use of polymerase chain reaction and microbial analysis.
• Patients completed the Hamilton Anxiety Rating (HAM-A) scale and were classified into groups. Those with HAM-A scores >14 were classified as being in the active state of GAD, and those with scores <7 were classified as being in the remissive state. 

Outcomes

• Among the samples collected, 8 bacterial taxa were found in different amounts in patients with GAD and healthy controls. Bacteroidetes, Ruminococcus gnavus, and Fusobacterium were increased in patients with GAD compared with controls, while Faecalibacterium, Eubacterium rectale, Sutterella, Lachnospira, and Butyricicoccus were increased in healthy controls.
• Bacterial variety was notably lower in the 12 patients who were treatment-naïve compared with the control group.
• There was no notable difference in microbial composition between patients in the active vs remissive state.

Conclusion

• Patients with GAD had less short chain fatty acid–producing bacteria (Faecalibacterium, Eubacterium rectale, Sutterella, Lachnospira, and Butyricicoccus) compared with controls. Decreased formation of short chain fatty acids could lead to GI barrier disruption. Fusobacterium and Ruminococcus were increased in patients with GAD. Fusobacterium can cause disease and be invasive when it disseminates within the body. The inflammatory characteristics of Fusobacterium contribute to the immunologic activation in GAD. Ruminococcus breaks down mucin, which could then increase GI permeability by mucous degradation of the GI lumen.

Changes in food processing and manufacturing have led to changes in our diets. Changes in our normal GI microbacterial flora could lead to increased gut permeability, bacterial dissemination, and subsequent systemic inflammation. Research has shown that the composition of the microbiota changes across the life span.⁹ A balanced intake of nutrients is important for both our physical and mental health and safeguards the basis of gut microbiome regulation. A well-regulated gut microbiome ensures low levels of inflammation in the brain and body. Lifestyle modifications and dietary coaching could be practical interventions for patients with psychiatric conditions.⁵ Current advances in technology now offer precise analyses of thousands of metabolites, enabling metabolomics to offer the promise of discovering new drug targets and biomarkers that may help pave a way to precision medicine.

The “human microbiota” describes all microorganisms within the human body, including bacteria, viruses, and eukaryotes. The related term “microbiome” refers to the complete catalog of these microbes and their genes.¹ There is a growing awareness that the human microbiota plays an important role in maintaining mental health, and that a disruption in its composition can contribute to manifestations of psychiatric disorders. A growing body of evidence has also linked mental health outcomes to the gut microbiome, suggesting that the gut microbiota can modulate the gut-brain axis.²

Numerous neurotransmitters, including dopamine, serotonin, gamma-aminobutyric acid, and acetylcholine, are produced in the gastrointestinal (GI) tract, and our diet is vital in sustaining and replenishing them. At the same time, our brain regulates our GI tract by secretion of hormones such as oxytocin, leptin, ghrelin, neuropeptide Y, corticotrophin-releasing factor, and a plethora of others. Dysregulation of this microbiome can lead to both physical and mental illnesses. Symptoms of psychiatric disorders, such as depression, psychosis, anxiety, and autism, can be a consequence of this dysregulation.²

Our diet can also modify the gut micro­organisms and therefore many of its metabolic pathways. More attention has been given to pre- and probiotics and their effects on DNA by epigenetic changes. One can quickly start to appreciate how this intricate crosstalk can lead to a variety of pathologic and psychiatric problems that have an adverse effect on autoimmune, inflammatory, metabolic, cognitive, and behavioral processes.^2,3

Thus far, links have mostly been reported in animal models, and human studies are limited.⁴ Researchers are just beginning to elucidate how the microbiota affect gut-brain signaling in humans. Such mechanisms may include alterations in microbial composition, immune activation, vagus nerve signaling, alterations in tryptophan metabolism, production of specific microbial neuroactive metabolites, and bacterial cell wall sugars.⁵ The microbiota-gut-brain axis plays a part in regulating/programming the hypothalamic-pituitary-adrenal (HPA) axis throughout the life span.³ The interactions between the gut microbiome, the immune system, and the CNS are regulated through pathways that involve endocrine functions (HPA axis), the immune system, and metabolic factors.^3,4 Recent research focusing on the gut microbiome has also given rise to international projects such as the Human Microbiome Project (Human Microbiome Project Consortium, 2012).³

Several studies have looked into psychiatry and inflammatory/immune pathways. Here we review 3 recent studies that have focused on the gut-brain axis (Table^6-8).

1. Rudzki L, Pawlak D, Pawlak K, et al. Immune suppression of IgG response against dairy proteins in major depression. BMC Psychiatry. 2017;17(1):268.

The aim of this study was to evaluate immunoglobulin G (IgG) response against 40 food products in patients with depression vs those in a control group, along with changes in inflammatory markers, psychological stress, and dietary variables.⁶

Study design

• N = 63, IgG levels against 44 food products, cortisol levels, tumor necrosis factor (TNF)-alpha, interleukin 6 (IL-6), and IL-1 beta levels were recorded. The psychological parameters of 34 participants with depression and 29 controls were compared using the Hamilton Depression Rating scale, (HAM-D-17), Perceived Stress scale, and Symptom Checklist scale. The study was conducted in Poland.

Continue to: Outcomes

Outcomes

• Patients who were depressed had lower IgG levels against dairy products compared to controls when there was high dairy consumption. However, there was no overall difference between patients and controls in mean IgG concentration against food products. 
• Patients who were depressed had higher levels of cortisol. Levels of cortisol had a positive correlation with HAM-D-17 score. Patients with depression had lower levels of TNF-alpha.

Conclusion

• Patients with depression had lower levels of IgG against dairy protein. Patients with depression had high cortisol levels but decreased levels of TNF-alpha, which could explain an immune suppression of IgG in these patients. There were no differences in IL-6 or IL-1beta levels.

Hypercortisolemia is present in approximately 60% of patients with depression. Elevated cortisol levels have a negative effect on lymphocyte function. B-lymphocytes (CD 10+ and CD 19+) are sensitive to glucocorticoids. Studies in mice have demonstrated that elevated glucocorticoid levels are associated with a 50% decrease in serum B-lymphocytes, and this can be explained by downregulation of c-myc protein, which plays a role in cell proliferation and cell survival. Glucocorticoids also decrease levels of protein kinases that are vital for the cell cycle to continue, and they upregulate p27 and p21, which are cell cycle inhibitors. Therefore, if high cortisol suppresses B-lymphocyte production, this can explain how patients with depression have low IgG levels, since B-lymphocytes differentiate into plasma cells that will produce antibodies.⁶

Depression can trigger an inflammatory response by increasing levels of inflammatory cytokines, acute phase reactants, and oxidative molecules. The inflammatory response can lead to intestinal wall disruption, and therefore bacteria can migrate across the GI barrier, along with food antigens, which could then lead to food antigen hypersensitivity.⁶

The significance of diet

Many studies have looked into specific types of diets, such as the Mediterranean diet, the ketogenic diet, and the addition of supplements such as probiotics, omega-3 fatty acids, zinc, and multivitamins.⁷ The Mediterranean diet is high in fiber, nuts, legumes, and fish.⁷ The ketogenic diet includes a controlled amount of fat, but is low in protein and carbohydrates.⁷ The main point is that a balanced diet can have a positive effect on mental health.⁷ The Mediterranean diet has shown to decrease the incidence of cardiovascular disease and lower the risk of depression.⁷ In animal studies, the ketogenic diet has improved anxiety, depression, and autism.⁷ Diet clearly affects gut microbiota and, as a consequence, the body’s level of inflammation.⁷ 

Continue to: The following review...

The following review highlighted the significance of diet on gut microbiome and mental health.⁷

2. Mörkl S, Wagner-Skacel J, Lahousen T, et al. The role of nutrition and the gut- brain axis in psychiatry: a review of the literature. Neuropsychobiology. 2018;17: 1-9.

Study design

• These researchers provided a narrative review of the significance of a healthy diet and nutritional supplements on the gut microbiome and the treatment of patients with psychiatric illness. 

Outcomes

• This review suggested dietary coaching as a nonpharmacologic treatment for patients with psychiatric illness. 

Conclusion

• The utilization of nutritional advice, along with medication management, therapy, and physical activity, can provide a holistic approach to the biopsychosocial treatment of patients with psychiatric illness.

This review also emphasized the poor dietary trends of Westernized countries, which include calorie-dense, genetically altered, processed meals. As Mörkl et al⁷ noted, we are overfed but undernourished. Mörkl et al⁷ reviewed studies that involve dietary coaching as part of the treatment plan of patients with mental illness. In one of these studies, patients who received nutritional advice and coaching over 6 weeks had a 40% to 50% decrease in depressive symptoms. These effects persisted for 2 more years. Mörkl et al⁷ also reviewed an Italian study that found that providing nutritional advice in patients with affective disorders and psychosis helped improve symptom severity and sleep.⁷

Continue to: Mörkl et al...

Mörkl et al⁷ also reviewed dietary supplements. Some studies have linked use of omega-3 fatty acids with improvement in affective disorders, Alzheimer’s disease, and posttraumatic stress disorder, as well as cardiovascular conditions. Omega-3 fatty acids may exert beneficial effects by enhancing brain-derived neurotrophic factor and neurogenesis as well as by decreasing inflammation.⁷

Zinc supplementation can also improve depression, as it has been linked to cytokine variation and hippocampal neuronal growth. Vitamin B₉ deficiency and vitamin D deficiency also have been associated with depression. Mörkl et al⁷ emphasized that a balanced diet that incorporates a variety of nutrients is more beneficial than supplementation of any individual vitamin alone.

Researchers have long emphasized the importance of a healthy balanced diet when treating patients with medical conditions such as cardiovascular or cerebrovascular diseases. Based on the studies Mörkl et al⁷ reviewed, the same emphasis should be communicated to our patients who suffer from psychiatric conditions.

The gut and anxiety

The gut microbiome has also been an area of research when studying generalized anxiety disorder (GAD).⁸

3. Jiang HY, Zhang X, Yu ZH, et al. Altered gut microbiota profile in patients with generalized anxiety disorder. J Psychiatr Res. 2018;104:130-136.

The aim of the study was to determine if there were changes in the composition of the gut microbiome in patients with GAD compared with healthy controls.⁸

Continue to: Study design

Study design

• A cross-sectional study of 76 patients in Zhejiang, China. Forty patients with GAD in the active state and 36 healthy controls were compared in terms of composition of GI microbacterial flora.
• Researchers also examined a subgroup of 12 patients who were treatment-naïve and 17 controls. Stool samples were collected from the 12 patients who were treatment-naïve before initiating medication.
• Researchers also conducted a prospective study in a subgroup of 9 patients with GAD in both the active state and remissive state. Two stool samples were collected from each patient—one during the active state of GAD and one during the remissive state—for a total of 18 samples. Stool samples analyzed with the use of polymerase chain reaction and microbial analysis.
• Patients completed the Hamilton Anxiety Rating (HAM-A) scale and were classified into groups. Those with HAM-A scores >14 were classified as being in the active state of GAD, and those with scores <7 were classified as being in the remissive state. 

Outcomes

• Among the samples collected, 8 bacterial taxa were found in different amounts in patients with GAD and healthy controls. Bacteroidetes, Ruminococcus gnavus, and Fusobacterium were increased in patients with GAD compared with controls, while Faecalibacterium, Eubacterium rectale, Sutterella, Lachnospira, and Butyricicoccus were increased in healthy controls.
• Bacterial variety was notably lower in the 12 patients who were treatment-naïve compared with the control group.
• There was no notable difference in microbial composition between patients in the active vs remissive state.

Conclusion

• Patients with GAD had less short chain fatty acid–producing bacteria (Faecalibacterium, Eubacterium rectale, Sutterella, Lachnospira, and Butyricicoccus) compared with controls. Decreased formation of short chain fatty acids could lead to GI barrier disruption. Fusobacterium and Ruminococcus were increased in patients with GAD. Fusobacterium can cause disease and be invasive when it disseminates within the body. The inflammatory characteristics of Fusobacterium contribute to the immunologic activation in GAD. Ruminococcus breaks down mucin, which could then increase GI permeability by mucous degradation of the GI lumen.

Changes in food processing and manufacturing have led to changes in our diets. Changes in our normal GI microbacterial flora could lead to increased gut permeability, bacterial dissemination, and subsequent systemic inflammation. Research has shown that the composition of the microbiota changes across the life span.⁹ A balanced intake of nutrients is important for both our physical and mental health and safeguards the basis of gut microbiome regulation. A well-regulated gut microbiome ensures low levels of inflammation in the brain and body. Lifestyle modifications and dietary coaching could be practical interventions for patients with psychiatric conditions.⁵ Current advances in technology now offer precise analyses of thousands of metabolites, enabling metabolomics to offer the promise of discovering new drug targets and biomarkers that may help pave a way to precision medicine.

The “human microbiota” describes all microorganisms within the human body, including bacteria, viruses, and eukaryotes. The related term “microbiome” refers to the complete catalog of these microbes and their genes.¹ There is a growing awareness that the human microbiota plays an important role in maintaining mental health, and that a disruption in its composition can contribute to manifestations of psychiatric disorders. A growing body of evidence has also linked mental health outcomes to the gut microbiome, suggesting that the gut microbiota can modulate the gut-brain axis.²

Numerous neurotransmitters, including dopamine, serotonin, gamma-aminobutyric acid, and acetylcholine, are produced in the gastrointestinal (GI) tract, and our diet is vital in sustaining and replenishing them. At the same time, our brain regulates our GI tract by secretion of hormones such as oxytocin, leptin, ghrelin, neuropeptide Y, corticotrophin-releasing factor, and a plethora of others. Dysregulation of this microbiome can lead to both physical and mental illnesses. Symptoms of psychiatric disorders, such as depression, psychosis, anxiety, and autism, can be a consequence of this dysregulation.²

Our diet can also modify the gut micro­organisms and therefore many of its metabolic pathways. More attention has been given to pre- and probiotics and their effects on DNA by epigenetic changes. One can quickly start to appreciate how this intricate crosstalk can lead to a variety of pathologic and psychiatric problems that have an adverse effect on autoimmune, inflammatory, metabolic, cognitive, and behavioral processes.^2,3

Thus far, links have mostly been reported in animal models, and human studies are limited.⁴ Researchers are just beginning to elucidate how the microbiota affect gut-brain signaling in humans. Such mechanisms may include alterations in microbial composition, immune activation, vagus nerve signaling, alterations in tryptophan metabolism, production of specific microbial neuroactive metabolites, and bacterial cell wall sugars.⁵ The microbiota-gut-brain axis plays a part in regulating/programming the hypothalamic-pituitary-adrenal (HPA) axis throughout the life span.³ The interactions between the gut microbiome, the immune system, and the CNS are regulated through pathways that involve endocrine functions (HPA axis), the immune system, and metabolic factors.^3,4 Recent research focusing on the gut microbiome has also given rise to international projects such as the Human Microbiome Project (Human Microbiome Project Consortium, 2012).³

Several studies have looked into psychiatry and inflammatory/immune pathways. Here we review 3 recent studies that have focused on the gut-brain axis (Table^6-8).

1. Rudzki L, Pawlak D, Pawlak K, et al. Immune suppression of IgG response against dairy proteins in major depression. BMC Psychiatry. 2017;17(1):268.

The aim of this study was to evaluate immunoglobulin G (IgG) response against 40 food products in patients with depression vs those in a control group, along with changes in inflammatory markers, psychological stress, and dietary variables.⁶

Study design

• N = 63, IgG levels against 44 food products, cortisol levels, tumor necrosis factor (TNF)-alpha, interleukin 6 (IL-6), and IL-1 beta levels were recorded. The psychological parameters of 34 participants with depression and 29 controls were compared using the Hamilton Depression Rating scale, (HAM-D-17), Perceived Stress scale, and Symptom Checklist scale. The study was conducted in Poland.

Continue to: Outcomes

Outcomes

• Patients who were depressed had lower IgG levels against dairy products compared to controls when there was high dairy consumption. However, there was no overall difference between patients and controls in mean IgG concentration against food products. 
• Patients who were depressed had higher levels of cortisol. Levels of cortisol had a positive correlation with HAM-D-17 score. Patients with depression had lower levels of TNF-alpha.

Conclusion

• Patients with depression had lower levels of IgG against dairy protein. Patients with depression had high cortisol levels but decreased levels of TNF-alpha, which could explain an immune suppression of IgG in these patients. There were no differences in IL-6 or IL-1beta levels.

Hypercortisolemia is present in approximately 60% of patients with depression. Elevated cortisol levels have a negative effect on lymphocyte function. B-lymphocytes (CD 10+ and CD 19+) are sensitive to glucocorticoids. Studies in mice have demonstrated that elevated glucocorticoid levels are associated with a 50% decrease in serum B-lymphocytes, and this can be explained by downregulation of c-myc protein, which plays a role in cell proliferation and cell survival. Glucocorticoids also decrease levels of protein kinases that are vital for the cell cycle to continue, and they upregulate p27 and p21, which are cell cycle inhibitors. Therefore, if high cortisol suppresses B-lymphocyte production, this can explain how patients with depression have low IgG levels, since B-lymphocytes differentiate into plasma cells that will produce antibodies.⁶

Depression can trigger an inflammatory response by increasing levels of inflammatory cytokines, acute phase reactants, and oxidative molecules. The inflammatory response can lead to intestinal wall disruption, and therefore bacteria can migrate across the GI barrier, along with food antigens, which could then lead to food antigen hypersensitivity.⁶

The significance of diet

Many studies have looked into specific types of diets, such as the Mediterranean diet, the ketogenic diet, and the addition of supplements such as probiotics, omega-3 fatty acids, zinc, and multivitamins.⁷ The Mediterranean diet is high in fiber, nuts, legumes, and fish.⁷ The ketogenic diet includes a controlled amount of fat, but is low in protein and carbohydrates.⁷ The main point is that a balanced diet can have a positive effect on mental health.⁷ The Mediterranean diet has shown to decrease the incidence of cardiovascular disease and lower the risk of depression.⁷ In animal studies, the ketogenic diet has improved anxiety, depression, and autism.⁷ Diet clearly affects gut microbiota and, as a consequence, the body’s level of inflammation.⁷ 

Continue to: The following review...

The following review highlighted the significance of diet on gut microbiome and mental health.⁷

2. Mörkl S, Wagner-Skacel J, Lahousen T, et al. The role of nutrition and the gut- brain axis in psychiatry: a review of the literature. Neuropsychobiology. 2018;17: 1-9.

Study design

• These researchers provided a narrative review of the significance of a healthy diet and nutritional supplements on the gut microbiome and the treatment of patients with psychiatric illness. 

Outcomes

• This review suggested dietary coaching as a nonpharmacologic treatment for patients with psychiatric illness. 

Conclusion

• The utilization of nutritional advice, along with medication management, therapy, and physical activity, can provide a holistic approach to the biopsychosocial treatment of patients with psychiatric illness.

This review also emphasized the poor dietary trends of Westernized countries, which include calorie-dense, genetically altered, processed meals. As Mörkl et al⁷ noted, we are overfed but undernourished. Mörkl et al⁷ reviewed studies that involve dietary coaching as part of the treatment plan of patients with mental illness. In one of these studies, patients who received nutritional advice and coaching over 6 weeks had a 40% to 50% decrease in depressive symptoms. These effects persisted for 2 more years. Mörkl et al⁷ also reviewed an Italian study that found that providing nutritional advice in patients with affective disorders and psychosis helped improve symptom severity and sleep.⁷

Continue to: Mörkl et al...

Mörkl et al⁷ also reviewed dietary supplements. Some studies have linked use of omega-3 fatty acids with improvement in affective disorders, Alzheimer’s disease, and posttraumatic stress disorder, as well as cardiovascular conditions. Omega-3 fatty acids may exert beneficial effects by enhancing brain-derived neurotrophic factor and neurogenesis as well as by decreasing inflammation.⁷

Zinc supplementation can also improve depression, as it has been linked to cytokine variation and hippocampal neuronal growth. Vitamin B₉ deficiency and vitamin D deficiency also have been associated with depression. Mörkl et al⁷ emphasized that a balanced diet that incorporates a variety of nutrients is more beneficial than supplementation of any individual vitamin alone.

Researchers have long emphasized the importance of a healthy balanced diet when treating patients with medical conditions such as cardiovascular or cerebrovascular diseases. Based on the studies Mörkl et al⁷ reviewed, the same emphasis should be communicated to our patients who suffer from psychiatric conditions.

The gut and anxiety

The gut microbiome has also been an area of research when studying generalized anxiety disorder (GAD).⁸

3. Jiang HY, Zhang X, Yu ZH, et al. Altered gut microbiota profile in patients with generalized anxiety disorder. J Psychiatr Res. 2018;104:130-136.

The aim of the study was to determine if there were changes in the composition of the gut microbiome in patients with GAD compared with healthy controls.⁸

Continue to: Study design

Study design

• A cross-sectional study of 76 patients in Zhejiang, China. Forty patients with GAD in the active state and 36 healthy controls were compared in terms of composition of GI microbacterial flora.
• Researchers also examined a subgroup of 12 patients who were treatment-naïve and 17 controls. Stool samples were collected from the 12 patients who were treatment-naïve before initiating medication.
• Researchers also conducted a prospective study in a subgroup of 9 patients with GAD in both the active state and remissive state. Two stool samples were collected from each patient—one during the active state of GAD and one during the remissive state—for a total of 18 samples. Stool samples analyzed with the use of polymerase chain reaction and microbial analysis.
• Patients completed the Hamilton Anxiety Rating (HAM-A) scale and were classified into groups. Those with HAM-A scores >14 were classified as being in the active state of GAD, and those with scores <7 were classified as being in the remissive state. 

Outcomes

• Among the samples collected, 8 bacterial taxa were found in different amounts in patients with GAD and healthy controls. Bacteroidetes, Ruminococcus gnavus, and Fusobacterium were increased in patients with GAD compared with controls, while Faecalibacterium, Eubacterium rectale, Sutterella, Lachnospira, and Butyricicoccus were increased in healthy controls.
• Bacterial variety was notably lower in the 12 patients who were treatment-naïve compared with the control group.
• There was no notable difference in microbial composition between patients in the active vs remissive state.

Conclusion

• Patients with GAD had less short chain fatty acid–producing bacteria (Faecalibacterium, Eubacterium rectale, Sutterella, Lachnospira, and Butyricicoccus) compared with controls. Decreased formation of short chain fatty acids could lead to GI barrier disruption. Fusobacterium and Ruminococcus were increased in patients with GAD. Fusobacterium can cause disease and be invasive when it disseminates within the body. The inflammatory characteristics of Fusobacterium contribute to the immunologic activation in GAD. Ruminococcus breaks down mucin, which could then increase GI permeability by mucous degradation of the GI lumen.

Changes in food processing and manufacturing have led to changes in our diets. Changes in our normal GI microbacterial flora could lead to increased gut permeability, bacterial dissemination, and subsequent systemic inflammation. Research has shown that the composition of the microbiota changes across the life span.⁹ A balanced intake of nutrients is important for both our physical and mental health and safeguards the basis of gut microbiome regulation. A well-regulated gut microbiome ensures low levels of inflammation in the brain and body. Lifestyle modifications and dietary coaching could be practical interventions for patients with psychiatric conditions.⁵ Current advances in technology now offer precise analyses of thousands of metabolites, enabling metabolomics to offer the promise of discovering new drug targets and biomarkers that may help pave a way to precision medicine.

Spinal cord injuries (SCI) are common in veteran populations.¹ Veterans with spinal cord injuries and disorders (SCI/D) also may have concurrent sleep disturbances. Spinal cord injury typically causes spasticity.^2,3 Hypersensitivity of the flexor reflex pathways is believed to cause painful muscle spasms in patients with SCI.⁴ Neuropathic pain at or below the level of the lesion also is common.

Restless legs syndrome (RLS) is a common sleep disorder that affects sleep quality and can occur concomitantly with spinal cord lesions.⁵ In about 80% of RLS cases, involuntary movements of legs across hip, knee, and ankle joints during sleep, known as periodic limb movement during sleep (PLMS), occurs.⁶ Several studies showed increased prevalence of PLMS in patients with SCI, and some case reports suggest an increased prevalence of RLS in this population.^7,8 One small study showed that 100% of patients with SCI had symptoms of RLS.⁶ Another study found that SCI could trigger PLMS.⁸

The pathophysiology of RLS and PLMS in patients with SCI is not fully understood, but case reports describing PLM in SCI patients points to a possible role of central pattern generators and the flexor reflex afferents in the pathophysiology of PLMS.^9,10 Changes of the tissue microstructure in the midbrain and upper cervical spinal cord have been described in patients with RLS.¹¹The objective of this study was to assess the prevalence of RLS in a veteran population with SCI/D and to determine possible neuroanatomical patterns involved in RLS and SCI/D.

Methods

The institutional review and ethical approval boards of the Minneapolis VA Health Care System approved the study. Within the VA system, 666 patients with SCI/D were identified using a national database. Of the 666 people, 316 were excluded, 199 were included, and 151 were deceased.

Patients aged between 18 and 65 years were included in the study. Charts of patients who had been discharged with the diagnosis of SCI from 2002 to 2008 were studied. All patients met the inclusion criteria of the International Restless Legs Syndrome Study Group diagnosis.

Exclusion criteria were as follows: Patients with evidence of brain pathology (eg, stroke), concurrent neurologic condition associated with RLS (Parkinson disease, spinocerebellar ataxia, peripheral neuropathy), concurrent psychiatric condition within the setting of treatment with dopamine antagonists, secondary causes of RLS (renal failure/uremia, iron deficiency, rheumatoid arthritis, and pregnancy) and a recent history of alcohol or drug misuse or current evidence of substance use of < 1 year.

A patient list was compiled that included the etiology of the SCI (vascular injury, multiple sclerosis [MS], trauma, unknown, and other), the level(s) and completeness of the SCI per radiology report, RLS pharmacotherapies, and pertinent medical history.

Axial T2-weighted images on magnetic resonance imaging (MRI) scans were retrospectively reviewed. Sagittal T1/T2-weighted and axial T2-weighted sequences were performed routinely on all patients with spinal cord lesions. The analysis included the extension of the lesion on both sagittal and axial distributions. The anatomic location of the cord lesion was categorized by the following: (1) pure gray matter (central cord); (2) white matter (dorsal [D], dorsolateral [DL], ventral [V], ventrolateral areas [VL]).

A questionnaire using standard diagnostic criteria for RLS was mailed to the 199 patients who met the inclusion criteria (Appendix A).

All analyses were carried out using StataCorp STATA 13 (College Station, TX). Descriptive statistics were used. The analyses were carried out using chi-square and Fisher exact tests. Differences between the groups were considered statistically significant at P < .05. The data were analyzed to obtain point prevalence among patients with SCI, and comparisons were made among the different subgroups.

Results

Of the 162 patients who chose to participate in the study, the sleep specialists confirmed 31 (19%) to have RLS, 112 (69%) were confirmed negative for RLS, and an additional 19 (12%) screened positive for RLS but were not confirmed to have RLS by the sleep specialists (Figure 1).

The etiology of SCI was subdivided into 4 groups: MS, trauma, vascular, and other/unknown. Within each group (– RLS vs + RLS), MS and trauma were the most common etiologies with 55% MS and 36% trauma in the + RLS group.

When comparing RLS among the spinal cord levels (cervical, thoracic, lumbar and cervical + thoracic), only the cervical + thoracic subgroup (18% + RLS vs 5% – RLS) showed a significant difference (Figure 2).

There was no significant difference found with the prevalence of RLS in the axial plane of the spinal cord lesions (ventral/ventro-lateral/central cord vs dorsal/dorsolateral) or by the completeness of spinal cord lesions, P = .76. There was a higher prevalence of incomplete cord injury, however, within each subgroup of RLS.

The Mann-Whitney test was used to analyze the burden of disease in both groups (+ RLS vs – RLS). Moderate level of burden was most frequently reported with a higher prevalence within the + RLS group. Of those receiving treatment for RLS, 71% were + RLS vs 46% – RLS with a P value of .01. Symptoms of RLS after cord injury were 89% + RLS vs 55% – RLS with a P value of .03.

Discussion

This study represents one of the first studies to determine the prevalence of RLS in veterans with spinal cord disease. Research in this area is important to raise awareness of RLS among the veteran population with and without SCI and disorders. Restless legs syndrome often escapes diagnosis because of difficulty understanding the patient’s descriptions of their sensations. In addition, RLS may cause debilitating symptoms of sleep deprivation, daytime sleepiness, discomfort, and fatigue, which often results in decreased quality of life (QOL). Proper screening and treatment may improve QOL.

A study by Kumru and colleagues showed a similar rate of RLS in patients with SCI and RLS symptoms presented in the first year after SCI as did this study (18% vs 19%, respectively).⁴ In that study, RLS was more common in patients with lesions in lumbosacral area. Kumru and colleagues also showed that a dopaminergic medication improved symptoms of RLS in this population, whereas this study did not explore treatment outcomes.⁴

The pathogenesis of RLS is not fully known, but hereditary factors, iron metabolism, and the brain dopaminergic system are thought to be involved.¹¹ It is hypothesized that spinal cord lesions allow the appearance of RLS symptoms and spinal leg movement generator by blocking descending inhibitory spinal pathways.¹² One hypothesis is that damage to A11 nuclei (the main source of dopamine in the spinal cord or its diencephalospinal tract in animals) causes hyperexcitability of the spinal cord and leads to PLM and RLS symptoms.¹³ As the axons of A11 nuclei are present along the whole span of the spinal cord, SCI/D in patients with RLS might interrupt this dopaminergic tract and produce the RLS symptoms.

Limitations

This study included only veterans, so the prevalence may not apply to the nonveteran SCI population. Also, the population mainly was male, and there was no accurate information on race. Ferritin levels of the patients were not checked and is a major factor in RLS. The reported onset of RLS after the SCI could be due to recall bias.

Conclusion

The prevalence of RLS in veterans with SCI is above that reported in the general population (19% vs 10%, respectively). Furthermore, those with RLS have symptoms that often started after the SCI (suggesting causality) and required therapy due to their level of RLS symptom burden. A spectrum of severity of symptoms is present among those with RLS, with 83% having moderate-to-severe RLS affecting their QOL.

Although there was not a statistically significant relationship between RLS and spinal cord lesion level, there was a slightly higher prevalence of RLS at the cervical and thoracic levels, which may be relevant for future studies. There was no difference found between the RLS subgroups with respect to the location of the lesion within the spinal cord; however, a larger sample size may be needed to determine whether this would reach statistical significance. Prompt search for symptoms of RLS in veterans with SCI is warranted to provide adequate treatment to improve sleep health and QOL in this population.

Spinal cord injuries (SCI) are common in veteran populations.¹ Veterans with spinal cord injuries and disorders (SCI/D) also may have concurrent sleep disturbances. Spinal cord injury typically causes spasticity.^2,3 Hypersensitivity of the flexor reflex pathways is believed to cause painful muscle spasms in patients with SCI.⁴ Neuropathic pain at or below the level of the lesion also is common.

Restless legs syndrome (RLS) is a common sleep disorder that affects sleep quality and can occur concomitantly with spinal cord lesions.⁵ In about 80% of RLS cases, involuntary movements of legs across hip, knee, and ankle joints during sleep, known as periodic limb movement during sleep (PLMS), occurs.⁶ Several studies showed increased prevalence of PLMS in patients with SCI, and some case reports suggest an increased prevalence of RLS in this population.^7,8 One small study showed that 100% of patients with SCI had symptoms of RLS.⁶ Another study found that SCI could trigger PLMS.⁸

The pathophysiology of RLS and PLMS in patients with SCI is not fully understood, but case reports describing PLM in SCI patients points to a possible role of central pattern generators and the flexor reflex afferents in the pathophysiology of PLMS.^9,10 Changes of the tissue microstructure in the midbrain and upper cervical spinal cord have been described in patients with RLS.¹¹The objective of this study was to assess the prevalence of RLS in a veteran population with SCI/D and to determine possible neuroanatomical patterns involved in RLS and SCI/D.

Methods

The institutional review and ethical approval boards of the Minneapolis VA Health Care System approved the study. Within the VA system, 666 patients with SCI/D were identified using a national database. Of the 666 people, 316 were excluded, 199 were included, and 151 were deceased.

Patients aged between 18 and 65 years were included in the study. Charts of patients who had been discharged with the diagnosis of SCI from 2002 to 2008 were studied. All patients met the inclusion criteria of the International Restless Legs Syndrome Study Group diagnosis.

Exclusion criteria were as follows: Patients with evidence of brain pathology (eg, stroke), concurrent neurologic condition associated with RLS (Parkinson disease, spinocerebellar ataxia, peripheral neuropathy), concurrent psychiatric condition within the setting of treatment with dopamine antagonists, secondary causes of RLS (renal failure/uremia, iron deficiency, rheumatoid arthritis, and pregnancy) and a recent history of alcohol or drug misuse or current evidence of substance use of < 1 year.

A patient list was compiled that included the etiology of the SCI (vascular injury, multiple sclerosis [MS], trauma, unknown, and other), the level(s) and completeness of the SCI per radiology report, RLS pharmacotherapies, and pertinent medical history.

Axial T2-weighted images on magnetic resonance imaging (MRI) scans were retrospectively reviewed. Sagittal T1/T2-weighted and axial T2-weighted sequences were performed routinely on all patients with spinal cord lesions. The analysis included the extension of the lesion on both sagittal and axial distributions. The anatomic location of the cord lesion was categorized by the following: (1) pure gray matter (central cord); (2) white matter (dorsal [D], dorsolateral [DL], ventral [V], ventrolateral areas [VL]).

A questionnaire using standard diagnostic criteria for RLS was mailed to the 199 patients who met the inclusion criteria (Appendix A).

All analyses were carried out using StataCorp STATA 13 (College Station, TX). Descriptive statistics were used. The analyses were carried out using chi-square and Fisher exact tests. Differences between the groups were considered statistically significant at P < .05. The data were analyzed to obtain point prevalence among patients with SCI, and comparisons were made among the different subgroups.

Results

Of the 162 patients who chose to participate in the study, the sleep specialists confirmed 31 (19%) to have RLS, 112 (69%) were confirmed negative for RLS, and an additional 19 (12%) screened positive for RLS but were not confirmed to have RLS by the sleep specialists (Figure 1).

The etiology of SCI was subdivided into 4 groups: MS, trauma, vascular, and other/unknown. Within each group (– RLS vs + RLS), MS and trauma were the most common etiologies with 55% MS and 36% trauma in the + RLS group.

When comparing RLS among the spinal cord levels (cervical, thoracic, lumbar and cervical + thoracic), only the cervical + thoracic subgroup (18% + RLS vs 5% – RLS) showed a significant difference (Figure 2).

There was no significant difference found with the prevalence of RLS in the axial plane of the spinal cord lesions (ventral/ventro-lateral/central cord vs dorsal/dorsolateral) or by the completeness of spinal cord lesions, P = .76. There was a higher prevalence of incomplete cord injury, however, within each subgroup of RLS.

The Mann-Whitney test was used to analyze the burden of disease in both groups (+ RLS vs – RLS). Moderate level of burden was most frequently reported with a higher prevalence within the + RLS group. Of those receiving treatment for RLS, 71% were + RLS vs 46% – RLS with a P value of .01. Symptoms of RLS after cord injury were 89% + RLS vs 55% – RLS with a P value of .03.

Discussion

This study represents one of the first studies to determine the prevalence of RLS in veterans with spinal cord disease. Research in this area is important to raise awareness of RLS among the veteran population with and without SCI and disorders. Restless legs syndrome often escapes diagnosis because of difficulty understanding the patient’s descriptions of their sensations. In addition, RLS may cause debilitating symptoms of sleep deprivation, daytime sleepiness, discomfort, and fatigue, which often results in decreased quality of life (QOL). Proper screening and treatment may improve QOL.

A study by Kumru and colleagues showed a similar rate of RLS in patients with SCI and RLS symptoms presented in the first year after SCI as did this study (18% vs 19%, respectively).⁴ In that study, RLS was more common in patients with lesions in lumbosacral area. Kumru and colleagues also showed that a dopaminergic medication improved symptoms of RLS in this population, whereas this study did not explore treatment outcomes.⁴

The pathogenesis of RLS is not fully known, but hereditary factors, iron metabolism, and the brain dopaminergic system are thought to be involved.¹¹ It is hypothesized that spinal cord lesions allow the appearance of RLS symptoms and spinal leg movement generator by blocking descending inhibitory spinal pathways.¹² One hypothesis is that damage to A11 nuclei (the main source of dopamine in the spinal cord or its diencephalospinal tract in animals) causes hyperexcitability of the spinal cord and leads to PLM and RLS symptoms.¹³ As the axons of A11 nuclei are present along the whole span of the spinal cord, SCI/D in patients with RLS might interrupt this dopaminergic tract and produce the RLS symptoms.

Limitations

This study included only veterans, so the prevalence may not apply to the nonveteran SCI population. Also, the population mainly was male, and there was no accurate information on race. Ferritin levels of the patients were not checked and is a major factor in RLS. The reported onset of RLS after the SCI could be due to recall bias.

Conclusion

The prevalence of RLS in veterans with SCI is above that reported in the general population (19% vs 10%, respectively). Furthermore, those with RLS have symptoms that often started after the SCI (suggesting causality) and required therapy due to their level of RLS symptom burden. A spectrum of severity of symptoms is present among those with RLS, with 83% having moderate-to-severe RLS affecting their QOL.

Although there was not a statistically significant relationship between RLS and spinal cord lesion level, there was a slightly higher prevalence of RLS at the cervical and thoracic levels, which may be relevant for future studies. There was no difference found between the RLS subgroups with respect to the location of the lesion within the spinal cord; however, a larger sample size may be needed to determine whether this would reach statistical significance. Prompt search for symptoms of RLS in veterans with SCI is warranted to provide adequate treatment to improve sleep health and QOL in this population.

Spinal cord injuries (SCI) are common in veteran populations.¹ Veterans with spinal cord injuries and disorders (SCI/D) also may have concurrent sleep disturbances. Spinal cord injury typically causes spasticity.^2,3 Hypersensitivity of the flexor reflex pathways is believed to cause painful muscle spasms in patients with SCI.⁴ Neuropathic pain at or below the level of the lesion also is common.

Restless legs syndrome (RLS) is a common sleep disorder that affects sleep quality and can occur concomitantly with spinal cord lesions.⁵ In about 80% of RLS cases, involuntary movements of legs across hip, knee, and ankle joints during sleep, known as periodic limb movement during sleep (PLMS), occurs.⁶ Several studies showed increased prevalence of PLMS in patients with SCI, and some case reports suggest an increased prevalence of RLS in this population.^7,8 One small study showed that 100% of patients with SCI had symptoms of RLS.⁶ Another study found that SCI could trigger PLMS.⁸

The pathophysiology of RLS and PLMS in patients with SCI is not fully understood, but case reports describing PLM in SCI patients points to a possible role of central pattern generators and the flexor reflex afferents in the pathophysiology of PLMS.^9,10 Changes of the tissue microstructure in the midbrain and upper cervical spinal cord have been described in patients with RLS.¹¹The objective of this study was to assess the prevalence of RLS in a veteran population with SCI/D and to determine possible neuroanatomical patterns involved in RLS and SCI/D.

Methods

The institutional review and ethical approval boards of the Minneapolis VA Health Care System approved the study. Within the VA system, 666 patients with SCI/D were identified using a national database. Of the 666 people, 316 were excluded, 199 were included, and 151 were deceased.

Patients aged between 18 and 65 years were included in the study. Charts of patients who had been discharged with the diagnosis of SCI from 2002 to 2008 were studied. All patients met the inclusion criteria of the International Restless Legs Syndrome Study Group diagnosis.

Exclusion criteria were as follows: Patients with evidence of brain pathology (eg, stroke), concurrent neurologic condition associated with RLS (Parkinson disease, spinocerebellar ataxia, peripheral neuropathy), concurrent psychiatric condition within the setting of treatment with dopamine antagonists, secondary causes of RLS (renal failure/uremia, iron deficiency, rheumatoid arthritis, and pregnancy) and a recent history of alcohol or drug misuse or current evidence of substance use of < 1 year.

A patient list was compiled that included the etiology of the SCI (vascular injury, multiple sclerosis [MS], trauma, unknown, and other), the level(s) and completeness of the SCI per radiology report, RLS pharmacotherapies, and pertinent medical history.

Axial T2-weighted images on magnetic resonance imaging (MRI) scans were retrospectively reviewed. Sagittal T1/T2-weighted and axial T2-weighted sequences were performed routinely on all patients with spinal cord lesions. The analysis included the extension of the lesion on both sagittal and axial distributions. The anatomic location of the cord lesion was categorized by the following: (1) pure gray matter (central cord); (2) white matter (dorsal [D], dorsolateral [DL], ventral [V], ventrolateral areas [VL]).

A questionnaire using standard diagnostic criteria for RLS was mailed to the 199 patients who met the inclusion criteria (Appendix A).

All analyses were carried out using StataCorp STATA 13 (College Station, TX). Descriptive statistics were used. The analyses were carried out using chi-square and Fisher exact tests. Differences between the groups were considered statistically significant at P < .05. The data were analyzed to obtain point prevalence among patients with SCI, and comparisons were made among the different subgroups.

Results

Of the 162 patients who chose to participate in the study, the sleep specialists confirmed 31 (19%) to have RLS, 112 (69%) were confirmed negative for RLS, and an additional 19 (12%) screened positive for RLS but were not confirmed to have RLS by the sleep specialists (Figure 1).

The etiology of SCI was subdivided into 4 groups: MS, trauma, vascular, and other/unknown. Within each group (– RLS vs + RLS), MS and trauma were the most common etiologies with 55% MS and 36% trauma in the + RLS group.

When comparing RLS among the spinal cord levels (cervical, thoracic, lumbar and cervical + thoracic), only the cervical + thoracic subgroup (18% + RLS vs 5% – RLS) showed a significant difference (Figure 2).

There was no significant difference found with the prevalence of RLS in the axial plane of the spinal cord lesions (ventral/ventro-lateral/central cord vs dorsal/dorsolateral) or by the completeness of spinal cord lesions, P = .76. There was a higher prevalence of incomplete cord injury, however, within each subgroup of RLS.

The Mann-Whitney test was used to analyze the burden of disease in both groups (+ RLS vs – RLS). Moderate level of burden was most frequently reported with a higher prevalence within the + RLS group. Of those receiving treatment for RLS, 71% were + RLS vs 46% – RLS with a P value of .01. Symptoms of RLS after cord injury were 89% + RLS vs 55% – RLS with a P value of .03.

Discussion

This study represents one of the first studies to determine the prevalence of RLS in veterans with spinal cord disease. Research in this area is important to raise awareness of RLS among the veteran population with and without SCI and disorders. Restless legs syndrome often escapes diagnosis because of difficulty understanding the patient’s descriptions of their sensations. In addition, RLS may cause debilitating symptoms of sleep deprivation, daytime sleepiness, discomfort, and fatigue, which often results in decreased quality of life (QOL). Proper screening and treatment may improve QOL.

A study by Kumru and colleagues showed a similar rate of RLS in patients with SCI and RLS symptoms presented in the first year after SCI as did this study (18% vs 19%, respectively).⁴ In that study, RLS was more common in patients with lesions in lumbosacral area. Kumru and colleagues also showed that a dopaminergic medication improved symptoms of RLS in this population, whereas this study did not explore treatment outcomes.⁴

The pathogenesis of RLS is not fully known, but hereditary factors, iron metabolism, and the brain dopaminergic system are thought to be involved.¹¹ It is hypothesized that spinal cord lesions allow the appearance of RLS symptoms and spinal leg movement generator by blocking descending inhibitory spinal pathways.¹² One hypothesis is that damage to A11 nuclei (the main source of dopamine in the spinal cord or its diencephalospinal tract in animals) causes hyperexcitability of the spinal cord and leads to PLM and RLS symptoms.¹³ As the axons of A11 nuclei are present along the whole span of the spinal cord, SCI/D in patients with RLS might interrupt this dopaminergic tract and produce the RLS symptoms.

Limitations

This study included only veterans, so the prevalence may not apply to the nonveteran SCI population. Also, the population mainly was male, and there was no accurate information on race. Ferritin levels of the patients were not checked and is a major factor in RLS. The reported onset of RLS after the SCI could be due to recall bias.

Conclusion

The prevalence of RLS in veterans with SCI is above that reported in the general population (19% vs 10%, respectively). Furthermore, those with RLS have symptoms that often started after the SCI (suggesting causality) and required therapy due to their level of RLS symptom burden. A spectrum of severity of symptoms is present among those with RLS, with 83% having moderate-to-severe RLS affecting their QOL.

Although there was not a statistically significant relationship between RLS and spinal cord lesion level, there was a slightly higher prevalence of RLS at the cervical and thoracic levels, which may be relevant for future studies. There was no difference found between the RLS subgroups with respect to the location of the lesion within the spinal cord; however, a larger sample size may be needed to determine whether this would reach statistical significance. Prompt search for symptoms of RLS in veterans with SCI is warranted to provide adequate treatment to improve sleep health and QOL in this population.

Sézary syndrome (SS) is an advanced leukemic form of cutaneous T-cell lymphoma (CTCL) that is characterized by generalized erythroderma and T-cell leukemia. Skin changes can include erythroderma, keratosis pilaris–like lesions, keratoderma, ectropion, alopecia, and nail changes.¹ Nail changes in SS patients frequently are overlooked and underreported; they vary greatly from patient to patient, and their incidence has not been widely evaluated in the literature.

In this retrospective study, we reviewed medical records from a previously collected CTCL clinic database at the University of Texas MD Anderson Cancer Center (Houston, Texas) and found nail abnormalities in 36 of 83 (43.4%) patients with a diagnosis of SS. Findings for 2 select cases are described in more detail; they were compared to prior case reports from the literature to establish a comprehensive list of nail irregularities that have been associated with SS.

Methods

We examined records from a previously collected CTCL clinic database at the University of Texas MD Anderson Cancer Center. This database was part of an institutional review board–approved protocol to prospectively collect data from patients with CTCL. Our search yielded 83 patients with SS who were seen between 2007 and 2014.

Results

Of the 83 cases reviewed from the CTCL database, 36 (43.4%) SS patients reported at least 1 nail abnormality on the fingernails or toenails. Patients ranged in age from 59 to 85 years and included 27 (75%) men and 9 (25%) women. Nail irregularities noted on physical examination are summarized in Table 1. More than half of the patients presented with nail thickening (58.3% [21/36]), dystrophy (55.6% [20/36]), or yellowing (55.6% [20/36]) of 1 or more nails. Other findings included 15 (41.7%) patients with subungual hyperkeratosis, 3 (8.3%) with Beau lines, and 1 (2.8%) with multiple oil spots consistent with salmon patches. Five (13.9%) patients had only 1 reported nail irregularity, and 1 (2.8%) patient had 6 irregularities. The average number of nail abnormalities per patient was 2.88 (range, 1–6). We selected 2 patients with extensive nail findings who represent the spectrum of nail findings in patients with SS.

Patient 1
A 71-year-old white man presented with a papular rash of 30 years’ duration. The eruption first occurred on the soles of the feet but progressed to generalized erythroderma. He was found to be colonized with methicillin-resistant Staphylococcus aureus. Over the next 9 months, the patient was diagnosed with SS at an outside institution and was treated with cyclophosphamide, hydroxydaunorubicin, vincristine, prednisone, gemcitabine, etoposide, methylprednisolone, cytarabine, cisplatin, topical steroids, and intravenous methotrexate with no apparent improvement. At presentation to our institution, physical examination revealed pruritus; alopecia; generalized lymphadenopathy; erythroderma; and irregular nail findings, including yellowing, thickened fingernails and toenails with subungual debris, and splinter hemorrhage (Figure 1). A thick plaque with perioral distribution as well as erosions on the face and feet were noted. The total body surface area (BSA) affected was 100% (patches, 91%; plaques, 9%).

At diagnosis at our institution, the patient’s white blood cell (WBC) count was 17,800/µL (reference range, 4000–11,000/µL), with 11% Sézary cells noted. Biopsy of a lymph node from the inguinal area indicated T-cell lymphoma with clonal T-cell receptor (TCR) β gene rearrangement. Biopsy of lesional skin in the right groin area showed an atypical T-cell lymphocytic infiltrate with a CD4:CD8 ratio of 2.9:1 and partial loss of CD7 expression, consistent with mycosis fungoides (MF)/SS stage IVA. At presentation to our institution, the WBC count was 12,700/µL with a neutrophil count of 47% (reference range, 42%–66%), lymphocyte count of 36% (reference range, 24%–44%), monocyte count of 4% (reference range, 2%–7%), platelet count of 427,000/µL (reference range, 150,000–350,000/µL), hemoglobin of 9.9 g/dL (reference range, 14.0–17.5 g/dL), and lactate dehydrogenase of 733 U/L (reference range, 135–214 U/L). Lymphocytes were positive for CD2, CD3, CD4, CD5, CD25, CD52, TCRα, TCRβ, and TCR VB17; partial for CD26; and negative for CD7, CD8, and CD57. At follow-up 1 month later, the CD4⁺CD26⁻ T-cell population was 56%, which was consistent with SS T-cell lymphoma.

Skin scrapings from the generalized keratoderma on the patient’s feet were positive for fungal hyphae under potassium hydroxide examination. Nail clippings showed compact keratin with periodic acid–Schiff–positive small yeast forms admixed with bacterial organisms, consistent with onychomycosis. At our institution, the patient received extracorporeal photopheresis, whirlpool therapy (a type of hydrotherapy), steroid wet wraps, and intravenous vancomycin for methicillin-resistant S aureus. He also received bexarotene, levothyroxine sodium, and fenofibrate. After antibiotics and 2 sessions of photopheresis, the total BSA improved from 100% to 33%. The feet and nails were treated with ciclopirox gel and terbinafine, but neither the keratoderma nor the nails improved.

Patient 2
An 84-year-old white man with B-cell chronic lymphocytic leukemia also was diagnosed with SS at an outside institution. One year later, he presented to our institution with mild pruritus and swelling of the lower left leg, which was diagnosed as deep vein thrombosis. There was bilateral scaling of the palms, with fissures present on the left palm. The fingernails showed dystrophy with Beau lines, and the toenails were dystrophic with onycholysis on the bilateral great toes (Figure 2). Patches were noted on most of the body, including the feet, with plaques limited to the hands; the total BSA affected was 80%. Flow cytometry showed an elevated Sézary cell count (CD4⁺CD26⁻) of 4700 cells/µL. Complete blood cell count with differential included a hemoglobin level of 11.4 g/dL, hematocrit level of 35.3% (reference range, 37%–47%), a platelet count of 217,000/µL, and a WBC count of 17,700/µL, of which 29% were neutrophils, 63% were lymphocytes, 6% were monocytes, and 2% were eosinophils. The lactate dehydrogenase level was elevated at 829 U/L. The patient had not been treated for chronic lymphocytic leukemia in the last 11 months due to adverse reactions to rituximab after 2 cycles of fludarabine, cyclophosphamide, and rituximab chemotherapy. First-line therapy for the patient was photopheresis for SS.

Comment

Nail changes are found in many cases of advanced-stage SS but rarely have been reported in the literature. A literature review of PubMed articles indexed for MEDLINE was conducted using the search terms Sézary, nail, onychodystrophy, cutaneous T-cell lymphoma, and CTCL. All results were reviewed for original reported cases of SS with at least 1 reported nail finding. A total of 7 reports^2-8 met these requirements with a total of 43 SS patients with reported nail findings, which are summarized in Tables 2 and 3.

Our findings are generally consistent with those previously described in the literature. Nail thickening, yellowing, subungual hyperkeratosis, dystrophy, and onycholysis are consistently some of the most common nail findings in patients with SS. In 2012, Martin and Duvic⁹ found that 52.9% (45/85) of SS patients with keratoderma on physical examination were positive for dermatophyte hyphae when skin scrapings were done under potassium hydroxide examination, a considerably greater incidence than in the general population (10%–20%). The nail changes seen in our SS patients were identical to those found in dermatophyte infections, including discoloration, subungual debris, nail thickening, onycholysis, and dystrophy.¹⁰ In patient 1, nail clippings were positive for onychomycosis, a common nail condition that is especially prevalent in older or immunocompromised patients.^9,10

Interestingly, findings not observed in the literature included salmon patches and Beau lines. Beau lines are horizontal depressions in the nail plate and often are indicative of temporary interruption of nail growth, such as due to an underlying disease process, severe illness, and/or chemotherapy.^11,12 In our review, patient 2 had clinical findings of Beau lines. Because the average time for fingernail regrowth is 3 to 6 months,¹³ it is reasonable to assume that physical findings associated with fludarabine, cyclophosphamide, and rituximab chemotherapy treatment would no longer be demonstrated 11 months after completion of therapy. On the other hand, paronychia was frequently observed by Damasco et al⁸ (63.2% [12/19] of their cases), yet it was not found in our database or the other literature reports we reviewed. Perhaps these differences are due to differences in patient populations and/or available therapies, lack of documentation, or small sample size and limited reports in the literature.

A common question is: Are the nail irregularities caused by the physical symptoms of advanced CTCL or by the underlying disease process in response to the atypical T cells? Erythroderma has been speculated to cause many of the clinical findings of nail abnormalities found in CTCL patients.^2,3 However, Fleming et al¹⁴ described an MF patient who experienced onychomadesis without erythroderma, which suggests that a different mechanism may cause these nail changes. The wide range of nail abnormalities in CTCL can cause problems with diagnosing the specific cause underlying the nail alteration.

To further complicate the issue, numerous therapies for CTCL also may cause nail changes, such as the previously described Beau lines. In 2010, Parmentier et al⁴ reported a patient with nail alterations that had been present for more than 1 year, with 9 of 10 fingernails demonstrating anonychia, onychomadesis, subungual distal hyperkeratosis, and onycholysis. In this case report, the authors were able to exclude phototherapy as the cause of onycholysis (visible separation of the nail plate from the nail bed) and other clinical nail findings in the SS patient based on the onset of nail changes prior to beginning psoralen plus UVA therapy and complete sparing of 1 finger.⁴ The findings in our patient 1, who had no history of psoralen plus UVA therapy at the time the irregular nail findings presented, supports this observation. Total skin electron beam therapy for MF also has been reported to cause temporary nail stasis and thus must be taken into account when considering nail changes in patients with MF/SS.¹⁵

A nail matrix biopsy may provide clues to the definitive cause of the clinically observed nail changes; however, this procedure typically is not performed due to patient concerns of postoperative complications including pain and nail dystrophy.¹⁶ Histopathology features were similar in reported nail biopsies of 2 SS patients.^3,4 Tosti et al³ reported that longitudinal biopsy showed a dense lymphocytic infiltrate of atypical lymphocytes with involuted nuclei and notable epidermotropism. Parmentier et al⁴ reported a longitudinal nail biopsy in an SS patient that presented with atypical lymphocytes, epidermotropism, and Pautrier microabscess formation. Immunostaining showed CD3 positivity within the distal nail matrix, nail bed, and hyponychium. One-third of the cells stained positive for CD4, while the majority stained positive for CD8. Most notably, the skin, nails, and blood showed identical clonal rearrangement of TCRγ.⁴ Nail matrix biopsies in MF patients rarely have been reported in the literature, but those that are available show similar features to those seen in SS patients. Harland et al¹⁷ summarized the findings of 4 case reports of CTCL patients that included nail biopsies by stating, “[a]ll histopathologic findings from nail biopsies showed a dense subepithelial infiltrate of lymphocytes with marked epitheliotropism.” These histopathologic abnormalities are akin to skin biopsies in MF patients, thus providing an essential link to the disease state of MF and the nail abnormalities found within SS patients.

Treatment of the nail problems found within SS is challenging due to limited research. Parmentier et al⁴ noted an SS patient who was treated with topical mechlorethamine applied directly to the nail. In this case, topical mechlorethamine was effective at treating onychomadesis, subungual distal hyperkeratosis, and onycholysis within 6 months.⁴ Another SS patient, who presented with thickening and yellowing of the nail, had reported a proximal nail plate that resolved after chemotherapy. The patient did not survive long enough to note complete improvement of the nail.³ In our study, patient 1 was treated with ciclopirox gel and terbinafine, which did not result in nail improvement. Nail treatments in SS patients have yet to show much improvement and thus need more research and focus in the literature.

Conclusion

Sézary syndrome is a rare CTCL that can present with clinical features that may be mistaken for other diseases. Nail abnormalities in SS patients may be related to fungal involvement, medical therapy, or the underlying disease process of SS. We report one of the largest populations of SS patients with specific reported nail abnormalities, thus expanding the possibilities of nail changes that accompany the disease. Continued research and studies involving SS can provide a better understanding of nail involvement and successful treatment of these clinical findings.

Sézary syndrome (SS) is an advanced leukemic form of cutaneous T-cell lymphoma (CTCL) that is characterized by generalized erythroderma and T-cell leukemia. Skin changes can include erythroderma, keratosis pilaris–like lesions, keratoderma, ectropion, alopecia, and nail changes.¹ Nail changes in SS patients frequently are overlooked and underreported; they vary greatly from patient to patient, and their incidence has not been widely evaluated in the literature.

In this retrospective study, we reviewed medical records from a previously collected CTCL clinic database at the University of Texas MD Anderson Cancer Center (Houston, Texas) and found nail abnormalities in 36 of 83 (43.4%) patients with a diagnosis of SS. Findings for 2 select cases are described in more detail; they were compared to prior case reports from the literature to establish a comprehensive list of nail irregularities that have been associated with SS.

Methods

We examined records from a previously collected CTCL clinic database at the University of Texas MD Anderson Cancer Center. This database was part of an institutional review board–approved protocol to prospectively collect data from patients with CTCL. Our search yielded 83 patients with SS who were seen between 2007 and 2014.

Results

Of the 83 cases reviewed from the CTCL database, 36 (43.4%) SS patients reported at least 1 nail abnormality on the fingernails or toenails. Patients ranged in age from 59 to 85 years and included 27 (75%) men and 9 (25%) women. Nail irregularities noted on physical examination are summarized in Table 1. More than half of the patients presented with nail thickening (58.3% [21/36]), dystrophy (55.6% [20/36]), or yellowing (55.6% [20/36]) of 1 or more nails. Other findings included 15 (41.7%) patients with subungual hyperkeratosis, 3 (8.3%) with Beau lines, and 1 (2.8%) with multiple oil spots consistent with salmon patches. Five (13.9%) patients had only 1 reported nail irregularity, and 1 (2.8%) patient had 6 irregularities. The average number of nail abnormalities per patient was 2.88 (range, 1–6). We selected 2 patients with extensive nail findings who represent the spectrum of nail findings in patients with SS.

Patient 1
A 71-year-old white man presented with a papular rash of 30 years’ duration. The eruption first occurred on the soles of the feet but progressed to generalized erythroderma. He was found to be colonized with methicillin-resistant Staphylococcus aureus. Over the next 9 months, the patient was diagnosed with SS at an outside institution and was treated with cyclophosphamide, hydroxydaunorubicin, vincristine, prednisone, gemcitabine, etoposide, methylprednisolone, cytarabine, cisplatin, topical steroids, and intravenous methotrexate with no apparent improvement. At presentation to our institution, physical examination revealed pruritus; alopecia; generalized lymphadenopathy; erythroderma; and irregular nail findings, including yellowing, thickened fingernails and toenails with subungual debris, and splinter hemorrhage (Figure 1). A thick plaque with perioral distribution as well as erosions on the face and feet were noted. The total body surface area (BSA) affected was 100% (patches, 91%; plaques, 9%).

At diagnosis at our institution, the patient’s white blood cell (WBC) count was 17,800/µL (reference range, 4000–11,000/µL), with 11% Sézary cells noted. Biopsy of a lymph node from the inguinal area indicated T-cell lymphoma with clonal T-cell receptor (TCR) β gene rearrangement. Biopsy of lesional skin in the right groin area showed an atypical T-cell lymphocytic infiltrate with a CD4:CD8 ratio of 2.9:1 and partial loss of CD7 expression, consistent with mycosis fungoides (MF)/SS stage IVA. At presentation to our institution, the WBC count was 12,700/µL with a neutrophil count of 47% (reference range, 42%–66%), lymphocyte count of 36% (reference range, 24%–44%), monocyte count of 4% (reference range, 2%–7%), platelet count of 427,000/µL (reference range, 150,000–350,000/µL), hemoglobin of 9.9 g/dL (reference range, 14.0–17.5 g/dL), and lactate dehydrogenase of 733 U/L (reference range, 135–214 U/L). Lymphocytes were positive for CD2, CD3, CD4, CD5, CD25, CD52, TCRα, TCRβ, and TCR VB17; partial for CD26; and negative for CD7, CD8, and CD57. At follow-up 1 month later, the CD4⁺CD26⁻ T-cell population was 56%, which was consistent with SS T-cell lymphoma.

Skin scrapings from the generalized keratoderma on the patient’s feet were positive for fungal hyphae under potassium hydroxide examination. Nail clippings showed compact keratin with periodic acid–Schiff–positive small yeast forms admixed with bacterial organisms, consistent with onychomycosis. At our institution, the patient received extracorporeal photopheresis, whirlpool therapy (a type of hydrotherapy), steroid wet wraps, and intravenous vancomycin for methicillin-resistant S aureus. He also received bexarotene, levothyroxine sodium, and fenofibrate. After antibiotics and 2 sessions of photopheresis, the total BSA improved from 100% to 33%. The feet and nails were treated with ciclopirox gel and terbinafine, but neither the keratoderma nor the nails improved.

Patient 2
An 84-year-old white man with B-cell chronic lymphocytic leukemia also was diagnosed with SS at an outside institution. One year later, he presented to our institution with mild pruritus and swelling of the lower left leg, which was diagnosed as deep vein thrombosis. There was bilateral scaling of the palms, with fissures present on the left palm. The fingernails showed dystrophy with Beau lines, and the toenails were dystrophic with onycholysis on the bilateral great toes (Figure 2). Patches were noted on most of the body, including the feet, with plaques limited to the hands; the total BSA affected was 80%. Flow cytometry showed an elevated Sézary cell count (CD4⁺CD26⁻) of 4700 cells/µL. Complete blood cell count with differential included a hemoglobin level of 11.4 g/dL, hematocrit level of 35.3% (reference range, 37%–47%), a platelet count of 217,000/µL, and a WBC count of 17,700/µL, of which 29% were neutrophils, 63% were lymphocytes, 6% were monocytes, and 2% were eosinophils. The lactate dehydrogenase level was elevated at 829 U/L. The patient had not been treated for chronic lymphocytic leukemia in the last 11 months due to adverse reactions to rituximab after 2 cycles of fludarabine, cyclophosphamide, and rituximab chemotherapy. First-line therapy for the patient was photopheresis for SS.

Comment

Nail changes are found in many cases of advanced-stage SS but rarely have been reported in the literature. A literature review of PubMed articles indexed for MEDLINE was conducted using the search terms Sézary, nail, onychodystrophy, cutaneous T-cell lymphoma, and CTCL. All results were reviewed for original reported cases of SS with at least 1 reported nail finding. A total of 7 reports^2-8 met these requirements with a total of 43 SS patients with reported nail findings, which are summarized in Tables 2 and 3.

Our findings are generally consistent with those previously described in the literature. Nail thickening, yellowing, subungual hyperkeratosis, dystrophy, and onycholysis are consistently some of the most common nail findings in patients with SS. In 2012, Martin and Duvic⁹ found that 52.9% (45/85) of SS patients with keratoderma on physical examination were positive for dermatophyte hyphae when skin scrapings were done under potassium hydroxide examination, a considerably greater incidence than in the general population (10%–20%). The nail changes seen in our SS patients were identical to those found in dermatophyte infections, including discoloration, subungual debris, nail thickening, onycholysis, and dystrophy.¹⁰ In patient 1, nail clippings were positive for onychomycosis, a common nail condition that is especially prevalent in older or immunocompromised patients.^9,10

Interestingly, findings not observed in the literature included salmon patches and Beau lines. Beau lines are horizontal depressions in the nail plate and often are indicative of temporary interruption of nail growth, such as due to an underlying disease process, severe illness, and/or chemotherapy.^11,12 In our review, patient 2 had clinical findings of Beau lines. Because the average time for fingernail regrowth is 3 to 6 months,¹³ it is reasonable to assume that physical findings associated with fludarabine, cyclophosphamide, and rituximab chemotherapy treatment would no longer be demonstrated 11 months after completion of therapy. On the other hand, paronychia was frequently observed by Damasco et al⁸ (63.2% [12/19] of their cases), yet it was not found in our database or the other literature reports we reviewed. Perhaps these differences are due to differences in patient populations and/or available therapies, lack of documentation, or small sample size and limited reports in the literature.

A common question is: Are the nail irregularities caused by the physical symptoms of advanced CTCL or by the underlying disease process in response to the atypical T cells? Erythroderma has been speculated to cause many of the clinical findings of nail abnormalities found in CTCL patients.^2,3 However, Fleming et al¹⁴ described an MF patient who experienced onychomadesis without erythroderma, which suggests that a different mechanism may cause these nail changes. The wide range of nail abnormalities in CTCL can cause problems with diagnosing the specific cause underlying the nail alteration.

To further complicate the issue, numerous therapies for CTCL also may cause nail changes, such as the previously described Beau lines. In 2010, Parmentier et al⁴ reported a patient with nail alterations that had been present for more than 1 year, with 9 of 10 fingernails demonstrating anonychia, onychomadesis, subungual distal hyperkeratosis, and onycholysis. In this case report, the authors were able to exclude phototherapy as the cause of onycholysis (visible separation of the nail plate from the nail bed) and other clinical nail findings in the SS patient based on the onset of nail changes prior to beginning psoralen plus UVA therapy and complete sparing of 1 finger.⁴ The findings in our patient 1, who had no history of psoralen plus UVA therapy at the time the irregular nail findings presented, supports this observation. Total skin electron beam therapy for MF also has been reported to cause temporary nail stasis and thus must be taken into account when considering nail changes in patients with MF/SS.¹⁵

A nail matrix biopsy may provide clues to the definitive cause of the clinically observed nail changes; however, this procedure typically is not performed due to patient concerns of postoperative complications including pain and nail dystrophy.¹⁶ Histopathology features were similar in reported nail biopsies of 2 SS patients.^3,4 Tosti et al³ reported that longitudinal biopsy showed a dense lymphocytic infiltrate of atypical lymphocytes with involuted nuclei and notable epidermotropism. Parmentier et al⁴ reported a longitudinal nail biopsy in an SS patient that presented with atypical lymphocytes, epidermotropism, and Pautrier microabscess formation. Immunostaining showed CD3 positivity within the distal nail matrix, nail bed, and hyponychium. One-third of the cells stained positive for CD4, while the majority stained positive for CD8. Most notably, the skin, nails, and blood showed identical clonal rearrangement of TCRγ.⁴ Nail matrix biopsies in MF patients rarely have been reported in the literature, but those that are available show similar features to those seen in SS patients. Harland et al¹⁷ summarized the findings of 4 case reports of CTCL patients that included nail biopsies by stating, “[a]ll histopathologic findings from nail biopsies showed a dense subepithelial infiltrate of lymphocytes with marked epitheliotropism.” These histopathologic abnormalities are akin to skin biopsies in MF patients, thus providing an essential link to the disease state of MF and the nail abnormalities found within SS patients.

Treatment of the nail problems found within SS is challenging due to limited research. Parmentier et al⁴ noted an SS patient who was treated with topical mechlorethamine applied directly to the nail. In this case, topical mechlorethamine was effective at treating onychomadesis, subungual distal hyperkeratosis, and onycholysis within 6 months.⁴ Another SS patient, who presented with thickening and yellowing of the nail, had reported a proximal nail plate that resolved after chemotherapy. The patient did not survive long enough to note complete improvement of the nail.³ In our study, patient 1 was treated with ciclopirox gel and terbinafine, which did not result in nail improvement. Nail treatments in SS patients have yet to show much improvement and thus need more research and focus in the literature.

Conclusion

Sézary syndrome is a rare CTCL that can present with clinical features that may be mistaken for other diseases. Nail abnormalities in SS patients may be related to fungal involvement, medical therapy, or the underlying disease process of SS. We report one of the largest populations of SS patients with specific reported nail abnormalities, thus expanding the possibilities of nail changes that accompany the disease. Continued research and studies involving SS can provide a better understanding of nail involvement and successful treatment of these clinical findings.

Sézary syndrome (SS) is an advanced leukemic form of cutaneous T-cell lymphoma (CTCL) that is characterized by generalized erythroderma and T-cell leukemia. Skin changes can include erythroderma, keratosis pilaris–like lesions, keratoderma, ectropion, alopecia, and nail changes.¹ Nail changes in SS patients frequently are overlooked and underreported; they vary greatly from patient to patient, and their incidence has not been widely evaluated in the literature.

In this retrospective study, we reviewed medical records from a previously collected CTCL clinic database at the University of Texas MD Anderson Cancer Center (Houston, Texas) and found nail abnormalities in 36 of 83 (43.4%) patients with a diagnosis of SS. Findings for 2 select cases are described in more detail; they were compared to prior case reports from the literature to establish a comprehensive list of nail irregularities that have been associated with SS.

Methods

We examined records from a previously collected CTCL clinic database at the University of Texas MD Anderson Cancer Center. This database was part of an institutional review board–approved protocol to prospectively collect data from patients with CTCL. Our search yielded 83 patients with SS who were seen between 2007 and 2014.

Results

Of the 83 cases reviewed from the CTCL database, 36 (43.4%) SS patients reported at least 1 nail abnormality on the fingernails or toenails. Patients ranged in age from 59 to 85 years and included 27 (75%) men and 9 (25%) women. Nail irregularities noted on physical examination are summarized in Table 1. More than half of the patients presented with nail thickening (58.3% [21/36]), dystrophy (55.6% [20/36]), or yellowing (55.6% [20/36]) of 1 or more nails. Other findings included 15 (41.7%) patients with subungual hyperkeratosis, 3 (8.3%) with Beau lines, and 1 (2.8%) with multiple oil spots consistent with salmon patches. Five (13.9%) patients had only 1 reported nail irregularity, and 1 (2.8%) patient had 6 irregularities. The average number of nail abnormalities per patient was 2.88 (range, 1–6). We selected 2 patients with extensive nail findings who represent the spectrum of nail findings in patients with SS.

Patient 1
A 71-year-old white man presented with a papular rash of 30 years’ duration. The eruption first occurred on the soles of the feet but progressed to generalized erythroderma. He was found to be colonized with methicillin-resistant Staphylococcus aureus. Over the next 9 months, the patient was diagnosed with SS at an outside institution and was treated with cyclophosphamide, hydroxydaunorubicin, vincristine, prednisone, gemcitabine, etoposide, methylprednisolone, cytarabine, cisplatin, topical steroids, and intravenous methotrexate with no apparent improvement. At presentation to our institution, physical examination revealed pruritus; alopecia; generalized lymphadenopathy; erythroderma; and irregular nail findings, including yellowing, thickened fingernails and toenails with subungual debris, and splinter hemorrhage (Figure 1). A thick plaque with perioral distribution as well as erosions on the face and feet were noted. The total body surface area (BSA) affected was 100% (patches, 91%; plaques, 9%).

At diagnosis at our institution, the patient’s white blood cell (WBC) count was 17,800/µL (reference range, 4000–11,000/µL), with 11% Sézary cells noted. Biopsy of a lymph node from the inguinal area indicated T-cell lymphoma with clonal T-cell receptor (TCR) β gene rearrangement. Biopsy of lesional skin in the right groin area showed an atypical T-cell lymphocytic infiltrate with a CD4:CD8 ratio of 2.9:1 and partial loss of CD7 expression, consistent with mycosis fungoides (MF)/SS stage IVA. At presentation to our institution, the WBC count was 12,700/µL with a neutrophil count of 47% (reference range, 42%–66%), lymphocyte count of 36% (reference range, 24%–44%), monocyte count of 4% (reference range, 2%–7%), platelet count of 427,000/µL (reference range, 150,000–350,000/µL), hemoglobin of 9.9 g/dL (reference range, 14.0–17.5 g/dL), and lactate dehydrogenase of 733 U/L (reference range, 135–214 U/L). Lymphocytes were positive for CD2, CD3, CD4, CD5, CD25, CD52, TCRα, TCRβ, and TCR VB17; partial for CD26; and negative for CD7, CD8, and CD57. At follow-up 1 month later, the CD4⁺CD26⁻ T-cell population was 56%, which was consistent with SS T-cell lymphoma.

Skin scrapings from the generalized keratoderma on the patient’s feet were positive for fungal hyphae under potassium hydroxide examination. Nail clippings showed compact keratin with periodic acid–Schiff–positive small yeast forms admixed with bacterial organisms, consistent with onychomycosis. At our institution, the patient received extracorporeal photopheresis, whirlpool therapy (a type of hydrotherapy), steroid wet wraps, and intravenous vancomycin for methicillin-resistant S aureus. He also received bexarotene, levothyroxine sodium, and fenofibrate. After antibiotics and 2 sessions of photopheresis, the total BSA improved from 100% to 33%. The feet and nails were treated with ciclopirox gel and terbinafine, but neither the keratoderma nor the nails improved.

Patient 2
An 84-year-old white man with B-cell chronic lymphocytic leukemia also was diagnosed with SS at an outside institution. One year later, he presented to our institution with mild pruritus and swelling of the lower left leg, which was diagnosed as deep vein thrombosis. There was bilateral scaling of the palms, with fissures present on the left palm. The fingernails showed dystrophy with Beau lines, and the toenails were dystrophic with onycholysis on the bilateral great toes (Figure 2). Patches were noted on most of the body, including the feet, with plaques limited to the hands; the total BSA affected was 80%. Flow cytometry showed an elevated Sézary cell count (CD4⁺CD26⁻) of 4700 cells/µL. Complete blood cell count with differential included a hemoglobin level of 11.4 g/dL, hematocrit level of 35.3% (reference range, 37%–47%), a platelet count of 217,000/µL, and a WBC count of 17,700/µL, of which 29% were neutrophils, 63% were lymphocytes, 6% were monocytes, and 2% were eosinophils. The lactate dehydrogenase level was elevated at 829 U/L. The patient had not been treated for chronic lymphocytic leukemia in the last 11 months due to adverse reactions to rituximab after 2 cycles of fludarabine, cyclophosphamide, and rituximab chemotherapy. First-line therapy for the patient was photopheresis for SS.

Comment

Nail changes are found in many cases of advanced-stage SS but rarely have been reported in the literature. A literature review of PubMed articles indexed for MEDLINE was conducted using the search terms Sézary, nail, onychodystrophy, cutaneous T-cell lymphoma, and CTCL. All results were reviewed for original reported cases of SS with at least 1 reported nail finding. A total of 7 reports^2-8 met these requirements with a total of 43 SS patients with reported nail findings, which are summarized in Tables 2 and 3.

Our findings are generally consistent with those previously described in the literature. Nail thickening, yellowing, subungual hyperkeratosis, dystrophy, and onycholysis are consistently some of the most common nail findings in patients with SS. In 2012, Martin and Duvic⁹ found that 52.9% (45/85) of SS patients with keratoderma on physical examination were positive for dermatophyte hyphae when skin scrapings were done under potassium hydroxide examination, a considerably greater incidence than in the general population (10%–20%). The nail changes seen in our SS patients were identical to those found in dermatophyte infections, including discoloration, subungual debris, nail thickening, onycholysis, and dystrophy.¹⁰ In patient 1, nail clippings were positive for onychomycosis, a common nail condition that is especially prevalent in older or immunocompromised patients.^9,10

Interestingly, findings not observed in the literature included salmon patches and Beau lines. Beau lines are horizontal depressions in the nail plate and often are indicative of temporary interruption of nail growth, such as due to an underlying disease process, severe illness, and/or chemotherapy.^11,12 In our review, patient 2 had clinical findings of Beau lines. Because the average time for fingernail regrowth is 3 to 6 months,¹³ it is reasonable to assume that physical findings associated with fludarabine, cyclophosphamide, and rituximab chemotherapy treatment would no longer be demonstrated 11 months after completion of therapy. On the other hand, paronychia was frequently observed by Damasco et al⁸ (63.2% [12/19] of their cases), yet it was not found in our database or the other literature reports we reviewed. Perhaps these differences are due to differences in patient populations and/or available therapies, lack of documentation, or small sample size and limited reports in the literature.

A common question is: Are the nail irregularities caused by the physical symptoms of advanced CTCL or by the underlying disease process in response to the atypical T cells? Erythroderma has been speculated to cause many of the clinical findings of nail abnormalities found in CTCL patients.^2,3 However, Fleming et al¹⁴ described an MF patient who experienced onychomadesis without erythroderma, which suggests that a different mechanism may cause these nail changes. The wide range of nail abnormalities in CTCL can cause problems with diagnosing the specific cause underlying the nail alteration.

To further complicate the issue, numerous therapies for CTCL also may cause nail changes, such as the previously described Beau lines. In 2010, Parmentier et al⁴ reported a patient with nail alterations that had been present for more than 1 year, with 9 of 10 fingernails demonstrating anonychia, onychomadesis, subungual distal hyperkeratosis, and onycholysis. In this case report, the authors were able to exclude phototherapy as the cause of onycholysis (visible separation of the nail plate from the nail bed) and other clinical nail findings in the SS patient based on the onset of nail changes prior to beginning psoralen plus UVA therapy and complete sparing of 1 finger.⁴ The findings in our patient 1, who had no history of psoralen plus UVA therapy at the time the irregular nail findings presented, supports this observation. Total skin electron beam therapy for MF also has been reported to cause temporary nail stasis and thus must be taken into account when considering nail changes in patients with MF/SS.¹⁵

A nail matrix biopsy may provide clues to the definitive cause of the clinically observed nail changes; however, this procedure typically is not performed due to patient concerns of postoperative complications including pain and nail dystrophy.¹⁶ Histopathology features were similar in reported nail biopsies of 2 SS patients.^3,4 Tosti et al³ reported that longitudinal biopsy showed a dense lymphocytic infiltrate of atypical lymphocytes with involuted nuclei and notable epidermotropism. Parmentier et al⁴ reported a longitudinal nail biopsy in an SS patient that presented with atypical lymphocytes, epidermotropism, and Pautrier microabscess formation. Immunostaining showed CD3 positivity within the distal nail matrix, nail bed, and hyponychium. One-third of the cells stained positive for CD4, while the majority stained positive for CD8. Most notably, the skin, nails, and blood showed identical clonal rearrangement of TCRγ.⁴ Nail matrix biopsies in MF patients rarely have been reported in the literature, but those that are available show similar features to those seen in SS patients. Harland et al¹⁷ summarized the findings of 4 case reports of CTCL patients that included nail biopsies by stating, “[a]ll histopathologic findings from nail biopsies showed a dense subepithelial infiltrate of lymphocytes with marked epitheliotropism.” These histopathologic abnormalities are akin to skin biopsies in MF patients, thus providing an essential link to the disease state of MF and the nail abnormalities found within SS patients.

Treatment of the nail problems found within SS is challenging due to limited research. Parmentier et al⁴ noted an SS patient who was treated with topical mechlorethamine applied directly to the nail. In this case, topical mechlorethamine was effective at treating onychomadesis, subungual distal hyperkeratosis, and onycholysis within 6 months.⁴ Another SS patient, who presented with thickening and yellowing of the nail, had reported a proximal nail plate that resolved after chemotherapy. The patient did not survive long enough to note complete improvement of the nail.³ In our study, patient 1 was treated with ciclopirox gel and terbinafine, which did not result in nail improvement. Nail treatments in SS patients have yet to show much improvement and thus need more research and focus in the literature.

Conclusion

Sézary syndrome is a rare CTCL that can present with clinical features that may be mistaken for other diseases. Nail abnormalities in SS patients may be related to fungal involvement, medical therapy, or the underlying disease process of SS. We report one of the largest populations of SS patients with specific reported nail abnormalities, thus expanding the possibilities of nail changes that accompany the disease. Continued research and studies involving SS can provide a better understanding of nail involvement and successful treatment of these clinical findings.

In recent years, driven by accelerating health care costs and desire for improved health care value, major specialty group guidelines have incorporated resource utilization and value calculations into their recommendations. High-value care has the characteristics of enhancing outcomes, safety, and patient satisfaction at a reasonable cost. As one example, the American College of Cardiology (ACC) recently published a consensus statement on its clinical practice guidelines with a specific focus on cost and value.¹ This guideline acknowledges the difficulty in incorporating value into clinical decision making but stresses a need for increased transparency and consistency to boost value in everyday practice.

Chest pain and related symptoms were listed as the second leading principle reasons for emergency department visits in the US in 2011 with 14% of patients undergoing cardiac enzyme testing.² The ACC guidelines advocate use of troponin as the preferred laboratory test for the initial evaluation of acute coronary syndrome (ACS). Fractionated creatine kinase (CK-MB) is an acceptable alternative only when a cardiac troponin test is not available.³ Furthermore, troponins should be obtained no more than 3 times for the initial evaluation of a single event, and further trending provides no additional benefit or prognostic information.

A recent study from an academic hospital showed that process improvement interventions focused on eliminating unnecessary cardiac enzyme testing led to a 1-year cost savings of $1.25 million while increasing the rate of ACS diagnosis.⁴ Common clinical practice at Naval Medical Center Portsmouth (NMCP) in Virginia still routinely includes both troponin as well as a CK panel comprised of CK, CK-MB, and a calculated CK-MB/CK index. Our study focuses on the implementation of quality improvement efforts described by Larochelle and colleagues at NMCP.⁴ The study aimed to determine the impact of implementing interventions designed to improve the ordering practices and reduce the cost of cardiac enzyme testing.

Methods

The primary focus of the intervention was on ordering practices of the emergency medicine department (EMD), internal medicine (IM) inpatient services, and cardiology inpatient services. Specific interventions were: (1) removal of the CK panel from the chest pain order set in the EMD electronic health record (EHR); (2) removal of the CK panel from the inpatient cardiology order set; (3) education of staff on the changes in CK panel utility via direct communication during IM academic seminars; (4) education of nursing staff ordering laboratory results on behalf of physicians on the cardiology service at the morning and evening huddles; and (5) addition of “max of 3 tests indicated” comment to the inpatient EHR ordering page of the troponin test. Acknowledging that the CK-MB has some utility to interventional cardiologists in the setting of confirmed ACS, the laboratory instituted an automated, reflexive order of the CK-MB panel only if the troponin tests were positive. This test was automatically run on the same vial originally sent to the lab to mitigate any additional delay in determining results.

Data Source

The process improvement interventions were considered exempt from institutional review board (IRB) approval; however, we obtained expedited IRB approval with waiver of consent for the research aspect of the project. We obtained clinical administrative data from the Military Health System Data Repository (MDR). We identified all adult patients aged ≥ 18 years who had a troponin test, CK-MB, or both drawn at NMCP on the following services: the EMD, IM, and cardiology. A troponin or CK-MB test was defined using Current Procedural Terminology (CPT) codes and unique Logical Observation Identifiers Names and Codes (LOINC).

Measures

The study was divided into 3 periods: the preintervention period from August 1, 2013 to July 31, 2014; the intervention period from August 1, 2014 to January 31, 2015; and the postintervention period February 1, 2015 to January 31, 2016.

The primary outcomes measured were the frequency of guideline concordance and total costs for tests ordered per month using the Centers for Medicare and Medicaid Services (CMS) clinical laboratory fee schedule of $13.40 for troponin and $16.17 for CK-MB.⁵Concordance was defined as ≤ 3 troponin tests and no CK-MB tests ordered during 1 encounter for a patient without an ACS diagnosis in the preceding 7 days. Due to faster cellular release kinetics of CK-MB compared with that of troponin, this test has utility in evaluating new or worsening chest pain in the setting of a recent myocardial infarction (MI). Therefore, we excluded any patient who had a MI within the preceding 7 days of an order for either CK-MB or troponin tests. Additionally, the number of tests, both CK-MB and troponin, ordered per patient encounter (hereafter referred to as an episode) were measured. Finally, we measured the monthly prevalence of ACS diagnosis and percentage of visits having that diagnosis.

Data Analysis

Descriptive statistics were used to calculate population demographics of age group, sex, beneficiary category, sponsor service, and clinical setting. Monthly data were grouped into the preintervention and postintervention periods. The analysis was performed using t tests to compare mean values and CIs before and after the intervention. Simple linear regression with attention to correlation was used to create best fit lines with confidence bands before and after the intervention. Interrupted time series (ITS) regression was used to describe all data points throughout the study. Consistency between these various methods was verified. Mean values and CIs were reported from the t tests. Statistical significance was reported when appropriate. Equations and confidence predictions on the simple linear regressions were produced and reported. These were used to identify values at the start, midpoint, and end of the pre- and postintervention periods.

Results

There were a total of 6,281 patients in the study population. More patients were seen during the postintervention period than in the preintervention period. The mean age of patients was slightly higher during the preintervention period (Table 1).

Guideline Concordance

To determine whether ordering practices for cardiac enzyme testing improved, we assessed the changes in the frequency of guideline concordance during the pre- and postintervention period. On average during the preintervention year, the percentage of tests ordered that met guideline concordance was 10.1% (95% CI, 7.4%-12.9%), increasing by 0.80% (95% CI, 0.17%-1.42%) each month. 

Costs

We assessed changes in total dollars spent on cardiac enzyme testing during the pre- and postintervention periods. During the preintervention year, $9,400 (95% CI, $8,700-$10,100) was spent on average each month, which did not change significantly throughout the period. During the postintervention year, the cost was stable at $5,000 (95% CI, $4,600-$5,300) on average each month, a reduction of $4,400 (95% CI, $3,700-$5,100) (Figure 2).

CK-MB and Troponin Tests per Patient

To further assess ordering practices for cardiac enzyme testing, we compared the changes in the monthly number of tests and the average number of CK-MB and troponin tests ordered per episode pre- and postintervention. On average during the preintervention year, 297 tests (95% CI, 278-315) were run per month, with an average of 1.21 CK tests (95% CI, 1.15-1.27) per episode (Table 2, Figure 3). 

The changes in troponin testing were not as dramatic. The counts of tests each month remained similar, with a preintervention year average of 341 (95% CI, 306-377) and postintervention year average of 310 (95% CI, 287-332), which were not statistically different. However, there was a statistically significant decrease in the number of tests per episode. During the preintervention year, 1.38 troponin tests (95% CI, 1.31-1.45) were ordered per patient on average. This dropped by 0.17 (95% CI, 0.09-0.24) to the postintervention average of 1.21 (95% CI, 1.17-1.25) (Table 2, Figure 4). 

ACS Prevalence

To determine whether there was an impact on ACS diagnoses, we looked at the numbers of ACS diagnoses and their prevalence among visits before and after the intervention. During the preintervention year, the average monthly number of diagnoses was 29.7 (95% CI, 26.1-33.2), and prevalence of ACS was 0.56% (95% CI, 0.48%-0.63%) of all episodes. Although the monthly rate was statistically decreasing by 0.022% (95% CI, 0.003-0.41), this has little meaning since the level of correlation (r² = 0.2522, not displayed) was poor due to the essentially nonexistent correlation in number of visits each month (r² = 0.0112, not displayed). During the postintervention year, the average number of diagnoses was 32.2 (95% CI, 27.9-36.6), and the prevalence of ACS was 0.62% (95% CI, 0.54-0.65). Neither of these values changed significantly between the pre- and postintervention period. All ICD-9 and ICD-10 diagnosis codes used for the analysis are available upon request from the authors.

Discussion

Our data demonstrate the ability of simple process improvement interventions to decrease unnecessary testing in the workup of ACS, increasing the rate of guideline concordant testing by > 70% at a single military treatment facility (MTF). In particular, with the now widespread use of EHR, the order set presents a high-yield target for process improvement in an easily implemented, durable fashion. We had expected to see some decrease in the efficacy of the intervention at a time of staff turnover in the summer of 2015 because ongoing dedicated teaching sessions were not performed. Despite that, the intervention remained effective without further dedicated teaching sessions. This outcome was certainly attributable to the hardwired interventions made (mainly via order sets), but possibly indicates an institutional memory that can take hold after an initial concerted effort is made.

We reduced the estimated preintervention annual cost of $113,000 by $53,000 (95% CI, $42,000-$64,000). Although on a much smaller scale than the study by Larochelle, our study represents a nearly 50% reduction in the total cost of initial testing for possible ACS and a > 80% reduction in unnecessary CK-MB testing.⁴ This result was achieved with no statistical change in the prevalence of ACS. The cost reduction does not account for the labor costs to clinically follow-up and address additional unnecessary lab results. The estimated cost of intervention was limited to the time required to educate residents, interns, and nursing staff as well as the implementation of the automated, reflexive laboratory results ordering process.

Unique to our study, we also demonstrated an intervention that satisfied all the major stakeholders in the ordering of these laboratory results. By instituting the reflexive ordering of CK-MB tests for positive troponins, we obtained the support of the facility’s interventional cardiology department, which finds value in that data. Appreciating the time-sensitive nature of an ACS diagnosis, the reflexive ordering minimized the delay in receiving these data while still greatly reducing the number of tests performed. That being said, if the current trend away from CK-MB in favor of exclusively testing troponin continues, removing the reflexive ordering for positive laboratory results protocol would be an easy follow-on intervention.

Limitations

Our study presented several limitations. First, reporting errors due to improper or insufficient medical coding as well as data entry errors may exist within the MDR; therefore, the results of this analysis may be over- or underestimated. Specifically, CPT codes for troponin and CK-MB were available only in 1 of the 2 data sets used for this study, which primarily contains outpatient patient encounters. For this reason, most of the laboratory testing comes from the EMD rather than from inpatient services. However, because we excluded all patients who eventually had an ACS diagnosis (patients who likely had more inpatient time and better indication for repeat troponin), we feel that our intervention was still thoroughly investigated. Second, the number of tests drawn per patient was significantly < 2, the expected minimum number of tests to rule out ACS in patients with appropriate symptoms.

This study was not designed to answer the source of variation from guidelines. Many patients had only 1 test, which we feel represents an opportunity for future study to identify other ways cardiac enzyme testing is being used clinically. These tests might be used for patients without convincing symptoms and signs of coronary syndromes or for patients with other primary problems. Third, by using the ITS analysis, we assumed that the outcome during each intervention period follows a linear pattern. However, changes may follow a nonlinear pattern over a long period. Finally, our intervention was limited to only a single MTF, which may limit generalizability to other facilities across military medicine. However, we feel this study should serve as a guide for other MTFs as well as US Department of Veterans Affairs facilities that could institute similar process improvements.

Conclusion

We made easily implemented and durable process improvement interventions that changed institution-wide ordering practices. These changes dramatically increased the rate of guideline-concordant testing, decreasing cost and furthering the goal of high-value medical care.

In recent years, driven by accelerating health care costs and desire for improved health care value, major specialty group guidelines have incorporated resource utilization and value calculations into their recommendations. High-value care has the characteristics of enhancing outcomes, safety, and patient satisfaction at a reasonable cost. As one example, the American College of Cardiology (ACC) recently published a consensus statement on its clinical practice guidelines with a specific focus on cost and value.¹ This guideline acknowledges the difficulty in incorporating value into clinical decision making but stresses a need for increased transparency and consistency to boost value in everyday practice.

Chest pain and related symptoms were listed as the second leading principle reasons for emergency department visits in the US in 2011 with 14% of patients undergoing cardiac enzyme testing.² The ACC guidelines advocate use of troponin as the preferred laboratory test for the initial evaluation of acute coronary syndrome (ACS). Fractionated creatine kinase (CK-MB) is an acceptable alternative only when a cardiac troponin test is not available.³ Furthermore, troponins should be obtained no more than 3 times for the initial evaluation of a single event, and further trending provides no additional benefit or prognostic information.

A recent study from an academic hospital showed that process improvement interventions focused on eliminating unnecessary cardiac enzyme testing led to a 1-year cost savings of $1.25 million while increasing the rate of ACS diagnosis.⁴ Common clinical practice at Naval Medical Center Portsmouth (NMCP) in Virginia still routinely includes both troponin as well as a CK panel comprised of CK, CK-MB, and a calculated CK-MB/CK index. Our study focuses on the implementation of quality improvement efforts described by Larochelle and colleagues at NMCP.⁴ The study aimed to determine the impact of implementing interventions designed to improve the ordering practices and reduce the cost of cardiac enzyme testing.

Methods

The primary focus of the intervention was on ordering practices of the emergency medicine department (EMD), internal medicine (IM) inpatient services, and cardiology inpatient services. Specific interventions were: (1) removal of the CK panel from the chest pain order set in the EMD electronic health record (EHR); (2) removal of the CK panel from the inpatient cardiology order set; (3) education of staff on the changes in CK panel utility via direct communication during IM academic seminars; (4) education of nursing staff ordering laboratory results on behalf of physicians on the cardiology service at the morning and evening huddles; and (5) addition of “max of 3 tests indicated” comment to the inpatient EHR ordering page of the troponin test. Acknowledging that the CK-MB has some utility to interventional cardiologists in the setting of confirmed ACS, the laboratory instituted an automated, reflexive order of the CK-MB panel only if the troponin tests were positive. This test was automatically run on the same vial originally sent to the lab to mitigate any additional delay in determining results.

Data Source

The process improvement interventions were considered exempt from institutional review board (IRB) approval; however, we obtained expedited IRB approval with waiver of consent for the research aspect of the project. We obtained clinical administrative data from the Military Health System Data Repository (MDR). We identified all adult patients aged ≥ 18 years who had a troponin test, CK-MB, or both drawn at NMCP on the following services: the EMD, IM, and cardiology. A troponin or CK-MB test was defined using Current Procedural Terminology (CPT) codes and unique Logical Observation Identifiers Names and Codes (LOINC).

Measures

The study was divided into 3 periods: the preintervention period from August 1, 2013 to July 31, 2014; the intervention period from August 1, 2014 to January 31, 2015; and the postintervention period February 1, 2015 to January 31, 2016.

The primary outcomes measured were the frequency of guideline concordance and total costs for tests ordered per month using the Centers for Medicare and Medicaid Services (CMS) clinical laboratory fee schedule of $13.40 for troponin and $16.17 for CK-MB.⁵Concordance was defined as ≤ 3 troponin tests and no CK-MB tests ordered during 1 encounter for a patient without an ACS diagnosis in the preceding 7 days. Due to faster cellular release kinetics of CK-MB compared with that of troponin, this test has utility in evaluating new or worsening chest pain in the setting of a recent myocardial infarction (MI). Therefore, we excluded any patient who had a MI within the preceding 7 days of an order for either CK-MB or troponin tests. Additionally, the number of tests, both CK-MB and troponin, ordered per patient encounter (hereafter referred to as an episode) were measured. Finally, we measured the monthly prevalence of ACS diagnosis and percentage of visits having that diagnosis.

Data Analysis

Descriptive statistics were used to calculate population demographics of age group, sex, beneficiary category, sponsor service, and clinical setting. Monthly data were grouped into the preintervention and postintervention periods. The analysis was performed using t tests to compare mean values and CIs before and after the intervention. Simple linear regression with attention to correlation was used to create best fit lines with confidence bands before and after the intervention. Interrupted time series (ITS) regression was used to describe all data points throughout the study. Consistency between these various methods was verified. Mean values and CIs were reported from the t tests. Statistical significance was reported when appropriate. Equations and confidence predictions on the simple linear regressions were produced and reported. These were used to identify values at the start, midpoint, and end of the pre- and postintervention periods.

Results

There were a total of 6,281 patients in the study population. More patients were seen during the postintervention period than in the preintervention period. The mean age of patients was slightly higher during the preintervention period (Table 1).

Guideline Concordance

To determine whether ordering practices for cardiac enzyme testing improved, we assessed the changes in the frequency of guideline concordance during the pre- and postintervention period. On average during the preintervention year, the percentage of tests ordered that met guideline concordance was 10.1% (95% CI, 7.4%-12.9%), increasing by 0.80% (95% CI, 0.17%-1.42%) each month. 

Costs

We assessed changes in total dollars spent on cardiac enzyme testing during the pre- and postintervention periods. During the preintervention year, $9,400 (95% CI, $8,700-$10,100) was spent on average each month, which did not change significantly throughout the period. During the postintervention year, the cost was stable at $5,000 (95% CI, $4,600-$5,300) on average each month, a reduction of $4,400 (95% CI, $3,700-$5,100) (Figure 2).

CK-MB and Troponin Tests per Patient

To further assess ordering practices for cardiac enzyme testing, we compared the changes in the monthly number of tests and the average number of CK-MB and troponin tests ordered per episode pre- and postintervention. On average during the preintervention year, 297 tests (95% CI, 278-315) were run per month, with an average of 1.21 CK tests (95% CI, 1.15-1.27) per episode (Table 2, Figure 3). 

The changes in troponin testing were not as dramatic. The counts of tests each month remained similar, with a preintervention year average of 341 (95% CI, 306-377) and postintervention year average of 310 (95% CI, 287-332), which were not statistically different. However, there was a statistically significant decrease in the number of tests per episode. During the preintervention year, 1.38 troponin tests (95% CI, 1.31-1.45) were ordered per patient on average. This dropped by 0.17 (95% CI, 0.09-0.24) to the postintervention average of 1.21 (95% CI, 1.17-1.25) (Table 2, Figure 4). 

ACS Prevalence

To determine whether there was an impact on ACS diagnoses, we looked at the numbers of ACS diagnoses and their prevalence among visits before and after the intervention. During the preintervention year, the average monthly number of diagnoses was 29.7 (95% CI, 26.1-33.2), and prevalence of ACS was 0.56% (95% CI, 0.48%-0.63%) of all episodes. Although the monthly rate was statistically decreasing by 0.022% (95% CI, 0.003-0.41), this has little meaning since the level of correlation (r² = 0.2522, not displayed) was poor due to the essentially nonexistent correlation in number of visits each month (r² = 0.0112, not displayed). During the postintervention year, the average number of diagnoses was 32.2 (95% CI, 27.9-36.6), and the prevalence of ACS was 0.62% (95% CI, 0.54-0.65). Neither of these values changed significantly between the pre- and postintervention period. All ICD-9 and ICD-10 diagnosis codes used for the analysis are available upon request from the authors.

Discussion

Our data demonstrate the ability of simple process improvement interventions to decrease unnecessary testing in the workup of ACS, increasing the rate of guideline concordant testing by > 70% at a single military treatment facility (MTF). In particular, with the now widespread use of EHR, the order set presents a high-yield target for process improvement in an easily implemented, durable fashion. We had expected to see some decrease in the efficacy of the intervention at a time of staff turnover in the summer of 2015 because ongoing dedicated teaching sessions were not performed. Despite that, the intervention remained effective without further dedicated teaching sessions. This outcome was certainly attributable to the hardwired interventions made (mainly via order sets), but possibly indicates an institutional memory that can take hold after an initial concerted effort is made.

We reduced the estimated preintervention annual cost of $113,000 by $53,000 (95% CI, $42,000-$64,000). Although on a much smaller scale than the study by Larochelle, our study represents a nearly 50% reduction in the total cost of initial testing for possible ACS and a > 80% reduction in unnecessary CK-MB testing.⁴ This result was achieved with no statistical change in the prevalence of ACS. The cost reduction does not account for the labor costs to clinically follow-up and address additional unnecessary lab results. The estimated cost of intervention was limited to the time required to educate residents, interns, and nursing staff as well as the implementation of the automated, reflexive laboratory results ordering process.

Unique to our study, we also demonstrated an intervention that satisfied all the major stakeholders in the ordering of these laboratory results. By instituting the reflexive ordering of CK-MB tests for positive troponins, we obtained the support of the facility’s interventional cardiology department, which finds value in that data. Appreciating the time-sensitive nature of an ACS diagnosis, the reflexive ordering minimized the delay in receiving these data while still greatly reducing the number of tests performed. That being said, if the current trend away from CK-MB in favor of exclusively testing troponin continues, removing the reflexive ordering for positive laboratory results protocol would be an easy follow-on intervention.

Limitations

Our study presented several limitations. First, reporting errors due to improper or insufficient medical coding as well as data entry errors may exist within the MDR; therefore, the results of this analysis may be over- or underestimated. Specifically, CPT codes for troponin and CK-MB were available only in 1 of the 2 data sets used for this study, which primarily contains outpatient patient encounters. For this reason, most of the laboratory testing comes from the EMD rather than from inpatient services. However, because we excluded all patients who eventually had an ACS diagnosis (patients who likely had more inpatient time and better indication for repeat troponin), we feel that our intervention was still thoroughly investigated. Second, the number of tests drawn per patient was significantly < 2, the expected minimum number of tests to rule out ACS in patients with appropriate symptoms.

This study was not designed to answer the source of variation from guidelines. Many patients had only 1 test, which we feel represents an opportunity for future study to identify other ways cardiac enzyme testing is being used clinically. These tests might be used for patients without convincing symptoms and signs of coronary syndromes or for patients with other primary problems. Third, by using the ITS analysis, we assumed that the outcome during each intervention period follows a linear pattern. However, changes may follow a nonlinear pattern over a long period. Finally, our intervention was limited to only a single MTF, which may limit generalizability to other facilities across military medicine. However, we feel this study should serve as a guide for other MTFs as well as US Department of Veterans Affairs facilities that could institute similar process improvements.

Conclusion

We made easily implemented and durable process improvement interventions that changed institution-wide ordering practices. These changes dramatically increased the rate of guideline-concordant testing, decreasing cost and furthering the goal of high-value medical care.

In recent years, driven by accelerating health care costs and desire for improved health care value, major specialty group guidelines have incorporated resource utilization and value calculations into their recommendations. High-value care has the characteristics of enhancing outcomes, safety, and patient satisfaction at a reasonable cost. As one example, the American College of Cardiology (ACC) recently published a consensus statement on its clinical practice guidelines with a specific focus on cost and value.¹ This guideline acknowledges the difficulty in incorporating value into clinical decision making but stresses a need for increased transparency and consistency to boost value in everyday practice.

Chest pain and related symptoms were listed as the second leading principle reasons for emergency department visits in the US in 2011 with 14% of patients undergoing cardiac enzyme testing.² The ACC guidelines advocate use of troponin as the preferred laboratory test for the initial evaluation of acute coronary syndrome (ACS). Fractionated creatine kinase (CK-MB) is an acceptable alternative only when a cardiac troponin test is not available.³ Furthermore, troponins should be obtained no more than 3 times for the initial evaluation of a single event, and further trending provides no additional benefit or prognostic information.

A recent study from an academic hospital showed that process improvement interventions focused on eliminating unnecessary cardiac enzyme testing led to a 1-year cost savings of $1.25 million while increasing the rate of ACS diagnosis.⁴ Common clinical practice at Naval Medical Center Portsmouth (NMCP) in Virginia still routinely includes both troponin as well as a CK panel comprised of CK, CK-MB, and a calculated CK-MB/CK index. Our study focuses on the implementation of quality improvement efforts described by Larochelle and colleagues at NMCP.⁴ The study aimed to determine the impact of implementing interventions designed to improve the ordering practices and reduce the cost of cardiac enzyme testing.

Methods

The primary focus of the intervention was on ordering practices of the emergency medicine department (EMD), internal medicine (IM) inpatient services, and cardiology inpatient services. Specific interventions were: (1) removal of the CK panel from the chest pain order set in the EMD electronic health record (EHR); (2) removal of the CK panel from the inpatient cardiology order set; (3) education of staff on the changes in CK panel utility via direct communication during IM academic seminars; (4) education of nursing staff ordering laboratory results on behalf of physicians on the cardiology service at the morning and evening huddles; and (5) addition of “max of 3 tests indicated” comment to the inpatient EHR ordering page of the troponin test. Acknowledging that the CK-MB has some utility to interventional cardiologists in the setting of confirmed ACS, the laboratory instituted an automated, reflexive order of the CK-MB panel only if the troponin tests were positive. This test was automatically run on the same vial originally sent to the lab to mitigate any additional delay in determining results.

Data Source

The process improvement interventions were considered exempt from institutional review board (IRB) approval; however, we obtained expedited IRB approval with waiver of consent for the research aspect of the project. We obtained clinical administrative data from the Military Health System Data Repository (MDR). We identified all adult patients aged ≥ 18 years who had a troponin test, CK-MB, or both drawn at NMCP on the following services: the EMD, IM, and cardiology. A troponin or CK-MB test was defined using Current Procedural Terminology (CPT) codes and unique Logical Observation Identifiers Names and Codes (LOINC).

Measures

The study was divided into 3 periods: the preintervention period from August 1, 2013 to July 31, 2014; the intervention period from August 1, 2014 to January 31, 2015; and the postintervention period February 1, 2015 to January 31, 2016.

The primary outcomes measured were the frequency of guideline concordance and total costs for tests ordered per month using the Centers for Medicare and Medicaid Services (CMS) clinical laboratory fee schedule of $13.40 for troponin and $16.17 for CK-MB.⁵Concordance was defined as ≤ 3 troponin tests and no CK-MB tests ordered during 1 encounter for a patient without an ACS diagnosis in the preceding 7 days. Due to faster cellular release kinetics of CK-MB compared with that of troponin, this test has utility in evaluating new or worsening chest pain in the setting of a recent myocardial infarction (MI). Therefore, we excluded any patient who had a MI within the preceding 7 days of an order for either CK-MB or troponin tests. Additionally, the number of tests, both CK-MB and troponin, ordered per patient encounter (hereafter referred to as an episode) were measured. Finally, we measured the monthly prevalence of ACS diagnosis and percentage of visits having that diagnosis.

Data Analysis

Descriptive statistics were used to calculate population demographics of age group, sex, beneficiary category, sponsor service, and clinical setting. Monthly data were grouped into the preintervention and postintervention periods. The analysis was performed using t tests to compare mean values and CIs before and after the intervention. Simple linear regression with attention to correlation was used to create best fit lines with confidence bands before and after the intervention. Interrupted time series (ITS) regression was used to describe all data points throughout the study. Consistency between these various methods was verified. Mean values and CIs were reported from the t tests. Statistical significance was reported when appropriate. Equations and confidence predictions on the simple linear regressions were produced and reported. These were used to identify values at the start, midpoint, and end of the pre- and postintervention periods.

Results

There were a total of 6,281 patients in the study population. More patients were seen during the postintervention period than in the preintervention period. The mean age of patients was slightly higher during the preintervention period (Table 1).

Guideline Concordance

To determine whether ordering practices for cardiac enzyme testing improved, we assessed the changes in the frequency of guideline concordance during the pre- and postintervention period. On average during the preintervention year, the percentage of tests ordered that met guideline concordance was 10.1% (95% CI, 7.4%-12.9%), increasing by 0.80% (95% CI, 0.17%-1.42%) each month. 

Costs

We assessed changes in total dollars spent on cardiac enzyme testing during the pre- and postintervention periods. During the preintervention year, $9,400 (95% CI, $8,700-$10,100) was spent on average each month, which did not change significantly throughout the period. During the postintervention year, the cost was stable at $5,000 (95% CI, $4,600-$5,300) on average each month, a reduction of $4,400 (95% CI, $3,700-$5,100) (Figure 2).

CK-MB and Troponin Tests per Patient

To further assess ordering practices for cardiac enzyme testing, we compared the changes in the monthly number of tests and the average number of CK-MB and troponin tests ordered per episode pre- and postintervention. On average during the preintervention year, 297 tests (95% CI, 278-315) were run per month, with an average of 1.21 CK tests (95% CI, 1.15-1.27) per episode (Table 2, Figure 3). 

The changes in troponin testing were not as dramatic. The counts of tests each month remained similar, with a preintervention year average of 341 (95% CI, 306-377) and postintervention year average of 310 (95% CI, 287-332), which were not statistically different. However, there was a statistically significant decrease in the number of tests per episode. During the preintervention year, 1.38 troponin tests (95% CI, 1.31-1.45) were ordered per patient on average. This dropped by 0.17 (95% CI, 0.09-0.24) to the postintervention average of 1.21 (95% CI, 1.17-1.25) (Table 2, Figure 4). 

ACS Prevalence

To determine whether there was an impact on ACS diagnoses, we looked at the numbers of ACS diagnoses and their prevalence among visits before and after the intervention. During the preintervention year, the average monthly number of diagnoses was 29.7 (95% CI, 26.1-33.2), and prevalence of ACS was 0.56% (95% CI, 0.48%-0.63%) of all episodes. Although the monthly rate was statistically decreasing by 0.022% (95% CI, 0.003-0.41), this has little meaning since the level of correlation (r² = 0.2522, not displayed) was poor due to the essentially nonexistent correlation in number of visits each month (r² = 0.0112, not displayed). During the postintervention year, the average number of diagnoses was 32.2 (95% CI, 27.9-36.6), and the prevalence of ACS was 0.62% (95% CI, 0.54-0.65). Neither of these values changed significantly between the pre- and postintervention period. All ICD-9 and ICD-10 diagnosis codes used for the analysis are available upon request from the authors.

Discussion

Our data demonstrate the ability of simple process improvement interventions to decrease unnecessary testing in the workup of ACS, increasing the rate of guideline concordant testing by > 70% at a single military treatment facility (MTF). In particular, with the now widespread use of EHR, the order set presents a high-yield target for process improvement in an easily implemented, durable fashion. We had expected to see some decrease in the efficacy of the intervention at a time of staff turnover in the summer of 2015 because ongoing dedicated teaching sessions were not performed. Despite that, the intervention remained effective without further dedicated teaching sessions. This outcome was certainly attributable to the hardwired interventions made (mainly via order sets), but possibly indicates an institutional memory that can take hold after an initial concerted effort is made.

We reduced the estimated preintervention annual cost of $113,000 by $53,000 (95% CI, $42,000-$64,000). Although on a much smaller scale than the study by Larochelle, our study represents a nearly 50% reduction in the total cost of initial testing for possible ACS and a > 80% reduction in unnecessary CK-MB testing.⁴ This result was achieved with no statistical change in the prevalence of ACS. The cost reduction does not account for the labor costs to clinically follow-up and address additional unnecessary lab results. The estimated cost of intervention was limited to the time required to educate residents, interns, and nursing staff as well as the implementation of the automated, reflexive laboratory results ordering process.

Unique to our study, we also demonstrated an intervention that satisfied all the major stakeholders in the ordering of these laboratory results. By instituting the reflexive ordering of CK-MB tests for positive troponins, we obtained the support of the facility’s interventional cardiology department, which finds value in that data. Appreciating the time-sensitive nature of an ACS diagnosis, the reflexive ordering minimized the delay in receiving these data while still greatly reducing the number of tests performed. That being said, if the current trend away from CK-MB in favor of exclusively testing troponin continues, removing the reflexive ordering for positive laboratory results protocol would be an easy follow-on intervention.

Limitations

Our study presented several limitations. First, reporting errors due to improper or insufficient medical coding as well as data entry errors may exist within the MDR; therefore, the results of this analysis may be over- or underestimated. Specifically, CPT codes for troponin and CK-MB were available only in 1 of the 2 data sets used for this study, which primarily contains outpatient patient encounters. For this reason, most of the laboratory testing comes from the EMD rather than from inpatient services. However, because we excluded all patients who eventually had an ACS diagnosis (patients who likely had more inpatient time and better indication for repeat troponin), we feel that our intervention was still thoroughly investigated. Second, the number of tests drawn per patient was significantly < 2, the expected minimum number of tests to rule out ACS in patients with appropriate symptoms.

This study was not designed to answer the source of variation from guidelines. Many patients had only 1 test, which we feel represents an opportunity for future study to identify other ways cardiac enzyme testing is being used clinically. These tests might be used for patients without convincing symptoms and signs of coronary syndromes or for patients with other primary problems. Third, by using the ITS analysis, we assumed that the outcome during each intervention period follows a linear pattern. However, changes may follow a nonlinear pattern over a long period. Finally, our intervention was limited to only a single MTF, which may limit generalizability to other facilities across military medicine. However, we feel this study should serve as a guide for other MTFs as well as US Department of Veterans Affairs facilities that could institute similar process improvements.

Conclusion

We made easily implemented and durable process improvement interventions that changed institution-wide ordering practices. These changes dramatically increased the rate of guideline-concordant testing, decreasing cost and furthering the goal of high-value medical care.

Postgraduate training for physician assistants (PAs) and nurse practitioners (NPs) is a rapidly evolving field. It has been estimated that the number of these advanced practice providers (APPs) almost doubled between 2000 and 2016 (from 15.3 to 28.2 per 100 physicians) and is expected to double again by 2030.¹ As APPs continue to become a progressively larger part of the healthcare workforce, medical organizations are seeking more comprehensive strategies to train and mentor them.² This has led to the development of formal postgraduate programs, often called APP fellowships.

Historically, postgraduate APP fellowships have functioned to help bridge the gap in clinical practice experience between physicians and APPs.³ This gap is evident in hours of clinical training. Whereas NPs are generally expected to complete 500-1,500 hours of clinical practice before graduating,⁴ and PAs are expected to complete 2,000 hours,⁵ most physicians will complete over 15,000 hours of clinical training by the end of residency.⁶ As increasing patient complexity continues to challenge the healthcare workforce,⁷ both the NP and the PA leadership have recommended increased training of graduates and outcome studies of formal postgraduate fellowships.^8,9 In 2007, there were over 60 of these programs in the United States,¹⁰ most of them offering training in surgical specialties.

First described in 2010 by the Mayo Clinic,¹¹ APP fellowships in hospital medicine are also being developed. These programs are built to improve the training of nonphysician hospitalists, who often work independently¹² and manage medically complex patients.¹³ However, little is known about the number or structure of these fellowships. The limited understanding of the current APP fellowship environment is partly due to the lack of an administrative body overseeing these programs.¹⁴ The Accreditation Review Commission on Education for the Physician Assistant (ARC-PA) pioneered a model in 2007 for postgraduate PA programs, but it has been held in abeyance since 2014.¹⁵ Both the American Nurses Credentialing Center and the National Nurse Practitioner Residency and Fellowship Training Consortium have fellowship accreditation review processes, but they are not specific to hospital medicine.¹⁶ The Society of Hospital Medicine (SHM) has several resources for the training of APPs;¹⁷ however, it neither reviews nor accredits fellowship programs. Without standards, guidelines, or active accrediting bodies, APP fellowships in hospital medicine are poorly understood and are of unknown efficacy. The purpose of this study was to identify and describe the active APP fellowships in hospital medicine.

This was a cross-sectional study of all APP adult and pediatric fellowships in hospital medicine, in the United States, that were identifiable through May 2018. Multiple methods were used to identify all active fellowships. First, all training programs offering a Hospital Medicine Fellowship in the ARC-PA and Association of Postgraduate PA Programs databases were noted. Second, questionnaires were given out at the NP/PA forum at the national SHM conference in 2018 to gather information on existing APP fellowships. Third, similar online requests to identify known programs were posted to the SHM web forum Hospital Medicine Exchange (HMX). Fourth, Internet searches were used to discover additional programs. Once those fellowships were identified, surveys were sent to their program directors (PDs). These surveys not only asked the PDs about their fellowship but also asked them to identify additional APP fellowships beyond those that we had captured. Once additional programs were identified, a second round of surveys was sent to their PDs. This was performed in an iterative fashion until no additional fellowships were discovered.

The survey tool was developed and validated internally in the AAMC Survey Development style¹⁸ and was influenced by prior validated surveys of postgraduate medical fellowships.^10,19-21 Each question was developed by a team that had expertise in survey design (Wright and Tackett), and two survey design team members were themselves PDs of APP fellowships in hospital medicine (Kisuule and Franco). The survey was revised iteratively by the team on the basis of meetings and pilot testing with PDs of other programs. All qualitative or descriptive questions had a free response option available to allow PDs to answer the survey accurately and exhaustively. The final version of the survey was approved by consensus of all authors. It consisted of 25 multiple choice questions which were created to gather information about the following key areas of APP hospital medicine fellowships: fellowship and learner characteristics, program rationales, curricula, and methods of fellow assessment.

A web-based survey format (Qualtrics) was used to distribute the questionnaire e-mail to the PDs. Follow up e-mail reminders were sent to all nonresponders to encourage full participation. Survey completion was voluntary; no financial incentives or gifts were offered. IRB approval was obtained at Johns Hopkins Bayview (IRB number 00181629). Descriptive statistics (proportions, means, and ranges as appropriate) were calculated for all variables. Stata 13 (StataCorp. 2013. Stata Statistical Software: Release 13. College Station, Texas. StataCorp LP) was used for data analysis.

In total, 11 fellowships were identified using our multimethod approach. We found four (36%) programs by utilizing existing online databases, two (18%) through the SHM questionnaire and HMX forum, three (27%) through internet searches, and the remaining two (18%) were referred to us by the other PDs who were surveyed. Of the programs surveyed, 10 were adult programs and one was a pediatric program. Surveys were sent to the PDs of the 11 fellowships, and all but one of them (10/11, 91%) responded. Respondent programs were given alphabetical designations A through J (Table). 

Fellowship and Individual Characteristics

Most programs have been in existence for five years or fewer. Eighty percent of the programs are about one year in duration; two outlier programs have fellowship lengths of six months and 18 months. The main hospital where training occurs has a mean of 496 beds (range 213 to 900). Ninety percent of the hospitals also have physician residency training programs. Sixty percent of programs enroll two to four fellows per year while 40% enroll five or more. The salary range paid by the programs is $55,000 to >$70,000, and half the programs pay more than $65,000.

The majority of fellows accepted into APP fellowships in hospital medicine are women. Eighty percent of fellows are 26-30 years old, and 90% of fellows have been out of NP or PA school for one year or less. Both NP and PA applicants are accepted in 80% of fellowships.

Program Rationales

All programs reported that training and retaining applicants is the main driver for developing their fellowship, and 50% of them offer financial incentives for retention upon successful completion of the program. Forty percent of PDs stated that there is an implicit or explicit understanding that successful completion of the fellowship would result in further employment. Over the last five years, 89% (range: 71%-100%) of graduates were asked to remain for a full-time position after program completion.

In addition to training and retention, building an interprofessional team (50%), managing patient volume (30%), and reducing overhead (20%) were also reported as rationales for program development. The majority of programs (80%) have fellows bill for clinical services, and five of those eight programs do so after their fellows become more clinically competent.

Curricula

Of the nine adult programs, 67% teach explicitly to SHM core competencies and 33% send their fellows to the SHM NP/PA Boot Camp. Thirty percent of fellowships partner formally with either a physician residency or a local PA program to develop educational content. Six of the nine programs with active physician residencies, including the pediatric fellowship, offer shared educational experiences for the residents and APPs.

There are notable differences in clinical rotations between the programs (Figure 1). No single rotation is universally required, although general hospital internal medicine is required in all adult fellowships. The majority (80%) of programs offer at least one elective. Six programs reported mandatory rotations outside the department of medicine, most commonly neurology or the stroke service (four programs). Only one program reported only general medicine rotations, with no subspecialty electives.

Methods of Fellow Assessment

Each program surveyed has a unique method of fellow assessment. Ninety percent of the programs use more than one method to assess their fellows. Faculty reviews are most commonly used and are conducted in all rotations in 80% of fellowships. Both self-assessment exercises and written examinations are used in some rotations by the majority of programs. Capstone projects are required infrequently (30%).

We found several commonalities between the fellowships surveyed. Many of the program characteristics, such as years in operation, salary, duration, and lack of accreditation, are quite similar. Most fellowships also have a similar rationale for building their programs and use resources from the SHM to inform their curricula. Fellows, on average, share several demographic characteristics, such as age, gender, and time out of schooling. Conversely, we found wide variability in clinical rotations, the general teaching structure, and methods of fellow evaluation.

There have been several publications detailing successful individual APP fellowships in medical subspecialties,²² psychiatry,²³ and surgical specialties,²⁴ all of which describe the benefits to the institution. One study found that physician hospitalists have a poor understanding of the training PAs undergo and would favor a standardized curriculum for PA hospitalists.²⁵ Another study compared all PA postgraduate training programs in emergency medicine;¹⁹ it also described a small number of relatively young programs with variable curricula and a need for standardization. Yet another paper¹⁰ surveyed postgraduate PA programs across all specialties; however, that study only captured two hospital medicine programs, and it was not focused on several key areas studied in this paper—such as the program rationale, curricular elements, and assessment.

It is noteworthy that every program surveyed was created with training and retention in mind, rather than other factors like decreasing overhead or managing patient volume. Training one’s own APPs so that they can learn on the job, come to understand expectations within a group, and witness the culture is extremely valuable. From a patient safety standpoint, it has been documented that physician hospitalists straight out of residency have a higher patient mortality compared with more experienced providers.²⁶ Given the findings that on a national level, the majority of hospitalist NPs and PAs practice autonomously or somewhat autonomously,¹² it is reasonable to assume that similar trends of more experienced providers delivering safer care would be expected for APPs, but this remains speculative. From a retention standpoint, it has been well described that high APP turnover is often due to decreased feelings of competence and confidence during their transition from trainees to medical providers.²⁷ APPs who have completed fellowships feel more confident and able to succeed in their field.²⁸ To this point, in one survey of hospitalist PAs, almost all reported that they would have been interested in completing a fellowship, even it meant a lower initial salary.²⁹Despite having the same general goals and using similar national resources, our study reveals that APP fellows are trained and assessed very differently between programs. This might represent an area of future growth in the field of hospitalist APP education. For physician learning, competency-based medical education (CBME) has emerged as a learner centric, outcomes-based model of teaching and assessment that emphasizes mastery of skills and progression through milestones.³⁰ Both the ACGME³¹ and the SHM³² have described core competencies that provide a framework within CBME for determining readiness for independent practice. While we were not surprised to find that each fellowship has its own unique method of determining readiness for practice, these findings suggest that graduates from different programs likely have very different skill sets and aptitude levels. In the future, an active accrediting body could offer guidance in defining hospitalist APP core competencies and help standardize education.

Several limitations to this study should be considered. While we used multiple strategies to locate as many fellowships as possible, it is unlikely that we successfully captured all existing programs, and new programs are being developed annually. We also relied on self-reported data from PDs. While we would expect PDs to provide accurate data, we could not externally validate their answers. Additionally, although our survey tool was reviewed extensively and validated internally, it was developed de novo for this study.

APP fellowships in hospital medicine have experienced marked growth since the first program was described in 2010. The majority of programs are 12 months long, operate in existing teaching centers, and are intended to further enhance the training and retention of newly graduated PAs and NPs. Despite their similarities, fellowships have striking variability in their methods of teaching and assessing their learners. Best practices have yet to be identified, and further study is required to determine how to standardize curricula across the board.

Acknowledgments

Disclosures

The authors report no conflicts of interest.

Funding

This project was supported by the Johns Hopkins School of Medicine Biostatistics, Epidemiology and Data Management (BEAD) Core. Dr. Wright is the Anne Gaines and G. Thomas Miller Professor of Medicine, which is supported through the Johns Hopkins’ Center for Innovative Medicine.

Postgraduate training for physician assistants (PAs) and nurse practitioners (NPs) is a rapidly evolving field. It has been estimated that the number of these advanced practice providers (APPs) almost doubled between 2000 and 2016 (from 15.3 to 28.2 per 100 physicians) and is expected to double again by 2030.¹ As APPs continue to become a progressively larger part of the healthcare workforce, medical organizations are seeking more comprehensive strategies to train and mentor them.² This has led to the development of formal postgraduate programs, often called APP fellowships.

Historically, postgraduate APP fellowships have functioned to help bridge the gap in clinical practice experience between physicians and APPs.³ This gap is evident in hours of clinical training. Whereas NPs are generally expected to complete 500-1,500 hours of clinical practice before graduating,⁴ and PAs are expected to complete 2,000 hours,⁵ most physicians will complete over 15,000 hours of clinical training by the end of residency.⁶ As increasing patient complexity continues to challenge the healthcare workforce,⁷ both the NP and the PA leadership have recommended increased training of graduates and outcome studies of formal postgraduate fellowships.^8,9 In 2007, there were over 60 of these programs in the United States,¹⁰ most of them offering training in surgical specialties.

First described in 2010 by the Mayo Clinic,¹¹ APP fellowships in hospital medicine are also being developed. These programs are built to improve the training of nonphysician hospitalists, who often work independently¹² and manage medically complex patients.¹³ However, little is known about the number or structure of these fellowships. The limited understanding of the current APP fellowship environment is partly due to the lack of an administrative body overseeing these programs.¹⁴ The Accreditation Review Commission on Education for the Physician Assistant (ARC-PA) pioneered a model in 2007 for postgraduate PA programs, but it has been held in abeyance since 2014.¹⁵ Both the American Nurses Credentialing Center and the National Nurse Practitioner Residency and Fellowship Training Consortium have fellowship accreditation review processes, but they are not specific to hospital medicine.¹⁶ The Society of Hospital Medicine (SHM) has several resources for the training of APPs;¹⁷ however, it neither reviews nor accredits fellowship programs. Without standards, guidelines, or active accrediting bodies, APP fellowships in hospital medicine are poorly understood and are of unknown efficacy. The purpose of this study was to identify and describe the active APP fellowships in hospital medicine.

This was a cross-sectional study of all APP adult and pediatric fellowships in hospital medicine, in the United States, that were identifiable through May 2018. Multiple methods were used to identify all active fellowships. First, all training programs offering a Hospital Medicine Fellowship in the ARC-PA and Association of Postgraduate PA Programs databases were noted. Second, questionnaires were given out at the NP/PA forum at the national SHM conference in 2018 to gather information on existing APP fellowships. Third, similar online requests to identify known programs were posted to the SHM web forum Hospital Medicine Exchange (HMX). Fourth, Internet searches were used to discover additional programs. Once those fellowships were identified, surveys were sent to their program directors (PDs). These surveys not only asked the PDs about their fellowship but also asked them to identify additional APP fellowships beyond those that we had captured. Once additional programs were identified, a second round of surveys was sent to their PDs. This was performed in an iterative fashion until no additional fellowships were discovered.

The survey tool was developed and validated internally in the AAMC Survey Development style¹⁸ and was influenced by prior validated surveys of postgraduate medical fellowships.^10,19-21 Each question was developed by a team that had expertise in survey design (Wright and Tackett), and two survey design team members were themselves PDs of APP fellowships in hospital medicine (Kisuule and Franco). The survey was revised iteratively by the team on the basis of meetings and pilot testing with PDs of other programs. All qualitative or descriptive questions had a free response option available to allow PDs to answer the survey accurately and exhaustively. The final version of the survey was approved by consensus of all authors. It consisted of 25 multiple choice questions which were created to gather information about the following key areas of APP hospital medicine fellowships: fellowship and learner characteristics, program rationales, curricula, and methods of fellow assessment.

A web-based survey format (Qualtrics) was used to distribute the questionnaire e-mail to the PDs. Follow up e-mail reminders were sent to all nonresponders to encourage full participation. Survey completion was voluntary; no financial incentives or gifts were offered. IRB approval was obtained at Johns Hopkins Bayview (IRB number 00181629). Descriptive statistics (proportions, means, and ranges as appropriate) were calculated for all variables. Stata 13 (StataCorp. 2013. Stata Statistical Software: Release 13. College Station, Texas. StataCorp LP) was used for data analysis.

In total, 11 fellowships were identified using our multimethod approach. We found four (36%) programs by utilizing existing online databases, two (18%) through the SHM questionnaire and HMX forum, three (27%) through internet searches, and the remaining two (18%) were referred to us by the other PDs who were surveyed. Of the programs surveyed, 10 were adult programs and one was a pediatric program. Surveys were sent to the PDs of the 11 fellowships, and all but one of them (10/11, 91%) responded. Respondent programs were given alphabetical designations A through J (Table). 

Fellowship and Individual Characteristics

Most programs have been in existence for five years or fewer. Eighty percent of the programs are about one year in duration; two outlier programs have fellowship lengths of six months and 18 months. The main hospital where training occurs has a mean of 496 beds (range 213 to 900). Ninety percent of the hospitals also have physician residency training programs. Sixty percent of programs enroll two to four fellows per year while 40% enroll five or more. The salary range paid by the programs is $55,000 to >$70,000, and half the programs pay more than $65,000.

The majority of fellows accepted into APP fellowships in hospital medicine are women. Eighty percent of fellows are 26-30 years old, and 90% of fellows have been out of NP or PA school for one year or less. Both NP and PA applicants are accepted in 80% of fellowships.

Program Rationales

All programs reported that training and retaining applicants is the main driver for developing their fellowship, and 50% of them offer financial incentives for retention upon successful completion of the program. Forty percent of PDs stated that there is an implicit or explicit understanding that successful completion of the fellowship would result in further employment. Over the last five years, 89% (range: 71%-100%) of graduates were asked to remain for a full-time position after program completion.

In addition to training and retention, building an interprofessional team (50%), managing patient volume (30%), and reducing overhead (20%) were also reported as rationales for program development. The majority of programs (80%) have fellows bill for clinical services, and five of those eight programs do so after their fellows become more clinically competent.

Curricula

Of the nine adult programs, 67% teach explicitly to SHM core competencies and 33% send their fellows to the SHM NP/PA Boot Camp. Thirty percent of fellowships partner formally with either a physician residency or a local PA program to develop educational content. Six of the nine programs with active physician residencies, including the pediatric fellowship, offer shared educational experiences for the residents and APPs.

There are notable differences in clinical rotations between the programs (Figure 1). No single rotation is universally required, although general hospital internal medicine is required in all adult fellowships. The majority (80%) of programs offer at least one elective. Six programs reported mandatory rotations outside the department of medicine, most commonly neurology or the stroke service (four programs). Only one program reported only general medicine rotations, with no subspecialty electives.

Methods of Fellow Assessment

Each program surveyed has a unique method of fellow assessment. Ninety percent of the programs use more than one method to assess their fellows. Faculty reviews are most commonly used and are conducted in all rotations in 80% of fellowships. Both self-assessment exercises and written examinations are used in some rotations by the majority of programs. Capstone projects are required infrequently (30%).

We found several commonalities between the fellowships surveyed. Many of the program characteristics, such as years in operation, salary, duration, and lack of accreditation, are quite similar. Most fellowships also have a similar rationale for building their programs and use resources from the SHM to inform their curricula. Fellows, on average, share several demographic characteristics, such as age, gender, and time out of schooling. Conversely, we found wide variability in clinical rotations, the general teaching structure, and methods of fellow evaluation.

There have been several publications detailing successful individual APP fellowships in medical subspecialties,²² psychiatry,²³ and surgical specialties,²⁴ all of which describe the benefits to the institution. One study found that physician hospitalists have a poor understanding of the training PAs undergo and would favor a standardized curriculum for PA hospitalists.²⁵ Another study compared all PA postgraduate training programs in emergency medicine;¹⁹ it also described a small number of relatively young programs with variable curricula and a need for standardization. Yet another paper¹⁰ surveyed postgraduate PA programs across all specialties; however, that study only captured two hospital medicine programs, and it was not focused on several key areas studied in this paper—such as the program rationale, curricular elements, and assessment.

It is noteworthy that every program surveyed was created with training and retention in mind, rather than other factors like decreasing overhead or managing patient volume. Training one’s own APPs so that they can learn on the job, come to understand expectations within a group, and witness the culture is extremely valuable. From a patient safety standpoint, it has been documented that physician hospitalists straight out of residency have a higher patient mortality compared with more experienced providers.²⁶ Given the findings that on a national level, the majority of hospitalist NPs and PAs practice autonomously or somewhat autonomously,¹² it is reasonable to assume that similar trends of more experienced providers delivering safer care would be expected for APPs, but this remains speculative. From a retention standpoint, it has been well described that high APP turnover is often due to decreased feelings of competence and confidence during their transition from trainees to medical providers.²⁷ APPs who have completed fellowships feel more confident and able to succeed in their field.²⁸ To this point, in one survey of hospitalist PAs, almost all reported that they would have been interested in completing a fellowship, even it meant a lower initial salary.²⁹Despite having the same general goals and using similar national resources, our study reveals that APP fellows are trained and assessed very differently between programs. This might represent an area of future growth in the field of hospitalist APP education. For physician learning, competency-based medical education (CBME) has emerged as a learner centric, outcomes-based model of teaching and assessment that emphasizes mastery of skills and progression through milestones.³⁰ Both the ACGME³¹ and the SHM³² have described core competencies that provide a framework within CBME for determining readiness for independent practice. While we were not surprised to find that each fellowship has its own unique method of determining readiness for practice, these findings suggest that graduates from different programs likely have very different skill sets and aptitude levels. In the future, an active accrediting body could offer guidance in defining hospitalist APP core competencies and help standardize education.

Several limitations to this study should be considered. While we used multiple strategies to locate as many fellowships as possible, it is unlikely that we successfully captured all existing programs, and new programs are being developed annually. We also relied on self-reported data from PDs. While we would expect PDs to provide accurate data, we could not externally validate their answers. Additionally, although our survey tool was reviewed extensively and validated internally, it was developed de novo for this study.

APP fellowships in hospital medicine have experienced marked growth since the first program was described in 2010. The majority of programs are 12 months long, operate in existing teaching centers, and are intended to further enhance the training and retention of newly graduated PAs and NPs. Despite their similarities, fellowships have striking variability in their methods of teaching and assessing their learners. Best practices have yet to be identified, and further study is required to determine how to standardize curricula across the board.

Acknowledgments

Disclosures

The authors report no conflicts of interest.

Funding

This project was supported by the Johns Hopkins School of Medicine Biostatistics, Epidemiology and Data Management (BEAD) Core. Dr. Wright is the Anne Gaines and G. Thomas Miller Professor of Medicine, which is supported through the Johns Hopkins’ Center for Innovative Medicine.

Postgraduate training for physician assistants (PAs) and nurse practitioners (NPs) is a rapidly evolving field. It has been estimated that the number of these advanced practice providers (APPs) almost doubled between 2000 and 2016 (from 15.3 to 28.2 per 100 physicians) and is expected to double again by 2030.¹ As APPs continue to become a progressively larger part of the healthcare workforce, medical organizations are seeking more comprehensive strategies to train and mentor them.² This has led to the development of formal postgraduate programs, often called APP fellowships.

Historically, postgraduate APP fellowships have functioned to help bridge the gap in clinical practice experience between physicians and APPs.³ This gap is evident in hours of clinical training. Whereas NPs are generally expected to complete 500-1,500 hours of clinical practice before graduating,⁴ and PAs are expected to complete 2,000 hours,⁵ most physicians will complete over 15,000 hours of clinical training by the end of residency.⁶ As increasing patient complexity continues to challenge the healthcare workforce,⁷ both the NP and the PA leadership have recommended increased training of graduates and outcome studies of formal postgraduate fellowships.^8,9 In 2007, there were over 60 of these programs in the United States,¹⁰ most of them offering training in surgical specialties.

First described in 2010 by the Mayo Clinic,¹¹ APP fellowships in hospital medicine are also being developed. These programs are built to improve the training of nonphysician hospitalists, who often work independently¹² and manage medically complex patients.¹³ However, little is known about the number or structure of these fellowships. The limited understanding of the current APP fellowship environment is partly due to the lack of an administrative body overseeing these programs.¹⁴ The Accreditation Review Commission on Education for the Physician Assistant (ARC-PA) pioneered a model in 2007 for postgraduate PA programs, but it has been held in abeyance since 2014.¹⁵ Both the American Nurses Credentialing Center and the National Nurse Practitioner Residency and Fellowship Training Consortium have fellowship accreditation review processes, but they are not specific to hospital medicine.¹⁶ The Society of Hospital Medicine (SHM) has several resources for the training of APPs;¹⁷ however, it neither reviews nor accredits fellowship programs. Without standards, guidelines, or active accrediting bodies, APP fellowships in hospital medicine are poorly understood and are of unknown efficacy. The purpose of this study was to identify and describe the active APP fellowships in hospital medicine.

This was a cross-sectional study of all APP adult and pediatric fellowships in hospital medicine, in the United States, that were identifiable through May 2018. Multiple methods were used to identify all active fellowships. First, all training programs offering a Hospital Medicine Fellowship in the ARC-PA and Association of Postgraduate PA Programs databases were noted. Second, questionnaires were given out at the NP/PA forum at the national SHM conference in 2018 to gather information on existing APP fellowships. Third, similar online requests to identify known programs were posted to the SHM web forum Hospital Medicine Exchange (HMX). Fourth, Internet searches were used to discover additional programs. Once those fellowships were identified, surveys were sent to their program directors (PDs). These surveys not only asked the PDs about their fellowship but also asked them to identify additional APP fellowships beyond those that we had captured. Once additional programs were identified, a second round of surveys was sent to their PDs. This was performed in an iterative fashion until no additional fellowships were discovered.

The survey tool was developed and validated internally in the AAMC Survey Development style¹⁸ and was influenced by prior validated surveys of postgraduate medical fellowships.^10,19-21 Each question was developed by a team that had expertise in survey design (Wright and Tackett), and two survey design team members were themselves PDs of APP fellowships in hospital medicine (Kisuule and Franco). The survey was revised iteratively by the team on the basis of meetings and pilot testing with PDs of other programs. All qualitative or descriptive questions had a free response option available to allow PDs to answer the survey accurately and exhaustively. The final version of the survey was approved by consensus of all authors. It consisted of 25 multiple choice questions which were created to gather information about the following key areas of APP hospital medicine fellowships: fellowship and learner characteristics, program rationales, curricula, and methods of fellow assessment.

A web-based survey format (Qualtrics) was used to distribute the questionnaire e-mail to the PDs. Follow up e-mail reminders were sent to all nonresponders to encourage full participation. Survey completion was voluntary; no financial incentives or gifts were offered. IRB approval was obtained at Johns Hopkins Bayview (IRB number 00181629). Descriptive statistics (proportions, means, and ranges as appropriate) were calculated for all variables. Stata 13 (StataCorp. 2013. Stata Statistical Software: Release 13. College Station, Texas. StataCorp LP) was used for data analysis.

In total, 11 fellowships were identified using our multimethod approach. We found four (36%) programs by utilizing existing online databases, two (18%) through the SHM questionnaire and HMX forum, three (27%) through internet searches, and the remaining two (18%) were referred to us by the other PDs who were surveyed. Of the programs surveyed, 10 were adult programs and one was a pediatric program. Surveys were sent to the PDs of the 11 fellowships, and all but one of them (10/11, 91%) responded. Respondent programs were given alphabetical designations A through J (Table). 

Fellowship and Individual Characteristics

Most programs have been in existence for five years or fewer. Eighty percent of the programs are about one year in duration; two outlier programs have fellowship lengths of six months and 18 months. The main hospital where training occurs has a mean of 496 beds (range 213 to 900). Ninety percent of the hospitals also have physician residency training programs. Sixty percent of programs enroll two to four fellows per year while 40% enroll five or more. The salary range paid by the programs is $55,000 to >$70,000, and half the programs pay more than $65,000.

The majority of fellows accepted into APP fellowships in hospital medicine are women. Eighty percent of fellows are 26-30 years old, and 90% of fellows have been out of NP or PA school for one year or less. Both NP and PA applicants are accepted in 80% of fellowships.

Program Rationales

All programs reported that training and retaining applicants is the main driver for developing their fellowship, and 50% of them offer financial incentives for retention upon successful completion of the program. Forty percent of PDs stated that there is an implicit or explicit understanding that successful completion of the fellowship would result in further employment. Over the last five years, 89% (range: 71%-100%) of graduates were asked to remain for a full-time position after program completion.

In addition to training and retention, building an interprofessional team (50%), managing patient volume (30%), and reducing overhead (20%) were also reported as rationales for program development. The majority of programs (80%) have fellows bill for clinical services, and five of those eight programs do so after their fellows become more clinically competent.

Curricula

Of the nine adult programs, 67% teach explicitly to SHM core competencies and 33% send their fellows to the SHM NP/PA Boot Camp. Thirty percent of fellowships partner formally with either a physician residency or a local PA program to develop educational content. Six of the nine programs with active physician residencies, including the pediatric fellowship, offer shared educational experiences for the residents and APPs.

There are notable differences in clinical rotations between the programs (Figure 1). No single rotation is universally required, although general hospital internal medicine is required in all adult fellowships. The majority (80%) of programs offer at least one elective. Six programs reported mandatory rotations outside the department of medicine, most commonly neurology or the stroke service (four programs). Only one program reported only general medicine rotations, with no subspecialty electives.

Methods of Fellow Assessment

Each program surveyed has a unique method of fellow assessment. Ninety percent of the programs use more than one method to assess their fellows. Faculty reviews are most commonly used and are conducted in all rotations in 80% of fellowships. Both self-assessment exercises and written examinations are used in some rotations by the majority of programs. Capstone projects are required infrequently (30%).

We found several commonalities between the fellowships surveyed. Many of the program characteristics, such as years in operation, salary, duration, and lack of accreditation, are quite similar. Most fellowships also have a similar rationale for building their programs and use resources from the SHM to inform their curricula. Fellows, on average, share several demographic characteristics, such as age, gender, and time out of schooling. Conversely, we found wide variability in clinical rotations, the general teaching structure, and methods of fellow evaluation.

There have been several publications detailing successful individual APP fellowships in medical subspecialties,²² psychiatry,²³ and surgical specialties,²⁴ all of which describe the benefits to the institution. One study found that physician hospitalists have a poor understanding of the training PAs undergo and would favor a standardized curriculum for PA hospitalists.²⁵ Another study compared all PA postgraduate training programs in emergency medicine;¹⁹ it also described a small number of relatively young programs with variable curricula and a need for standardization. Yet another paper¹⁰ surveyed postgraduate PA programs across all specialties; however, that study only captured two hospital medicine programs, and it was not focused on several key areas studied in this paper—such as the program rationale, curricular elements, and assessment.

It is noteworthy that every program surveyed was created with training and retention in mind, rather than other factors like decreasing overhead or managing patient volume. Training one’s own APPs so that they can learn on the job, come to understand expectations within a group, and witness the culture is extremely valuable. From a patient safety standpoint, it has been documented that physician hospitalists straight out of residency have a higher patient mortality compared with more experienced providers.²⁶ Given the findings that on a national level, the majority of hospitalist NPs and PAs practice autonomously or somewhat autonomously,¹² it is reasonable to assume that similar trends of more experienced providers delivering safer care would be expected for APPs, but this remains speculative. From a retention standpoint, it has been well described that high APP turnover is often due to decreased feelings of competence and confidence during their transition from trainees to medical providers.²⁷ APPs who have completed fellowships feel more confident and able to succeed in their field.²⁸ To this point, in one survey of hospitalist PAs, almost all reported that they would have been interested in completing a fellowship, even it meant a lower initial salary.²⁹Despite having the same general goals and using similar national resources, our study reveals that APP fellows are trained and assessed very differently between programs. This might represent an area of future growth in the field of hospitalist APP education. For physician learning, competency-based medical education (CBME) has emerged as a learner centric, outcomes-based model of teaching and assessment that emphasizes mastery of skills and progression through milestones.³⁰ Both the ACGME³¹ and the SHM³² have described core competencies that provide a framework within CBME for determining readiness for independent practice. While we were not surprised to find that each fellowship has its own unique method of determining readiness for practice, these findings suggest that graduates from different programs likely have very different skill sets and aptitude levels. In the future, an active accrediting body could offer guidance in defining hospitalist APP core competencies and help standardize education.

Several limitations to this study should be considered. While we used multiple strategies to locate as many fellowships as possible, it is unlikely that we successfully captured all existing programs, and new programs are being developed annually. We also relied on self-reported data from PDs. While we would expect PDs to provide accurate data, we could not externally validate their answers. Additionally, although our survey tool was reviewed extensively and validated internally, it was developed de novo for this study.

APP fellowships in hospital medicine have experienced marked growth since the first program was described in 2010. The majority of programs are 12 months long, operate in existing teaching centers, and are intended to further enhance the training and retention of newly graduated PAs and NPs. Despite their similarities, fellowships have striking variability in their methods of teaching and assessing their learners. Best practices have yet to be identified, and further study is required to determine how to standardize curricula across the board.

Acknowledgments

Disclosures

The authors report no conflicts of interest.

Funding

This project was supported by the Johns Hopkins School of Medicine Biostatistics, Epidemiology and Data Management (BEAD) Core. Dr. Wright is the Anne Gaines and G. Thomas Miller Professor of Medicine, which is supported through the Johns Hopkins’ Center for Innovative Medicine.