Article

Key clinical point: An increased waist-to-height ratio (WHtR), indicating central obesity, was associated with an increased incidence of migraine, particularly in individuals aged <60 years. 

Major findings: Each unit increase in WHtR was associated with a 70% increase in the incidence of migraines (odds ratio [OR], 1.70; 95% CI, 1.04-2.78). Individuals in the highest WHtR quartile (WHtR values, 0.64-1.01) had a 13% greater incidence of migraines than those in the lowest quartile (WHtR values, 0.35-0.52; OR, 1.13; 95% CI, 0.99-1.28). For individuals aged <60 years, each unit increase in WHtR was associated with an 82% increased risk for migraine (P < .01); however, WHtR was negatively associated with migraine risk among those aged ≥60 years. 

Study details: This cross-sectional study analyzed data from the National Health and Nutrition Examination Survey, including 13,344 participants, of whom 2764 had migraines. Disclosure: The study did not receive any funding. The authors declared no conflicts of interest. 

Source: Jin J, Zheng Y, Gao T, Lin X, Li S, Huang C. Associations between the waist-to-height ratio index and migraine: A cross-section study of the NHANES 1999–2004. PLoS ONE. Published online October 23, 2024. Source

Key clinical point: An increased waist-to-height ratio (WHtR), indicating central obesity, was associated with an increased incidence of migraine, particularly in individuals aged <60 years. 

Major findings: Each unit increase in WHtR was associated with a 70% increase in the incidence of migraines (odds ratio [OR], 1.70; 95% CI, 1.04-2.78). Individuals in the highest WHtR quartile (WHtR values, 0.64-1.01) had a 13% greater incidence of migraines than those in the lowest quartile (WHtR values, 0.35-0.52; OR, 1.13; 95% CI, 0.99-1.28). For individuals aged <60 years, each unit increase in WHtR was associated with an 82% increased risk for migraine (P < .01); however, WHtR was negatively associated with migraine risk among those aged ≥60 years. 

Study details: This cross-sectional study analyzed data from the National Health and Nutrition Examination Survey, including 13,344 participants, of whom 2764 had migraines. Disclosure: The study did not receive any funding. The authors declared no conflicts of interest. 

Source: Jin J, Zheng Y, Gao T, Lin X, Li S, Huang C. Associations between the waist-to-height ratio index and migraine: A cross-section study of the NHANES 1999–2004. PLoS ONE. Published online October 23, 2024. Source

Key clinical point: An increased waist-to-height ratio (WHtR), indicating central obesity, was associated with an increased incidence of migraine, particularly in individuals aged <60 years. 

Major findings: Each unit increase in WHtR was associated with a 70% increase in the incidence of migraines (odds ratio [OR], 1.70; 95% CI, 1.04-2.78). Individuals in the highest WHtR quartile (WHtR values, 0.64-1.01) had a 13% greater incidence of migraines than those in the lowest quartile (WHtR values, 0.35-0.52; OR, 1.13; 95% CI, 0.99-1.28). For individuals aged <60 years, each unit increase in WHtR was associated with an 82% increased risk for migraine (P < .01); however, WHtR was negatively associated with migraine risk among those aged ≥60 years. 

Study details: This cross-sectional study analyzed data from the National Health and Nutrition Examination Survey, including 13,344 participants, of whom 2764 had migraines. Disclosure: The study did not receive any funding. The authors declared no conflicts of interest. 

Source: Jin J, Zheng Y, Gao T, Lin X, Li S, Huang C. Associations between the waist-to-height ratio index and migraine: A cross-section study of the NHANES 1999–2004. PLoS ONE. Published online October 23, 2024. Source

Supporting meaningful research that has a positive impact on the health and quality of life of veterans is a priority of the US Department of Veterans Affairs Office of Research and Development.¹ For nearly a century, VA researchers have been conducting high quality studies. To continue this trajectory, it is imperative to attract, train, and retain exceptional investigators while nurturing their development throughout their careers.²

Mentorship is defined as guidance provided by an experienced and trusted party to another (usually junior) individual with the intent of helping the person succeed. It benefits the mentee, mentor, and their institutions.³ Mentorship is crucial for personal and professional development as well as productivity, which may help reduce clinician burnout.^4-7 Conversely, a lack of mentorship could have negative effects on work satisfaction and stagnate career progression.⁸

Mentorship is vital for developing and advancing a VA investigator’s research agenda. Funding, grant writing, and research design were among the most discussed topics in a large comprehensive mentorship program for academic faculty.⁹ However, there are several known barriers to effective research mentorship; among them include a lack of resources, time constraints, and competing clinical priorities.^10,11

Finding time for effective one-on-one research mentoring is difficult within the time constraints of clinical duties; a group mentorship model may help overcome this barrier. Group mentorship can aid in personal and professional development because no single mentor can effectively meet every mentoring need of an individual.¹² Group mentorship also allows for the exchange of ideas among individuals with different backgrounds and the ability to utilize the strengths of each member of the group. For example, a member may have methodological expertise, while another may be skilled in grantsmanship. A team of mentors may be more beneficial for both the mentors (eg, establish a more manageable workload) and the mentee (eg, gains a broader perspective of expertise) when compared to having a single mentor.³

Peer mentorship within the group setting may also yield additional benefits. For example, having a supportive peer group may help reduce stress levels and burnout, while also improving overall well-being.^3,13 Formal mentorship programs do not frequently discuss concerns such as work-life balance, so including peers as mentors may help fill this void.⁹ Peer mentorship has also been found to be beneficial in providing mentees with pooled resources and shared learning.^12,13 This article describes the components, benefits, impacts, and challenges of a group research mentorship program for VA clinicians interested in conducting VA-relevant research.

Program Description

The VA Clinical Research Mentorship Program was initiated at the VA Ann Arbor Healthcare System (VAAAHS) in October 2015 by the Chief of Medicine to assist VA clinician investigators with developing and submitting VA clinical science and health services research grant applications. The program offers group and one-on-one consultation services through the expertise of 2 experienced investigators/faculty mentors who also serve as program directors, each of whom devote about 3 to 5 hours per month to activities associated with the mentorship program (eg, attending the meeting, reviewing materials sent by mentees, and one-on-one discussions with mentees).

The program also fostered peer-led mentorship. This encourages all attendees to provide feedback during group sessions and communication by mentees outside the group sessions. An experienced project manager serves as program coordinator and contributes about 4 hours per month for activities such as attending, scheduling, and sending reminders for each meeting, distributing handouts, reviewing materials, and answering mentee’s questions via email. A statistician and additional research staff (ie, an epidemiologist and research assistant) do not attend the recurring meetings, but are available for offline consultation as needed. The program runs on a 12-month cycle with regular meetings occurring twice monthly during the 9-month academic period. Resources to support the program, primarily program director(s) and project coordinator effort, are provided by the Chief of Medicine and through the VAAAHS affiliated VA Health Systems Research (formerly Health Services Research & Development) Center of Innovation.

Invitations for new mentees are sent annually. Mentees expressing interest in the program outside of its annual recruitment period are evaluated for inclusion on a rolling basis. Recruitment begins with the program coordinator sending email notifications to all VAAAHS Medicine Service faculty, section chiefs, and division chiefs at the VAAAHS academic affiliate. Recipients are encouraged to distribute the announcement to eligible applicants and refer them to the application materials for entry consideration into the program. The application consists of the applicant’s curriculum vitae and a 1-page summary that includes a description of their research area of interest, how it is relevant to the VA, in addition to an idea for a research study, its potential significance, and proposed methodology. Applicant materials are reviewed by the program coordinator and program directors. The applicants are evaluated using a simple scoring approach that focuses on the applicant’s research area and agenda, past research training, past research productivity, potential for obtaining VA funding, and whether they have sufficient research time.

Program eligibility initially required being a physician with ≥ 1/8 VA appointment from the Medicine Service. However, clinicians with clinical appointments from other VA services are also accepted for participation as needed. Applicants must have previous research experience and have a career goal to obtain external funding for conducting and publishing original research. Those who have previously served as a principal investigator on a funded VA grant proposal are not eligible as new applicants but can remain in the program as peer mentors. The number of annual applicants varies and ranges from 1 to 11; on average, about 90% of applicants receive invitations to join the program.

Sessions

The program holds recurring meetings twice monthly for 1 hour during the 9-month academic year. However, program directors are available year-round, and mentees are encouraged to communicate questions or concerns via email during nonacademic months. Prior to the COVID-19 pandemic, all meetings were held in-person. However, the group pivoted to virtual meetings and continues to utilize this format. The dedicated program coordinator is responsible for coordinating meetings and distributing meeting materials.

Each session is informal, flexible, and supportive. Attendance is not enforced, and mentees are allowed to join meetings as their schedules permit; however, program directors and program coordinator attend each meeting. In advance of each session, the program coordinator sends out a call for agenda items to all active members invited to discuss any research related items. Each mentee presents their ideas to lead the discussion for their portion of the meeting with no defined format required.

A variety of topics are covered including, but not limited to: (1) grant-specific concerns (eg, questions related to specific aim pages, grantsmanship, postsubmission comments from reviewers, or postaward logistics); (2) research procedures (eg, questions related to methodological practices or institutional review board concerns); (3) manuscript or presentation preparation; and (4) careerrelated issues. The program coordinator distributes handouts prior to meetings and mentees may record their presentations. These handouts may include, but are not limited to, specific aims pages, analytical plans, grant solicitations, and PowerPoint presentations. If a resource that can benefit the entire group is mentioned during the meeting, the program coordinator is responsible for distribution.

The program follows a group facilitated discussion format. Program directors facilitate each meeting, but input is encouraged from all attendees. This model allows for mentees to learn from the faculty mentors as well as peer mentees in a simultaneous and efficient fashion. Group discussions foster collective problem solving, peer support, and resource sharing that would not be possible through individualized mentorship. Participants have access to varied expertise during each session which reduces the need to seek specialized help elsewhere. Participants are also encouraged to contact the program directors or research staff for consultation as needed. Some one-on-one consultations have transitioned to a more sustained and ongoing mentorship relationship between a program director and mentee, but most are often brief email exchanges or a single meeting.

Participants

Since its inception in 2015, 35 clinicians have enrolled in the program. The mentees are equally distributed by sex and practice in a variety of disciplines including gastroenterology, hematology/oncology, cardiology, and general medicine (Table 1). Mentees have submitted 33 grant proposals addressing a variety of health care issues to a diverse group of federal and nonfederal funding agencies (Table 2). As of May 15, 2024, 19 (58%) of the submitted applications have been funded.

Many factors contribute to a successfully funded grant application, and several mentees report that participating in the mentorship program was helpful. For example, a mentee became the first lead investigator for a VA Cooperative Studies Program funded at VAAAHS. The VA Cooperative Studies Program, a division of the Office of Research and Development, plans and conducts large multicenter clinical trials and epidemiological studies within the VA via a vast network of clinician investigators, statisticians, and other key research experts.¹⁴

Several program mentees have also received VA Clinical Science Research and Development Career Development Awards. The VA Career Development program supports investigators during their early research careers with a goal of retaining talented researchers committed to improving the health and care of veterans.¹⁵

Survey Responses

Mentee productivity and updates are tracked through direct mentee input, as requested by the program coordinator. Since 2022, participants could complete an end-of-year survey based on an assessment tool used in a VAAAHS nonresearch mentorship program.¹⁶ The survey, distributed to mentees and program directors, requests feedback on logistics (eg, if the meeting was a good use of time and barriers to attendance); perceptions of effectiveness (eg, ability to discuss agenda items, helpfulness with setting and reaching research goals, and quality of mentors’ feedback); and the impact of the mentoring program on work satisfaction and clinician burnout. Respondents are also encouraged to leave open-ended qualitative feedback.

To date the survey has elicited 19 responses. Seventeen (89%) indicated that they agree or strongly agree the meetings were an effective use of their time and 11 (58%) indicated that they were able to discuss all or most of the items they wanted to during the meeting. Sixteen respondents (84%) agreed the program helped them set and achieve their research goals and 14 respondents (74%) agreed the feedback they received during the meeting was specific, actionable, and focused on how to improve their research agenda. Seventeen respondents (89%) agreed the program increased their work satisfaction, while 13 respondents (68%) felt the program reduced levels of clinician burnout.

As attendance was not mandatory, the survey asked participants how often they attended meetings during the past year. Responses were mixed: 4 (21%) respondents attended regularly (12 to 16 times per year) and 8 (42%) attended most sessions (8 to 11 times per year). Noted barriers to attendance included conflicts with patient care activities and conflicts with other high priority meetings.

Mentees also provided qualitive feedback regarding the program. They highlighted the supportive environment, valuable expertise of the mentors, and usefulness of obtaining tailored feedback from the group. “This group is an amazing resource to anyone developing a research career,” a mentee noted, adding that the program directors “fostered an incredibly supportive group where research ideas and methodology can be explored in a nonthreatening and creative environment.”

Conclusions

This mentorship program aims to help aspiring VA clinician investigators develop and submit competitive research grant applications. The addition of the program to the existing robust research environments at VAAAHS and its academic affiliate appears to have contributed to this success, with 58% of applications submitted by program mentees receiving funding.

In addition to funding success, we also found that most participants have a favorable impression of the program. Of the participants who responded to the program evaluation survey, nearly all indicated the program was an effective use of their time. The program also appeared to increase work satisfaction and reduce levels of clinician burnout. Barriers to attendance were also noted, with the most frequent being scheduling conflicts.

This program’s format includes facilitated group discussion as well as peer mentorship. This collaborative structure allows for an efficient and rich learning experience. Feedback from multiple perspectives encourages natural networking and relationship building. Incorporating the collective wisdom of the faculty mentors and peer mentees is beneficial; it not only empowers the mentees but also enriches the experience for the mentors. This program can serve as a model for other VA facilities—or non-VA academic medical centers—to enhance their research programs.

Supporting meaningful research that has a positive impact on the health and quality of life of veterans is a priority of the US Department of Veterans Affairs Office of Research and Development.¹ For nearly a century, VA researchers have been conducting high quality studies. To continue this trajectory, it is imperative to attract, train, and retain exceptional investigators while nurturing their development throughout their careers.²

Mentorship is defined as guidance provided by an experienced and trusted party to another (usually junior) individual with the intent of helping the person succeed. It benefits the mentee, mentor, and their institutions.³ Mentorship is crucial for personal and professional development as well as productivity, which may help reduce clinician burnout.^4-7 Conversely, a lack of mentorship could have negative effects on work satisfaction and stagnate career progression.⁸

Mentorship is vital for developing and advancing a VA investigator’s research agenda. Funding, grant writing, and research design were among the most discussed topics in a large comprehensive mentorship program for academic faculty.⁹ However, there are several known barriers to effective research mentorship; among them include a lack of resources, time constraints, and competing clinical priorities.^10,11

Finding time for effective one-on-one research mentoring is difficult within the time constraints of clinical duties; a group mentorship model may help overcome this barrier. Group mentorship can aid in personal and professional development because no single mentor can effectively meet every mentoring need of an individual.¹² Group mentorship also allows for the exchange of ideas among individuals with different backgrounds and the ability to utilize the strengths of each member of the group. For example, a member may have methodological expertise, while another may be skilled in grantsmanship. A team of mentors may be more beneficial for both the mentors (eg, establish a more manageable workload) and the mentee (eg, gains a broader perspective of expertise) when compared to having a single mentor.³

Peer mentorship within the group setting may also yield additional benefits. For example, having a supportive peer group may help reduce stress levels and burnout, while also improving overall well-being.^3,13 Formal mentorship programs do not frequently discuss concerns such as work-life balance, so including peers as mentors may help fill this void.⁹ Peer mentorship has also been found to be beneficial in providing mentees with pooled resources and shared learning.^12,13 This article describes the components, benefits, impacts, and challenges of a group research mentorship program for VA clinicians interested in conducting VA-relevant research.

Program Description

The VA Clinical Research Mentorship Program was initiated at the VA Ann Arbor Healthcare System (VAAAHS) in October 2015 by the Chief of Medicine to assist VA clinician investigators with developing and submitting VA clinical science and health services research grant applications. The program offers group and one-on-one consultation services through the expertise of 2 experienced investigators/faculty mentors who also serve as program directors, each of whom devote about 3 to 5 hours per month to activities associated with the mentorship program (eg, attending the meeting, reviewing materials sent by mentees, and one-on-one discussions with mentees).

The program also fostered peer-led mentorship. This encourages all attendees to provide feedback during group sessions and communication by mentees outside the group sessions. An experienced project manager serves as program coordinator and contributes about 4 hours per month for activities such as attending, scheduling, and sending reminders for each meeting, distributing handouts, reviewing materials, and answering mentee’s questions via email. A statistician and additional research staff (ie, an epidemiologist and research assistant) do not attend the recurring meetings, but are available for offline consultation as needed. The program runs on a 12-month cycle with regular meetings occurring twice monthly during the 9-month academic period. Resources to support the program, primarily program director(s) and project coordinator effort, are provided by the Chief of Medicine and through the VAAAHS affiliated VA Health Systems Research (formerly Health Services Research & Development) Center of Innovation.

Invitations for new mentees are sent annually. Mentees expressing interest in the program outside of its annual recruitment period are evaluated for inclusion on a rolling basis. Recruitment begins with the program coordinator sending email notifications to all VAAAHS Medicine Service faculty, section chiefs, and division chiefs at the VAAAHS academic affiliate. Recipients are encouraged to distribute the announcement to eligible applicants and refer them to the application materials for entry consideration into the program. The application consists of the applicant’s curriculum vitae and a 1-page summary that includes a description of their research area of interest, how it is relevant to the VA, in addition to an idea for a research study, its potential significance, and proposed methodology. Applicant materials are reviewed by the program coordinator and program directors. The applicants are evaluated using a simple scoring approach that focuses on the applicant’s research area and agenda, past research training, past research productivity, potential for obtaining VA funding, and whether they have sufficient research time.

Program eligibility initially required being a physician with ≥ 1/8 VA appointment from the Medicine Service. However, clinicians with clinical appointments from other VA services are also accepted for participation as needed. Applicants must have previous research experience and have a career goal to obtain external funding for conducting and publishing original research. Those who have previously served as a principal investigator on a funded VA grant proposal are not eligible as new applicants but can remain in the program as peer mentors. The number of annual applicants varies and ranges from 1 to 11; on average, about 90% of applicants receive invitations to join the program.

Sessions

The program holds recurring meetings twice monthly for 1 hour during the 9-month academic year. However, program directors are available year-round, and mentees are encouraged to communicate questions or concerns via email during nonacademic months. Prior to the COVID-19 pandemic, all meetings were held in-person. However, the group pivoted to virtual meetings and continues to utilize this format. The dedicated program coordinator is responsible for coordinating meetings and distributing meeting materials.

Each session is informal, flexible, and supportive. Attendance is not enforced, and mentees are allowed to join meetings as their schedules permit; however, program directors and program coordinator attend each meeting. In advance of each session, the program coordinator sends out a call for agenda items to all active members invited to discuss any research related items. Each mentee presents their ideas to lead the discussion for their portion of the meeting with no defined format required.

A variety of topics are covered including, but not limited to: (1) grant-specific concerns (eg, questions related to specific aim pages, grantsmanship, postsubmission comments from reviewers, or postaward logistics); (2) research procedures (eg, questions related to methodological practices or institutional review board concerns); (3) manuscript or presentation preparation; and (4) careerrelated issues. The program coordinator distributes handouts prior to meetings and mentees may record their presentations. These handouts may include, but are not limited to, specific aims pages, analytical plans, grant solicitations, and PowerPoint presentations. If a resource that can benefit the entire group is mentioned during the meeting, the program coordinator is responsible for distribution.

The program follows a group facilitated discussion format. Program directors facilitate each meeting, but input is encouraged from all attendees. This model allows for mentees to learn from the faculty mentors as well as peer mentees in a simultaneous and efficient fashion. Group discussions foster collective problem solving, peer support, and resource sharing that would not be possible through individualized mentorship. Participants have access to varied expertise during each session which reduces the need to seek specialized help elsewhere. Participants are also encouraged to contact the program directors or research staff for consultation as needed. Some one-on-one consultations have transitioned to a more sustained and ongoing mentorship relationship between a program director and mentee, but most are often brief email exchanges or a single meeting.

Participants

Since its inception in 2015, 35 clinicians have enrolled in the program. The mentees are equally distributed by sex and practice in a variety of disciplines including gastroenterology, hematology/oncology, cardiology, and general medicine (Table 1). Mentees have submitted 33 grant proposals addressing a variety of health care issues to a diverse group of federal and nonfederal funding agencies (Table 2). As of May 15, 2024, 19 (58%) of the submitted applications have been funded.

Many factors contribute to a successfully funded grant application, and several mentees report that participating in the mentorship program was helpful. For example, a mentee became the first lead investigator for a VA Cooperative Studies Program funded at VAAAHS. The VA Cooperative Studies Program, a division of the Office of Research and Development, plans and conducts large multicenter clinical trials and epidemiological studies within the VA via a vast network of clinician investigators, statisticians, and other key research experts.¹⁴

Several program mentees have also received VA Clinical Science Research and Development Career Development Awards. The VA Career Development program supports investigators during their early research careers with a goal of retaining talented researchers committed to improving the health and care of veterans.¹⁵

Survey Responses

Mentee productivity and updates are tracked through direct mentee input, as requested by the program coordinator. Since 2022, participants could complete an end-of-year survey based on an assessment tool used in a VAAAHS nonresearch mentorship program.¹⁶ The survey, distributed to mentees and program directors, requests feedback on logistics (eg, if the meeting was a good use of time and barriers to attendance); perceptions of effectiveness (eg, ability to discuss agenda items, helpfulness with setting and reaching research goals, and quality of mentors’ feedback); and the impact of the mentoring program on work satisfaction and clinician burnout. Respondents are also encouraged to leave open-ended qualitative feedback.

To date the survey has elicited 19 responses. Seventeen (89%) indicated that they agree or strongly agree the meetings were an effective use of their time and 11 (58%) indicated that they were able to discuss all or most of the items they wanted to during the meeting. Sixteen respondents (84%) agreed the program helped them set and achieve their research goals and 14 respondents (74%) agreed the feedback they received during the meeting was specific, actionable, and focused on how to improve their research agenda. Seventeen respondents (89%) agreed the program increased their work satisfaction, while 13 respondents (68%) felt the program reduced levels of clinician burnout.

As attendance was not mandatory, the survey asked participants how often they attended meetings during the past year. Responses were mixed: 4 (21%) respondents attended regularly (12 to 16 times per year) and 8 (42%) attended most sessions (8 to 11 times per year). Noted barriers to attendance included conflicts with patient care activities and conflicts with other high priority meetings.

Mentees also provided qualitive feedback regarding the program. They highlighted the supportive environment, valuable expertise of the mentors, and usefulness of obtaining tailored feedback from the group. “This group is an amazing resource to anyone developing a research career,” a mentee noted, adding that the program directors “fostered an incredibly supportive group where research ideas and methodology can be explored in a nonthreatening and creative environment.”

Conclusions

This mentorship program aims to help aspiring VA clinician investigators develop and submit competitive research grant applications. The addition of the program to the existing robust research environments at VAAAHS and its academic affiliate appears to have contributed to this success, with 58% of applications submitted by program mentees receiving funding.

In addition to funding success, we also found that most participants have a favorable impression of the program. Of the participants who responded to the program evaluation survey, nearly all indicated the program was an effective use of their time. The program also appeared to increase work satisfaction and reduce levels of clinician burnout. Barriers to attendance were also noted, with the most frequent being scheduling conflicts.

This program’s format includes facilitated group discussion as well as peer mentorship. This collaborative structure allows for an efficient and rich learning experience. Feedback from multiple perspectives encourages natural networking and relationship building. Incorporating the collective wisdom of the faculty mentors and peer mentees is beneficial; it not only empowers the mentees but also enriches the experience for the mentors. This program can serve as a model for other VA facilities—or non-VA academic medical centers—to enhance their research programs.

Supporting meaningful research that has a positive impact on the health and quality of life of veterans is a priority of the US Department of Veterans Affairs Office of Research and Development.¹ For nearly a century, VA researchers have been conducting high quality studies. To continue this trajectory, it is imperative to attract, train, and retain exceptional investigators while nurturing their development throughout their careers.²

Mentorship is defined as guidance provided by an experienced and trusted party to another (usually junior) individual with the intent of helping the person succeed. It benefits the mentee, mentor, and their institutions.³ Mentorship is crucial for personal and professional development as well as productivity, which may help reduce clinician burnout.^4-7 Conversely, a lack of mentorship could have negative effects on work satisfaction and stagnate career progression.⁸

Mentorship is vital for developing and advancing a VA investigator’s research agenda. Funding, grant writing, and research design were among the most discussed topics in a large comprehensive mentorship program for academic faculty.⁹ However, there are several known barriers to effective research mentorship; among them include a lack of resources, time constraints, and competing clinical priorities.^10,11

Finding time for effective one-on-one research mentoring is difficult within the time constraints of clinical duties; a group mentorship model may help overcome this barrier. Group mentorship can aid in personal and professional development because no single mentor can effectively meet every mentoring need of an individual.¹² Group mentorship also allows for the exchange of ideas among individuals with different backgrounds and the ability to utilize the strengths of each member of the group. For example, a member may have methodological expertise, while another may be skilled in grantsmanship. A team of mentors may be more beneficial for both the mentors (eg, establish a more manageable workload) and the mentee (eg, gains a broader perspective of expertise) when compared to having a single mentor.³

Peer mentorship within the group setting may also yield additional benefits. For example, having a supportive peer group may help reduce stress levels and burnout, while also improving overall well-being.^3,13 Formal mentorship programs do not frequently discuss concerns such as work-life balance, so including peers as mentors may help fill this void.⁹ Peer mentorship has also been found to be beneficial in providing mentees with pooled resources and shared learning.^12,13 This article describes the components, benefits, impacts, and challenges of a group research mentorship program for VA clinicians interested in conducting VA-relevant research.

Program Description

The VA Clinical Research Mentorship Program was initiated at the VA Ann Arbor Healthcare System (VAAAHS) in October 2015 by the Chief of Medicine to assist VA clinician investigators with developing and submitting VA clinical science and health services research grant applications. The program offers group and one-on-one consultation services through the expertise of 2 experienced investigators/faculty mentors who also serve as program directors, each of whom devote about 3 to 5 hours per month to activities associated with the mentorship program (eg, attending the meeting, reviewing materials sent by mentees, and one-on-one discussions with mentees).

The program also fostered peer-led mentorship. This encourages all attendees to provide feedback during group sessions and communication by mentees outside the group sessions. An experienced project manager serves as program coordinator and contributes about 4 hours per month for activities such as attending, scheduling, and sending reminders for each meeting, distributing handouts, reviewing materials, and answering mentee’s questions via email. A statistician and additional research staff (ie, an epidemiologist and research assistant) do not attend the recurring meetings, but are available for offline consultation as needed. The program runs on a 12-month cycle with regular meetings occurring twice monthly during the 9-month academic period. Resources to support the program, primarily program director(s) and project coordinator effort, are provided by the Chief of Medicine and through the VAAAHS affiliated VA Health Systems Research (formerly Health Services Research & Development) Center of Innovation.

Invitations for new mentees are sent annually. Mentees expressing interest in the program outside of its annual recruitment period are evaluated for inclusion on a rolling basis. Recruitment begins with the program coordinator sending email notifications to all VAAAHS Medicine Service faculty, section chiefs, and division chiefs at the VAAAHS academic affiliate. Recipients are encouraged to distribute the announcement to eligible applicants and refer them to the application materials for entry consideration into the program. The application consists of the applicant’s curriculum vitae and a 1-page summary that includes a description of their research area of interest, how it is relevant to the VA, in addition to an idea for a research study, its potential significance, and proposed methodology. Applicant materials are reviewed by the program coordinator and program directors. The applicants are evaluated using a simple scoring approach that focuses on the applicant’s research area and agenda, past research training, past research productivity, potential for obtaining VA funding, and whether they have sufficient research time.

Program eligibility initially required being a physician with ≥ 1/8 VA appointment from the Medicine Service. However, clinicians with clinical appointments from other VA services are also accepted for participation as needed. Applicants must have previous research experience and have a career goal to obtain external funding for conducting and publishing original research. Those who have previously served as a principal investigator on a funded VA grant proposal are not eligible as new applicants but can remain in the program as peer mentors. The number of annual applicants varies and ranges from 1 to 11; on average, about 90% of applicants receive invitations to join the program.

Sessions

The program holds recurring meetings twice monthly for 1 hour during the 9-month academic year. However, program directors are available year-round, and mentees are encouraged to communicate questions or concerns via email during nonacademic months. Prior to the COVID-19 pandemic, all meetings were held in-person. However, the group pivoted to virtual meetings and continues to utilize this format. The dedicated program coordinator is responsible for coordinating meetings and distributing meeting materials.

Each session is informal, flexible, and supportive. Attendance is not enforced, and mentees are allowed to join meetings as their schedules permit; however, program directors and program coordinator attend each meeting. In advance of each session, the program coordinator sends out a call for agenda items to all active members invited to discuss any research related items. Each mentee presents their ideas to lead the discussion for their portion of the meeting with no defined format required.

A variety of topics are covered including, but not limited to: (1) grant-specific concerns (eg, questions related to specific aim pages, grantsmanship, postsubmission comments from reviewers, or postaward logistics); (2) research procedures (eg, questions related to methodological practices or institutional review board concerns); (3) manuscript or presentation preparation; and (4) careerrelated issues. The program coordinator distributes handouts prior to meetings and mentees may record their presentations. These handouts may include, but are not limited to, specific aims pages, analytical plans, grant solicitations, and PowerPoint presentations. If a resource that can benefit the entire group is mentioned during the meeting, the program coordinator is responsible for distribution.

The program follows a group facilitated discussion format. Program directors facilitate each meeting, but input is encouraged from all attendees. This model allows for mentees to learn from the faculty mentors as well as peer mentees in a simultaneous and efficient fashion. Group discussions foster collective problem solving, peer support, and resource sharing that would not be possible through individualized mentorship. Participants have access to varied expertise during each session which reduces the need to seek specialized help elsewhere. Participants are also encouraged to contact the program directors or research staff for consultation as needed. Some one-on-one consultations have transitioned to a more sustained and ongoing mentorship relationship between a program director and mentee, but most are often brief email exchanges or a single meeting.

Participants

Since its inception in 2015, 35 clinicians have enrolled in the program. The mentees are equally distributed by sex and practice in a variety of disciplines including gastroenterology, hematology/oncology, cardiology, and general medicine (Table 1). Mentees have submitted 33 grant proposals addressing a variety of health care issues to a diverse group of federal and nonfederal funding agencies (Table 2). As of May 15, 2024, 19 (58%) of the submitted applications have been funded.

Many factors contribute to a successfully funded grant application, and several mentees report that participating in the mentorship program was helpful. For example, a mentee became the first lead investigator for a VA Cooperative Studies Program funded at VAAAHS. The VA Cooperative Studies Program, a division of the Office of Research and Development, plans and conducts large multicenter clinical trials and epidemiological studies within the VA via a vast network of clinician investigators, statisticians, and other key research experts.¹⁴

Several program mentees have also received VA Clinical Science Research and Development Career Development Awards. The VA Career Development program supports investigators during their early research careers with a goal of retaining talented researchers committed to improving the health and care of veterans.¹⁵

Survey Responses

Mentee productivity and updates are tracked through direct mentee input, as requested by the program coordinator. Since 2022, participants could complete an end-of-year survey based on an assessment tool used in a VAAAHS nonresearch mentorship program.¹⁶ The survey, distributed to mentees and program directors, requests feedback on logistics (eg, if the meeting was a good use of time and barriers to attendance); perceptions of effectiveness (eg, ability to discuss agenda items, helpfulness with setting and reaching research goals, and quality of mentors’ feedback); and the impact of the mentoring program on work satisfaction and clinician burnout. Respondents are also encouraged to leave open-ended qualitative feedback.

To date the survey has elicited 19 responses. Seventeen (89%) indicated that they agree or strongly agree the meetings were an effective use of their time and 11 (58%) indicated that they were able to discuss all or most of the items they wanted to during the meeting. Sixteen respondents (84%) agreed the program helped them set and achieve their research goals and 14 respondents (74%) agreed the feedback they received during the meeting was specific, actionable, and focused on how to improve their research agenda. Seventeen respondents (89%) agreed the program increased their work satisfaction, while 13 respondents (68%) felt the program reduced levels of clinician burnout.

As attendance was not mandatory, the survey asked participants how often they attended meetings during the past year. Responses were mixed: 4 (21%) respondents attended regularly (12 to 16 times per year) and 8 (42%) attended most sessions (8 to 11 times per year). Noted barriers to attendance included conflicts with patient care activities and conflicts with other high priority meetings.

Mentees also provided qualitive feedback regarding the program. They highlighted the supportive environment, valuable expertise of the mentors, and usefulness of obtaining tailored feedback from the group. “This group is an amazing resource to anyone developing a research career,” a mentee noted, adding that the program directors “fostered an incredibly supportive group where research ideas and methodology can be explored in a nonthreatening and creative environment.”

Conclusions

This mentorship program aims to help aspiring VA clinician investigators develop and submit competitive research grant applications. The addition of the program to the existing robust research environments at VAAAHS and its academic affiliate appears to have contributed to this success, with 58% of applications submitted by program mentees receiving funding.

In addition to funding success, we also found that most participants have a favorable impression of the program. Of the participants who responded to the program evaluation survey, nearly all indicated the program was an effective use of their time. The program also appeared to increase work satisfaction and reduce levels of clinician burnout. Barriers to attendance were also noted, with the most frequent being scheduling conflicts.

This program’s format includes facilitated group discussion as well as peer mentorship. This collaborative structure allows for an efficient and rich learning experience. Feedback from multiple perspectives encourages natural networking and relationship building. Incorporating the collective wisdom of the faculty mentors and peer mentees is beneficial; it not only empowers the mentees but also enriches the experience for the mentors. This program can serve as a model for other VA facilities—or non-VA academic medical centers—to enhance their research programs.

As Veterans Day approaches, stores and restaurants will offer discounts and free meals to veterans. Children will write thank you letters, and citizens nationwide will raise flags to honor and thank veterans. We can never repay those who lost their life, health, or livelihood in defense of the nation. Since the American Revolution, and in gratitude for that incalculable debt, the US government, on behalf of the American public, has seen fit to grant a host of benefits and services to those who wore the uniform.^2,3 Among the best known are health care, burial services, compensation and pensions, home loans, and the GI Bill.

Less recognized yet arguably essential for the fair and consistent provision of these entitlements is a legal principle: the veteran’s canon. A canon is a system of rules or maxims used to interpret legal instruments, such as statutes. They are not rules but serve as a “principle that guides the interpretation of the text.”⁴ Since I am not a lawyer, I will undoubtedly oversimplify this legal principle, but I hope to get enough right to explain why the veteran’s canon should matter to federal health care professionals.

At its core, the veteran’s canon means that when the US Department of Veterans Affairs (VA) and a veteran have a legal dispute about VA benefits, the courts will give deference to the veteran. Underscoring that any ambiguity in the statute is resolved in the veteran’s favor, the canon is known in legal circles as the Gardner deference. This is a reference to a 1994 case in which a Korean War veteran underwent surgery in a VA facility for a herniated disc he alleged caused pain and weakness in his left lower extremity.⁵ Gardner argued that federal statutes 38 USC § 1151 underlying corresponding VA regulation 38 CFR § 3.358(c)(3) granted disability benefits to veterans injured during VA treatment. The VA denied the disability claim, contending the regulation restricted compensation to veterans whose injury was the fault of the VA; thus, the disability had to have been the result of negligent treatment or an unforeseen therapeutic accident.⁵

The case wound its way through various appeals boards and courts until the Supreme Court of the United States (SCOTUS) ruled that the statute’s context left no ambiguity, and that any care provided under VA auspices was covered under the statute. What is important for this column is that the justices opined that had ambiguity been present, it would have legally necessitated, “applying the rule that interpretive doubt is to be resolved in the veteran’s favor.”⁵ In Gardner’s case, the courts reaffirmed nearly 80 years of judicial precedent upholding the veteran’s canon.

Thirty years later, Rudisill v McDonough again questioned the veteran’s canon.⁶ Educational benefits, namely the GI Bill, were the issue in this case. Rudisill served during 3 different periods in the US Army, totaling 8 years. Two educational programs overlapped during Rudisill’s tenure in the military: the Montgomery GI Bill and Post-9/11 Veterans Educational Assistance Act. Rudisill had used a portion of his Montgomery benefits to fund his undergraduate education and now wished to use the more extensive Post-9/11 assistance to finance his graduate degree. Rudisill and the VA disagreed about when his combined benefits would be capped, either at 36 or 48 months. After working its way through appeals courts, SCOTUS was again called upon for judgment.

The justices found that Rudisill qualified under both programs and could use them in any order he wished up to the cap. The majority found no ambiguity in the statute; however, if interpretation was required, the majority of justices indicated that the veteran’s canon would have supported Rudisill. While this sounds like good news for veterans, 2 justices authored a dissenting opinion that questioned the constitutional grounding of the veteran’s canon, noting that the “canon appears to have developed almost by accident.”⁶ The minority opinion suggested that when the veteran’s canon allocates resources to pay for specific veteran benefits, other interests and groups are deprived of those same resources, resulting in potential inequity.⁷

The potential ethical import and clinical impact of striking down the veteran’s canon is serious. It is especially concerning given that in a recent case, the SCOTUS ruling struck down another legal interpretation that also benefited the VA and ultimately veterans: the Chevron deference.⁸ This precedent held that when a legal dispute arises about the meaning of a specific federal agency regulation or policy, the courts should defer to the federal agency’s presumably superior understanding of the matter. The principle places the locus of decision-making with the subject-matter experts of the respective agency rather than the courts.

Ironically, given the legislative purposes of both interpretive principles, their overturning would likely introduce much more uncertainty, variation, and unpredictability in cases involving veteran benefits. This is bad news for both veterans and the VA. Veterans might not prevail as often in court when they have a reasonable claim, leading to more aggressive challenges. In response, the VA would have a heavier and more costly burden of administrative proof to defend sound decisions.⁹ Recently, the VA has tried to reduce the backlog of claims. The inability to have legal recourse to Chevron or Gardener could result in even more delay in adjudicating veterans’ claims that enable them to access benefits and services, already an object of congressional pressure.¹⁰

Courts will continue to debate the issue with another judicial test of the canon on the current SCOTUS docket (Bufkin v McDonough).¹¹ The veteran’s canon was put in place to equalize the power differential between the VA and the veteran: in administrative language, to make it more likely than not that the veteran would prevail when regulations were ambiguous. There are many legal and political rationales for veteran’s canon, including enabling veterans to file claims for service-connected illnesses. The veteran’s cannon helped Vietnam War-era veterans receive VA care while researchers were still studying the sequela of Agent Orange exposure. ¹² The legislative purpose of the veteran’s canon is the same as that of all VA benefits and services commemorated on Veterans Day. As expressed by SCOTUS justices in the wake of World War II, the benefit statutes should be “liberally construed for the benefit of those who left private life to serve their country in its hour of greatest need.”¹³

As Veterans Day approaches, stores and restaurants will offer discounts and free meals to veterans. Children will write thank you letters, and citizens nationwide will raise flags to honor and thank veterans. We can never repay those who lost their life, health, or livelihood in defense of the nation. Since the American Revolution, and in gratitude for that incalculable debt, the US government, on behalf of the American public, has seen fit to grant a host of benefits and services to those who wore the uniform.^2,3 Among the best known are health care, burial services, compensation and pensions, home loans, and the GI Bill.

Less recognized yet arguably essential for the fair and consistent provision of these entitlements is a legal principle: the veteran’s canon. A canon is a system of rules or maxims used to interpret legal instruments, such as statutes. They are not rules but serve as a “principle that guides the interpretation of the text.”⁴ Since I am not a lawyer, I will undoubtedly oversimplify this legal principle, but I hope to get enough right to explain why the veteran’s canon should matter to federal health care professionals.

At its core, the veteran’s canon means that when the US Department of Veterans Affairs (VA) and a veteran have a legal dispute about VA benefits, the courts will give deference to the veteran. Underscoring that any ambiguity in the statute is resolved in the veteran’s favor, the canon is known in legal circles as the Gardner deference. This is a reference to a 1994 case in which a Korean War veteran underwent surgery in a VA facility for a herniated disc he alleged caused pain and weakness in his left lower extremity.⁵ Gardner argued that federal statutes 38 USC § 1151 underlying corresponding VA regulation 38 CFR § 3.358(c)(3) granted disability benefits to veterans injured during VA treatment. The VA denied the disability claim, contending the regulation restricted compensation to veterans whose injury was the fault of the VA; thus, the disability had to have been the result of negligent treatment or an unforeseen therapeutic accident.⁵

The case wound its way through various appeals boards and courts until the Supreme Court of the United States (SCOTUS) ruled that the statute’s context left no ambiguity, and that any care provided under VA auspices was covered under the statute. What is important for this column is that the justices opined that had ambiguity been present, it would have legally necessitated, “applying the rule that interpretive doubt is to be resolved in the veteran’s favor.”⁵ In Gardner’s case, the courts reaffirmed nearly 80 years of judicial precedent upholding the veteran’s canon.

Thirty years later, Rudisill v McDonough again questioned the veteran’s canon.⁶ Educational benefits, namely the GI Bill, were the issue in this case. Rudisill served during 3 different periods in the US Army, totaling 8 years. Two educational programs overlapped during Rudisill’s tenure in the military: the Montgomery GI Bill and Post-9/11 Veterans Educational Assistance Act. Rudisill had used a portion of his Montgomery benefits to fund his undergraduate education and now wished to use the more extensive Post-9/11 assistance to finance his graduate degree. Rudisill and the VA disagreed about when his combined benefits would be capped, either at 36 or 48 months. After working its way through appeals courts, SCOTUS was again called upon for judgment.

The justices found that Rudisill qualified under both programs and could use them in any order he wished up to the cap. The majority found no ambiguity in the statute; however, if interpretation was required, the majority of justices indicated that the veteran’s canon would have supported Rudisill. While this sounds like good news for veterans, 2 justices authored a dissenting opinion that questioned the constitutional grounding of the veteran’s canon, noting that the “canon appears to have developed almost by accident.”⁶ The minority opinion suggested that when the veteran’s canon allocates resources to pay for specific veteran benefits, other interests and groups are deprived of those same resources, resulting in potential inequity.⁷

The potential ethical import and clinical impact of striking down the veteran’s canon is serious. It is especially concerning given that in a recent case, the SCOTUS ruling struck down another legal interpretation that also benefited the VA and ultimately veterans: the Chevron deference.⁸ This precedent held that when a legal dispute arises about the meaning of a specific federal agency regulation or policy, the courts should defer to the federal agency’s presumably superior understanding of the matter. The principle places the locus of decision-making with the subject-matter experts of the respective agency rather than the courts.

Ironically, given the legislative purposes of both interpretive principles, their overturning would likely introduce much more uncertainty, variation, and unpredictability in cases involving veteran benefits. This is bad news for both veterans and the VA. Veterans might not prevail as often in court when they have a reasonable claim, leading to more aggressive challenges. In response, the VA would have a heavier and more costly burden of administrative proof to defend sound decisions.⁹ Recently, the VA has tried to reduce the backlog of claims. The inability to have legal recourse to Chevron or Gardener could result in even more delay in adjudicating veterans’ claims that enable them to access benefits and services, already an object of congressional pressure.¹⁰

Courts will continue to debate the issue with another judicial test of the canon on the current SCOTUS docket (Bufkin v McDonough).¹¹ The veteran’s canon was put in place to equalize the power differential between the VA and the veteran: in administrative language, to make it more likely than not that the veteran would prevail when regulations were ambiguous. There are many legal and political rationales for veteran’s canon, including enabling veterans to file claims for service-connected illnesses. The veteran’s cannon helped Vietnam War-era veterans receive VA care while researchers were still studying the sequela of Agent Orange exposure. ¹² The legislative purpose of the veteran’s canon is the same as that of all VA benefits and services commemorated on Veterans Day. As expressed by SCOTUS justices in the wake of World War II, the benefit statutes should be “liberally construed for the benefit of those who left private life to serve their country in its hour of greatest need.”¹³

As Veterans Day approaches, stores and restaurants will offer discounts and free meals to veterans. Children will write thank you letters, and citizens nationwide will raise flags to honor and thank veterans. We can never repay those who lost their life, health, or livelihood in defense of the nation. Since the American Revolution, and in gratitude for that incalculable debt, the US government, on behalf of the American public, has seen fit to grant a host of benefits and services to those who wore the uniform.^2,3 Among the best known are health care, burial services, compensation and pensions, home loans, and the GI Bill.

Less recognized yet arguably essential for the fair and consistent provision of these entitlements is a legal principle: the veteran’s canon. A canon is a system of rules or maxims used to interpret legal instruments, such as statutes. They are not rules but serve as a “principle that guides the interpretation of the text.”⁴ Since I am not a lawyer, I will undoubtedly oversimplify this legal principle, but I hope to get enough right to explain why the veteran’s canon should matter to federal health care professionals.

At its core, the veteran’s canon means that when the US Department of Veterans Affairs (VA) and a veteran have a legal dispute about VA benefits, the courts will give deference to the veteran. Underscoring that any ambiguity in the statute is resolved in the veteran’s favor, the canon is known in legal circles as the Gardner deference. This is a reference to a 1994 case in which a Korean War veteran underwent surgery in a VA facility for a herniated disc he alleged caused pain and weakness in his left lower extremity.⁵ Gardner argued that federal statutes 38 USC § 1151 underlying corresponding VA regulation 38 CFR § 3.358(c)(3) granted disability benefits to veterans injured during VA treatment. The VA denied the disability claim, contending the regulation restricted compensation to veterans whose injury was the fault of the VA; thus, the disability had to have been the result of negligent treatment or an unforeseen therapeutic accident.⁵

The case wound its way through various appeals boards and courts until the Supreme Court of the United States (SCOTUS) ruled that the statute’s context left no ambiguity, and that any care provided under VA auspices was covered under the statute. What is important for this column is that the justices opined that had ambiguity been present, it would have legally necessitated, “applying the rule that interpretive doubt is to be resolved in the veteran’s favor.”⁵ In Gardner’s case, the courts reaffirmed nearly 80 years of judicial precedent upholding the veteran’s canon.

Thirty years later, Rudisill v McDonough again questioned the veteran’s canon.⁶ Educational benefits, namely the GI Bill, were the issue in this case. Rudisill served during 3 different periods in the US Army, totaling 8 years. Two educational programs overlapped during Rudisill’s tenure in the military: the Montgomery GI Bill and Post-9/11 Veterans Educational Assistance Act. Rudisill had used a portion of his Montgomery benefits to fund his undergraduate education and now wished to use the more extensive Post-9/11 assistance to finance his graduate degree. Rudisill and the VA disagreed about when his combined benefits would be capped, either at 36 or 48 months. After working its way through appeals courts, SCOTUS was again called upon for judgment.

The justices found that Rudisill qualified under both programs and could use them in any order he wished up to the cap. The majority found no ambiguity in the statute; however, if interpretation was required, the majority of justices indicated that the veteran’s canon would have supported Rudisill. While this sounds like good news for veterans, 2 justices authored a dissenting opinion that questioned the constitutional grounding of the veteran’s canon, noting that the “canon appears to have developed almost by accident.”⁶ The minority opinion suggested that when the veteran’s canon allocates resources to pay for specific veteran benefits, other interests and groups are deprived of those same resources, resulting in potential inequity.⁷

The potential ethical import and clinical impact of striking down the veteran’s canon is serious. It is especially concerning given that in a recent case, the SCOTUS ruling struck down another legal interpretation that also benefited the VA and ultimately veterans: the Chevron deference.⁸ This precedent held that when a legal dispute arises about the meaning of a specific federal agency regulation or policy, the courts should defer to the federal agency’s presumably superior understanding of the matter. The principle places the locus of decision-making with the subject-matter experts of the respective agency rather than the courts.

Ironically, given the legislative purposes of both interpretive principles, their overturning would likely introduce much more uncertainty, variation, and unpredictability in cases involving veteran benefits. This is bad news for both veterans and the VA. Veterans might not prevail as often in court when they have a reasonable claim, leading to more aggressive challenges. In response, the VA would have a heavier and more costly burden of administrative proof to defend sound decisions.⁹ Recently, the VA has tried to reduce the backlog of claims. The inability to have legal recourse to Chevron or Gardener could result in even more delay in adjudicating veterans’ claims that enable them to access benefits and services, already an object of congressional pressure.¹⁰

Courts will continue to debate the issue with another judicial test of the canon on the current SCOTUS docket (Bufkin v McDonough).¹¹ The veteran’s canon was put in place to equalize the power differential between the VA and the veteran: in administrative language, to make it more likely than not that the veteran would prevail when regulations were ambiguous. There are many legal and political rationales for veteran’s canon, including enabling veterans to file claims for service-connected illnesses. The veteran’s cannon helped Vietnam War-era veterans receive VA care while researchers were still studying the sequela of Agent Orange exposure. ¹² The legislative purpose of the veteran’s canon is the same as that of all VA benefits and services commemorated on Veterans Day. As expressed by SCOTUS justices in the wake of World War II, the benefit statutes should be “liberally construed for the benefit of those who left private life to serve their country in its hour of greatest need.”¹³

The COVID-19 pandemic affected diverse workplaces globally, leading to temporary and permanent changes across the health care landscape. Included among the impacted areas of care were epilepsy and electroencephalogram (EEG) clinicians and services. Surveys among epilepsy specialists and neurophysiologists conducted at the onset of the pandemic to evaluate working conditions include analyses from the American Epilepsy Society (AES), the National Association of Epilepsy Centers (NAEC), the International League Against Epilepsy, and an Italian national survey.^1-4 These investigations revealed reductions in epilepsy monitoring unit (EMU) admissions (23% decline), epilepsy surgery (6% decline), inpatient EEG (22% of respondents reported decline), and patients having difficulty accessing epilepsy professionals (28% of respondents reported decline) or obtaining medications (20% of respondents reported decline).^1-3

While such research provided evidence for changes to epilepsy care in 2020, there are limited data on subsequent adaptations during the pandemic. These studies did not incorporate data on the spread of COVID-19 or administrative workload numbers to analyze service delivery beyond self reports. This study aimed to address this gap in the literature by highlighting results from longitudinal national surveys conducted at the Epilepsy Centers of Excellence (ECoE), a specialty care service within the Veterans Health Administration (VHA), which annually serves > 9 million veterans.⁵ The ECoE represents epileptologists and neurophysiologists across the United States at the 17 primary facilities that were established at the time of this survey (2 ECoEs have been added since survey completion) in 4 geographical regions and for which other regional facilities refer patients for diagnostic services or specialty care.⁶

National surveys were conducted among the ECoE directors regarding adaptations made from May 2020 to June 2022 to provide a comprehensive account of limitations they experienced and how adjustments have been made to improve patient care. Survey responses were compared to administrative workload numbers and COVID-19 spread data from the Centers for Disease Control and Prevention (CDC) to provide a comprehensive analysis of performance during the pandemic.

METHODS

Data were collected as part of a quality improvement initiative by the VHA ECoE; institutional review board approval was not required. An 18-item survey covering 5 broad domains was sent to ECoE directors 4 separate times to accumulate data from 4 time periods: May to June 2020 (T1); December 2020 to February 2021 (T2); July to August 2021 (T3); and June to July 2022 (T4). These periods correspond to the following phases of the pandemic: T1, onset of pandemic; T2, vaccine availability; T3, Delta variant predominant; T4, Omicron variant predominant.

Data on the spread of COVID-19 were collected from the CDC archived dataset, US COVID-19 County Level of Community Transmission Historical Changes (Table 1).⁷ Administrative workload (patient counts) for EEG, EMU, and outpatient clinics were extracted from VHA administrative databases for the participating sites for the months prior to each survey: T1, April 2020; T2, November 2020; T3, June 2021; and T4, May 2022 (Table 2).

Survey Structure and Content

The survey was developed by the ECoE and was not validated prior to its use due to the time-sensitive nature of gathering information during the pandemic. The first survey (T1) was an emailed spreadsheet with open-ended questions to gauge availability of services (ie, outpatient clinic, EEG, EMU), assess whether safety precautions were being introduced, and understand whether national or local guidelines were thought to be helpful. Responses from this and subsequent surveys were standardized into yes/no and multiple choice formats. Subsequent surveys were administered online using a Research Electronic Data Capture tool.^8,9

Availability of outpatient epilepsy services across the 4 time periods were categorized as unlimited (in-person with no restrictions), limited (in-person with restrictions), planned (not currently performed but scheduled for the near future), and unavailable (no in-person services offered) (eAppendices 1-6, available in article PDF).

Statistical Analyses

Analyses were performed to compare survey responses to workload and CDC data on COVID-19 community spread. The following associations were examined: (1) CDC COVID-19 spread vs respondents’ perception of spread; (2) respondents’ perception of spread vs availability of services; (3) CDC COVID-19 spread vs availability of services; (4) respondents’ perception of spread vs workload; and (5) CDC COVID-19 spread vs workload. Availability of services was dichotomized for analyses, with limited or fully available services classified as available. As services were mostly open at T3 regardless of the spread of the virus, and the CDC COVID-19 spread classification for all sites was severe or high at T2 and T4, corresponding associations were not tested at these time points. For associations 1 through 3, Fisher exact tests were used; for associations 4 and 5, Mann-Whitney U tests (where the COVID-19 spread fell into 2 categories) and Kruskal-Wallis tests (for 3 categories of COVID-19 spread) were performed. All tests were 2-tailed and performed at 0.05 error rate. Bonferroni corrections were applied to adjust P values for multiple hypotheses tests.

RESULTS

From the 17 sites invited, responses at each time point were obtained from 13 (T1),17 (T2), 15 (T3), and 16 (T4) centers. There was no significant association between self-reported COVID-19 spread and CDC classification of COVID spread. There were no associations between COVID-19 community spread (respondent reported or CDC severity level) and outpatient clinic availability (self-reported or workload captured). At T3, a positive association was found between the CDC spread level and workload (P = .008), but this was not significant after Bonferroni correction (P = .06).

EEG availability surpassed EMU availability at all time points, although EMU services made some recovery at T3 and T4. No associations were found between COVID-19 community spread (self-reported or CDC severity level) and outpatient EEG or EMU availability (self-reported or workload captured). At T3, there was a positive association between EEG workload and CDC COVID-19 severity level (P = .04), but this was not significant after Bonferroni correction (P = .30). 

For outpatient EEG, staff and patient mask use were universally implemented by T2, while the use of full personal protective equipment (PPE) occurred at a subset of sites (T2, 6/17 [35%]; T3, 3/15 [20%]; T4: 4/16 [25%]). COVID-19 testing was rarely implemented prior to outpatient EEG (T1, 0 sites; T2, 1 site; T3, 1 site; T4, 0 sites). Within the EMU, safety precautions including COVID-19 testing, patient mask usage, staff mask usage, and aerosolization demonstrated a sustained majority usage across the 4 surveys.

National and Local Guidelines

The open-ended survey at T1 asked site directors, “Should there be national recommendations on how EEGs and related procedures should be done during the pandemic or should this be left to local conditions?” Responses were mixed, with 5 respondents desiring a national standard, 4 respondents favoring a local response, and 4 respondents believing a national standard should be in place but with modifications based on local outbreak levels and needs.

Surveys performed at T2 through T4 asked, “Which of the following do you feel was/will be helpful in adapting to COVID-19–related changes?” Overall, there was substantial agreement that guidelines were helpful. Most sites anticipated permanent changes in enhanced safety precautions and telehealth.

DISCUSSION

This longitudinal study across 4 time points describes how epilepsy services within the VHA and ECoE adapted to the COVID-19 pandemic. The first survey, conducted 2 months after COVID-19 was declared a pandemic, allowed a comparison with other concurrent US national surveys.^1,2,10 The subsequent surveys describe longitudinal adaptations to balance patient and staff safety with service availability and is a unique feature of the current report. Results demonstrate flexibility and adaptability by the ECoEs surveyed, which surprisingly did not show significant associations between CDC COVID-19 spread data and administrative workload data.

Trends in Availability of Services

The most significant impact of COVID-19 restrictions was during T1. There were no significant relationships between service availability/workload and objective CDC COVID-19 spread levels or subjective self-reported COVID-19 spread. Respondents’ perceptions of local COVID-19 spread showed no association with CDC COVID-19 spread data. It appears that subjective perception of spread may be unreliable and factors other than actual or perceived COVID-19 spread were likely driving patterns for service availability.

In-person outpatient visits were most impacted at T1, similar to other civilian surveys, with only 1 site reporting in-person outpatient visits without limitations.^1,2 These numbers significantly changed by T2, with all sites offering either limited or unlimited in-person visits. While the surveys did not evaluate factors leading to this rapid recovery, it may be related to the availability of COVID-19 vaccinations within the VHA during this time.¹¹ The US Department of Veterans Affairs was the first federal agency to mandate employee vaccination.¹² By the most recent time point (T4), all responding sites offered outpatient visits. Outpatient EEGs followed a similar trend, with T1 being the most restrictive and full, unrestricted outpatient EEGs available by T3. 

Fiscal year (FY) trends from ECoE annual reports suggest that encounters slowly recovered over the course of the pandemic. In FY 2019 there were 13,143 outpatient encounters and 6394 EEGs, which dropped to 8097 outpatient encounters and 4432 EEGs in FY 2020 before rising to 8489 outpatient encounters and 5604 EEGs in FY 2021 and 9772 outpatient encounters and 5062 EEGs in FY 2022. Thus, while clinicians described availability of services, patients may have remained hesitant or were otherwise unable to fulfill in-person appointments. The increased availability of home EEG (145 encounter days in 2021 and 436 encounter days in 2022) may be filling this gap. 

In contrast to outpatient clinics and EEG, EMU availability showed relatively slower reimplementation. In the last survey, about 30% of sites were still not offering EMU or had limited services. Early trends regarding reduced staffing and patient reluctance for elective admission cited in other surveys may have also affected EMU availability within the VHA.^2,13 Consistent with trends in availability, ECoE annual report data suggest EMU patient participation was about one-half of prepandemic rates: 3069 encounters in FY 2019 dropped to 1614 encounters in 2020. By 2021, rates were about two-thirds of prepandemic rates with 2058 encounters in 2021 and 2101 encounters in 2022.

Early survey results (T1) from this study echo trends from other surveys. In the AES survey (April to June 2020), about a quarter of respondents (22%) reported doing fewer EEG studies than usual. The Italian national survey (April 2020) revealed reduced presurgical evaluations (81%), ambulatory EEG (78%), standard EEG (5%) and long-term EEG (32%).⁴ In the NAEC survey (end of 2020)—which roughly corresponded to T2—outpatient EEGs were still < 75% of pre-COVID levels in one-half of the centers.

National and Local Guidelines

Both national and local guidelines were perceived as useful by most respondents, with national guidelines being more beneficial. This aligns with the NAEC survey, where there was a perceived need for detailed recommendations for PPE and COVID-19 testing of patients, visitors, and staff. Based on national and local guidelines, ECoE implemented safety procedures, as reflected in responses. Staff masking procedures appeared to be the most widely adopted for all services, while the use of full PPE waned as the pandemic progressed. COVID-19 testing was rarely used for routine outpatient visits but common in EMU admissions. This is similar to a survey conducted by the American Academy of Neurology which found full PPE implementation intermittently in outpatient settings and more frequently in inpatient settings.¹⁴

Telehealth Attitudes

While most sites anticipated permanent implementation of safety precautions and telehealth, the latter was consistently reported as more likely to be sustained. The VHA had a large and well-developed system of telehealth services that considerably predated the pandemic.^15,16 Through this established infrastructure, remote services were quickly increased across the VHA.^17-19 This telehealth structure was supplemented by the ability of VHA clinicians to practice across state lines, following a 2018 federal rule.²⁰ The AES survey noted the VHA ECoE's longstanding experience with telehealth as a model for telemedicine use in providing direct patient care, remote EEG analysis, and clinician-to-clinician consultation.¹

Trends in the number of telehealth patients seen, observed through patterns in ECoE annual reports are consistent with positive views toward this method of service provision. Specifically, these annual reports capture trends in Video Telehealth Clinic (local station), Video Telehealth Clinic (different station), Home Video Telehealth, Telephone Clinic, and eConsults. Though video telehealth at in-person stations had a precipitous drop in 2020 that continued to wane in subsequent years (898 encounters in 2019; 455 encounters in 2020; 90 encounters in 2021; 88 encounters in 2022), use of home video telehealth rose over time (143 encounters in 2019; 1003 encounters in 2020; 3206 encounters in 2021; 3315 encounters in 2022). Use of telephone services rose drastically in 2020 but has since become a less frequently used service method (2636 in 2019; 5923 in 2020; 5319 in 2021; 3704 in 2022).

Limitations

While the survey encouraged a high response rate, this limited its scope and interpretability. While the availability of services was evaluated, the underlying reasons were not queried. Follow-up questions about barriers to reopening may have allowed for a better understanding of why some services, such as EMU, continued to operate suboptimally later in the pandemic. Similarly, asking about unique strategies or barriers for telehealth would have allowed for a better understanding of its current and future use. We hypothesize that staffing changes during the pandemic may have influenced the availability of services, but changes to staffing were not assessed via the survey and were not readily available via other sources (eg, ECoE annual reports) at the time of publication. An additional limitation is the lack of comparable surveys in the literature for time points T2 to T4, as most analogous surveys were performed early in 2020.

Conclusions

This longitudinal study performed at 4 time points during the COVID-19 pandemic is the first to offer a comprehensive picture of changes to epilepsy and EEG services over time, given that other similar surveys lacked follow-up. Results reveal a significant limitation of services at VHA ECoE shortly after the onset of the pandemic, with return to near-complete operational status 2 years later. While safety precautions and telehealth are predicted to continue, telehealth is perceived as a more permanent change in services.

The COVID-19 pandemic affected diverse workplaces globally, leading to temporary and permanent changes across the health care landscape. Included among the impacted areas of care were epilepsy and electroencephalogram (EEG) clinicians and services. Surveys among epilepsy specialists and neurophysiologists conducted at the onset of the pandemic to evaluate working conditions include analyses from the American Epilepsy Society (AES), the National Association of Epilepsy Centers (NAEC), the International League Against Epilepsy, and an Italian national survey.^1-4 These investigations revealed reductions in epilepsy monitoring unit (EMU) admissions (23% decline), epilepsy surgery (6% decline), inpatient EEG (22% of respondents reported decline), and patients having difficulty accessing epilepsy professionals (28% of respondents reported decline) or obtaining medications (20% of respondents reported decline).^1-3

While such research provided evidence for changes to epilepsy care in 2020, there are limited data on subsequent adaptations during the pandemic. These studies did not incorporate data on the spread of COVID-19 or administrative workload numbers to analyze service delivery beyond self reports. This study aimed to address this gap in the literature by highlighting results from longitudinal national surveys conducted at the Epilepsy Centers of Excellence (ECoE), a specialty care service within the Veterans Health Administration (VHA), which annually serves > 9 million veterans.⁵ The ECoE represents epileptologists and neurophysiologists across the United States at the 17 primary facilities that were established at the time of this survey (2 ECoEs have been added since survey completion) in 4 geographical regions and for which other regional facilities refer patients for diagnostic services or specialty care.⁶

National surveys were conducted among the ECoE directors regarding adaptations made from May 2020 to June 2022 to provide a comprehensive account of limitations they experienced and how adjustments have been made to improve patient care. Survey responses were compared to administrative workload numbers and COVID-19 spread data from the Centers for Disease Control and Prevention (CDC) to provide a comprehensive analysis of performance during the pandemic.

METHODS

Data were collected as part of a quality improvement initiative by the VHA ECoE; institutional review board approval was not required. An 18-item survey covering 5 broad domains was sent to ECoE directors 4 separate times to accumulate data from 4 time periods: May to June 2020 (T1); December 2020 to February 2021 (T2); July to August 2021 (T3); and June to July 2022 (T4). These periods correspond to the following phases of the pandemic: T1, onset of pandemic; T2, vaccine availability; T3, Delta variant predominant; T4, Omicron variant predominant.

Data on the spread of COVID-19 were collected from the CDC archived dataset, US COVID-19 County Level of Community Transmission Historical Changes (Table 1).⁷ Administrative workload (patient counts) for EEG, EMU, and outpatient clinics were extracted from VHA administrative databases for the participating sites for the months prior to each survey: T1, April 2020; T2, November 2020; T3, June 2021; and T4, May 2022 (Table 2).

Survey Structure and Content

The survey was developed by the ECoE and was not validated prior to its use due to the time-sensitive nature of gathering information during the pandemic. The first survey (T1) was an emailed spreadsheet with open-ended questions to gauge availability of services (ie, outpatient clinic, EEG, EMU), assess whether safety precautions were being introduced, and understand whether national or local guidelines were thought to be helpful. Responses from this and subsequent surveys were standardized into yes/no and multiple choice formats. Subsequent surveys were administered online using a Research Electronic Data Capture tool.^8,9

Availability of outpatient epilepsy services across the 4 time periods were categorized as unlimited (in-person with no restrictions), limited (in-person with restrictions), planned (not currently performed but scheduled for the near future), and unavailable (no in-person services offered) (eAppendices 1-6, available in article PDF).

Statistical Analyses

Analyses were performed to compare survey responses to workload and CDC data on COVID-19 community spread. The following associations were examined: (1) CDC COVID-19 spread vs respondents’ perception of spread; (2) respondents’ perception of spread vs availability of services; (3) CDC COVID-19 spread vs availability of services; (4) respondents’ perception of spread vs workload; and (5) CDC COVID-19 spread vs workload. Availability of services was dichotomized for analyses, with limited or fully available services classified as available. As services were mostly open at T3 regardless of the spread of the virus, and the CDC COVID-19 spread classification for all sites was severe or high at T2 and T4, corresponding associations were not tested at these time points. For associations 1 through 3, Fisher exact tests were used; for associations 4 and 5, Mann-Whitney U tests (where the COVID-19 spread fell into 2 categories) and Kruskal-Wallis tests (for 3 categories of COVID-19 spread) were performed. All tests were 2-tailed and performed at 0.05 error rate. Bonferroni corrections were applied to adjust P values for multiple hypotheses tests.

RESULTS

From the 17 sites invited, responses at each time point were obtained from 13 (T1),17 (T2), 15 (T3), and 16 (T4) centers. There was no significant association between self-reported COVID-19 spread and CDC classification of COVID spread. There were no associations between COVID-19 community spread (respondent reported or CDC severity level) and outpatient clinic availability (self-reported or workload captured). At T3, a positive association was found between the CDC spread level and workload (P = .008), but this was not significant after Bonferroni correction (P = .06).

EEG availability surpassed EMU availability at all time points, although EMU services made some recovery at T3 and T4. No associations were found between COVID-19 community spread (self-reported or CDC severity level) and outpatient EEG or EMU availability (self-reported or workload captured). At T3, there was a positive association between EEG workload and CDC COVID-19 severity level (P = .04), but this was not significant after Bonferroni correction (P = .30). 

For outpatient EEG, staff and patient mask use were universally implemented by T2, while the use of full personal protective equipment (PPE) occurred at a subset of sites (T2, 6/17 [35%]; T3, 3/15 [20%]; T4: 4/16 [25%]). COVID-19 testing was rarely implemented prior to outpatient EEG (T1, 0 sites; T2, 1 site; T3, 1 site; T4, 0 sites). Within the EMU, safety precautions including COVID-19 testing, patient mask usage, staff mask usage, and aerosolization demonstrated a sustained majority usage across the 4 surveys.

National and Local Guidelines

The open-ended survey at T1 asked site directors, “Should there be national recommendations on how EEGs and related procedures should be done during the pandemic or should this be left to local conditions?” Responses were mixed, with 5 respondents desiring a national standard, 4 respondents favoring a local response, and 4 respondents believing a national standard should be in place but with modifications based on local outbreak levels and needs.

Surveys performed at T2 through T4 asked, “Which of the following do you feel was/will be helpful in adapting to COVID-19–related changes?” Overall, there was substantial agreement that guidelines were helpful. Most sites anticipated permanent changes in enhanced safety precautions and telehealth.

DISCUSSION

This longitudinal study across 4 time points describes how epilepsy services within the VHA and ECoE adapted to the COVID-19 pandemic. The first survey, conducted 2 months after COVID-19 was declared a pandemic, allowed a comparison with other concurrent US national surveys.^1,2,10 The subsequent surveys describe longitudinal adaptations to balance patient and staff safety with service availability and is a unique feature of the current report. Results demonstrate flexibility and adaptability by the ECoEs surveyed, which surprisingly did not show significant associations between CDC COVID-19 spread data and administrative workload data.

Trends in Availability of Services

The most significant impact of COVID-19 restrictions was during T1. There were no significant relationships between service availability/workload and objective CDC COVID-19 spread levels or subjective self-reported COVID-19 spread. Respondents’ perceptions of local COVID-19 spread showed no association with CDC COVID-19 spread data. It appears that subjective perception of spread may be unreliable and factors other than actual or perceived COVID-19 spread were likely driving patterns for service availability.

In-person outpatient visits were most impacted at T1, similar to other civilian surveys, with only 1 site reporting in-person outpatient visits without limitations.^1,2 These numbers significantly changed by T2, with all sites offering either limited or unlimited in-person visits. While the surveys did not evaluate factors leading to this rapid recovery, it may be related to the availability of COVID-19 vaccinations within the VHA during this time.¹¹ The US Department of Veterans Affairs was the first federal agency to mandate employee vaccination.¹² By the most recent time point (T4), all responding sites offered outpatient visits. Outpatient EEGs followed a similar trend, with T1 being the most restrictive and full, unrestricted outpatient EEGs available by T3. 

Fiscal year (FY) trends from ECoE annual reports suggest that encounters slowly recovered over the course of the pandemic. In FY 2019 there were 13,143 outpatient encounters and 6394 EEGs, which dropped to 8097 outpatient encounters and 4432 EEGs in FY 2020 before rising to 8489 outpatient encounters and 5604 EEGs in FY 2021 and 9772 outpatient encounters and 5062 EEGs in FY 2022. Thus, while clinicians described availability of services, patients may have remained hesitant or were otherwise unable to fulfill in-person appointments. The increased availability of home EEG (145 encounter days in 2021 and 436 encounter days in 2022) may be filling this gap. 

In contrast to outpatient clinics and EEG, EMU availability showed relatively slower reimplementation. In the last survey, about 30% of sites were still not offering EMU or had limited services. Early trends regarding reduced staffing and patient reluctance for elective admission cited in other surveys may have also affected EMU availability within the VHA.^2,13 Consistent with trends in availability, ECoE annual report data suggest EMU patient participation was about one-half of prepandemic rates: 3069 encounters in FY 2019 dropped to 1614 encounters in 2020. By 2021, rates were about two-thirds of prepandemic rates with 2058 encounters in 2021 and 2101 encounters in 2022.

Early survey results (T1) from this study echo trends from other surveys. In the AES survey (April to June 2020), about a quarter of respondents (22%) reported doing fewer EEG studies than usual. The Italian national survey (April 2020) revealed reduced presurgical evaluations (81%), ambulatory EEG (78%), standard EEG (5%) and long-term EEG (32%).⁴ In the NAEC survey (end of 2020)—which roughly corresponded to T2—outpatient EEGs were still < 75% of pre-COVID levels in one-half of the centers.

National and Local Guidelines

Both national and local guidelines were perceived as useful by most respondents, with national guidelines being more beneficial. This aligns with the NAEC survey, where there was a perceived need for detailed recommendations for PPE and COVID-19 testing of patients, visitors, and staff. Based on national and local guidelines, ECoE implemented safety procedures, as reflected in responses. Staff masking procedures appeared to be the most widely adopted for all services, while the use of full PPE waned as the pandemic progressed. COVID-19 testing was rarely used for routine outpatient visits but common in EMU admissions. This is similar to a survey conducted by the American Academy of Neurology which found full PPE implementation intermittently in outpatient settings and more frequently in inpatient settings.¹⁴

Telehealth Attitudes

While most sites anticipated permanent implementation of safety precautions and telehealth, the latter was consistently reported as more likely to be sustained. The VHA had a large and well-developed system of telehealth services that considerably predated the pandemic.^15,16 Through this established infrastructure, remote services were quickly increased across the VHA.^17-19 This telehealth structure was supplemented by the ability of VHA clinicians to practice across state lines, following a 2018 federal rule.²⁰ The AES survey noted the VHA ECoE's longstanding experience with telehealth as a model for telemedicine use in providing direct patient care, remote EEG analysis, and clinician-to-clinician consultation.¹

Trends in the number of telehealth patients seen, observed through patterns in ECoE annual reports are consistent with positive views toward this method of service provision. Specifically, these annual reports capture trends in Video Telehealth Clinic (local station), Video Telehealth Clinic (different station), Home Video Telehealth, Telephone Clinic, and eConsults. Though video telehealth at in-person stations had a precipitous drop in 2020 that continued to wane in subsequent years (898 encounters in 2019; 455 encounters in 2020; 90 encounters in 2021; 88 encounters in 2022), use of home video telehealth rose over time (143 encounters in 2019; 1003 encounters in 2020; 3206 encounters in 2021; 3315 encounters in 2022). Use of telephone services rose drastically in 2020 but has since become a less frequently used service method (2636 in 2019; 5923 in 2020; 5319 in 2021; 3704 in 2022).

Limitations

While the survey encouraged a high response rate, this limited its scope and interpretability. While the availability of services was evaluated, the underlying reasons were not queried. Follow-up questions about barriers to reopening may have allowed for a better understanding of why some services, such as EMU, continued to operate suboptimally later in the pandemic. Similarly, asking about unique strategies or barriers for telehealth would have allowed for a better understanding of its current and future use. We hypothesize that staffing changes during the pandemic may have influenced the availability of services, but changes to staffing were not assessed via the survey and were not readily available via other sources (eg, ECoE annual reports) at the time of publication. An additional limitation is the lack of comparable surveys in the literature for time points T2 to T4, as most analogous surveys were performed early in 2020.

Conclusions

This longitudinal study performed at 4 time points during the COVID-19 pandemic is the first to offer a comprehensive picture of changes to epilepsy and EEG services over time, given that other similar surveys lacked follow-up. Results reveal a significant limitation of services at VHA ECoE shortly after the onset of the pandemic, with return to near-complete operational status 2 years later. While safety precautions and telehealth are predicted to continue, telehealth is perceived as a more permanent change in services.

The COVID-19 pandemic affected diverse workplaces globally, leading to temporary and permanent changes across the health care landscape. Included among the impacted areas of care were epilepsy and electroencephalogram (EEG) clinicians and services. Surveys among epilepsy specialists and neurophysiologists conducted at the onset of the pandemic to evaluate working conditions include analyses from the American Epilepsy Society (AES), the National Association of Epilepsy Centers (NAEC), the International League Against Epilepsy, and an Italian national survey.^1-4 These investigations revealed reductions in epilepsy monitoring unit (EMU) admissions (23% decline), epilepsy surgery (6% decline), inpatient EEG (22% of respondents reported decline), and patients having difficulty accessing epilepsy professionals (28% of respondents reported decline) or obtaining medications (20% of respondents reported decline).^1-3

While such research provided evidence for changes to epilepsy care in 2020, there are limited data on subsequent adaptations during the pandemic. These studies did not incorporate data on the spread of COVID-19 or administrative workload numbers to analyze service delivery beyond self reports. This study aimed to address this gap in the literature by highlighting results from longitudinal national surveys conducted at the Epilepsy Centers of Excellence (ECoE), a specialty care service within the Veterans Health Administration (VHA), which annually serves > 9 million veterans.⁵ The ECoE represents epileptologists and neurophysiologists across the United States at the 17 primary facilities that were established at the time of this survey (2 ECoEs have been added since survey completion) in 4 geographical regions and for which other regional facilities refer patients for diagnostic services or specialty care.⁶

National surveys were conducted among the ECoE directors regarding adaptations made from May 2020 to June 2022 to provide a comprehensive account of limitations they experienced and how adjustments have been made to improve patient care. Survey responses were compared to administrative workload numbers and COVID-19 spread data from the Centers for Disease Control and Prevention (CDC) to provide a comprehensive analysis of performance during the pandemic.

METHODS

Data were collected as part of a quality improvement initiative by the VHA ECoE; institutional review board approval was not required. An 18-item survey covering 5 broad domains was sent to ECoE directors 4 separate times to accumulate data from 4 time periods: May to June 2020 (T1); December 2020 to February 2021 (T2); July to August 2021 (T3); and June to July 2022 (T4). These periods correspond to the following phases of the pandemic: T1, onset of pandemic; T2, vaccine availability; T3, Delta variant predominant; T4, Omicron variant predominant.

Data on the spread of COVID-19 were collected from the CDC archived dataset, US COVID-19 County Level of Community Transmission Historical Changes (Table 1).⁷ Administrative workload (patient counts) for EEG, EMU, and outpatient clinics were extracted from VHA administrative databases for the participating sites for the months prior to each survey: T1, April 2020; T2, November 2020; T3, June 2021; and T4, May 2022 (Table 2).

Survey Structure and Content

The survey was developed by the ECoE and was not validated prior to its use due to the time-sensitive nature of gathering information during the pandemic. The first survey (T1) was an emailed spreadsheet with open-ended questions to gauge availability of services (ie, outpatient clinic, EEG, EMU), assess whether safety precautions were being introduced, and understand whether national or local guidelines were thought to be helpful. Responses from this and subsequent surveys were standardized into yes/no and multiple choice formats. Subsequent surveys were administered online using a Research Electronic Data Capture tool.^8,9

Availability of outpatient epilepsy services across the 4 time periods were categorized as unlimited (in-person with no restrictions), limited (in-person with restrictions), planned (not currently performed but scheduled for the near future), and unavailable (no in-person services offered) (eAppendices 1-6, available in article PDF).

Statistical Analyses

Analyses were performed to compare survey responses to workload and CDC data on COVID-19 community spread. The following associations were examined: (1) CDC COVID-19 spread vs respondents’ perception of spread; (2) respondents’ perception of spread vs availability of services; (3) CDC COVID-19 spread vs availability of services; (4) respondents’ perception of spread vs workload; and (5) CDC COVID-19 spread vs workload. Availability of services was dichotomized for analyses, with limited or fully available services classified as available. As services were mostly open at T3 regardless of the spread of the virus, and the CDC COVID-19 spread classification for all sites was severe or high at T2 and T4, corresponding associations were not tested at these time points. For associations 1 through 3, Fisher exact tests were used; for associations 4 and 5, Mann-Whitney U tests (where the COVID-19 spread fell into 2 categories) and Kruskal-Wallis tests (for 3 categories of COVID-19 spread) were performed. All tests were 2-tailed and performed at 0.05 error rate. Bonferroni corrections were applied to adjust P values for multiple hypotheses tests.

RESULTS

From the 17 sites invited, responses at each time point were obtained from 13 (T1),17 (T2), 15 (T3), and 16 (T4) centers. There was no significant association between self-reported COVID-19 spread and CDC classification of COVID spread. There were no associations between COVID-19 community spread (respondent reported or CDC severity level) and outpatient clinic availability (self-reported or workload captured). At T3, a positive association was found between the CDC spread level and workload (P = .008), but this was not significant after Bonferroni correction (P = .06).

EEG availability surpassed EMU availability at all time points, although EMU services made some recovery at T3 and T4. No associations were found between COVID-19 community spread (self-reported or CDC severity level) and outpatient EEG or EMU availability (self-reported or workload captured). At T3, there was a positive association between EEG workload and CDC COVID-19 severity level (P = .04), but this was not significant after Bonferroni correction (P = .30). 

For outpatient EEG, staff and patient mask use were universally implemented by T2, while the use of full personal protective equipment (PPE) occurred at a subset of sites (T2, 6/17 [35%]; T3, 3/15 [20%]; T4: 4/16 [25%]). COVID-19 testing was rarely implemented prior to outpatient EEG (T1, 0 sites; T2, 1 site; T3, 1 site; T4, 0 sites). Within the EMU, safety precautions including COVID-19 testing, patient mask usage, staff mask usage, and aerosolization demonstrated a sustained majority usage across the 4 surveys.

National and Local Guidelines

The open-ended survey at T1 asked site directors, “Should there be national recommendations on how EEGs and related procedures should be done during the pandemic or should this be left to local conditions?” Responses were mixed, with 5 respondents desiring a national standard, 4 respondents favoring a local response, and 4 respondents believing a national standard should be in place but with modifications based on local outbreak levels and needs.

Surveys performed at T2 through T4 asked, “Which of the following do you feel was/will be helpful in adapting to COVID-19–related changes?” Overall, there was substantial agreement that guidelines were helpful. Most sites anticipated permanent changes in enhanced safety precautions and telehealth.

DISCUSSION

This longitudinal study across 4 time points describes how epilepsy services within the VHA and ECoE adapted to the COVID-19 pandemic. The first survey, conducted 2 months after COVID-19 was declared a pandemic, allowed a comparison with other concurrent US national surveys.^1,2,10 The subsequent surveys describe longitudinal adaptations to balance patient and staff safety with service availability and is a unique feature of the current report. Results demonstrate flexibility and adaptability by the ECoEs surveyed, which surprisingly did not show significant associations between CDC COVID-19 spread data and administrative workload data.

Trends in Availability of Services

The most significant impact of COVID-19 restrictions was during T1. There were no significant relationships between service availability/workload and objective CDC COVID-19 spread levels or subjective self-reported COVID-19 spread. Respondents’ perceptions of local COVID-19 spread showed no association with CDC COVID-19 spread data. It appears that subjective perception of spread may be unreliable and factors other than actual or perceived COVID-19 spread were likely driving patterns for service availability.

In-person outpatient visits were most impacted at T1, similar to other civilian surveys, with only 1 site reporting in-person outpatient visits without limitations.^1,2 These numbers significantly changed by T2, with all sites offering either limited or unlimited in-person visits. While the surveys did not evaluate factors leading to this rapid recovery, it may be related to the availability of COVID-19 vaccinations within the VHA during this time.¹¹ The US Department of Veterans Affairs was the first federal agency to mandate employee vaccination.¹² By the most recent time point (T4), all responding sites offered outpatient visits. Outpatient EEGs followed a similar trend, with T1 being the most restrictive and full, unrestricted outpatient EEGs available by T3. 

Fiscal year (FY) trends from ECoE annual reports suggest that encounters slowly recovered over the course of the pandemic. In FY 2019 there were 13,143 outpatient encounters and 6394 EEGs, which dropped to 8097 outpatient encounters and 4432 EEGs in FY 2020 before rising to 8489 outpatient encounters and 5604 EEGs in FY 2021 and 9772 outpatient encounters and 5062 EEGs in FY 2022. Thus, while clinicians described availability of services, patients may have remained hesitant or were otherwise unable to fulfill in-person appointments. The increased availability of home EEG (145 encounter days in 2021 and 436 encounter days in 2022) may be filling this gap. 

In contrast to outpatient clinics and EEG, EMU availability showed relatively slower reimplementation. In the last survey, about 30% of sites were still not offering EMU or had limited services. Early trends regarding reduced staffing and patient reluctance for elective admission cited in other surveys may have also affected EMU availability within the VHA.^2,13 Consistent with trends in availability, ECoE annual report data suggest EMU patient participation was about one-half of prepandemic rates: 3069 encounters in FY 2019 dropped to 1614 encounters in 2020. By 2021, rates were about two-thirds of prepandemic rates with 2058 encounters in 2021 and 2101 encounters in 2022.

Early survey results (T1) from this study echo trends from other surveys. In the AES survey (April to June 2020), about a quarter of respondents (22%) reported doing fewer EEG studies than usual. The Italian national survey (April 2020) revealed reduced presurgical evaluations (81%), ambulatory EEG (78%), standard EEG (5%) and long-term EEG (32%).⁴ In the NAEC survey (end of 2020)—which roughly corresponded to T2—outpatient EEGs were still < 75% of pre-COVID levels in one-half of the centers.

National and Local Guidelines

Both national and local guidelines were perceived as useful by most respondents, with national guidelines being more beneficial. This aligns with the NAEC survey, where there was a perceived need for detailed recommendations for PPE and COVID-19 testing of patients, visitors, and staff. Based on national and local guidelines, ECoE implemented safety procedures, as reflected in responses. Staff masking procedures appeared to be the most widely adopted for all services, while the use of full PPE waned as the pandemic progressed. COVID-19 testing was rarely used for routine outpatient visits but common in EMU admissions. This is similar to a survey conducted by the American Academy of Neurology which found full PPE implementation intermittently in outpatient settings and more frequently in inpatient settings.¹⁴

Telehealth Attitudes

While most sites anticipated permanent implementation of safety precautions and telehealth, the latter was consistently reported as more likely to be sustained. The VHA had a large and well-developed system of telehealth services that considerably predated the pandemic.^15,16 Through this established infrastructure, remote services were quickly increased across the VHA.^17-19 This telehealth structure was supplemented by the ability of VHA clinicians to practice across state lines, following a 2018 federal rule.²⁰ The AES survey noted the VHA ECoE's longstanding experience with telehealth as a model for telemedicine use in providing direct patient care, remote EEG analysis, and clinician-to-clinician consultation.¹

Trends in the number of telehealth patients seen, observed through patterns in ECoE annual reports are consistent with positive views toward this method of service provision. Specifically, these annual reports capture trends in Video Telehealth Clinic (local station), Video Telehealth Clinic (different station), Home Video Telehealth, Telephone Clinic, and eConsults. Though video telehealth at in-person stations had a precipitous drop in 2020 that continued to wane in subsequent years (898 encounters in 2019; 455 encounters in 2020; 90 encounters in 2021; 88 encounters in 2022), use of home video telehealth rose over time (143 encounters in 2019; 1003 encounters in 2020; 3206 encounters in 2021; 3315 encounters in 2022). Use of telephone services rose drastically in 2020 but has since become a less frequently used service method (2636 in 2019; 5923 in 2020; 5319 in 2021; 3704 in 2022).

Limitations

While the survey encouraged a high response rate, this limited its scope and interpretability. While the availability of services was evaluated, the underlying reasons were not queried. Follow-up questions about barriers to reopening may have allowed for a better understanding of why some services, such as EMU, continued to operate suboptimally later in the pandemic. Similarly, asking about unique strategies or barriers for telehealth would have allowed for a better understanding of its current and future use. We hypothesize that staffing changes during the pandemic may have influenced the availability of services, but changes to staffing were not assessed via the survey and were not readily available via other sources (eg, ECoE annual reports) at the time of publication. An additional limitation is the lack of comparable surveys in the literature for time points T2 to T4, as most analogous surveys were performed early in 2020.

Conclusions

This longitudinal study performed at 4 time points during the COVID-19 pandemic is the first to offer a comprehensive picture of changes to epilepsy and EEG services over time, given that other similar surveys lacked follow-up. Results reveal a significant limitation of services at VHA ECoE shortly after the onset of the pandemic, with return to near-complete operational status 2 years later. While safety precautions and telehealth are predicted to continue, telehealth is perceived as a more permanent change in services.

It is estimated that the prevalence of human hookworm infection is approximately 450 million individuals worldwide, representing a substantial global disease burden.¹ The annual global public health burden ranges from approximately 2 million to 4 million disability-adjusted life-years and $10 billion to $140 billion in hookwormrelated costs.² In this article, we discuss the lifecycle, transmission, and disease burden of cutaneous larva migrans (CLM) as well as prevention and treatment strategies.

Background

The Ancylostomatidae nematode family comprises at least 68 known species of hookworm that infect more than 110 different species of mammals.³ Many of these parasites are able to infect more than 1 primary host species, but from a disease perspective they can be classified as either anthropophilic, with humans as the intended host, or zoonotic, with humans as an incidental host. It is important to make this distinction because, though the lifecycles and biology of hookworm species generally are similar, the manifestations of incidental human infection from zoonotic hookworms are different from those of anthropophilic hookworms. Of the anthropophilic species, Necator americanus and Ancylostoma duodenale predominate. In the instance of zoonotic hookworm, dog-infecting A caninum and cat- and doginfecting A braziliense and Uncinaria stenocephala are common causes of incidental human disease.³

The life cycle of Ancylostomatidae organisms is astounding. Through millions of years of co-evolution with mammals,⁴ these parasitic worms have developed perhaps one of the most circuitous paths to propagate themselves in the natural world. Hookworms start their arduous journey as eggs deposited in soil, sand, and ground vegetation from the feces of infected animals.⁵ Approximately 1 day after the eggs are deposited, they hatch and begin the larval stage, during which they become infective 1 to 5 weeks later. At this point, the larvae become sensitive to their environment, responding to rising temperatures, increasing carbon dioxide levels, and vibrations in the soil—all of which suggest the presence of a potential host and contribute to a concordant increase in undulatory movement of the larvae.^5,6 Here, the most vulnerable tissues include the uncovered soles, palms, and buttocks of host mammals that come into contact with contaminated soil. In an undulating fashion and guided by temperature cues, the larvae locate the skin of the host and utilize a mixture of enzymes including hyaluronidases, metalloprotease, and other proteases to penetrate the epidermis.⁷ Anthropophilic hookworms such as N americanus and A duodenale will enter the circulatory system; from there, the hookworms migrate through the right-sided cardiopulmonary circuit and eventually ascend into the pulmonary vasculature.⁸ They then penetrate the lung capillary beds and parenchyma to reach the alveoli, ascend the respiratory tree, and, with the help of the mucociliary escalator, reach the esophagus, where they are swallowed by the host. In the gastrointestinal tract, adult hookworms consume host blood, mate, and lay eggs over a period of approximately 1 to 3 years if left untreated.⁹ Eggs are laid into the lower gastrointestinal tract, and the journey begins again in feces contacting ground or soil.

Geographic Distribution

Hookworms are found in almost all regions of the world, with species-specific distributions that highlight tropical and subtropical regions. Necator americanus and A duodenale are the most common hookworm species, with the former found predominantly in Southeast Asia and Latin America and the latter in Asia-Pacific regions.¹⁰ The highest prevalence of hookworms is in Southeast Asia followed by Sub-Saharan Africa, and the unique climate and soil composition of a region help determine the best environments for specific species of hookworm to thrive.¹¹ In addition, socioeconomics and social determinants of health play a big role in the spread of hookworms, as hygiene practices (eg, wearing clean shoes and clothing, bathing), infrastructure (eg, clean water and streets), and anthelmintic campaigns help reduce transmission.¹² Soil-transmitted helminths were once endemic to the southeastern United States, with some reports of approximately 40% of individuals infected in the south in the early 1900s.¹³ Anthelmintic campaigns such as water, sanitation, and hygiene programs as well as deworming of humans and livestock have proven effective in reducing the prevalence of helminth disease in industrialized nations.^13,14 However, zoonotic infections remain a problem in these regions, and in some parts of the United States more than 40% of sampled cats and dogs harbored species such as A braziliense.¹⁵

Clinical Manifestation

Initial hookworm infection often goes unnoticed because symptoms can range in severity, but it is characterized by transient ground itch—a local pruritic, erythematous, and papular eruption that develops in response to epidermal penetration.¹⁶ Because the larvae must traverse the host from skin to target organs for reproduction over several weeks, iron-deficiency anemia will manifest much later than signs of the initial penetration. In the case of incidental infection from zoonotic Ancylostomatidae organisms, the misguided larvae result in CLM, an often intensely pruritic skin condition that will self-resolve in 2 to 8 weeks with eventual death of the larvae.⁵

Diagnosis and Pathology of Disease

Zoonotic Hookworm—The major presenting sign of zoonotic hookworm infection is CLM. The diagnosis of CLM usually is made clinically, as the larvae themselves are 0.5 mm thick to 10 mm long (Figure 1) and usually extend several centimeters beyond the dermal lesion, with dermoscopy having limited utility.¹⁷ Patients may begin to experience itching as little as 1 hour after hookworm penetration of the skin.¹⁸ Once in contact with the skin, the hookworms’ hyaluronidases and proteases are capable of breaking through the epidermis, but zoonotic hookworms typically are unable to penetrate the basal layer of the human epidermis and remain entombed between the stratum granulosum and stratum corneum. With the exception of rare cases of direct or indirect pulmonary involvement resulting in Löffler syndrome,¹⁹ the larvae will die within weeks to months, and symptoms will subsequently resolve.

Although the infection generally is self-limiting, the dermatologic manifestations of CLM can be severe and warrant intervention. The lesions start as small reddish papules at the site of penetration (Figure 2), then the hallmark elevated, migrating, serpiginous, urticarial rash develops (Figure 3). Cutaneous larva migrans generally manifests unilaterally and is both erythematous and intensely pruritic. As the larvae migrate, they leave behind 1- to 5-cm tunneled creeping eruptions in their wake. The lesions, which can manifest with pain or be painless, may develop eczematous, bullous, follicular, or impetiginized appearances.²⁰ Atypical manifestations include folliculitis and urticarial plaques.¹⁷

Anthropophilic Hookworm—The lifecycles of N americanus and A duodenale are completed in human infection. Dermatologic manifestations are transient with the development of ground itch at the site of epidermal penetration. The hookworms employ collagenases that allow penetration of the basal layer of the skin, and eosinophilia develops as the parasites travel from the skin to the small intestine. Once attached to the gastrointestinal lumen, blood meals and proteolytic enzymes result in iron-deficiency anemia in the host and may lead to weakness, fatigue, and low birth weights in pregnant patients. With prolonged infection or heavy parasitic burden, patients can develop hypoproteinemia, anasarca, and yellowing of the skin known as chlorosis.¹¹ A clinical diagnosis can be made by examining patient stool samples for eggs, and definitive characterization can be made using molecular tools such as polymerase chain reaction.^21,22

Common to hookworm infections is the immune reaction, which promotes inflammation with localized eosinophilia and mastocytosis.¹¹ In a clinical biopsy specimen of gut—usually obtained through esophagogastroduodenoscopy— T-helper (Th) 2–type immune (IL-4, IL-5, IL-9 and IL-13), regulatory Th10 (IL-10 and transcription growth factor β), and some evidence of Th1 (interferon gamma and IL-2) cytokines are present, but little evidence of Th17-type immune response was found.²³ It is believed that in zoonotic infections, antiparasitic IgE from basophils are somewhat successful at trapping the helminths in the epidermis, but in the anthropophilic species, IgE and Th2 responses are ineffective at clearing the parasite from the gut, and the defeated immune system transitions to a host-tolerance approach of limiting infection.¹¹ It is now believed that this natural armistice can be manipulated into a potential therapy against autoimmune and inflammatory conditions. Intentional infection with zoonotic whipworm or hookworm has been proposed as a mechanism of switching Th1 and Th2 responses to host-tolerant mechanisms in conditions such as Crohn disease and celiac disease,²⁴ and it has even been hypothesized that prior hookworm infection may reduce the chance of developing allergic conditions such as eczema.²⁵

Treatment and Prevention

The World Health Organization and Centers for Disease Control and Prevention recommend a single oral dose of 400 mg albendazole for adults or 10 to 15 mg/kg in children for CLM. A single dose of ivermectin at 12 mg in adults or 150 μg/kg in children can be used as an alternative where albendazole is not available.¹¹ Topical applications of thiabendazole 10% to 15% under occlusion or 3 times daily for 15 days without occlusion also can manage CLM, and pruritus can be treated with topical corticosteroids for symptomatic relief. Oral albendazole 400 mg twice daily or mebendazole 100 mg twice daily for 3 days or a single 500-mg dose, as well as 11 mg/kg (up to a maximum of 1 g) oral pyrantel pamoate once daily for 3 days can be used to treat intestinal hookworm infection, though it should be avoided in pregnancy. Iron deficiency should be managed with supplementation.¹¹

Prevention of hookworm infection is focused around 2 broad public health efforts: mass drug administration programs and the water, sanitation, and hygiene program. In mass drug administration, treatments such as benzimidazoles are given in mass to communities affected by endemic hookworm as a single dose to reduce the burden of disease. Together, these strategies effectively eliminated hookworms in many developed nations, but areas of resurgence are beginning to surface worldwide. With changes in climate, emerging drug resistance, and socioeconomic disparities, particularly affecting the southeast, a resurgence of hookworm has occurred in the United States.²⁶ One recent study demonstrated that almost one-third (19/55) of children sampled in an impoverished area of rural Alabama had hookworm eggs in their stool.²⁷ Furthermore, pets serve not only as zoonotic reservoirs for CLM recurrence but also as vehicles for the evolution of drug-resistant strains, leading some to call for a ban of animals from beaches and playgrounds as well as tightly controlled veterinary programs.^5,28 Ubiquitous benzimidazole use in livestock has led to bendazole-resistant strains, and it is likely that with continued and poorly adherent drug use, more zoonotic and anthropophilic drug-resistant strains of hookworm will emerge.^29,30

Conclusion

The burden of hookworm infection and CLM is substantial in parts of the United States. Dermatologists play a critical role in the recognition and management of hookworm infection for both treatment of affected patients and the subsequent prevention of its spread. As drug-resistant strains evolve, clinicians, public health officials, and scientists need to continue to work together to prevent and treat hookworm infection.

It is estimated that the prevalence of human hookworm infection is approximately 450 million individuals worldwide, representing a substantial global disease burden.¹ The annual global public health burden ranges from approximately 2 million to 4 million disability-adjusted life-years and $10 billion to $140 billion in hookwormrelated costs.² In this article, we discuss the lifecycle, transmission, and disease burden of cutaneous larva migrans (CLM) as well as prevention and treatment strategies.

Background

The Ancylostomatidae nematode family comprises at least 68 known species of hookworm that infect more than 110 different species of mammals.³ Many of these parasites are able to infect more than 1 primary host species, but from a disease perspective they can be classified as either anthropophilic, with humans as the intended host, or zoonotic, with humans as an incidental host. It is important to make this distinction because, though the lifecycles and biology of hookworm species generally are similar, the manifestations of incidental human infection from zoonotic hookworms are different from those of anthropophilic hookworms. Of the anthropophilic species, Necator americanus and Ancylostoma duodenale predominate. In the instance of zoonotic hookworm, dog-infecting A caninum and cat- and doginfecting A braziliense and Uncinaria stenocephala are common causes of incidental human disease.³

The life cycle of Ancylostomatidae organisms is astounding. Through millions of years of co-evolution with mammals,⁴ these parasitic worms have developed perhaps one of the most circuitous paths to propagate themselves in the natural world. Hookworms start their arduous journey as eggs deposited in soil, sand, and ground vegetation from the feces of infected animals.⁵ Approximately 1 day after the eggs are deposited, they hatch and begin the larval stage, during which they become infective 1 to 5 weeks later. At this point, the larvae become sensitive to their environment, responding to rising temperatures, increasing carbon dioxide levels, and vibrations in the soil—all of which suggest the presence of a potential host and contribute to a concordant increase in undulatory movement of the larvae.^5,6 Here, the most vulnerable tissues include the uncovered soles, palms, and buttocks of host mammals that come into contact with contaminated soil. In an undulating fashion and guided by temperature cues, the larvae locate the skin of the host and utilize a mixture of enzymes including hyaluronidases, metalloprotease, and other proteases to penetrate the epidermis.⁷ Anthropophilic hookworms such as N americanus and A duodenale will enter the circulatory system; from there, the hookworms migrate through the right-sided cardiopulmonary circuit and eventually ascend into the pulmonary vasculature.⁸ They then penetrate the lung capillary beds and parenchyma to reach the alveoli, ascend the respiratory tree, and, with the help of the mucociliary escalator, reach the esophagus, where they are swallowed by the host. In the gastrointestinal tract, adult hookworms consume host blood, mate, and lay eggs over a period of approximately 1 to 3 years if left untreated.⁹ Eggs are laid into the lower gastrointestinal tract, and the journey begins again in feces contacting ground or soil.

Geographic Distribution

Hookworms are found in almost all regions of the world, with species-specific distributions that highlight tropical and subtropical regions. Necator americanus and A duodenale are the most common hookworm species, with the former found predominantly in Southeast Asia and Latin America and the latter in Asia-Pacific regions.¹⁰ The highest prevalence of hookworms is in Southeast Asia followed by Sub-Saharan Africa, and the unique climate and soil composition of a region help determine the best environments for specific species of hookworm to thrive.¹¹ In addition, socioeconomics and social determinants of health play a big role in the spread of hookworms, as hygiene practices (eg, wearing clean shoes and clothing, bathing), infrastructure (eg, clean water and streets), and anthelmintic campaigns help reduce transmission.¹² Soil-transmitted helminths were once endemic to the southeastern United States, with some reports of approximately 40% of individuals infected in the south in the early 1900s.¹³ Anthelmintic campaigns such as water, sanitation, and hygiene programs as well as deworming of humans and livestock have proven effective in reducing the prevalence of helminth disease in industrialized nations.^13,14 However, zoonotic infections remain a problem in these regions, and in some parts of the United States more than 40% of sampled cats and dogs harbored species such as A braziliense.¹⁵

Clinical Manifestation

Initial hookworm infection often goes unnoticed because symptoms can range in severity, but it is characterized by transient ground itch—a local pruritic, erythematous, and papular eruption that develops in response to epidermal penetration.¹⁶ Because the larvae must traverse the host from skin to target organs for reproduction over several weeks, iron-deficiency anemia will manifest much later than signs of the initial penetration. In the case of incidental infection from zoonotic Ancylostomatidae organisms, the misguided larvae result in CLM, an often intensely pruritic skin condition that will self-resolve in 2 to 8 weeks with eventual death of the larvae.⁵

Diagnosis and Pathology of Disease

Zoonotic Hookworm—The major presenting sign of zoonotic hookworm infection is CLM. The diagnosis of CLM usually is made clinically, as the larvae themselves are 0.5 mm thick to 10 mm long (Figure 1) and usually extend several centimeters beyond the dermal lesion, with dermoscopy having limited utility.¹⁷ Patients may begin to experience itching as little as 1 hour after hookworm penetration of the skin.¹⁸ Once in contact with the skin, the hookworms’ hyaluronidases and proteases are capable of breaking through the epidermis, but zoonotic hookworms typically are unable to penetrate the basal layer of the human epidermis and remain entombed between the stratum granulosum and stratum corneum. With the exception of rare cases of direct or indirect pulmonary involvement resulting in Löffler syndrome,¹⁹ the larvae will die within weeks to months, and symptoms will subsequently resolve.

Although the infection generally is self-limiting, the dermatologic manifestations of CLM can be severe and warrant intervention. The lesions start as small reddish papules at the site of penetration (Figure 2), then the hallmark elevated, migrating, serpiginous, urticarial rash develops (Figure 3). Cutaneous larva migrans generally manifests unilaterally and is both erythematous and intensely pruritic. As the larvae migrate, they leave behind 1- to 5-cm tunneled creeping eruptions in their wake. The lesions, which can manifest with pain or be painless, may develop eczematous, bullous, follicular, or impetiginized appearances.²⁰ Atypical manifestations include folliculitis and urticarial plaques.¹⁷

Anthropophilic Hookworm—The lifecycles of N americanus and A duodenale are completed in human infection. Dermatologic manifestations are transient with the development of ground itch at the site of epidermal penetration. The hookworms employ collagenases that allow penetration of the basal layer of the skin, and eosinophilia develops as the parasites travel from the skin to the small intestine. Once attached to the gastrointestinal lumen, blood meals and proteolytic enzymes result in iron-deficiency anemia in the host and may lead to weakness, fatigue, and low birth weights in pregnant patients. With prolonged infection or heavy parasitic burden, patients can develop hypoproteinemia, anasarca, and yellowing of the skin known as chlorosis.¹¹ A clinical diagnosis can be made by examining patient stool samples for eggs, and definitive characterization can be made using molecular tools such as polymerase chain reaction.^21,22

Common to hookworm infections is the immune reaction, which promotes inflammation with localized eosinophilia and mastocytosis.¹¹ In a clinical biopsy specimen of gut—usually obtained through esophagogastroduodenoscopy— T-helper (Th) 2–type immune (IL-4, IL-5, IL-9 and IL-13), regulatory Th10 (IL-10 and transcription growth factor β), and some evidence of Th1 (interferon gamma and IL-2) cytokines are present, but little evidence of Th17-type immune response was found.²³ It is believed that in zoonotic infections, antiparasitic IgE from basophils are somewhat successful at trapping the helminths in the epidermis, but in the anthropophilic species, IgE and Th2 responses are ineffective at clearing the parasite from the gut, and the defeated immune system transitions to a host-tolerance approach of limiting infection.¹¹ It is now believed that this natural armistice can be manipulated into a potential therapy against autoimmune and inflammatory conditions. Intentional infection with zoonotic whipworm or hookworm has been proposed as a mechanism of switching Th1 and Th2 responses to host-tolerant mechanisms in conditions such as Crohn disease and celiac disease,²⁴ and it has even been hypothesized that prior hookworm infection may reduce the chance of developing allergic conditions such as eczema.²⁵

Treatment and Prevention

The World Health Organization and Centers for Disease Control and Prevention recommend a single oral dose of 400 mg albendazole for adults or 10 to 15 mg/kg in children for CLM. A single dose of ivermectin at 12 mg in adults or 150 μg/kg in children can be used as an alternative where albendazole is not available.¹¹ Topical applications of thiabendazole 10% to 15% under occlusion or 3 times daily for 15 days without occlusion also can manage CLM, and pruritus can be treated with topical corticosteroids for symptomatic relief. Oral albendazole 400 mg twice daily or mebendazole 100 mg twice daily for 3 days or a single 500-mg dose, as well as 11 mg/kg (up to a maximum of 1 g) oral pyrantel pamoate once daily for 3 days can be used to treat intestinal hookworm infection, though it should be avoided in pregnancy. Iron deficiency should be managed with supplementation.¹¹

Prevention of hookworm infection is focused around 2 broad public health efforts: mass drug administration programs and the water, sanitation, and hygiene program. In mass drug administration, treatments such as benzimidazoles are given in mass to communities affected by endemic hookworm as a single dose to reduce the burden of disease. Together, these strategies effectively eliminated hookworms in many developed nations, but areas of resurgence are beginning to surface worldwide. With changes in climate, emerging drug resistance, and socioeconomic disparities, particularly affecting the southeast, a resurgence of hookworm has occurred in the United States.²⁶ One recent study demonstrated that almost one-third (19/55) of children sampled in an impoverished area of rural Alabama had hookworm eggs in their stool.²⁷ Furthermore, pets serve not only as zoonotic reservoirs for CLM recurrence but also as vehicles for the evolution of drug-resistant strains, leading some to call for a ban of animals from beaches and playgrounds as well as tightly controlled veterinary programs.^5,28 Ubiquitous benzimidazole use in livestock has led to bendazole-resistant strains, and it is likely that with continued and poorly adherent drug use, more zoonotic and anthropophilic drug-resistant strains of hookworm will emerge.^29,30

Conclusion

The burden of hookworm infection and CLM is substantial in parts of the United States. Dermatologists play a critical role in the recognition and management of hookworm infection for both treatment of affected patients and the subsequent prevention of its spread. As drug-resistant strains evolve, clinicians, public health officials, and scientists need to continue to work together to prevent and treat hookworm infection.

It is estimated that the prevalence of human hookworm infection is approximately 450 million individuals worldwide, representing a substantial global disease burden.¹ The annual global public health burden ranges from approximately 2 million to 4 million disability-adjusted life-years and $10 billion to $140 billion in hookwormrelated costs.² In this article, we discuss the lifecycle, transmission, and disease burden of cutaneous larva migrans (CLM) as well as prevention and treatment strategies.

Background

The Ancylostomatidae nematode family comprises at least 68 known species of hookworm that infect more than 110 different species of mammals.³ Many of these parasites are able to infect more than 1 primary host species, but from a disease perspective they can be classified as either anthropophilic, with humans as the intended host, or zoonotic, with humans as an incidental host. It is important to make this distinction because, though the lifecycles and biology of hookworm species generally are similar, the manifestations of incidental human infection from zoonotic hookworms are different from those of anthropophilic hookworms. Of the anthropophilic species, Necator americanus and Ancylostoma duodenale predominate. In the instance of zoonotic hookworm, dog-infecting A caninum and cat- and doginfecting A braziliense and Uncinaria stenocephala are common causes of incidental human disease.³

The life cycle of Ancylostomatidae organisms is astounding. Through millions of years of co-evolution with mammals,⁴ these parasitic worms have developed perhaps one of the most circuitous paths to propagate themselves in the natural world. Hookworms start their arduous journey as eggs deposited in soil, sand, and ground vegetation from the feces of infected animals.⁵ Approximately 1 day after the eggs are deposited, they hatch and begin the larval stage, during which they become infective 1 to 5 weeks later. At this point, the larvae become sensitive to their environment, responding to rising temperatures, increasing carbon dioxide levels, and vibrations in the soil—all of which suggest the presence of a potential host and contribute to a concordant increase in undulatory movement of the larvae.^5,6 Here, the most vulnerable tissues include the uncovered soles, palms, and buttocks of host mammals that come into contact with contaminated soil. In an undulating fashion and guided by temperature cues, the larvae locate the skin of the host and utilize a mixture of enzymes including hyaluronidases, metalloprotease, and other proteases to penetrate the epidermis.⁷ Anthropophilic hookworms such as N americanus and A duodenale will enter the circulatory system; from there, the hookworms migrate through the right-sided cardiopulmonary circuit and eventually ascend into the pulmonary vasculature.⁸ They then penetrate the lung capillary beds and parenchyma to reach the alveoli, ascend the respiratory tree, and, with the help of the mucociliary escalator, reach the esophagus, where they are swallowed by the host. In the gastrointestinal tract, adult hookworms consume host blood, mate, and lay eggs over a period of approximately 1 to 3 years if left untreated.⁹ Eggs are laid into the lower gastrointestinal tract, and the journey begins again in feces contacting ground or soil.

Geographic Distribution

Hookworms are found in almost all regions of the world, with species-specific distributions that highlight tropical and subtropical regions. Necator americanus and A duodenale are the most common hookworm species, with the former found predominantly in Southeast Asia and Latin America and the latter in Asia-Pacific regions.¹⁰ The highest prevalence of hookworms is in Southeast Asia followed by Sub-Saharan Africa, and the unique climate and soil composition of a region help determine the best environments for specific species of hookworm to thrive.¹¹ In addition, socioeconomics and social determinants of health play a big role in the spread of hookworms, as hygiene practices (eg, wearing clean shoes and clothing, bathing), infrastructure (eg, clean water and streets), and anthelmintic campaigns help reduce transmission.¹² Soil-transmitted helminths were once endemic to the southeastern United States, with some reports of approximately 40% of individuals infected in the south in the early 1900s.¹³ Anthelmintic campaigns such as water, sanitation, and hygiene programs as well as deworming of humans and livestock have proven effective in reducing the prevalence of helminth disease in industrialized nations.^13,14 However, zoonotic infections remain a problem in these regions, and in some parts of the United States more than 40% of sampled cats and dogs harbored species such as A braziliense.¹⁵

Clinical Manifestation

Initial hookworm infection often goes unnoticed because symptoms can range in severity, but it is characterized by transient ground itch—a local pruritic, erythematous, and papular eruption that develops in response to epidermal penetration.¹⁶ Because the larvae must traverse the host from skin to target organs for reproduction over several weeks, iron-deficiency anemia will manifest much later than signs of the initial penetration. In the case of incidental infection from zoonotic Ancylostomatidae organisms, the misguided larvae result in CLM, an often intensely pruritic skin condition that will self-resolve in 2 to 8 weeks with eventual death of the larvae.⁵

Diagnosis and Pathology of Disease

Zoonotic Hookworm—The major presenting sign of zoonotic hookworm infection is CLM. The diagnosis of CLM usually is made clinically, as the larvae themselves are 0.5 mm thick to 10 mm long (Figure 1) and usually extend several centimeters beyond the dermal lesion, with dermoscopy having limited utility.¹⁷ Patients may begin to experience itching as little as 1 hour after hookworm penetration of the skin.¹⁸ Once in contact with the skin, the hookworms’ hyaluronidases and proteases are capable of breaking through the epidermis, but zoonotic hookworms typically are unable to penetrate the basal layer of the human epidermis and remain entombed between the stratum granulosum and stratum corneum. With the exception of rare cases of direct or indirect pulmonary involvement resulting in Löffler syndrome,¹⁹ the larvae will die within weeks to months, and symptoms will subsequently resolve.

Although the infection generally is self-limiting, the dermatologic manifestations of CLM can be severe and warrant intervention. The lesions start as small reddish papules at the site of penetration (Figure 2), then the hallmark elevated, migrating, serpiginous, urticarial rash develops (Figure 3). Cutaneous larva migrans generally manifests unilaterally and is both erythematous and intensely pruritic. As the larvae migrate, they leave behind 1- to 5-cm tunneled creeping eruptions in their wake. The lesions, which can manifest with pain or be painless, may develop eczematous, bullous, follicular, or impetiginized appearances.²⁰ Atypical manifestations include folliculitis and urticarial plaques.¹⁷

Anthropophilic Hookworm—The lifecycles of N americanus and A duodenale are completed in human infection. Dermatologic manifestations are transient with the development of ground itch at the site of epidermal penetration. The hookworms employ collagenases that allow penetration of the basal layer of the skin, and eosinophilia develops as the parasites travel from the skin to the small intestine. Once attached to the gastrointestinal lumen, blood meals and proteolytic enzymes result in iron-deficiency anemia in the host and may lead to weakness, fatigue, and low birth weights in pregnant patients. With prolonged infection or heavy parasitic burden, patients can develop hypoproteinemia, anasarca, and yellowing of the skin known as chlorosis.¹¹ A clinical diagnosis can be made by examining patient stool samples for eggs, and definitive characterization can be made using molecular tools such as polymerase chain reaction.^21,22

Common to hookworm infections is the immune reaction, which promotes inflammation with localized eosinophilia and mastocytosis.¹¹ In a clinical biopsy specimen of gut—usually obtained through esophagogastroduodenoscopy— T-helper (Th) 2–type immune (IL-4, IL-5, IL-9 and IL-13), regulatory Th10 (IL-10 and transcription growth factor β), and some evidence of Th1 (interferon gamma and IL-2) cytokines are present, but little evidence of Th17-type immune response was found.²³ It is believed that in zoonotic infections, antiparasitic IgE from basophils are somewhat successful at trapping the helminths in the epidermis, but in the anthropophilic species, IgE and Th2 responses are ineffective at clearing the parasite from the gut, and the defeated immune system transitions to a host-tolerance approach of limiting infection.¹¹ It is now believed that this natural armistice can be manipulated into a potential therapy against autoimmune and inflammatory conditions. Intentional infection with zoonotic whipworm or hookworm has been proposed as a mechanism of switching Th1 and Th2 responses to host-tolerant mechanisms in conditions such as Crohn disease and celiac disease,²⁴ and it has even been hypothesized that prior hookworm infection may reduce the chance of developing allergic conditions such as eczema.²⁵

Treatment and Prevention

The World Health Organization and Centers for Disease Control and Prevention recommend a single oral dose of 400 mg albendazole for adults or 10 to 15 mg/kg in children for CLM. A single dose of ivermectin at 12 mg in adults or 150 μg/kg in children can be used as an alternative where albendazole is not available.¹¹ Topical applications of thiabendazole 10% to 15% under occlusion or 3 times daily for 15 days without occlusion also can manage CLM, and pruritus can be treated with topical corticosteroids for symptomatic relief. Oral albendazole 400 mg twice daily or mebendazole 100 mg twice daily for 3 days or a single 500-mg dose, as well as 11 mg/kg (up to a maximum of 1 g) oral pyrantel pamoate once daily for 3 days can be used to treat intestinal hookworm infection, though it should be avoided in pregnancy. Iron deficiency should be managed with supplementation.¹¹

Prevention of hookworm infection is focused around 2 broad public health efforts: mass drug administration programs and the water, sanitation, and hygiene program. In mass drug administration, treatments such as benzimidazoles are given in mass to communities affected by endemic hookworm as a single dose to reduce the burden of disease. Together, these strategies effectively eliminated hookworms in many developed nations, but areas of resurgence are beginning to surface worldwide. With changes in climate, emerging drug resistance, and socioeconomic disparities, particularly affecting the southeast, a resurgence of hookworm has occurred in the United States.²⁶ One recent study demonstrated that almost one-third (19/55) of children sampled in an impoverished area of rural Alabama had hookworm eggs in their stool.²⁷ Furthermore, pets serve not only as zoonotic reservoirs for CLM recurrence but also as vehicles for the evolution of drug-resistant strains, leading some to call for a ban of animals from beaches and playgrounds as well as tightly controlled veterinary programs.^5,28 Ubiquitous benzimidazole use in livestock has led to bendazole-resistant strains, and it is likely that with continued and poorly adherent drug use, more zoonotic and anthropophilic drug-resistant strains of hookworm will emerge.^29,30

Conclusion

The burden of hookworm infection and CLM is substantial in parts of the United States. Dermatologists play a critical role in the recognition and management of hookworm infection for both treatment of affected patients and the subsequent prevention of its spread. As drug-resistant strains evolve, clinicians, public health officials, and scientists need to continue to work together to prevent and treat hookworm infection.

Approximately 25% of Americans have at least one skin condition, and 20% are estimated to develop skin cancer during their lifetime.^1,2 However, 40% of the US population lives in areas underserved by dermatologists. ³ The severity and mortality of skin cancers such as melanoma and mycosis fungoides have been positively associated with minoritized race, lack of health insurance, and unstable housing status.^4-6 Patients who receive health care at free clinics often are of a racial or ethnic minoritized social group, are uninsured, and/or lack stable housing; this underserved group also includes recent immigrants to the United States who have limited English proficiency (LEP).⁷ Only 25% of free clinics offer specialty care services such as dermatology.^7,8

Of the 42 free clinics and Federally Qualified Health Centers in Pittsburgh, Pennsylvania, the Birmingham Free Clinic (BFC) is one of the few that offers specialty care services including dermatology.⁹ Founded in 1994, the BFC serves as a safety net for Pittsburgh’s medically underserved population, offering primary and acute care, medication access, and social services. From January 2020 to May 2022, the BFC offered 27 dermatology clinics that provided approximately 100 people with comprehensive care including full-body skin examinations, dermatologic diagnoses and treatments, minor procedures, and dermatopathology services.

In this study, we compared the BFC dermatology patient care model with that of the dermatology department at the University of Pittsburgh Medical Center (UPMC), an insurance-based tertiary referral health care system in western Pennsylvania. By analyzing the demographics, dermatologic diagnoses, and management strategies of both the BFC and UPMC, we gained an understanding of how these patient care models differ and how they can be improved to care for diverse patient populations.

Methods

A retrospective chart review of dermatology patients seen in person at the BFC and UPMC during the period from January 2020 to May 2022 was performed. The UPMC group included patients seen by 3 general dermatologists (including A.J.J.) at matched time points. Data were collected from patients’ first in-person visit during the study period. Variables of interest included patient age, sex, race, ethnicity, primary language, zip code, health insurance status, distance to clinic (estimated using Google Maps to calculate the shortest driving distance from the patient’s zip code to the clinic), history of skin cancer, dermatologic diagnoses, and management strategies. These variables were not collected for patients who cancelled or noshowed their first in-person appointments. All patient charts and notes corresponding to the date and visit of interest were accessed through the electronic medical record (EMR). Patient data were de-identified and stored in a password-protected spreadsheet. Comparisons between the BFC and UPMC patient populations were performed using X² tests of independence, Fisher exact tests, and Mann-Whitney U tests via SPSS software (IBM). Statistical significance was set at P<.05.

Results

Patient Characteristics—Our analysis included 76 initial appointments at the BFC and 322 at UPMC (Table 1). The mean age for patients at the BFC and UPMC was 39.6 years and 47.8 years, respectively (P=.001). Males accounted for 39 (51.3%) and 112 (34.8%) of BFC and UPMC patients, respectively (P=.008); 2 (0.6%) patients from UPMC were transgender. Of the BFC and UPMC patients, 44.7% (34/76) and 0.9% (3/322) were Hispanic, respectively (P<.001). With regard to race, 52.6% (40/76) of BFC patients were White, 19.7% (15/76) were Black, 6.6% (5/76) were Asian/Pacific Islander (Chinese, 1.3% [1/76]; other Asian, 5.3% [4/76]), and 21.1% (16/76) were American Indian/other/unspecified (American Indian, 1.3% [1/76]; other, 13.2% [10/76]; unspecified, 6.6% [5/76]). At UPMC, 61.2% (197/322) of patients were White, 28.0% (90/322) were Black, 5.3% (17/322) were Asian/Pacific Islander (Chinese, 1.2% [4/322]; Indian [Asian], 1.9% [6/322]; Japanese, 0.3% [1/322]; other Asian, 1.6% [5/322]; other Asian/American Indian, 0.3% [1/322]), and 5.6% (18/322) were American Indian/other/ unspecified (American Indian, 0.3% [1/322]; other, 0.3% [1/322]; unspecified, 5.0% [16/322]). Overall, the BFC patient population was more ethnically and racially diverse than that of UPMC (P<.001).

Forty-six percent (35/76) of BFC patients and 4.3% (14/322) of UPMC patients had LEP (P<.001). Primary languages among BFC patients were 53.9% (41/76) English, 40.8% (31/76) Spanish, and 5.2% (4/76) other/ unspecified (Chinese, 1.3% [1/76]; Indonesian, 2.6% [2/76]; unspecified, 1.3% [1/76]). Primary languages among UPMC patients were 95.7% (308/322) English and 4.3% (14/322) other/unspecified (Chinese, 0.6% [2/322]; Nepali, 0.6% [2/322]; Pali, 0.3% [1/322]; Russian, 0.3% [1/322]; unspecified, 2.5% [8/322]). There were notable differences in insurance status at the BFC vs UPMC (P<.001), with more UPMC patients having private insurance (52.8% [170/322] vs 11.8% [9/76]) and more BFC patients being uninsured (52.8% [51/76] vs 1.9% [6/322]). There was no significant difference in distance to clinic between the 2 groups (P=.183). More UPMC patients had a history of skin cancer (P=.003). More patients at the BFC were no-shows for their appointments (P<.001), and UPMC patients more frequently canceled their appointments (P<.001).

Dermatologic Diagnoses—The most commonly diagnosed dermatologic conditions at the BFC were dermatitis (23.7% [18/76]), neoplasm of uncertain behavior (15.8% [12/76]), alopecia (11.8% [9/76]), and acne (10.5% [8/76]) (Table 2). The most commonly diagnosed conditions at UPMC were nevi (26.4% [85/322]), dermatitis (22.7% [73/322]), seborrheic keratosis (21.7% [70/322]), and skin cancer screening (21.4% [70/322]). Neoplasm of uncertain behavior was more common in BFC vs UPMC patients (P=.040), while UPMC patients were more frequently diagnosed with nevi (P<.001), seborrheic keratosis (P<.001), and skin cancer screening (P<.001). There was no significant difference between the incidence of skin cancer diagnoses in the BFC (1.3% [1/76]) and UPMC (0.6% [2/76]) patient populations (P=.471). Among the biopsied neoplasms, there was also no significant difference in malignant (BFC, 50.0% [5/10]; UPMC, 32.0% [8/25]) and benign (BFC, 50.0% [5/10]; UPMC, 36.0% [9/25]) neoplasms diagnosed at each clinic (P=.444).

Management Strategies—Systemic antibiotics were more frequently prescribed (P<.001) and laboratory testing/ imaging were more frequently ordered (P=.005) at the BFC vs UPMC (Table 3). Patients at the BFC also more frequently required emergency insurance (P=.036). Patients at UPMC were more frequently recommended sunscreen (P=.003) and received education about skin cancer signs by review of the ABCDEs of melanoma (P<.001), sun-protective behaviors (P=.001), and skin examination frequency (P<.001). Notes in the EMR for UPMC patients more frequently specified patient followup instructions (P<.001).

Comment

As of 2020, the city of Pittsburgh had an estimated population of nearly 303,000 based on US Census data.10 Its population is predominantly White (62.7%) followed by Black/African American (22.8%) and Asian (6.5%); 5.9% identify as 2 or more races. Approximately 3.8% identify as Hispanic or Latino. More than 11% of the Pittsburgh population aged 5 years and older speaks a language other than English as their primary language, including Spanish (2.3%), other Indo-European languages (3.9%), and Asian and Pacific Island languages (3.5%).¹¹ More than 5% of the Pittsburgh population does not have health insurance.¹²

The BFC is located in Pittsburgh’s South Side area, while one of UPMC’s primary dermatology clinics is located in the Oakland district; however, most patients who seek care at these clinics live outside these areas. Our study results indicated that the BFC and UPMC serve distinct groups of people within the Pittsburgh population. The BFC patient population was younger with a higher percentage of patients who were male, Hispanic, racially diverse, and with LEP compared with the UPMC patient population. In this clinical setting, the BFC health care team engages with people from diverse backgrounds and requires greater interpreter and medical support services.

The BFC largely is supported by volunteers, UPMC, grants, and philanthropy. Dermatology clinics are staffed by paid and volunteer team members. Paid team members include 1 nurse and 1 access lead who operates the front desk and registration. Volunteer team members include 1 board-certified dermatologist from UPMC (A.J.J.), or an affiliate clinic and 1 or 2 of each of the following: UPMC dermatology residents, medical or undergraduate students from the University of Pittsburgh, AmeriCorps national service members, and student or community medical interpreters. The onsite pharmacy is run by volunteer faculty, resident, and student pharmacists from the University of Pittsburgh. Dermatology clinics are half-day clinics that occur monthly. Volunteers for each clinic are recruited approximately 1 month in advance.

Dermatology patients at the BFC are referred from the BFC general medicine clinic and nearby Federally Qualified Health Center s for simple to complex medical and surgical dermatologic skin conditions. Each BFC dermatology clinic schedules an average of 7 patients per clinic and places other patients on a wait-list unless more urgent triage is needed. Patients are notified when they are scheduled via phone or text message, and they receive a reminder call or text 1 or 2 days prior to their appointment that also asks them to confirm attendance. Patients with LEP are called with an interpreter and also may receive text reminders that can be translated using Google Translate. Patients are instructed to notify the BFC if they need to cancel or reschedule their appointment. At the end of each visit, patients are given an after-visit summary that lists follow-up instructions, medications prescribed during the visit, and upcoming appointments. The BFC offers bus tickets to help patients get to their appointments. In rare cases, the BFC may pay for a car service to drive patients to and from the clinic.

Dermatology clinics at UPMC use scheduling and self-scheduling systems through which patients can make appointments at a location of their choice with any available board-certified dermatologist or physician assistant. Patients receive a reminder phone call 3 days prior to their appointment instructing them to call the office if they are unable to keep their appointment. Patients signed up for the online portal also receive a reminder message and an option to confirm or cancel their appointment. Patients with cell phone numbers in the UPMC system receive a text message approximately 2 days prior to their appointment that allows them to preregister and pay their copayment in advance. They receive another text 20 minutes prior to their appointment with an option for contactless check-in. At the conclusion of their visit, patients can schedule a follow-up appointment and receive a printed copy of their after-visit summary that provides information about follow-up instructions, prescribed medications, and upcoming visits. They may alternatively access this summary via the online patient portal. Patients are not provided transportation to UPMC clinics, but they are offered parking validation.

Among the most common dermatologic diagnoses for each group, BFC patients presented for treatment of more acute dermatologic conditions, while UPMC patients presented for more benign and preventive-care conditions. This difference may be attributable to the BFC’s referral and triage system, wherein patients with more urgent problems are given scheduling priority. This patient care model contrasts with UPMC’s scheduling process in which no known formal triage system is utilized. Interestingly, there was no difference in skin cancer incidence despite a higher percentage of preventive skin cancer screenings at UPMC.

Patients at the BFC more often required emergency insurance for surgical interventions, which is consistent with the higher percentage of uninsured individuals in this population. Patients at UPMC more frequently were recommended sunscreen and were educated about skin cancer, sun protection, and skin examination, in part due to this group’s more extensive history of skin cancer and frequent presentation for skin cancer screenings. At the same time, educational materials for skin care at both the BFC and UPMC are populated into the EMR in English, whereas materials in other languages are less readily available.

Our retrospective study had several limitations. Demographic information that relied on clinic-dependent intake questionnaires may be limited due to variable intake processes and patients opting out of self-reporting. By comparing patient populations between 2 clinics, confounding variables such as location and hours of operation may impact the patient demographics recorded at the BFC vs UPMC. Resources and staff availability may affect the management strategies and follow-up care offered by each clinic. Our study period also was unique in that COVID-19 may have affected resources, staffing, scheduling, and logistics at both clinics.

Based on the aforementioned differences between the BFC and UPMC patient characteristics, care models should be strategically designed to support the needs of diverse populations. The BFC patient care model appropriately focuses on communication skills with patients with LEP by using interpreter services. Providing more skin care education and follow-up instructions in patients’ primary languages will help them develop a better understanding of their skin conditions. Another key asset of the BFC patient care model is its provision of social services such as transportation and insurance assistance.

To improve the UPMC patient care model, providing patients with bus tickets and car services may potentially reduce appointment cancellations. Using interpreter services to call and text appointment reminders, as well as interpreter resources to facilitate patient visits and patient instructions, also can mitigate language barriers for patients with LEP. Implementing a triage system into the UPMC scheduling system may help patients with more urgent skin conditions to be seen in a timely manner.

Other investigators have analyzed costs of care and proven the value of dermatologic services at free clinics to guide allocation of supplies and resources, demonstrating an area for future investigation at the BFC.¹³ A cost analysis of care provided at the BFC compared to UPMC could inform us about the value of the BFC’s services.

Conclusion

The dermatology clinics at the BFC and UPMC have distinct demographics, diagnoses, and management strategies to provide an inclusive patient care model. The services provided by both clinics are necessary to ensure that people in Pittsburgh have access to dermatologic care regardless of social barriers (eg, lack of health insurance, LEP). To achieve greater accessibility and health equity, dermatologic care at the BFC and UPMC can be improved by strengthening communication with people with LEP, providing skin care education, and offering social and scheduling services.

Approximately 25% of Americans have at least one skin condition, and 20% are estimated to develop skin cancer during their lifetime.^1,2 However, 40% of the US population lives in areas underserved by dermatologists. ³ The severity and mortality of skin cancers such as melanoma and mycosis fungoides have been positively associated with minoritized race, lack of health insurance, and unstable housing status.^4-6 Patients who receive health care at free clinics often are of a racial or ethnic minoritized social group, are uninsured, and/or lack stable housing; this underserved group also includes recent immigrants to the United States who have limited English proficiency (LEP).⁷ Only 25% of free clinics offer specialty care services such as dermatology.^7,8

Of the 42 free clinics and Federally Qualified Health Centers in Pittsburgh, Pennsylvania, the Birmingham Free Clinic (BFC) is one of the few that offers specialty care services including dermatology.⁹ Founded in 1994, the BFC serves as a safety net for Pittsburgh’s medically underserved population, offering primary and acute care, medication access, and social services. From January 2020 to May 2022, the BFC offered 27 dermatology clinics that provided approximately 100 people with comprehensive care including full-body skin examinations, dermatologic diagnoses and treatments, minor procedures, and dermatopathology services.

In this study, we compared the BFC dermatology patient care model with that of the dermatology department at the University of Pittsburgh Medical Center (UPMC), an insurance-based tertiary referral health care system in western Pennsylvania. By analyzing the demographics, dermatologic diagnoses, and management strategies of both the BFC and UPMC, we gained an understanding of how these patient care models differ and how they can be improved to care for diverse patient populations.

Methods

A retrospective chart review of dermatology patients seen in person at the BFC and UPMC during the period from January 2020 to May 2022 was performed. The UPMC group included patients seen by 3 general dermatologists (including A.J.J.) at matched time points. Data were collected from patients’ first in-person visit during the study period. Variables of interest included patient age, sex, race, ethnicity, primary language, zip code, health insurance status, distance to clinic (estimated using Google Maps to calculate the shortest driving distance from the patient’s zip code to the clinic), history of skin cancer, dermatologic diagnoses, and management strategies. These variables were not collected for patients who cancelled or noshowed their first in-person appointments. All patient charts and notes corresponding to the date and visit of interest were accessed through the electronic medical record (EMR). Patient data were de-identified and stored in a password-protected spreadsheet. Comparisons between the BFC and UPMC patient populations were performed using X² tests of independence, Fisher exact tests, and Mann-Whitney U tests via SPSS software (IBM). Statistical significance was set at P<.05.

Results

Patient Characteristics—Our analysis included 76 initial appointments at the BFC and 322 at UPMC (Table 1). The mean age for patients at the BFC and UPMC was 39.6 years and 47.8 years, respectively (P=.001). Males accounted for 39 (51.3%) and 112 (34.8%) of BFC and UPMC patients, respectively (P=.008); 2 (0.6%) patients from UPMC were transgender. Of the BFC and UPMC patients, 44.7% (34/76) and 0.9% (3/322) were Hispanic, respectively (P<.001). With regard to race, 52.6% (40/76) of BFC patients were White, 19.7% (15/76) were Black, 6.6% (5/76) were Asian/Pacific Islander (Chinese, 1.3% [1/76]; other Asian, 5.3% [4/76]), and 21.1% (16/76) were American Indian/other/unspecified (American Indian, 1.3% [1/76]; other, 13.2% [10/76]; unspecified, 6.6% [5/76]). At UPMC, 61.2% (197/322) of patients were White, 28.0% (90/322) were Black, 5.3% (17/322) were Asian/Pacific Islander (Chinese, 1.2% [4/322]; Indian [Asian], 1.9% [6/322]; Japanese, 0.3% [1/322]; other Asian, 1.6% [5/322]; other Asian/American Indian, 0.3% [1/322]), and 5.6% (18/322) were American Indian/other/ unspecified (American Indian, 0.3% [1/322]; other, 0.3% [1/322]; unspecified, 5.0% [16/322]). Overall, the BFC patient population was more ethnically and racially diverse than that of UPMC (P<.001).

Forty-six percent (35/76) of BFC patients and 4.3% (14/322) of UPMC patients had LEP (P<.001). Primary languages among BFC patients were 53.9% (41/76) English, 40.8% (31/76) Spanish, and 5.2% (4/76) other/ unspecified (Chinese, 1.3% [1/76]; Indonesian, 2.6% [2/76]; unspecified, 1.3% [1/76]). Primary languages among UPMC patients were 95.7% (308/322) English and 4.3% (14/322) other/unspecified (Chinese, 0.6% [2/322]; Nepali, 0.6% [2/322]; Pali, 0.3% [1/322]; Russian, 0.3% [1/322]; unspecified, 2.5% [8/322]). There were notable differences in insurance status at the BFC vs UPMC (P<.001), with more UPMC patients having private insurance (52.8% [170/322] vs 11.8% [9/76]) and more BFC patients being uninsured (52.8% [51/76] vs 1.9% [6/322]). There was no significant difference in distance to clinic between the 2 groups (P=.183). More UPMC patients had a history of skin cancer (P=.003). More patients at the BFC were no-shows for their appointments (P<.001), and UPMC patients more frequently canceled their appointments (P<.001).

Dermatologic Diagnoses—The most commonly diagnosed dermatologic conditions at the BFC were dermatitis (23.7% [18/76]), neoplasm of uncertain behavior (15.8% [12/76]), alopecia (11.8% [9/76]), and acne (10.5% [8/76]) (Table 2). The most commonly diagnosed conditions at UPMC were nevi (26.4% [85/322]), dermatitis (22.7% [73/322]), seborrheic keratosis (21.7% [70/322]), and skin cancer screening (21.4% [70/322]). Neoplasm of uncertain behavior was more common in BFC vs UPMC patients (P=.040), while UPMC patients were more frequently diagnosed with nevi (P<.001), seborrheic keratosis (P<.001), and skin cancer screening (P<.001). There was no significant difference between the incidence of skin cancer diagnoses in the BFC (1.3% [1/76]) and UPMC (0.6% [2/76]) patient populations (P=.471). Among the biopsied neoplasms, there was also no significant difference in malignant (BFC, 50.0% [5/10]; UPMC, 32.0% [8/25]) and benign (BFC, 50.0% [5/10]; UPMC, 36.0% [9/25]) neoplasms diagnosed at each clinic (P=.444).

Management Strategies—Systemic antibiotics were more frequently prescribed (P<.001) and laboratory testing/ imaging were more frequently ordered (P=.005) at the BFC vs UPMC (Table 3). Patients at the BFC also more frequently required emergency insurance (P=.036). Patients at UPMC were more frequently recommended sunscreen (P=.003) and received education about skin cancer signs by review of the ABCDEs of melanoma (P<.001), sun-protective behaviors (P=.001), and skin examination frequency (P<.001). Notes in the EMR for UPMC patients more frequently specified patient followup instructions (P<.001).

Comment

As of 2020, the city of Pittsburgh had an estimated population of nearly 303,000 based on US Census data.10 Its population is predominantly White (62.7%) followed by Black/African American (22.8%) and Asian (6.5%); 5.9% identify as 2 or more races. Approximately 3.8% identify as Hispanic or Latino. More than 11% of the Pittsburgh population aged 5 years and older speaks a language other than English as their primary language, including Spanish (2.3%), other Indo-European languages (3.9%), and Asian and Pacific Island languages (3.5%).¹¹ More than 5% of the Pittsburgh population does not have health insurance.¹²

The BFC is located in Pittsburgh’s South Side area, while one of UPMC’s primary dermatology clinics is located in the Oakland district; however, most patients who seek care at these clinics live outside these areas. Our study results indicated that the BFC and UPMC serve distinct groups of people within the Pittsburgh population. The BFC patient population was younger with a higher percentage of patients who were male, Hispanic, racially diverse, and with LEP compared with the UPMC patient population. In this clinical setting, the BFC health care team engages with people from diverse backgrounds and requires greater interpreter and medical support services.

The BFC largely is supported by volunteers, UPMC, grants, and philanthropy. Dermatology clinics are staffed by paid and volunteer team members. Paid team members include 1 nurse and 1 access lead who operates the front desk and registration. Volunteer team members include 1 board-certified dermatologist from UPMC (A.J.J.), or an affiliate clinic and 1 or 2 of each of the following: UPMC dermatology residents, medical or undergraduate students from the University of Pittsburgh, AmeriCorps national service members, and student or community medical interpreters. The onsite pharmacy is run by volunteer faculty, resident, and student pharmacists from the University of Pittsburgh. Dermatology clinics are half-day clinics that occur monthly. Volunteers for each clinic are recruited approximately 1 month in advance.

Dermatology patients at the BFC are referred from the BFC general medicine clinic and nearby Federally Qualified Health Center s for simple to complex medical and surgical dermatologic skin conditions. Each BFC dermatology clinic schedules an average of 7 patients per clinic and places other patients on a wait-list unless more urgent triage is needed. Patients are notified when they are scheduled via phone or text message, and they receive a reminder call or text 1 or 2 days prior to their appointment that also asks them to confirm attendance. Patients with LEP are called with an interpreter and also may receive text reminders that can be translated using Google Translate. Patients are instructed to notify the BFC if they need to cancel or reschedule their appointment. At the end of each visit, patients are given an after-visit summary that lists follow-up instructions, medications prescribed during the visit, and upcoming appointments. The BFC offers bus tickets to help patients get to their appointments. In rare cases, the BFC may pay for a car service to drive patients to and from the clinic.

Dermatology clinics at UPMC use scheduling and self-scheduling systems through which patients can make appointments at a location of their choice with any available board-certified dermatologist or physician assistant. Patients receive a reminder phone call 3 days prior to their appointment instructing them to call the office if they are unable to keep their appointment. Patients signed up for the online portal also receive a reminder message and an option to confirm or cancel their appointment. Patients with cell phone numbers in the UPMC system receive a text message approximately 2 days prior to their appointment that allows them to preregister and pay their copayment in advance. They receive another text 20 minutes prior to their appointment with an option for contactless check-in. At the conclusion of their visit, patients can schedule a follow-up appointment and receive a printed copy of their after-visit summary that provides information about follow-up instructions, prescribed medications, and upcoming visits. They may alternatively access this summary via the online patient portal. Patients are not provided transportation to UPMC clinics, but they are offered parking validation.

Among the most common dermatologic diagnoses for each group, BFC patients presented for treatment of more acute dermatologic conditions, while UPMC patients presented for more benign and preventive-care conditions. This difference may be attributable to the BFC’s referral and triage system, wherein patients with more urgent problems are given scheduling priority. This patient care model contrasts with UPMC’s scheduling process in which no known formal triage system is utilized. Interestingly, there was no difference in skin cancer incidence despite a higher percentage of preventive skin cancer screenings at UPMC.

Patients at the BFC more often required emergency insurance for surgical interventions, which is consistent with the higher percentage of uninsured individuals in this population. Patients at UPMC more frequently were recommended sunscreen and were educated about skin cancer, sun protection, and skin examination, in part due to this group’s more extensive history of skin cancer and frequent presentation for skin cancer screenings. At the same time, educational materials for skin care at both the BFC and UPMC are populated into the EMR in English, whereas materials in other languages are less readily available.

Our retrospective study had several limitations. Demographic information that relied on clinic-dependent intake questionnaires may be limited due to variable intake processes and patients opting out of self-reporting. By comparing patient populations between 2 clinics, confounding variables such as location and hours of operation may impact the patient demographics recorded at the BFC vs UPMC. Resources and staff availability may affect the management strategies and follow-up care offered by each clinic. Our study period also was unique in that COVID-19 may have affected resources, staffing, scheduling, and logistics at both clinics.

Based on the aforementioned differences between the BFC and UPMC patient characteristics, care models should be strategically designed to support the needs of diverse populations. The BFC patient care model appropriately focuses on communication skills with patients with LEP by using interpreter services. Providing more skin care education and follow-up instructions in patients’ primary languages will help them develop a better understanding of their skin conditions. Another key asset of the BFC patient care model is its provision of social services such as transportation and insurance assistance.

To improve the UPMC patient care model, providing patients with bus tickets and car services may potentially reduce appointment cancellations. Using interpreter services to call and text appointment reminders, as well as interpreter resources to facilitate patient visits and patient instructions, also can mitigate language barriers for patients with LEP. Implementing a triage system into the UPMC scheduling system may help patients with more urgent skin conditions to be seen in a timely manner.

Other investigators have analyzed costs of care and proven the value of dermatologic services at free clinics to guide allocation of supplies and resources, demonstrating an area for future investigation at the BFC.¹³ A cost analysis of care provided at the BFC compared to UPMC could inform us about the value of the BFC’s services.

Conclusion

The dermatology clinics at the BFC and UPMC have distinct demographics, diagnoses, and management strategies to provide an inclusive patient care model. The services provided by both clinics are necessary to ensure that people in Pittsburgh have access to dermatologic care regardless of social barriers (eg, lack of health insurance, LEP). To achieve greater accessibility and health equity, dermatologic care at the BFC and UPMC can be improved by strengthening communication with people with LEP, providing skin care education, and offering social and scheduling services.

Approximately 25% of Americans have at least one skin condition, and 20% are estimated to develop skin cancer during their lifetime.^1,2 However, 40% of the US population lives in areas underserved by dermatologists. ³ The severity and mortality of skin cancers such as melanoma and mycosis fungoides have been positively associated with minoritized race, lack of health insurance, and unstable housing status.^4-6 Patients who receive health care at free clinics often are of a racial or ethnic minoritized social group, are uninsured, and/or lack stable housing; this underserved group also includes recent immigrants to the United States who have limited English proficiency (LEP).⁷ Only 25% of free clinics offer specialty care services such as dermatology.^7,8

Of the 42 free clinics and Federally Qualified Health Centers in Pittsburgh, Pennsylvania, the Birmingham Free Clinic (BFC) is one of the few that offers specialty care services including dermatology.⁹ Founded in 1994, the BFC serves as a safety net for Pittsburgh’s medically underserved population, offering primary and acute care, medication access, and social services. From January 2020 to May 2022, the BFC offered 27 dermatology clinics that provided approximately 100 people with comprehensive care including full-body skin examinations, dermatologic diagnoses and treatments, minor procedures, and dermatopathology services.

In this study, we compared the BFC dermatology patient care model with that of the dermatology department at the University of Pittsburgh Medical Center (UPMC), an insurance-based tertiary referral health care system in western Pennsylvania. By analyzing the demographics, dermatologic diagnoses, and management strategies of both the BFC and UPMC, we gained an understanding of how these patient care models differ and how they can be improved to care for diverse patient populations.

Methods

A retrospective chart review of dermatology patients seen in person at the BFC and UPMC during the period from January 2020 to May 2022 was performed. The UPMC group included patients seen by 3 general dermatologists (including A.J.J.) at matched time points. Data were collected from patients’ first in-person visit during the study period. Variables of interest included patient age, sex, race, ethnicity, primary language, zip code, health insurance status, distance to clinic (estimated using Google Maps to calculate the shortest driving distance from the patient’s zip code to the clinic), history of skin cancer, dermatologic diagnoses, and management strategies. These variables were not collected for patients who cancelled or noshowed their first in-person appointments. All patient charts and notes corresponding to the date and visit of interest were accessed through the electronic medical record (EMR). Patient data were de-identified and stored in a password-protected spreadsheet. Comparisons between the BFC and UPMC patient populations were performed using X² tests of independence, Fisher exact tests, and Mann-Whitney U tests via SPSS software (IBM). Statistical significance was set at P<.05.

Results

Patient Characteristics—Our analysis included 76 initial appointments at the BFC and 322 at UPMC (Table 1). The mean age for patients at the BFC and UPMC was 39.6 years and 47.8 years, respectively (P=.001). Males accounted for 39 (51.3%) and 112 (34.8%) of BFC and UPMC patients, respectively (P=.008); 2 (0.6%) patients from UPMC were transgender. Of the BFC and UPMC patients, 44.7% (34/76) and 0.9% (3/322) were Hispanic, respectively (P<.001). With regard to race, 52.6% (40/76) of BFC patients were White, 19.7% (15/76) were Black, 6.6% (5/76) were Asian/Pacific Islander (Chinese, 1.3% [1/76]; other Asian, 5.3% [4/76]), and 21.1% (16/76) were American Indian/other/unspecified (American Indian, 1.3% [1/76]; other, 13.2% [10/76]; unspecified, 6.6% [5/76]). At UPMC, 61.2% (197/322) of patients were White, 28.0% (90/322) were Black, 5.3% (17/322) were Asian/Pacific Islander (Chinese, 1.2% [4/322]; Indian [Asian], 1.9% [6/322]; Japanese, 0.3% [1/322]; other Asian, 1.6% [5/322]; other Asian/American Indian, 0.3% [1/322]), and 5.6% (18/322) were American Indian/other/ unspecified (American Indian, 0.3% [1/322]; other, 0.3% [1/322]; unspecified, 5.0% [16/322]). Overall, the BFC patient population was more ethnically and racially diverse than that of UPMC (P<.001).

Forty-six percent (35/76) of BFC patients and 4.3% (14/322) of UPMC patients had LEP (P<.001). Primary languages among BFC patients were 53.9% (41/76) English, 40.8% (31/76) Spanish, and 5.2% (4/76) other/ unspecified (Chinese, 1.3% [1/76]; Indonesian, 2.6% [2/76]; unspecified, 1.3% [1/76]). Primary languages among UPMC patients were 95.7% (308/322) English and 4.3% (14/322) other/unspecified (Chinese, 0.6% [2/322]; Nepali, 0.6% [2/322]; Pali, 0.3% [1/322]; Russian, 0.3% [1/322]; unspecified, 2.5% [8/322]). There were notable differences in insurance status at the BFC vs UPMC (P<.001), with more UPMC patients having private insurance (52.8% [170/322] vs 11.8% [9/76]) and more BFC patients being uninsured (52.8% [51/76] vs 1.9% [6/322]). There was no significant difference in distance to clinic between the 2 groups (P=.183). More UPMC patients had a history of skin cancer (P=.003). More patients at the BFC were no-shows for their appointments (P<.001), and UPMC patients more frequently canceled their appointments (P<.001).

Dermatologic Diagnoses—The most commonly diagnosed dermatologic conditions at the BFC were dermatitis (23.7% [18/76]), neoplasm of uncertain behavior (15.8% [12/76]), alopecia (11.8% [9/76]), and acne (10.5% [8/76]) (Table 2). The most commonly diagnosed conditions at UPMC were nevi (26.4% [85/322]), dermatitis (22.7% [73/322]), seborrheic keratosis (21.7% [70/322]), and skin cancer screening (21.4% [70/322]). Neoplasm of uncertain behavior was more common in BFC vs UPMC patients (P=.040), while UPMC patients were more frequently diagnosed with nevi (P<.001), seborrheic keratosis (P<.001), and skin cancer screening (P<.001). There was no significant difference between the incidence of skin cancer diagnoses in the BFC (1.3% [1/76]) and UPMC (0.6% [2/76]) patient populations (P=.471). Among the biopsied neoplasms, there was also no significant difference in malignant (BFC, 50.0% [5/10]; UPMC, 32.0% [8/25]) and benign (BFC, 50.0% [5/10]; UPMC, 36.0% [9/25]) neoplasms diagnosed at each clinic (P=.444).

Management Strategies—Systemic antibiotics were more frequently prescribed (P<.001) and laboratory testing/ imaging were more frequently ordered (P=.005) at the BFC vs UPMC (Table 3). Patients at the BFC also more frequently required emergency insurance (P=.036). Patients at UPMC were more frequently recommended sunscreen (P=.003) and received education about skin cancer signs by review of the ABCDEs of melanoma (P<.001), sun-protective behaviors (P=.001), and skin examination frequency (P<.001). Notes in the EMR for UPMC patients more frequently specified patient followup instructions (P<.001).

Comment

As of 2020, the city of Pittsburgh had an estimated population of nearly 303,000 based on US Census data.10 Its population is predominantly White (62.7%) followed by Black/African American (22.8%) and Asian (6.5%); 5.9% identify as 2 or more races. Approximately 3.8% identify as Hispanic or Latino. More than 11% of the Pittsburgh population aged 5 years and older speaks a language other than English as their primary language, including Spanish (2.3%), other Indo-European languages (3.9%), and Asian and Pacific Island languages (3.5%).¹¹ More than 5% of the Pittsburgh population does not have health insurance.¹²

The BFC is located in Pittsburgh’s South Side area, while one of UPMC’s primary dermatology clinics is located in the Oakland district; however, most patients who seek care at these clinics live outside these areas. Our study results indicated that the BFC and UPMC serve distinct groups of people within the Pittsburgh population. The BFC patient population was younger with a higher percentage of patients who were male, Hispanic, racially diverse, and with LEP compared with the UPMC patient population. In this clinical setting, the BFC health care team engages with people from diverse backgrounds and requires greater interpreter and medical support services.

The BFC largely is supported by volunteers, UPMC, grants, and philanthropy. Dermatology clinics are staffed by paid and volunteer team members. Paid team members include 1 nurse and 1 access lead who operates the front desk and registration. Volunteer team members include 1 board-certified dermatologist from UPMC (A.J.J.), or an affiliate clinic and 1 or 2 of each of the following: UPMC dermatology residents, medical or undergraduate students from the University of Pittsburgh, AmeriCorps national service members, and student or community medical interpreters. The onsite pharmacy is run by volunteer faculty, resident, and student pharmacists from the University of Pittsburgh. Dermatology clinics are half-day clinics that occur monthly. Volunteers for each clinic are recruited approximately 1 month in advance.

Dermatology patients at the BFC are referred from the BFC general medicine clinic and nearby Federally Qualified Health Center s for simple to complex medical and surgical dermatologic skin conditions. Each BFC dermatology clinic schedules an average of 7 patients per clinic and places other patients on a wait-list unless more urgent triage is needed. Patients are notified when they are scheduled via phone or text message, and they receive a reminder call or text 1 or 2 days prior to their appointment that also asks them to confirm attendance. Patients with LEP are called with an interpreter and also may receive text reminders that can be translated using Google Translate. Patients are instructed to notify the BFC if they need to cancel or reschedule their appointment. At the end of each visit, patients are given an after-visit summary that lists follow-up instructions, medications prescribed during the visit, and upcoming appointments. The BFC offers bus tickets to help patients get to their appointments. In rare cases, the BFC may pay for a car service to drive patients to and from the clinic.

Dermatology clinics at UPMC use scheduling and self-scheduling systems through which patients can make appointments at a location of their choice with any available board-certified dermatologist or physician assistant. Patients receive a reminder phone call 3 days prior to their appointment instructing them to call the office if they are unable to keep their appointment. Patients signed up for the online portal also receive a reminder message and an option to confirm or cancel their appointment. Patients with cell phone numbers in the UPMC system receive a text message approximately 2 days prior to their appointment that allows them to preregister and pay their copayment in advance. They receive another text 20 minutes prior to their appointment with an option for contactless check-in. At the conclusion of their visit, patients can schedule a follow-up appointment and receive a printed copy of their after-visit summary that provides information about follow-up instructions, prescribed medications, and upcoming visits. They may alternatively access this summary via the online patient portal. Patients are not provided transportation to UPMC clinics, but they are offered parking validation.

Among the most common dermatologic diagnoses for each group, BFC patients presented for treatment of more acute dermatologic conditions, while UPMC patients presented for more benign and preventive-care conditions. This difference may be attributable to the BFC’s referral and triage system, wherein patients with more urgent problems are given scheduling priority. This patient care model contrasts with UPMC’s scheduling process in which no known formal triage system is utilized. Interestingly, there was no difference in skin cancer incidence despite a higher percentage of preventive skin cancer screenings at UPMC.

Patients at the BFC more often required emergency insurance for surgical interventions, which is consistent with the higher percentage of uninsured individuals in this population. Patients at UPMC more frequently were recommended sunscreen and were educated about skin cancer, sun protection, and skin examination, in part due to this group’s more extensive history of skin cancer and frequent presentation for skin cancer screenings. At the same time, educational materials for skin care at both the BFC and UPMC are populated into the EMR in English, whereas materials in other languages are less readily available.

Our retrospective study had several limitations. Demographic information that relied on clinic-dependent intake questionnaires may be limited due to variable intake processes and patients opting out of self-reporting. By comparing patient populations between 2 clinics, confounding variables such as location and hours of operation may impact the patient demographics recorded at the BFC vs UPMC. Resources and staff availability may affect the management strategies and follow-up care offered by each clinic. Our study period also was unique in that COVID-19 may have affected resources, staffing, scheduling, and logistics at both clinics.

Based on the aforementioned differences between the BFC and UPMC patient characteristics, care models should be strategically designed to support the needs of diverse populations. The BFC patient care model appropriately focuses on communication skills with patients with LEP by using interpreter services. Providing more skin care education and follow-up instructions in patients’ primary languages will help them develop a better understanding of their skin conditions. Another key asset of the BFC patient care model is its provision of social services such as transportation and insurance assistance.

To improve the UPMC patient care model, providing patients with bus tickets and car services may potentially reduce appointment cancellations. Using interpreter services to call and text appointment reminders, as well as interpreter resources to facilitate patient visits and patient instructions, also can mitigate language barriers for patients with LEP. Implementing a triage system into the UPMC scheduling system may help patients with more urgent skin conditions to be seen in a timely manner.

Other investigators have analyzed costs of care and proven the value of dermatologic services at free clinics to guide allocation of supplies and resources, demonstrating an area for future investigation at the BFC.¹³ A cost analysis of care provided at the BFC compared to UPMC could inform us about the value of the BFC’s services.

Conclusion

The dermatology clinics at the BFC and UPMC have distinct demographics, diagnoses, and management strategies to provide an inclusive patient care model. The services provided by both clinics are necessary to ensure that people in Pittsburgh have access to dermatologic care regardless of social barriers (eg, lack of health insurance, LEP). To achieve greater accessibility and health equity, dermatologic care at the BFC and UPMC can be improved by strengthening communication with people with LEP, providing skin care education, and offering social and scheduling services.

Hot Topics in Primary Care 2024 is a resource that explores the newest developments in primary care topics that impact your daily clinical practice. 

Click on the link below to access the entire supplement. 

• Case Studies in Continuous Glucose Monitoring
• Detection and Diagnosis of Early Symptomatic 
  Alzheimer’s Disease in Primary Care
• Elevating the Importance of Asthma Care in the United States
• Hypercortisolism is More Common Than You Think – 
  Here’s How to Find It
• Improving COPD Management at Transitions of Care
• Improving Patient-Centric COPD Management 
• The Role of Finerenone in Optimizing Cardiovascular-
  Kidney-Metabolic Health: Everything PCPs Should Know
• What Primary Care Clinicians Need to Know About Once-Weekly Insulins

This supplement offers the opportunity to earn a total of 3 continuing medical education (CME) credits. Credit is awarded for the successful completion of the evaluation after reading the article. The links can be found within the supplement on the first page of each article where CME credits are offered.

Click here to read the 2024 Hot Topics in Primary Care

Hot Topics in Primary Care 2024 is a resource that explores the newest developments in primary care topics that impact your daily clinical practice. 

Click on the link below to access the entire supplement. 

• Case Studies in Continuous Glucose Monitoring
• Detection and Diagnosis of Early Symptomatic 
  Alzheimer’s Disease in Primary Care
• Elevating the Importance of Asthma Care in the United States
• Hypercortisolism is More Common Than You Think – 
  Here’s How to Find It
• Improving COPD Management at Transitions of Care
• Improving Patient-Centric COPD Management 
• The Role of Finerenone in Optimizing Cardiovascular-
  Kidney-Metabolic Health: Everything PCPs Should Know
• What Primary Care Clinicians Need to Know About Once-Weekly Insulins

This supplement offers the opportunity to earn a total of 3 continuing medical education (CME) credits. Credit is awarded for the successful completion of the evaluation after reading the article. The links can be found within the supplement on the first page of each article where CME credits are offered.

Click here to read the 2024 Hot Topics in Primary Care

Hot Topics in Primary Care 2024 is a resource that explores the newest developments in primary care topics that impact your daily clinical practice. 

Click on the link below to access the entire supplement. 

• Case Studies in Continuous Glucose Monitoring
• Detection and Diagnosis of Early Symptomatic 
  Alzheimer’s Disease in Primary Care
• Elevating the Importance of Asthma Care in the United States
• Hypercortisolism is More Common Than You Think – 
  Here’s How to Find It
• Improving COPD Management at Transitions of Care
• Improving Patient-Centric COPD Management 
• The Role of Finerenone in Optimizing Cardiovascular-
  Kidney-Metabolic Health: Everything PCPs Should Know
• What Primary Care Clinicians Need to Know About Once-Weekly Insulins

This supplement offers the opportunity to earn a total of 3 continuing medical education (CME) credits. Credit is awarded for the successful completion of the evaluation after reading the article. The links can be found within the supplement on the first page of each article where CME credits are offered.

Click here to read the 2024 Hot Topics in Primary Care

In assessing chronic medical disorders that adversely affect quality of life, including disorders such as malaria, acquired immunodeficiency syndrome, and sickle cell anemia, the World Health Organization has ranked migraine as the number 1 disorder for women. The diagnosis of chronic migraine (CM) implies an established history of migraine, including 15 or more days of headache per month. This clinical variant accounts for a disproportionate share of the public health burden generally imposed by migraine. One of the great paradoxes of migraine is that, despite this outsized burden, as few as 5% of the millions of Americans with CM seek medical attention, are diagnosed accurately, and receive appropriate therapy.

Why is this? The answer lies in part with the relatively recent formal recognition of CM as a distinct primary headache disorder. Not until the 2006 revision of the International Classification of Headache Disorders were widely accepted diagnostic criteria for CM available. Without such criteria in place, meaningful clinical research (and therapeutic research in particular) is impossible. As recently as 2009, we consequently lacked any evidence-based therapies for suppression of CM. In what may be a therapeutic revolution unrivaled in almost any other area of clinical medicine, within 15 years we have identified and introduced into general clinical practice no fewer than 6 new evidence-based therapies for suppression of CM that are safe, generally well tolerated, and effective. Those therapies are onabotulinumtoxin A, an orally administered atogepant, the 3 subcutaneously self-administered anti–calcitonin gene-related peptide (CGRP) monoclonal antibodies (erenumab, galcanezumab, and fremanezumab), and the intravenously administered anti-CGRP monoclonal antibody eptinenzumab. A seventh and older therapy, topiramate, could be included in that group, except tolerability issues reduce its utility.

As we presently lack adequate active vs active comparator trials, at this point there is no “best” among these 6 CM therapies. Each is effective in reducing migraine burden substantially in a large proportion of patients. In many cases, the onset of the positive treatment response is gratifyingly rapid. Are they disease-modifying therapies (DMTs) that, in addition to suppressing symptoms, can alter the disorder’s biologic underpinnings and its long-term course? Put another way, does their benefit extend beyond the period of time the patient is actively on treatment? Given their relative newness, the answer is unknown. However, preliminary evidence suggests that onabotulinumtoxin A may qualify as a DMT. The US Food and Drug Administration indicated onabotulinumtoxin A for CM in 2010, and it is the first of the group introduced into clinical practice.

Complicating clinical use of these 6 therapies for suppression of CM have been various proscriptions emanating from the health care insurance industry. Chief among these has been the oft-encountered mandate that, for 1 of these 6 evidence-based treatments to be authorized for coverage, the patient with CM first must fail an “adequate” trial of multiple older, generic, and less costly therapies commonly used for prophylaxis in episodic migraine. But, with the exception of frequently difficult-to-tolerate topiramate, these therapies have no meaningful evidence base for use in CM. Furthermore, some limited evidence suggests that the efficacy of the evidence-based treatments may be less robust in patients whose CM has been longstanding. Therefore, spending many months prescribing a succession of therapies lacking an evidence base seems counterproductive and contrary to the patient’s best interests, when evidence-based therapies are available. In addition, the higher cost of evidence-based therapies may be offset by their reduction in the direct medical costs that result from clinical improvement.

The American Headache Society (AHS) recently published a position statement recommending that the CGRP-targeting medications─both the small molecule gepants and the large molecule monoclonal antibodies─be considered first-line treatment for migraine prevention. An estimated 40% of patients with migraine require prevention therapy. The soundness of the AHS recommendation is especially evident in the sizeable subset of that population with CM; for treatment of CM, this list of recommended first-line therapies logically can be extended to include onabotulinumtoxin A. In their consensus statement the AHS acknowledged the relatively higher cost of these therapies but noted the resulting reduction in direct and indirect costs could justify their use.

Migraine ranks near or at the top of the list of chronic medical disorders that adversely impact public health. CM, a common variant of migraine, is particularly adept at eroding quality of life. Presently available are 6 safe and typically well-tolerated therapies, known to be effective for treating CM and potentially capable of modifying its long-term course. Although the cost of these exceeds the cost of older generic therapies commonly used for migraine prophylaxis, those older therapies generally lack any meaningful evidence base for use in CM. In addition, prescription of these older therapies may delay easing the CM patient’s migraine burden without any associated reduction in long-term net cost. Both medically and financially, a strong case can be made for designating these 6 medications as first-line therapy for CM. Insurers, the ball is now in your court.

In assessing chronic medical disorders that adversely affect quality of life, including disorders such as malaria, acquired immunodeficiency syndrome, and sickle cell anemia, the World Health Organization has ranked migraine as the number 1 disorder for women. The diagnosis of chronic migraine (CM) implies an established history of migraine, including 15 or more days of headache per month. This clinical variant accounts for a disproportionate share of the public health burden generally imposed by migraine. One of the great paradoxes of migraine is that, despite this outsized burden, as few as 5% of the millions of Americans with CM seek medical attention, are diagnosed accurately, and receive appropriate therapy.

Why is this? The answer lies in part with the relatively recent formal recognition of CM as a distinct primary headache disorder. Not until the 2006 revision of the International Classification of Headache Disorders were widely accepted diagnostic criteria for CM available. Without such criteria in place, meaningful clinical research (and therapeutic research in particular) is impossible. As recently as 2009, we consequently lacked any evidence-based therapies for suppression of CM. In what may be a therapeutic revolution unrivaled in almost any other area of clinical medicine, within 15 years we have identified and introduced into general clinical practice no fewer than 6 new evidence-based therapies for suppression of CM that are safe, generally well tolerated, and effective. Those therapies are onabotulinumtoxin A, an orally administered atogepant, the 3 subcutaneously self-administered anti–calcitonin gene-related peptide (CGRP) monoclonal antibodies (erenumab, galcanezumab, and fremanezumab), and the intravenously administered anti-CGRP monoclonal antibody eptinenzumab. A seventh and older therapy, topiramate, could be included in that group, except tolerability issues reduce its utility.

As we presently lack adequate active vs active comparator trials, at this point there is no “best” among these 6 CM therapies. Each is effective in reducing migraine burden substantially in a large proportion of patients. In many cases, the onset of the positive treatment response is gratifyingly rapid. Are they disease-modifying therapies (DMTs) that, in addition to suppressing symptoms, can alter the disorder’s biologic underpinnings and its long-term course? Put another way, does their benefit extend beyond the period of time the patient is actively on treatment? Given their relative newness, the answer is unknown. However, preliminary evidence suggests that onabotulinumtoxin A may qualify as a DMT. The US Food and Drug Administration indicated onabotulinumtoxin A for CM in 2010, and it is the first of the group introduced into clinical practice.

Complicating clinical use of these 6 therapies for suppression of CM have been various proscriptions emanating from the health care insurance industry. Chief among these has been the oft-encountered mandate that, for 1 of these 6 evidence-based treatments to be authorized for coverage, the patient with CM first must fail an “adequate” trial of multiple older, generic, and less costly therapies commonly used for prophylaxis in episodic migraine. But, with the exception of frequently difficult-to-tolerate topiramate, these therapies have no meaningful evidence base for use in CM. Furthermore, some limited evidence suggests that the efficacy of the evidence-based treatments may be less robust in patients whose CM has been longstanding. Therefore, spending many months prescribing a succession of therapies lacking an evidence base seems counterproductive and contrary to the patient’s best interests, when evidence-based therapies are available. In addition, the higher cost of evidence-based therapies may be offset by their reduction in the direct medical costs that result from clinical improvement.

The American Headache Society (AHS) recently published a position statement recommending that the CGRP-targeting medications─both the small molecule gepants and the large molecule monoclonal antibodies─be considered first-line treatment for migraine prevention. An estimated 40% of patients with migraine require prevention therapy. The soundness of the AHS recommendation is especially evident in the sizeable subset of that population with CM; for treatment of CM, this list of recommended first-line therapies logically can be extended to include onabotulinumtoxin A. In their consensus statement the AHS acknowledged the relatively higher cost of these therapies but noted the resulting reduction in direct and indirect costs could justify their use.

Migraine ranks near or at the top of the list of chronic medical disorders that adversely impact public health. CM, a common variant of migraine, is particularly adept at eroding quality of life. Presently available are 6 safe and typically well-tolerated therapies, known to be effective for treating CM and potentially capable of modifying its long-term course. Although the cost of these exceeds the cost of older generic therapies commonly used for migraine prophylaxis, those older therapies generally lack any meaningful evidence base for use in CM. In addition, prescription of these older therapies may delay easing the CM patient’s migraine burden without any associated reduction in long-term net cost. Both medically and financially, a strong case can be made for designating these 6 medications as first-line therapy for CM. Insurers, the ball is now in your court.

In assessing chronic medical disorders that adversely affect quality of life, including disorders such as malaria, acquired immunodeficiency syndrome, and sickle cell anemia, the World Health Organization has ranked migraine as the number 1 disorder for women. The diagnosis of chronic migraine (CM) implies an established history of migraine, including 15 or more days of headache per month. This clinical variant accounts for a disproportionate share of the public health burden generally imposed by migraine. One of the great paradoxes of migraine is that, despite this outsized burden, as few as 5% of the millions of Americans with CM seek medical attention, are diagnosed accurately, and receive appropriate therapy.

Why is this? The answer lies in part with the relatively recent formal recognition of CM as a distinct primary headache disorder. Not until the 2006 revision of the International Classification of Headache Disorders were widely accepted diagnostic criteria for CM available. Without such criteria in place, meaningful clinical research (and therapeutic research in particular) is impossible. As recently as 2009, we consequently lacked any evidence-based therapies for suppression of CM. In what may be a therapeutic revolution unrivaled in almost any other area of clinical medicine, within 15 years we have identified and introduced into general clinical practice no fewer than 6 new evidence-based therapies for suppression of CM that are safe, generally well tolerated, and effective. Those therapies are onabotulinumtoxin A, an orally administered atogepant, the 3 subcutaneously self-administered anti–calcitonin gene-related peptide (CGRP) monoclonal antibodies (erenumab, galcanezumab, and fremanezumab), and the intravenously administered anti-CGRP monoclonal antibody eptinenzumab. A seventh and older therapy, topiramate, could be included in that group, except tolerability issues reduce its utility.

As we presently lack adequate active vs active comparator trials, at this point there is no “best” among these 6 CM therapies. Each is effective in reducing migraine burden substantially in a large proportion of patients. In many cases, the onset of the positive treatment response is gratifyingly rapid. Are they disease-modifying therapies (DMTs) that, in addition to suppressing symptoms, can alter the disorder’s biologic underpinnings and its long-term course? Put another way, does their benefit extend beyond the period of time the patient is actively on treatment? Given their relative newness, the answer is unknown. However, preliminary evidence suggests that onabotulinumtoxin A may qualify as a DMT. The US Food and Drug Administration indicated onabotulinumtoxin A for CM in 2010, and it is the first of the group introduced into clinical practice.

Complicating clinical use of these 6 therapies for suppression of CM have been various proscriptions emanating from the health care insurance industry. Chief among these has been the oft-encountered mandate that, for 1 of these 6 evidence-based treatments to be authorized for coverage, the patient with CM first must fail an “adequate” trial of multiple older, generic, and less costly therapies commonly used for prophylaxis in episodic migraine. But, with the exception of frequently difficult-to-tolerate topiramate, these therapies have no meaningful evidence base for use in CM. Furthermore, some limited evidence suggests that the efficacy of the evidence-based treatments may be less robust in patients whose CM has been longstanding. Therefore, spending many months prescribing a succession of therapies lacking an evidence base seems counterproductive and contrary to the patient’s best interests, when evidence-based therapies are available. In addition, the higher cost of evidence-based therapies may be offset by their reduction in the direct medical costs that result from clinical improvement.

The American Headache Society (AHS) recently published a position statement recommending that the CGRP-targeting medications─both the small molecule gepants and the large molecule monoclonal antibodies─be considered first-line treatment for migraine prevention. An estimated 40% of patients with migraine require prevention therapy. The soundness of the AHS recommendation is especially evident in the sizeable subset of that population with CM; for treatment of CM, this list of recommended first-line therapies logically can be extended to include onabotulinumtoxin A. In their consensus statement the AHS acknowledged the relatively higher cost of these therapies but noted the resulting reduction in direct and indirect costs could justify their use.

Migraine ranks near or at the top of the list of chronic medical disorders that adversely impact public health. CM, a common variant of migraine, is particularly adept at eroding quality of life. Presently available are 6 safe and typically well-tolerated therapies, known to be effective for treating CM and potentially capable of modifying its long-term course. Although the cost of these exceeds the cost of older generic therapies commonly used for migraine prophylaxis, those older therapies generally lack any meaningful evidence base for use in CM. In addition, prescription of these older therapies may delay easing the CM patient’s migraine burden without any associated reduction in long-term net cost. Both medically and financially, a strong case can be made for designating these 6 medications as first-line therapy for CM. Insurers, the ball is now in your court.