Lisa L. Aquino, MD, MS

Dermatology has become one of the most competitive, if not the most competitive, medical specialties to enter. It attracts the brightest and most accomplished medical students who have excelled not only in the classroom and clinical setting but also in the research setting. Many successful applicants take a substantial amount of time off to pursue research and publish articles.

Despite the competitive nature of the specialty, it is well known that a marked shortage of academic dermatologists has existed for more than 30 years.^1-3 In fact, the number of graduates from US dermatology residency programs who pursue academic careers has progressively declined.⁴ Nearly all dermatology residents have a strong academic background; however, many residents opt to pursue private practice instead of a career in academia.^5-9 This trend has implications not only for future dermatology research but also for the teaching and training of future generations of dermatologists.⁸

To address this shortage, it is important to recruit dermatology residents who have a genuine interest in pursuing academic careers. Unfortunately, many residency applicants may overinflate their interest in academics to boost their chances of acceptance.⁴ Additionally, it has been shown that dermatology residents who were interested in academic careers at the time of application to the program often lost interest during residency.¹⁰

Because it can be difficult to determine a resident’s true interest in an academic career at the time of application and his/her initial interest may wean during residency, it may be more helpful to encourage dermatology residency programs to create environments that will produce residents who are more enthusiastic about and more likely to pursue careers in academia. A lack of mentorship has been shown to be associated with a loss of interest in academic careers during residency.¹⁰ If better mentorship opportunities were provided, then perhaps dermatology residents would be more likely to pursue careers in teaching and research.

A 2006 study by Wu et al⁵ demonstrated that various program characteristics were associated with the pursuit of academic careers among dermatology residents. The number of faculty members and the number of full-time faculty publications at a given residency program were most strongly correlated with the number of residents who pursued academic careers, which suggested that having a large faculty and encouraging dermatology residents to publish research during residency may motivate their pursuit of academic careers in dermatology.⁵ The current study was designed to replicate these data and respond to limitations in the original study.

Methods

Data were collected from all accredited dermatology residency programs in the United States as of December 31, 2008. The names of all full-time faculty members at these resident programs were obtained, and it was determined where each faculty member attended dermatology residency. The number of graduates who became full-time clinical or research faculty members and the number of graduates who became chairs or chiefs were counted. Residency programs excluded from these analyses included The University of Texas at Austin, University of Texas Medical Branch, and University of Connecticut, which commenced in 2008, as well as Kaiser Permanente Southern California, which commenced in 2010. Residency programs that were started after 2004 were excluded from the study, as it was thought that these programs may not have graduated a sufficient number of residents for assessment. Military residency programs also were excluded, as graduates from these programs often do not freely choose their careers after residency, and the National Institutes of Health (NIH) dermatology residency program was excluded because it is not a traditional 3-year residency program.

The primary end point was the ratio of full-time faculty members graduated to the total number of graduates from each dermatology residency program. Based on a prior study by Wu et al⁵ in 2006, it was believed that several program variables might affect pursuit of academic careers among dermatology residents, including total number of full-time faculty members, total number of residents, NIH funding received (in dollars) in 2008 (http://www.report.nih.gov/award/index.cfm), Dermatology Foundation (DF) funding received (in dollars) in 2008 (http://www.dermatology foundation.org/rap/), number of publications from full-time faculty members in 2008 (http://www.ncbi.nlm.nih.gov/pubmed/), number of full-time faculty lectures given at annual meetings of 5 societies in 2008 (American Academy of Dermatology, the Society for Investigative Dermatology, the American Society of Dermatopathology, the Society for Pediatric Dermatology, and American Society for Dermatologic Surgery), number of faculty members on the editorial boards of 6 major dermatology journals (Journal of the American Academy of Dermatology, Journal of Investigative Dermatology, Archives of Dermatology [currently known as JAMA Dermatology], Dermatologic Surgery, Pediatric Dermatology, and Journal of Cutaneous Pathology), and status as a department of dermatology or a division of internal medicine. The association between the ratio of number of full-time faculty members to number of residents for each residency program were determined for each of the outcome variables because they were believed to serve as an indicator of mentorship.

Data regarding faculty and residents were obtained from program Web sites and inquiries from individual programs. The year 1974 was used as a cutoff for the total number of graduates from each program. For faculty members who split time between 2 residency programs, each program was given credit for the duration of time spent at that program. If it was not clear how long the faculty member spent at each program, a credit of 1.5 years was given, which is half the duration of a dermatology residency. Faculty members who held a PhD only and those who completed their residencies in non-US dermatology residency programs were excluded from the outcome variables. To avoid duplicate faculty publications, collections for each residency program were created within PubMed (ie, if 2 authors from the same program coauthored an article, it was only counted once toward the total number of faculty publications from that program).

Descriptive exploratory statistical analysis in the form of a correlation matrix was completed to determine the most strongly positive and negative variables that were correlated with the ratio of graduating full-time faculty to estimated total graduates. Variables also were correlated with the secondary outcomes of ratio of graduating department chairs/chiefs to estimated total number of graduates and ratio of graduating program directors to estimated total number of graduates. Spearman rank correlation coefficients and P values were reported. Additionally, a 2-sample t test was performed to compare primary and secondary outcome variables between dermatology department versus division of dermatology under the department of internal medicine. Data were analyzed using SAS version 9.2. The institutional review board at Kaiser Permanente Southern California approved this study.

Results

Due to space considerations, analyses are based on data that are not published in this article. Data regarding the characteristics of each residency program are available from the authors.

Data from 103 dermatology residency programs were included in the analysis. Of these programs, 43% had received NIH funding in 2008 and 22% had received DF funding. Two-thirds of programs had at least 1 faculty member on the editorial boards of 6 major dermatology journals; 38% had at least 2 faculty members and 9% had at least 5 faculty members on editorial boards. One-third of programs had no faculty members on these editorial boards. Sixty-nine percent of programs had 1 or more lectures given by full-time faculty members at annual society meetings in 2008; 48% of these programs had 1 to 5 lectures, 17% had 6 to 10, and 5% had more than 10. Thirty-one percent had no faculty members lecture at these meetings. Ninety-six percent of programs had 1 or more publications from full-time faculty members in 2008; 54% of programs had 1 to 20 publications, 24% had 21 to 40 publications, 14% had 41 to 60 publications, 5% had 61 to 80 publications, and 3% had more than 81 publications. Four percent of programs had no publications. Seventy-seven percent of programs were classified as departments and 23% were classified as divisions.

Factors Correlated With Producing Full-time Faculty

The Spearman rank correlation coefficient and P value were reported for each variable (Table 1). All coefficients were positive, signifying a positive correlation. Values closer to 1 were indicative of stronger correlations. P<.05 indicated statistically significant correlations for all factors investigated. The most strongly correlated factor was the ratio of faculty to residents in 2008, followed by number of full-time faculty, number of full-time faculty publications, number of lectures from full-time faculty, and number of faculty on editorial boards. The amount of NIH and DF funding received as well as total number of residents in 2008 also were correlated.

A 2-sample t test was performed to compare the number of graduates pursuing careers in academia from departments versus divisions, but the results were not statistically significant (P=.92).

Ranking Individual Programs With the Highest Number of Graduates Pursuing Academic Careers

The top 5 dermatology residency programs with the highest ratio of full-time faculty members graduated to the estimated total number of graduates were Brown University (Providence, Rhode Island), Loyola University (Chicago, Illinois), UMDNJ-Robert Wood Johnson Medical School at Camden (Camden, New Jersey), Harvard University (Cambridge, Massachusetts), and the University of Vermont (Burlington, Vermont)(Table 2). If faculty members spent time at more than one residency program, they were given credit for time spent at each program. For this reason, not all numbers are integers.

Ranking Individual Programs With the Highest Number of Full-time Faculty Members Who Completed Residency at the Same Institution

The top 10 institutions with the highest percentage of full-time faculty members who completed their residency at the same institution are shown in Table 3. Most of the programs with higher percentages are small programs. Fourteen programs had no faculty members who completed residency at the same program (data not shown).

Comment

Although this study focused on US dermatology residency programs, a shortage of academic dermatologists has been noted worldwide.^11-13 By determining residency program variables associated with the pursuit of academic careers, individual programs may be able to make changes that would encourage residents to pursue careers in academia following graduation.

The factor most strongly correlated with graduates pursuing a career in academics was the ratio of faculty to residents in 2008. Also highly correlated was total number of full-time faculty members. We hypothesize that programs with more faculty members may provide better mentorship for residents. A study by Reck et al¹⁰ demonstrated that a lack of mentorship was associated with residents’ loss of interest in academics. A survey of residency program directors demonstrated that mentorship played a role in career development and that it was important for residents to have mentors.¹⁴ Given the shortage of academic faculty, increasing the number of faculty may not be feasible for individual programs; however, assuming that more faculty members is a marker for mentorship, there are many ways that individual programs may improve mentorship. There are ample opportunities in clinics to demonstrate to residents the value of research to patient care.¹³ Residency programs also could establish mentorship programs, pairing residents with individual faculty members who share similar interests. Some programs currently have such mentorship programs but many do not.

Also strongly correlated with the number of graduates pursuing careers in academia was the number of full-time faculty publications. It is presumed that these programs also have published extensively in the past, which may have positively influenced the programs’ residents toward academics. This factor can be easily addressed among individual residency programs, and in fact many residency programs do encourage or require residents to publish during their residency. Exposure to the process of collecting data and writing manuscripts can bolster a resident’s interest in academics. In a prior study by Wu et al,⁵ publications were counted multiple times if multiple faculty members were authors. This limitation was addressed by only counting each manuscript once.

Other variables correlated with the number of graduates pursuing academic careers included number of lectures from full-time faculty members at annual society meetings, number of full-time faculty on editorial boards, and amount of NIH and DF funding received. All of these variables represent the importance of establishing an academic environment and promoting dermatology research. A 2009 study by Lim and Kimball,¹⁵ which also evaluated factors associated with pursuing a career in academics, demonstrated that the number of publications prior to residency and volunteerism were associated with an academic career choice. Residents with an MD/PhD were more likely to pursue a career in academics, which also was demonstrated by a similar study in 2008.¹⁶

Wu et al⁵ demonstrated similar results using data from 2001 to 2004; however, amount of NIH and DF funding received was found to be negatively correlated with graduates pursuing a career in academics in the original study. This finding was surprising but was not replicated in the current study. In the current study, the amount of funding (in dollars) rather than number of grants was analyzed because it was felt that the amount of funding received was a better reflection of the quantity of research being conducted. The prior study examined the relationship between the number of grants received and the number of graduates pursuing a career in academics. It is unclear why results varied in the 2 studies, but the most recent data are consistent with our hypothesis that increased research funding is associated with more residents pursuing careers in academia.

The current study also has some limitations. It is a retrospective observational study looking at data from 2008. Choosing data from another study period may have provided different results; however, it is reassuring that the data from the current study were very similar to a prior study looking at data from 2001 to 2004.⁵ 

It is assumed that residency program characteristics remained constant over time. It was also assumed that the number of residents in a program remained constant over time. These assumptions were necessary to draw conclusions about the data. It may be that some programs have changed substantially over time, which was not accounted for in the current study. To estimate the total number of graduates, it was assumed that faculty did not practice for more than 35 years, which may not have been true; if the faculty member practiced more than 35 years, it would have altered our estimates. It is also assumed that data on residency programs’ Web sites at time of data collection were updated and accurate. It is likely that data were not 100% accurate on all Web sites at the time of data collection. However, verification of accuracy of data would have been cumbersome, and it would have been difficult to get participation from all programs.   

This study does not differentiate the total number of new faculty members who join a residency program and those who are retained for many years. Encouraging residents to pursue a career in academics may increase the number of academic faculty.¹⁷ However, efforts must also be placed on retention of faculty members because many of the newly graduated residents who enter academics ultimately leave.^18,19

Conclusion

In conclusion, our data suggest that programs with more faculty members may encourage residents to enter careers in academia following graduation. Additionally, the number of publications increases the likelihood of residents pursuing academic careers. By providing mentorship and research opportunities to residents, perhaps residency programs can encourage their graduates to become academic dermatologists. A program’s overall academic environment, including faculty lectures at annual society meetings, faculty on editorial boards, and increased research funding are all associated with graduates pursuing careers in academics.

Ackowledgements—We thank all of the program coordinators, full-time faculty members, program directors, chairs, and chiefs who were kind enough to field our questions if we had any missing data about their programs. We would like to thank Mary H. Black, PhD (Los Angeles, California), for her contribution to the analytical plan.

APPENDIX

Dermatology has become one of the most competitive, if not the most competitive, medical specialties to enter. It attracts the brightest and most accomplished medical students who have excelled not only in the classroom and clinical setting but also in the research setting. Many successful applicants take a substantial amount of time off to pursue research and publish articles.

Despite the competitive nature of the specialty, it is well known that a marked shortage of academic dermatologists has existed for more than 30 years.^1-3 In fact, the number of graduates from US dermatology residency programs who pursue academic careers has progressively declined.⁴ Nearly all dermatology residents have a strong academic background; however, many residents opt to pursue private practice instead of a career in academia.^5-9 This trend has implications not only for future dermatology research but also for the teaching and training of future generations of dermatologists.⁸

To address this shortage, it is important to recruit dermatology residents who have a genuine interest in pursuing academic careers. Unfortunately, many residency applicants may overinflate their interest in academics to boost their chances of acceptance.⁴ Additionally, it has been shown that dermatology residents who were interested in academic careers at the time of application to the program often lost interest during residency.¹⁰

Because it can be difficult to determine a resident’s true interest in an academic career at the time of application and his/her initial interest may wean during residency, it may be more helpful to encourage dermatology residency programs to create environments that will produce residents who are more enthusiastic about and more likely to pursue careers in academia. A lack of mentorship has been shown to be associated with a loss of interest in academic careers during residency.¹⁰ If better mentorship opportunities were provided, then perhaps dermatology residents would be more likely to pursue careers in teaching and research.

A 2006 study by Wu et al⁵ demonstrated that various program characteristics were associated with the pursuit of academic careers among dermatology residents. The number of faculty members and the number of full-time faculty publications at a given residency program were most strongly correlated with the number of residents who pursued academic careers, which suggested that having a large faculty and encouraging dermatology residents to publish research during residency may motivate their pursuit of academic careers in dermatology.⁵ The current study was designed to replicate these data and respond to limitations in the original study.

Methods

Data were collected from all accredited dermatology residency programs in the United States as of December 31, 2008. The names of all full-time faculty members at these resident programs were obtained, and it was determined where each faculty member attended dermatology residency. The number of graduates who became full-time clinical or research faculty members and the number of graduates who became chairs or chiefs were counted. Residency programs excluded from these analyses included The University of Texas at Austin, University of Texas Medical Branch, and University of Connecticut, which commenced in 2008, as well as Kaiser Permanente Southern California, which commenced in 2010. Residency programs that were started after 2004 were excluded from the study, as it was thought that these programs may not have graduated a sufficient number of residents for assessment. Military residency programs also were excluded, as graduates from these programs often do not freely choose their careers after residency, and the National Institutes of Health (NIH) dermatology residency program was excluded because it is not a traditional 3-year residency program.

The primary end point was the ratio of full-time faculty members graduated to the total number of graduates from each dermatology residency program. Based on a prior study by Wu et al⁵ in 2006, it was believed that several program variables might affect pursuit of academic careers among dermatology residents, including total number of full-time faculty members, total number of residents, NIH funding received (in dollars) in 2008 (http://www.report.nih.gov/award/index.cfm), Dermatology Foundation (DF) funding received (in dollars) in 2008 (http://www.dermatology foundation.org/rap/), number of publications from full-time faculty members in 2008 (http://www.ncbi.nlm.nih.gov/pubmed/), number of full-time faculty lectures given at annual meetings of 5 societies in 2008 (American Academy of Dermatology, the Society for Investigative Dermatology, the American Society of Dermatopathology, the Society for Pediatric Dermatology, and American Society for Dermatologic Surgery), number of faculty members on the editorial boards of 6 major dermatology journals (Journal of the American Academy of Dermatology, Journal of Investigative Dermatology, Archives of Dermatology [currently known as JAMA Dermatology], Dermatologic Surgery, Pediatric Dermatology, and Journal of Cutaneous Pathology), and status as a department of dermatology or a division of internal medicine. The association between the ratio of number of full-time faculty members to number of residents for each residency program were determined for each of the outcome variables because they were believed to serve as an indicator of mentorship.

Data regarding faculty and residents were obtained from program Web sites and inquiries from individual programs. The year 1974 was used as a cutoff for the total number of graduates from each program. For faculty members who split time between 2 residency programs, each program was given credit for the duration of time spent at that program. If it was not clear how long the faculty member spent at each program, a credit of 1.5 years was given, which is half the duration of a dermatology residency. Faculty members who held a PhD only and those who completed their residencies in non-US dermatology residency programs were excluded from the outcome variables. To avoid duplicate faculty publications, collections for each residency program were created within PubMed (ie, if 2 authors from the same program coauthored an article, it was only counted once toward the total number of faculty publications from that program).

Descriptive exploratory statistical analysis in the form of a correlation matrix was completed to determine the most strongly positive and negative variables that were correlated with the ratio of graduating full-time faculty to estimated total graduates. Variables also were correlated with the secondary outcomes of ratio of graduating department chairs/chiefs to estimated total number of graduates and ratio of graduating program directors to estimated total number of graduates. Spearman rank correlation coefficients and P values were reported. Additionally, a 2-sample t test was performed to compare primary and secondary outcome variables between dermatology department versus division of dermatology under the department of internal medicine. Data were analyzed using SAS version 9.2. The institutional review board at Kaiser Permanente Southern California approved this study.

Results

Due to space considerations, analyses are based on data that are not published in this article. Data regarding the characteristics of each residency program are available from the authors.

Data from 103 dermatology residency programs were included in the analysis. Of these programs, 43% had received NIH funding in 2008 and 22% had received DF funding. Two-thirds of programs had at least 1 faculty member on the editorial boards of 6 major dermatology journals; 38% had at least 2 faculty members and 9% had at least 5 faculty members on editorial boards. One-third of programs had no faculty members on these editorial boards. Sixty-nine percent of programs had 1 or more lectures given by full-time faculty members at annual society meetings in 2008; 48% of these programs had 1 to 5 lectures, 17% had 6 to 10, and 5% had more than 10. Thirty-one percent had no faculty members lecture at these meetings. Ninety-six percent of programs had 1 or more publications from full-time faculty members in 2008; 54% of programs had 1 to 20 publications, 24% had 21 to 40 publications, 14% had 41 to 60 publications, 5% had 61 to 80 publications, and 3% had more than 81 publications. Four percent of programs had no publications. Seventy-seven percent of programs were classified as departments and 23% were classified as divisions.

Factors Correlated With Producing Full-time Faculty

The Spearman rank correlation coefficient and P value were reported for each variable (Table 1). All coefficients were positive, signifying a positive correlation. Values closer to 1 were indicative of stronger correlations. P<.05 indicated statistically significant correlations for all factors investigated. The most strongly correlated factor was the ratio of faculty to residents in 2008, followed by number of full-time faculty, number of full-time faculty publications, number of lectures from full-time faculty, and number of faculty on editorial boards. The amount of NIH and DF funding received as well as total number of residents in 2008 also were correlated.

A 2-sample t test was performed to compare the number of graduates pursuing careers in academia from departments versus divisions, but the results were not statistically significant (P=.92).

Ranking Individual Programs With the Highest Number of Graduates Pursuing Academic Careers

The top 5 dermatology residency programs with the highest ratio of full-time faculty members graduated to the estimated total number of graduates were Brown University (Providence, Rhode Island), Loyola University (Chicago, Illinois), UMDNJ-Robert Wood Johnson Medical School at Camden (Camden, New Jersey), Harvard University (Cambridge, Massachusetts), and the University of Vermont (Burlington, Vermont)(Table 2). If faculty members spent time at more than one residency program, they were given credit for time spent at each program. For this reason, not all numbers are integers.

Ranking Individual Programs With the Highest Number of Full-time Faculty Members Who Completed Residency at the Same Institution

The top 10 institutions with the highest percentage of full-time faculty members who completed their residency at the same institution are shown in Table 3. Most of the programs with higher percentages are small programs. Fourteen programs had no faculty members who completed residency at the same program (data not shown).

Comment

Although this study focused on US dermatology residency programs, a shortage of academic dermatologists has been noted worldwide.^11-13 By determining residency program variables associated with the pursuit of academic careers, individual programs may be able to make changes that would encourage residents to pursue careers in academia following graduation.

The factor most strongly correlated with graduates pursuing a career in academics was the ratio of faculty to residents in 2008. Also highly correlated was total number of full-time faculty members. We hypothesize that programs with more faculty members may provide better mentorship for residents. A study by Reck et al¹⁰ demonstrated that a lack of mentorship was associated with residents’ loss of interest in academics. A survey of residency program directors demonstrated that mentorship played a role in career development and that it was important for residents to have mentors.¹⁴ Given the shortage of academic faculty, increasing the number of faculty may not be feasible for individual programs; however, assuming that more faculty members is a marker for mentorship, there are many ways that individual programs may improve mentorship. There are ample opportunities in clinics to demonstrate to residents the value of research to patient care.¹³ Residency programs also could establish mentorship programs, pairing residents with individual faculty members who share similar interests. Some programs currently have such mentorship programs but many do not.

Also strongly correlated with the number of graduates pursuing careers in academia was the number of full-time faculty publications. It is presumed that these programs also have published extensively in the past, which may have positively influenced the programs’ residents toward academics. This factor can be easily addressed among individual residency programs, and in fact many residency programs do encourage or require residents to publish during their residency. Exposure to the process of collecting data and writing manuscripts can bolster a resident’s interest in academics. In a prior study by Wu et al,⁵ publications were counted multiple times if multiple faculty members were authors. This limitation was addressed by only counting each manuscript once.

Other variables correlated with the number of graduates pursuing academic careers included number of lectures from full-time faculty members at annual society meetings, number of full-time faculty on editorial boards, and amount of NIH and DF funding received. All of these variables represent the importance of establishing an academic environment and promoting dermatology research. A 2009 study by Lim and Kimball,¹⁵ which also evaluated factors associated with pursuing a career in academics, demonstrated that the number of publications prior to residency and volunteerism were associated with an academic career choice. Residents with an MD/PhD were more likely to pursue a career in academics, which also was demonstrated by a similar study in 2008.¹⁶

Wu et al⁵ demonstrated similar results using data from 2001 to 2004; however, amount of NIH and DF funding received was found to be negatively correlated with graduates pursuing a career in academics in the original study. This finding was surprising but was not replicated in the current study. In the current study, the amount of funding (in dollars) rather than number of grants was analyzed because it was felt that the amount of funding received was a better reflection of the quantity of research being conducted. The prior study examined the relationship between the number of grants received and the number of graduates pursuing a career in academics. It is unclear why results varied in the 2 studies, but the most recent data are consistent with our hypothesis that increased research funding is associated with more residents pursuing careers in academia.

The current study also has some limitations. It is a retrospective observational study looking at data from 2008. Choosing data from another study period may have provided different results; however, it is reassuring that the data from the current study were very similar to a prior study looking at data from 2001 to 2004.⁵ 

It is assumed that residency program characteristics remained constant over time. It was also assumed that the number of residents in a program remained constant over time. These assumptions were necessary to draw conclusions about the data. It may be that some programs have changed substantially over time, which was not accounted for in the current study. To estimate the total number of graduates, it was assumed that faculty did not practice for more than 35 years, which may not have been true; if the faculty member practiced more than 35 years, it would have altered our estimates. It is also assumed that data on residency programs’ Web sites at time of data collection were updated and accurate. It is likely that data were not 100% accurate on all Web sites at the time of data collection. However, verification of accuracy of data would have been cumbersome, and it would have been difficult to get participation from all programs.   

This study does not differentiate the total number of new faculty members who join a residency program and those who are retained for many years. Encouraging residents to pursue a career in academics may increase the number of academic faculty.¹⁷ However, efforts must also be placed on retention of faculty members because many of the newly graduated residents who enter academics ultimately leave.^18,19

Conclusion

In conclusion, our data suggest that programs with more faculty members may encourage residents to enter careers in academia following graduation. Additionally, the number of publications increases the likelihood of residents pursuing academic careers. By providing mentorship and research opportunities to residents, perhaps residency programs can encourage their graduates to become academic dermatologists. A program’s overall academic environment, including faculty lectures at annual society meetings, faculty on editorial boards, and increased research funding are all associated with graduates pursuing careers in academics.

Ackowledgements—We thank all of the program coordinators, full-time faculty members, program directors, chairs, and chiefs who were kind enough to field our questions if we had any missing data about their programs. We would like to thank Mary H. Black, PhD (Los Angeles, California), for her contribution to the analytical plan.

APPENDIX

Dermatology has become one of the most competitive, if not the most competitive, medical specialties to enter. It attracts the brightest and most accomplished medical students who have excelled not only in the classroom and clinical setting but also in the research setting. Many successful applicants take a substantial amount of time off to pursue research and publish articles.

Despite the competitive nature of the specialty, it is well known that a marked shortage of academic dermatologists has existed for more than 30 years.^1-3 In fact, the number of graduates from US dermatology residency programs who pursue academic careers has progressively declined.⁴ Nearly all dermatology residents have a strong academic background; however, many residents opt to pursue private practice instead of a career in academia.^5-9 This trend has implications not only for future dermatology research but also for the teaching and training of future generations of dermatologists.⁸

To address this shortage, it is important to recruit dermatology residents who have a genuine interest in pursuing academic careers. Unfortunately, many residency applicants may overinflate their interest in academics to boost their chances of acceptance.⁴ Additionally, it has been shown that dermatology residents who were interested in academic careers at the time of application to the program often lost interest during residency.¹⁰

Because it can be difficult to determine a resident’s true interest in an academic career at the time of application and his/her initial interest may wean during residency, it may be more helpful to encourage dermatology residency programs to create environments that will produce residents who are more enthusiastic about and more likely to pursue careers in academia. A lack of mentorship has been shown to be associated with a loss of interest in academic careers during residency.¹⁰ If better mentorship opportunities were provided, then perhaps dermatology residents would be more likely to pursue careers in teaching and research.

A 2006 study by Wu et al⁵ demonstrated that various program characteristics were associated with the pursuit of academic careers among dermatology residents. The number of faculty members and the number of full-time faculty publications at a given residency program were most strongly correlated with the number of residents who pursued academic careers, which suggested that having a large faculty and encouraging dermatology residents to publish research during residency may motivate their pursuit of academic careers in dermatology.⁵ The current study was designed to replicate these data and respond to limitations in the original study.

Methods

Data were collected from all accredited dermatology residency programs in the United States as of December 31, 2008. The names of all full-time faculty members at these resident programs were obtained, and it was determined where each faculty member attended dermatology residency. The number of graduates who became full-time clinical or research faculty members and the number of graduates who became chairs or chiefs were counted. Residency programs excluded from these analyses included The University of Texas at Austin, University of Texas Medical Branch, and University of Connecticut, which commenced in 2008, as well as Kaiser Permanente Southern California, which commenced in 2010. Residency programs that were started after 2004 were excluded from the study, as it was thought that these programs may not have graduated a sufficient number of residents for assessment. Military residency programs also were excluded, as graduates from these programs often do not freely choose their careers after residency, and the National Institutes of Health (NIH) dermatology residency program was excluded because it is not a traditional 3-year residency program.

The primary end point was the ratio of full-time faculty members graduated to the total number of graduates from each dermatology residency program. Based on a prior study by Wu et al⁵ in 2006, it was believed that several program variables might affect pursuit of academic careers among dermatology residents, including total number of full-time faculty members, total number of residents, NIH funding received (in dollars) in 2008 (http://www.report.nih.gov/award/index.cfm), Dermatology Foundation (DF) funding received (in dollars) in 2008 (http://www.dermatology foundation.org/rap/), number of publications from full-time faculty members in 2008 (http://www.ncbi.nlm.nih.gov/pubmed/), number of full-time faculty lectures given at annual meetings of 5 societies in 2008 (American Academy of Dermatology, the Society for Investigative Dermatology, the American Society of Dermatopathology, the Society for Pediatric Dermatology, and American Society for Dermatologic Surgery), number of faculty members on the editorial boards of 6 major dermatology journals (Journal of the American Academy of Dermatology, Journal of Investigative Dermatology, Archives of Dermatology [currently known as JAMA Dermatology], Dermatologic Surgery, Pediatric Dermatology, and Journal of Cutaneous Pathology), and status as a department of dermatology or a division of internal medicine. The association between the ratio of number of full-time faculty members to number of residents for each residency program were determined for each of the outcome variables because they were believed to serve as an indicator of mentorship.

Data regarding faculty and residents were obtained from program Web sites and inquiries from individual programs. The year 1974 was used as a cutoff for the total number of graduates from each program. For faculty members who split time between 2 residency programs, each program was given credit for the duration of time spent at that program. If it was not clear how long the faculty member spent at each program, a credit of 1.5 years was given, which is half the duration of a dermatology residency. Faculty members who held a PhD only and those who completed their residencies in non-US dermatology residency programs were excluded from the outcome variables. To avoid duplicate faculty publications, collections for each residency program were created within PubMed (ie, if 2 authors from the same program coauthored an article, it was only counted once toward the total number of faculty publications from that program).

Descriptive exploratory statistical analysis in the form of a correlation matrix was completed to determine the most strongly positive and negative variables that were correlated with the ratio of graduating full-time faculty to estimated total graduates. Variables also were correlated with the secondary outcomes of ratio of graduating department chairs/chiefs to estimated total number of graduates and ratio of graduating program directors to estimated total number of graduates. Spearman rank correlation coefficients and P values were reported. Additionally, a 2-sample t test was performed to compare primary and secondary outcome variables between dermatology department versus division of dermatology under the department of internal medicine. Data were analyzed using SAS version 9.2. The institutional review board at Kaiser Permanente Southern California approved this study.

Results

Due to space considerations, analyses are based on data that are not published in this article. Data regarding the characteristics of each residency program are available from the authors.

Data from 103 dermatology residency programs were included in the analysis. Of these programs, 43% had received NIH funding in 2008 and 22% had received DF funding. Two-thirds of programs had at least 1 faculty member on the editorial boards of 6 major dermatology journals; 38% had at least 2 faculty members and 9% had at least 5 faculty members on editorial boards. One-third of programs had no faculty members on these editorial boards. Sixty-nine percent of programs had 1 or more lectures given by full-time faculty members at annual society meetings in 2008; 48% of these programs had 1 to 5 lectures, 17% had 6 to 10, and 5% had more than 10. Thirty-one percent had no faculty members lecture at these meetings. Ninety-six percent of programs had 1 or more publications from full-time faculty members in 2008; 54% of programs had 1 to 20 publications, 24% had 21 to 40 publications, 14% had 41 to 60 publications, 5% had 61 to 80 publications, and 3% had more than 81 publications. Four percent of programs had no publications. Seventy-seven percent of programs were classified as departments and 23% were classified as divisions.

Factors Correlated With Producing Full-time Faculty

The Spearman rank correlation coefficient and P value were reported for each variable (Table 1). All coefficients were positive, signifying a positive correlation. Values closer to 1 were indicative of stronger correlations. P<.05 indicated statistically significant correlations for all factors investigated. The most strongly correlated factor was the ratio of faculty to residents in 2008, followed by number of full-time faculty, number of full-time faculty publications, number of lectures from full-time faculty, and number of faculty on editorial boards. The amount of NIH and DF funding received as well as total number of residents in 2008 also were correlated.

A 2-sample t test was performed to compare the number of graduates pursuing careers in academia from departments versus divisions, but the results were not statistically significant (P=.92).

Ranking Individual Programs With the Highest Number of Graduates Pursuing Academic Careers

The top 5 dermatology residency programs with the highest ratio of full-time faculty members graduated to the estimated total number of graduates were Brown University (Providence, Rhode Island), Loyola University (Chicago, Illinois), UMDNJ-Robert Wood Johnson Medical School at Camden (Camden, New Jersey), Harvard University (Cambridge, Massachusetts), and the University of Vermont (Burlington, Vermont)(Table 2). If faculty members spent time at more than one residency program, they were given credit for time spent at each program. For this reason, not all numbers are integers.

Ranking Individual Programs With the Highest Number of Full-time Faculty Members Who Completed Residency at the Same Institution

The top 10 institutions with the highest percentage of full-time faculty members who completed their residency at the same institution are shown in Table 3. Most of the programs with higher percentages are small programs. Fourteen programs had no faculty members who completed residency at the same program (data not shown).

Comment

Although this study focused on US dermatology residency programs, a shortage of academic dermatologists has been noted worldwide.^11-13 By determining residency program variables associated with the pursuit of academic careers, individual programs may be able to make changes that would encourage residents to pursue careers in academia following graduation.

The factor most strongly correlated with graduates pursuing a career in academics was the ratio of faculty to residents in 2008. Also highly correlated was total number of full-time faculty members. We hypothesize that programs with more faculty members may provide better mentorship for residents. A study by Reck et al¹⁰ demonstrated that a lack of mentorship was associated with residents’ loss of interest in academics. A survey of residency program directors demonstrated that mentorship played a role in career development and that it was important for residents to have mentors.¹⁴ Given the shortage of academic faculty, increasing the number of faculty may not be feasible for individual programs; however, assuming that more faculty members is a marker for mentorship, there are many ways that individual programs may improve mentorship. There are ample opportunities in clinics to demonstrate to residents the value of research to patient care.¹³ Residency programs also could establish mentorship programs, pairing residents with individual faculty members who share similar interests. Some programs currently have such mentorship programs but many do not.

Also strongly correlated with the number of graduates pursuing careers in academia was the number of full-time faculty publications. It is presumed that these programs also have published extensively in the past, which may have positively influenced the programs’ residents toward academics. This factor can be easily addressed among individual residency programs, and in fact many residency programs do encourage or require residents to publish during their residency. Exposure to the process of collecting data and writing manuscripts can bolster a resident’s interest in academics. In a prior study by Wu et al,⁵ publications were counted multiple times if multiple faculty members were authors. This limitation was addressed by only counting each manuscript once.

Other variables correlated with the number of graduates pursuing academic careers included number of lectures from full-time faculty members at annual society meetings, number of full-time faculty on editorial boards, and amount of NIH and DF funding received. All of these variables represent the importance of establishing an academic environment and promoting dermatology research. A 2009 study by Lim and Kimball,¹⁵ which also evaluated factors associated with pursuing a career in academics, demonstrated that the number of publications prior to residency and volunteerism were associated with an academic career choice. Residents with an MD/PhD were more likely to pursue a career in academics, which also was demonstrated by a similar study in 2008.¹⁶

Wu et al⁵ demonstrated similar results using data from 2001 to 2004; however, amount of NIH and DF funding received was found to be negatively correlated with graduates pursuing a career in academics in the original study. This finding was surprising but was not replicated in the current study. In the current study, the amount of funding (in dollars) rather than number of grants was analyzed because it was felt that the amount of funding received was a better reflection of the quantity of research being conducted. The prior study examined the relationship between the number of grants received and the number of graduates pursuing a career in academics. It is unclear why results varied in the 2 studies, but the most recent data are consistent with our hypothesis that increased research funding is associated with more residents pursuing careers in academia.

The current study also has some limitations. It is a retrospective observational study looking at data from 2008. Choosing data from another study period may have provided different results; however, it is reassuring that the data from the current study were very similar to a prior study looking at data from 2001 to 2004.⁵ 

It is assumed that residency program characteristics remained constant over time. It was also assumed that the number of residents in a program remained constant over time. These assumptions were necessary to draw conclusions about the data. It may be that some programs have changed substantially over time, which was not accounted for in the current study. To estimate the total number of graduates, it was assumed that faculty did not practice for more than 35 years, which may not have been true; if the faculty member practiced more than 35 years, it would have altered our estimates. It is also assumed that data on residency programs’ Web sites at time of data collection were updated and accurate. It is likely that data were not 100% accurate on all Web sites at the time of data collection. However, verification of accuracy of data would have been cumbersome, and it would have been difficult to get participation from all programs.   

This study does not differentiate the total number of new faculty members who join a residency program and those who are retained for many years. Encouraging residents to pursue a career in academics may increase the number of academic faculty.¹⁷ However, efforts must also be placed on retention of faculty members because many of the newly graduated residents who enter academics ultimately leave.^18,19

Conclusion

In conclusion, our data suggest that programs with more faculty members may encourage residents to enter careers in academia following graduation. Additionally, the number of publications increases the likelihood of residents pursuing academic careers. By providing mentorship and research opportunities to residents, perhaps residency programs can encourage their graduates to become academic dermatologists. A program’s overall academic environment, including faculty lectures at annual society meetings, faculty on editorial boards, and increased research funding are all associated with graduates pursuing careers in academics.

Ackowledgements—We thank all of the program coordinators, full-time faculty members, program directors, chairs, and chiefs who were kind enough to field our questions if we had any missing data about their programs. We would like to thank Mary H. Black, PhD (Los Angeles, California), for her contribution to the analytical plan.

APPENDIX

Leadership is widely recognized as a key component in the role of a physician,¹ which is especially true in dermatology, a specialty that faces severe challenges in the recruitment and retention of academic faculty.² A study of the dermatology workforce found that academic institutions are more likely to be seeking to hire new faculty² and that many dermatology residency programs often are looking to replace chairpersons (chairs) and/or chiefs.³ Although fewer dermatology residents are pursuing academic careers than careers in private practice, full-time faculty members also are leaving their academic posts. This shift is demonstrated by the younger mean age of academic dermatologists² and the increased rate of departure from academia prior to pursuing more formalized leadership roles.⁴

It has been suggested that the number of full-time faculty and number of faculty publications positively influence graduates of dermatology residency programs to pursue academic careers; however, variables affecting the likelihood of graduates of dermatology residency programs becoming academic leaders later in their career have not been well studied.³ The purpose of this study is to determine the factors that influence the development of program chairs/chiefs and program directors (PDs) of dermatology residency programs.

Methods

Data were collected from all accredited dermatology residency programs in the United States as of December 31, 2008. Residency programs that were started after 2004 were excluded from the study, as it was thought that these programs may not have graduated a sufficient number of residents for assessment. Military residency programs also were excluded, as graduates from these programs often do not freely choose their careers after residency.

Primary end points were the number of chairs/chiefs and PDs who had graduated from each dermatology residency program. Variables included the number of years the program had been in existence, status of the program as a department or division, number of full-time faculty members, number of residents, National Institutes of Health funding received in 2008 (http://www.report.nih.gov/award/index.cfm), Dermatology Foundation (DF) funding received (http://www.dermatologyfoundation.org/rap/), number of publications from full-time faculty members in 2008 (http://www.ncbi.nlm.nih.gov/pubmed/), number of faculty lectures given at annual meetings of 5 societies in 2008 (American Academy of Dermatology, the Society for Investigative Dermatology, the American Society of Dermatopathology, the Society for Pediatric Dermatology, and American Society for Dermatologic Surgery), and the number of faculty members on the editorial boards of 6 major dermatology journals (Journal of the American Academy of Dermatology, Journal of Investigative Dermatology, Archives of Dermatology [currently known as JAMA Dermatology], Dermatologic Surgery, Pediatric Dermatology, and Journal of Cutaneous Pathology). Data regarding faculty and residents were obtained from program Web sites and inquiries from individual programs. The year 1974 was used as a cutoff for the number of years a program had been in existence. Years of existence of a program was controlled for in the analysis. The ratio of faculty to residents was calculated per year and categorized as 4 or more or less than 4 to minimize the effect of changing program size over the years. For faculty members who split time between 2 residency programs, each program was given credit for the duration of time spent at that program. Faculty members who hold a PhD only and those who completed their residencies in non-US dermatology residency programs were excluded from the outcome variables. To avoid duplicate faculty publications, collections for each residency program were created within PubMed (ie, if 2 authors from the same program coauthored an article, it was only counted once toward the total number of faculty publications from that program).

Because the data were skewed (ie, there were a large number of programs with 0 graduating chairs/chiefs and PDs), nonparametric analyses were utilized. Logistic regression was used to calculate the odds of producing chairs/chiefs or PDs (yes vs no). Multiple logistic regression helped to determine those variables that were most closely associated with odds of graduating a chair/chief or PD. Variables with a significance level of P<.10 were considered in the multiple logistic regression, and backward selection was used to determine a model. Multiple linear regression was used to determine correlation coefficients for each of the variables and the number of chairs/chiefs or PDs graduated, controlling for the estimated number of graduates from the program and number of years the program had been in existence. Analyses for graduating chairs/chiefs and PDs were conducted separately. The final significance level used was P<.05. Data were analyzed using SAS version 9.3. This study was approved by the institutional review board at Kaiser Permanente Southern California.

Results

Data from 103 dermatology residency programs were included in the analysis. Of these programs, 47 had graduated at least 1 chair/chief and 55 had graduated at least 1 PD. Among the programs graduating any chairs/chiefs, they produced an average of 2.04 chairs/chiefs and 1.86 PDs. The 5 dermatology residency programs that graduated the highest total number of chairs/chiefs and PDs were Harvard University (Cambridge, Massachusetts), the University of Michigan (Ann Arbor, Michigan), New York University (New York, New York), Yale-New Haven Hospital (New Haven, Connecticut), and the University of Minnesota (Minneapolis, Minnesota).

Factors that had the highest effect on the odds of a program graduating a chair/chief included the ratio of faculty to residents per year, presence of DF funding in 2008, number of years program was in existence, number of residents, number of full-time faculty, and number of full-time faculty on editorial boards of 6 major dermatology journals (Table 1). When controlling for each of these variables in the final multivariable analysis, programs with 4 or more faculty per resident had 3.31 times the odds of producing a chair/chief (95% confidence interval [CI], 1.14-9.66; P=.028).

Factors that had the highest effect on the odds of a program graduating a PD included status as department versus division, ratio of faculty to residents per year, presence of DF funding in 2008, number of lectures given by full-time faculty members at annual society meetings, number of residents, number of years program was in existence, number of full-time faculty, and number of publications from full-time faculty members (Table 2). The most significant factor associated with graduating PDs after controlling for other variables was the number of publications from full-time faculty members. The odds increased by 3.2% for every 1 additional publication and 32% for every 10 additional publications (95% CI, 1.01-1.06; P=.026).

Multiple linear regression demonstrated a positive relationship between the number of graduating chairs/chiefs and total full-time faculty members (R²=0.26; P=.034) and ratio of full-time faculty to residents (R²=0.29; P<.001). Marginally significant correlations were seen between the number of PDs and ratio of full-time faculty to residents (R²=0.32; P=.05) as well as the number of publications from full-time faculty members (R²=0.32; P=.05).

Comment

The ratio of full-time faculty to residents increased a program’s odds of graduating a chair/chief. More faculty members may lead to more opportunities for mentorship of residents and young faculty. Mentors are widely perceived to be integral to the learning and development of residents, not only in dermatology⁵ but across all specialties.⁶ Mentors also have been noted to play a key role in bolstering and maintaining interest in academics,⁷ which is true not only with regard to recruiting new residents but for retaining young faculty members. In a study (N=109) that examined factors associated with residents’ loss of interest in academic careers, half of the participants reported a lack of effective mentors, role models, and professional guidance.⁸ Mentors provide teaching, supervision, and advice, especially with regard to research and career paths.⁹ A large number of faculty members provides more opportunities for direct mentorship and offers residents more exposure to research, specialty clinics, and academic philosophies, which may positively influence and even inspire academic pursuits and leadership.³

Although the solution to producing future chairs/chiefs and PDs may lie in faculty guidance, finding and retaining faculty members as mentors amidst a shortage of academic dermatologists presents an underlying issue.³ In addition to a lack of mentorship, residents cite bureaucracy, salary differentials, and location to explain a loss of interest in academic careers.⁸ Several programs have been developed to address the recruitment of dermatology residents for academic careers, including combined medical-dermatology programs, 2+2 programs (2 years of clinical residency plus 2 additional research years), clinical research fellowships,¹⁰ and the Society for Investigative Dermatology’s Dermatology Resident Retreat for Future Academicians (http://www.sidnet.org/fortraineesandresidents).¹¹ Perhaps recruitment should even start at the medical student level. In light of the academic strength of the current pool of dermatology residency applicants,¹² training programs should continue to screen for applicants with sincere interests in academia.¹³ Students with more research and publications may be more likely to pursue academic careers, in accordance with prior studies of dermatology trainees.^3,14 Studies also have shown that graduates of foreign dermatology residencies¹⁵ and individuals who hold both MD and PhD degrees may be more likely to enter into academic careers.¹⁶

For creating future chairs/chiefs and PDs, retention of young faculty in academics is as important as recruiting residents.¹⁷ At the mid-career level, the decline of funds for research has generated pressure for academic physicians to see increasing numbers of patients, leaving insufficient time for the many duties that accompany academic posts,² including teaching and publishing. Other reasons that faculty members leave their posts before 40 years of age include financial and family concerns¹⁸ as well as the desire for more autonomy.⁴ Formalized training is seen with the American Academy of Dermatology’s Academic Dermatology Leadership Program (https://www.aad.org/members/leadership-institute/mentoring/aad-mentoring-opportunities/academic-dermatology-leadership-program-mentee), which promotes advanced leadership training to recent graduates.⁵ Other methods include support of young faculty with mentorship; grant applications; and administration at the department, hospital, and government levels.¹⁷ Recruitment of faculty from private practice may represent another potential source of faculty who wish to pursue more scholarly endeavors.⁴ Teaching has been cited as a primary reason for faculty members to remain in academia,¹⁸ and thus time for teaching must be protected. Such a strategy is in accordance with our findings that amount of annual DF funding received, number of full-time faculty publications, number of faculty members on editorial boards of major dermatology journals, and number of lectures given by full-time faculty factors at annual society meetings are positively associated with the odds of producing chairs/chiefs or PDs. In particular, the number of full-time faculty publications is directly related to increased odds of graduates becoming PDs. Residents and young faculty members who take part in research and attend national conferences may find inspiration or develop a passion for academic leadership.

A limitation to this study is that the ratio of faculty to graduated residents for some programs likely has fluctuated over the last 35 years. This study assumed that certain programs remained generally small or large during the course of their existence, which was controlled by using the ratio between faculty and residents. Additionally, the number of years that a program has been in existence influences the likelihood of that program to graduate higher numbers of residents who become chairs/chiefs or PDs. As a result, we used multiple linear regression to control for the number of residents and number of years that a program had been in existence. Finally, while the relationship between academic leaders and research may be explained by the increased likelihood of faculty with more funding, publications, or lectures to be selected for leadership roles, this finding supports the notion that research can contribute to leadership. This analysis identifies modifiable factors among residency training programs to improve the odds of graduating future academic leaders.

Conclusion

As the present study shows, the ratio of faculty to residents and the number of full-time faculty publications are key to graduating academic leaders in dermatology. Retention of faculty as leaders in academic dermatology is as crucial to the field as recruitment of residents into academic dermatology. Mentorship should be highly encouraged through the creation of formal programs and should not end at the resident level. Emphasizing the intellectually stimulating aspects of academia and providing administrative resources may help decrease the burden of academic duties, allowing the pursuit of teaching and research and ultimately the resources to become candidates for leadership positions in academia.

Leadership is widely recognized as a key component in the role of a physician,¹ which is especially true in dermatology, a specialty that faces severe challenges in the recruitment and retention of academic faculty.² A study of the dermatology workforce found that academic institutions are more likely to be seeking to hire new faculty² and that many dermatology residency programs often are looking to replace chairpersons (chairs) and/or chiefs.³ Although fewer dermatology residents are pursuing academic careers than careers in private practice, full-time faculty members also are leaving their academic posts. This shift is demonstrated by the younger mean age of academic dermatologists² and the increased rate of departure from academia prior to pursuing more formalized leadership roles.⁴

It has been suggested that the number of full-time faculty and number of faculty publications positively influence graduates of dermatology residency programs to pursue academic careers; however, variables affecting the likelihood of graduates of dermatology residency programs becoming academic leaders later in their career have not been well studied.³ The purpose of this study is to determine the factors that influence the development of program chairs/chiefs and program directors (PDs) of dermatology residency programs.

Methods

Data were collected from all accredited dermatology residency programs in the United States as of December 31, 2008. Residency programs that were started after 2004 were excluded from the study, as it was thought that these programs may not have graduated a sufficient number of residents for assessment. Military residency programs also were excluded, as graduates from these programs often do not freely choose their careers after residency.

Primary end points were the number of chairs/chiefs and PDs who had graduated from each dermatology residency program. Variables included the number of years the program had been in existence, status of the program as a department or division, number of full-time faculty members, number of residents, National Institutes of Health funding received in 2008 (http://www.report.nih.gov/award/index.cfm), Dermatology Foundation (DF) funding received (http://www.dermatologyfoundation.org/rap/), number of publications from full-time faculty members in 2008 (http://www.ncbi.nlm.nih.gov/pubmed/), number of faculty lectures given at annual meetings of 5 societies in 2008 (American Academy of Dermatology, the Society for Investigative Dermatology, the American Society of Dermatopathology, the Society for Pediatric Dermatology, and American Society for Dermatologic Surgery), and the number of faculty members on the editorial boards of 6 major dermatology journals (Journal of the American Academy of Dermatology, Journal of Investigative Dermatology, Archives of Dermatology [currently known as JAMA Dermatology], Dermatologic Surgery, Pediatric Dermatology, and Journal of Cutaneous Pathology). Data regarding faculty and residents were obtained from program Web sites and inquiries from individual programs. The year 1974 was used as a cutoff for the number of years a program had been in existence. Years of existence of a program was controlled for in the analysis. The ratio of faculty to residents was calculated per year and categorized as 4 or more or less than 4 to minimize the effect of changing program size over the years. For faculty members who split time between 2 residency programs, each program was given credit for the duration of time spent at that program. Faculty members who hold a PhD only and those who completed their residencies in non-US dermatology residency programs were excluded from the outcome variables. To avoid duplicate faculty publications, collections for each residency program were created within PubMed (ie, if 2 authors from the same program coauthored an article, it was only counted once toward the total number of faculty publications from that program).

Because the data were skewed (ie, there were a large number of programs with 0 graduating chairs/chiefs and PDs), nonparametric analyses were utilized. Logistic regression was used to calculate the odds of producing chairs/chiefs or PDs (yes vs no). Multiple logistic regression helped to determine those variables that were most closely associated with odds of graduating a chair/chief or PD. Variables with a significance level of P<.10 were considered in the multiple logistic regression, and backward selection was used to determine a model. Multiple linear regression was used to determine correlation coefficients for each of the variables and the number of chairs/chiefs or PDs graduated, controlling for the estimated number of graduates from the program and number of years the program had been in existence. Analyses for graduating chairs/chiefs and PDs were conducted separately. The final significance level used was P<.05. Data were analyzed using SAS version 9.3. This study was approved by the institutional review board at Kaiser Permanente Southern California.

Results

Data from 103 dermatology residency programs were included in the analysis. Of these programs, 47 had graduated at least 1 chair/chief and 55 had graduated at least 1 PD. Among the programs graduating any chairs/chiefs, they produced an average of 2.04 chairs/chiefs and 1.86 PDs. The 5 dermatology residency programs that graduated the highest total number of chairs/chiefs and PDs were Harvard University (Cambridge, Massachusetts), the University of Michigan (Ann Arbor, Michigan), New York University (New York, New York), Yale-New Haven Hospital (New Haven, Connecticut), and the University of Minnesota (Minneapolis, Minnesota).

Factors that had the highest effect on the odds of a program graduating a chair/chief included the ratio of faculty to residents per year, presence of DF funding in 2008, number of years program was in existence, number of residents, number of full-time faculty, and number of full-time faculty on editorial boards of 6 major dermatology journals (Table 1). When controlling for each of these variables in the final multivariable analysis, programs with 4 or more faculty per resident had 3.31 times the odds of producing a chair/chief (95% confidence interval [CI], 1.14-9.66; P=.028).

Factors that had the highest effect on the odds of a program graduating a PD included status as department versus division, ratio of faculty to residents per year, presence of DF funding in 2008, number of lectures given by full-time faculty members at annual society meetings, number of residents, number of years program was in existence, number of full-time faculty, and number of publications from full-time faculty members (Table 2). The most significant factor associated with graduating PDs after controlling for other variables was the number of publications from full-time faculty members. The odds increased by 3.2% for every 1 additional publication and 32% for every 10 additional publications (95% CI, 1.01-1.06; P=.026).

Multiple linear regression demonstrated a positive relationship between the number of graduating chairs/chiefs and total full-time faculty members (R²=0.26; P=.034) and ratio of full-time faculty to residents (R²=0.29; P<.001). Marginally significant correlations were seen between the number of PDs and ratio of full-time faculty to residents (R²=0.32; P=.05) as well as the number of publications from full-time faculty members (R²=0.32; P=.05).

Comment

The ratio of full-time faculty to residents increased a program’s odds of graduating a chair/chief. More faculty members may lead to more opportunities for mentorship of residents and young faculty. Mentors are widely perceived to be integral to the learning and development of residents, not only in dermatology⁵ but across all specialties.⁶ Mentors also have been noted to play a key role in bolstering and maintaining interest in academics,⁷ which is true not only with regard to recruiting new residents but for retaining young faculty members. In a study (N=109) that examined factors associated with residents’ loss of interest in academic careers, half of the participants reported a lack of effective mentors, role models, and professional guidance.⁸ Mentors provide teaching, supervision, and advice, especially with regard to research and career paths.⁹ A large number of faculty members provides more opportunities for direct mentorship and offers residents more exposure to research, specialty clinics, and academic philosophies, which may positively influence and even inspire academic pursuits and leadership.³

Although the solution to producing future chairs/chiefs and PDs may lie in faculty guidance, finding and retaining faculty members as mentors amidst a shortage of academic dermatologists presents an underlying issue.³ In addition to a lack of mentorship, residents cite bureaucracy, salary differentials, and location to explain a loss of interest in academic careers.⁸ Several programs have been developed to address the recruitment of dermatology residents for academic careers, including combined medical-dermatology programs, 2+2 programs (2 years of clinical residency plus 2 additional research years), clinical research fellowships,¹⁰ and the Society for Investigative Dermatology’s Dermatology Resident Retreat for Future Academicians (http://www.sidnet.org/fortraineesandresidents).¹¹ Perhaps recruitment should even start at the medical student level. In light of the academic strength of the current pool of dermatology residency applicants,¹² training programs should continue to screen for applicants with sincere interests in academia.¹³ Students with more research and publications may be more likely to pursue academic careers, in accordance with prior studies of dermatology trainees.^3,14 Studies also have shown that graduates of foreign dermatology residencies¹⁵ and individuals who hold both MD and PhD degrees may be more likely to enter into academic careers.¹⁶

For creating future chairs/chiefs and PDs, retention of young faculty in academics is as important as recruiting residents.¹⁷ At the mid-career level, the decline of funds for research has generated pressure for academic physicians to see increasing numbers of patients, leaving insufficient time for the many duties that accompany academic posts,² including teaching and publishing. Other reasons that faculty members leave their posts before 40 years of age include financial and family concerns¹⁸ as well as the desire for more autonomy.⁴ Formalized training is seen with the American Academy of Dermatology’s Academic Dermatology Leadership Program (https://www.aad.org/members/leadership-institute/mentoring/aad-mentoring-opportunities/academic-dermatology-leadership-program-mentee), which promotes advanced leadership training to recent graduates.⁵ Other methods include support of young faculty with mentorship; grant applications; and administration at the department, hospital, and government levels.¹⁷ Recruitment of faculty from private practice may represent another potential source of faculty who wish to pursue more scholarly endeavors.⁴ Teaching has been cited as a primary reason for faculty members to remain in academia,¹⁸ and thus time for teaching must be protected. Such a strategy is in accordance with our findings that amount of annual DF funding received, number of full-time faculty publications, number of faculty members on editorial boards of major dermatology journals, and number of lectures given by full-time faculty factors at annual society meetings are positively associated with the odds of producing chairs/chiefs or PDs. In particular, the number of full-time faculty publications is directly related to increased odds of graduates becoming PDs. Residents and young faculty members who take part in research and attend national conferences may find inspiration or develop a passion for academic leadership.

A limitation to this study is that the ratio of faculty to graduated residents for some programs likely has fluctuated over the last 35 years. This study assumed that certain programs remained generally small or large during the course of their existence, which was controlled by using the ratio between faculty and residents. Additionally, the number of years that a program has been in existence influences the likelihood of that program to graduate higher numbers of residents who become chairs/chiefs or PDs. As a result, we used multiple linear regression to control for the number of residents and number of years that a program had been in existence. Finally, while the relationship between academic leaders and research may be explained by the increased likelihood of faculty with more funding, publications, or lectures to be selected for leadership roles, this finding supports the notion that research can contribute to leadership. This analysis identifies modifiable factors among residency training programs to improve the odds of graduating future academic leaders.

Conclusion

As the present study shows, the ratio of faculty to residents and the number of full-time faculty publications are key to graduating academic leaders in dermatology. Retention of faculty as leaders in academic dermatology is as crucial to the field as recruitment of residents into academic dermatology. Mentorship should be highly encouraged through the creation of formal programs and should not end at the resident level. Emphasizing the intellectually stimulating aspects of academia and providing administrative resources may help decrease the burden of academic duties, allowing the pursuit of teaching and research and ultimately the resources to become candidates for leadership positions in academia.

Leadership is widely recognized as a key component in the role of a physician,¹ which is especially true in dermatology, a specialty that faces severe challenges in the recruitment and retention of academic faculty.² A study of the dermatology workforce found that academic institutions are more likely to be seeking to hire new faculty² and that many dermatology residency programs often are looking to replace chairpersons (chairs) and/or chiefs.³ Although fewer dermatology residents are pursuing academic careers than careers in private practice, full-time faculty members also are leaving their academic posts. This shift is demonstrated by the younger mean age of academic dermatologists² and the increased rate of departure from academia prior to pursuing more formalized leadership roles.⁴

It has been suggested that the number of full-time faculty and number of faculty publications positively influence graduates of dermatology residency programs to pursue academic careers; however, variables affecting the likelihood of graduates of dermatology residency programs becoming academic leaders later in their career have not been well studied.³ The purpose of this study is to determine the factors that influence the development of program chairs/chiefs and program directors (PDs) of dermatology residency programs.

Methods

Data were collected from all accredited dermatology residency programs in the United States as of December 31, 2008. Residency programs that were started after 2004 were excluded from the study, as it was thought that these programs may not have graduated a sufficient number of residents for assessment. Military residency programs also were excluded, as graduates from these programs often do not freely choose their careers after residency.

Primary end points were the number of chairs/chiefs and PDs who had graduated from each dermatology residency program. Variables included the number of years the program had been in existence, status of the program as a department or division, number of full-time faculty members, number of residents, National Institutes of Health funding received in 2008 (http://www.report.nih.gov/award/index.cfm), Dermatology Foundation (DF) funding received (http://www.dermatologyfoundation.org/rap/), number of publications from full-time faculty members in 2008 (http://www.ncbi.nlm.nih.gov/pubmed/), number of faculty lectures given at annual meetings of 5 societies in 2008 (American Academy of Dermatology, the Society for Investigative Dermatology, the American Society of Dermatopathology, the Society for Pediatric Dermatology, and American Society for Dermatologic Surgery), and the number of faculty members on the editorial boards of 6 major dermatology journals (Journal of the American Academy of Dermatology, Journal of Investigative Dermatology, Archives of Dermatology [currently known as JAMA Dermatology], Dermatologic Surgery, Pediatric Dermatology, and Journal of Cutaneous Pathology). Data regarding faculty and residents were obtained from program Web sites and inquiries from individual programs. The year 1974 was used as a cutoff for the number of years a program had been in existence. Years of existence of a program was controlled for in the analysis. The ratio of faculty to residents was calculated per year and categorized as 4 or more or less than 4 to minimize the effect of changing program size over the years. For faculty members who split time between 2 residency programs, each program was given credit for the duration of time spent at that program. Faculty members who hold a PhD only and those who completed their residencies in non-US dermatology residency programs were excluded from the outcome variables. To avoid duplicate faculty publications, collections for each residency program were created within PubMed (ie, if 2 authors from the same program coauthored an article, it was only counted once toward the total number of faculty publications from that program).

Because the data were skewed (ie, there were a large number of programs with 0 graduating chairs/chiefs and PDs), nonparametric analyses were utilized. Logistic regression was used to calculate the odds of producing chairs/chiefs or PDs (yes vs no). Multiple logistic regression helped to determine those variables that were most closely associated with odds of graduating a chair/chief or PD. Variables with a significance level of P<.10 were considered in the multiple logistic regression, and backward selection was used to determine a model. Multiple linear regression was used to determine correlation coefficients for each of the variables and the number of chairs/chiefs or PDs graduated, controlling for the estimated number of graduates from the program and number of years the program had been in existence. Analyses for graduating chairs/chiefs and PDs were conducted separately. The final significance level used was P<.05. Data were analyzed using SAS version 9.3. This study was approved by the institutional review board at Kaiser Permanente Southern California.

Results

Data from 103 dermatology residency programs were included in the analysis. Of these programs, 47 had graduated at least 1 chair/chief and 55 had graduated at least 1 PD. Among the programs graduating any chairs/chiefs, they produced an average of 2.04 chairs/chiefs and 1.86 PDs. The 5 dermatology residency programs that graduated the highest total number of chairs/chiefs and PDs were Harvard University (Cambridge, Massachusetts), the University of Michigan (Ann Arbor, Michigan), New York University (New York, New York), Yale-New Haven Hospital (New Haven, Connecticut), and the University of Minnesota (Minneapolis, Minnesota).

Factors that had the highest effect on the odds of a program graduating a chair/chief included the ratio of faculty to residents per year, presence of DF funding in 2008, number of years program was in existence, number of residents, number of full-time faculty, and number of full-time faculty on editorial boards of 6 major dermatology journals (Table 1). When controlling for each of these variables in the final multivariable analysis, programs with 4 or more faculty per resident had 3.31 times the odds of producing a chair/chief (95% confidence interval [CI], 1.14-9.66; P=.028).

Factors that had the highest effect on the odds of a program graduating a PD included status as department versus division, ratio of faculty to residents per year, presence of DF funding in 2008, number of lectures given by full-time faculty members at annual society meetings, number of residents, number of years program was in existence, number of full-time faculty, and number of publications from full-time faculty members (Table 2). The most significant factor associated with graduating PDs after controlling for other variables was the number of publications from full-time faculty members. The odds increased by 3.2% for every 1 additional publication and 32% for every 10 additional publications (95% CI, 1.01-1.06; P=.026).

Multiple linear regression demonstrated a positive relationship between the number of graduating chairs/chiefs and total full-time faculty members (R²=0.26; P=.034) and ratio of full-time faculty to residents (R²=0.29; P<.001). Marginally significant correlations were seen between the number of PDs and ratio of full-time faculty to residents (R²=0.32; P=.05) as well as the number of publications from full-time faculty members (R²=0.32; P=.05).

Comment

The ratio of full-time faculty to residents increased a program’s odds of graduating a chair/chief. More faculty members may lead to more opportunities for mentorship of residents and young faculty. Mentors are widely perceived to be integral to the learning and development of residents, not only in dermatology⁵ but across all specialties.⁶ Mentors also have been noted to play a key role in bolstering and maintaining interest in academics,⁷ which is true not only with regard to recruiting new residents but for retaining young faculty members. In a study (N=109) that examined factors associated with residents’ loss of interest in academic careers, half of the participants reported a lack of effective mentors, role models, and professional guidance.⁸ Mentors provide teaching, supervision, and advice, especially with regard to research and career paths.⁹ A large number of faculty members provides more opportunities for direct mentorship and offers residents more exposure to research, specialty clinics, and academic philosophies, which may positively influence and even inspire academic pursuits and leadership.³

Although the solution to producing future chairs/chiefs and PDs may lie in faculty guidance, finding and retaining faculty members as mentors amidst a shortage of academic dermatologists presents an underlying issue.³ In addition to a lack of mentorship, residents cite bureaucracy, salary differentials, and location to explain a loss of interest in academic careers.⁸ Several programs have been developed to address the recruitment of dermatology residents for academic careers, including combined medical-dermatology programs, 2+2 programs (2 years of clinical residency plus 2 additional research years), clinical research fellowships,¹⁰ and the Society for Investigative Dermatology’s Dermatology Resident Retreat for Future Academicians (http://www.sidnet.org/fortraineesandresidents).¹¹ Perhaps recruitment should even start at the medical student level. In light of the academic strength of the current pool of dermatology residency applicants,¹² training programs should continue to screen for applicants with sincere interests in academia.¹³ Students with more research and publications may be more likely to pursue academic careers, in accordance with prior studies of dermatology trainees.^3,14 Studies also have shown that graduates of foreign dermatology residencies¹⁵ and individuals who hold both MD and PhD degrees may be more likely to enter into academic careers.¹⁶

For creating future chairs/chiefs and PDs, retention of young faculty in academics is as important as recruiting residents.¹⁷ At the mid-career level, the decline of funds for research has generated pressure for academic physicians to see increasing numbers of patients, leaving insufficient time for the many duties that accompany academic posts,² including teaching and publishing. Other reasons that faculty members leave their posts before 40 years of age include financial and family concerns¹⁸ as well as the desire for more autonomy.⁴ Formalized training is seen with the American Academy of Dermatology’s Academic Dermatology Leadership Program (https://www.aad.org/members/leadership-institute/mentoring/aad-mentoring-opportunities/academic-dermatology-leadership-program-mentee), which promotes advanced leadership training to recent graduates.⁵ Other methods include support of young faculty with mentorship; grant applications; and administration at the department, hospital, and government levels.¹⁷ Recruitment of faculty from private practice may represent another potential source of faculty who wish to pursue more scholarly endeavors.⁴ Teaching has been cited as a primary reason for faculty members to remain in academia,¹⁸ and thus time for teaching must be protected. Such a strategy is in accordance with our findings that amount of annual DF funding received, number of full-time faculty publications, number of faculty members on editorial boards of major dermatology journals, and number of lectures given by full-time faculty factors at annual society meetings are positively associated with the odds of producing chairs/chiefs or PDs. In particular, the number of full-time faculty publications is directly related to increased odds of graduates becoming PDs. Residents and young faculty members who take part in research and attend national conferences may find inspiration or develop a passion for academic leadership.

A limitation to this study is that the ratio of faculty to graduated residents for some programs likely has fluctuated over the last 35 years. This study assumed that certain programs remained generally small or large during the course of their existence, which was controlled by using the ratio between faculty and residents. Additionally, the number of years that a program has been in existence influences the likelihood of that program to graduate higher numbers of residents who become chairs/chiefs or PDs. As a result, we used multiple linear regression to control for the number of residents and number of years that a program had been in existence. Finally, while the relationship between academic leaders and research may be explained by the increased likelihood of faculty with more funding, publications, or lectures to be selected for leadership roles, this finding supports the notion that research can contribute to leadership. This analysis identifies modifiable factors among residency training programs to improve the odds of graduating future academic leaders.

Conclusion

As the present study shows, the ratio of faculty to residents and the number of full-time faculty publications are key to graduating academic leaders in dermatology. Retention of faculty as leaders in academic dermatology is as crucial to the field as recruitment of residents into academic dermatology. Mentorship should be highly encouraged through the creation of formal programs and should not end at the resident level. Emphasizing the intellectually stimulating aspects of academia and providing administrative resources may help decrease the burden of academic duties, allowing the pursuit of teaching and research and ultimately the resources to become candidates for leadership positions in academia.

Many hospital-based adult specialties are routinely ranked annually by US News & World Report. Its 2014-2015 rankings included cancer; cardiology and heart surgery; diabetes and endocrinology; ear, nose, and throat; gastroenterology and gastrointestinal surgery; geriatrics; gynecology; nephrology; neurology and neurosurgery; ophthalmology; orthopedics; pulmonology; psychiatry; rehabilitation; rheumatology; and urology.¹ Only teaching hospitals that saw a high volume of patients were included in the rankings. Ophthalmology, psychiatry, rehabilitation, and rheumatology were ranked based on reputation among specialists; the others were ranked based on objective data (eg, hospital mortality rates, nursing ratios) in addition to a reputation score from a survey of physicians.¹

Dermatology has never been included in the annual US News & World Report rankings. In the past, dermatology residency programs have been ranked solely according to the amount of annual funding received from the National Institutes of Health (NIH).² Wu et al³ expanded the scope by creating an algorithm used to rank dermatology residency programs based on scholarly achievement. Included were a number of NIH grants as well as 4 other factors—publications in 2001-2004, Dermatology Foundation (DF) grants from 2001-2004, faculty lectures in 2004 delivered at national conferences, and number of full-time faculty members who were on the editorial boards of the top 3 US dermatology journals and the top 4 subspecialty journals—deemed important by the authors.³

The current study refines the prior algorithm by creating a weighted ranking algorithm and using criteria that the authors considered to be more meaningful than the original criteria used.³ Specifically, the current study considered the amount of NIH and DF funding received versus the number of grants received, and less importance was given to the number of faculty members on editorial boards and DF funding relative to other criteria. We used publicly available data from Web searches to conduct this study.

Methods

The overall ranking algorithm was designed based on the methodology used by the Institute of Higher Education, Shanghai Jiao Tong University, to rank universities in the annual Academic Ranking of World Universities, which is published annually and uses a weighted ranking algorithm that includes academic and research performance factors to evaluate universities worldwide.⁴

The names of all dermatology residency programs in the United States were obtained as of December 31, 2008, from FREIDA Online using the search term dermatology; the names of all full-time faculty members at these residency programs and number of residents also were obtained by searching the programs’ Web sites.

For another related study investigating the relationship between residency program characteristics and residents pursuing a career in dermatology, the following data were obtained: total number of full-time faculty members at the program; total number of residents; amount of NIH funding received in 2008 (http://report.nih.gov/award/index.cfm); amount of DF funding received in 2008 (http://dermatologyfoundation.org/pdf/pubs/DF_2008_Annual_Report.pdf.); number of publications by full-time faculty members in 2008 (http://www.ncbi.nlm.nih.gov/pubmed/); number of faculty lectures given at annual meetings of 5 societies in 2008 (American Academy of Dermatology, the Society for Investigative Dermatology, the American Society of Dermatopathology, the Society for Pediatric Dermatology, and the American Society for Dermatologic Surgery); the number of full-time faculty members who were on the editorial boards of 6 dermatology journals (Journal of Investigative Dermatology, Archives of Dermatology [currently known as JAMA Dermatology], Journal of the American Academy of Dermatology, Dermatologic Surgery, Journal of Cutaneous Pathology, and Pediatric Dermatology); and whether a program was housed within an institution’s department of dermatology or division of internal medicine.

The data were summed for all faculty members at a given program. To avoid duplicate faculty publications, collections for each residency program were created within PubMed (ie, if 2 authors from the same program coauthored an article, it was only counted once toward the total number of faculty publications from that program).

The dermatology residency programs that were excluded from this analysis included University of Texas at Austin, University of Texas Medical Branch, and the University of Connecticut, which were started in 2008, as well as Kaiser Permanente Southern California, which was started in 2010. The combined Boston University and Tufts University dermatology residency program was established prior to 2008 and therefore was counted as such in the analysis. The program at Harbor-UCLA Medical Center was not included in the analysis because the program was reestablished in 2004 and we could not assume that the prior program had similar attributes. Military residency programs also were excluded from the analysis, as residents are assigned to faculty positions upon graduation. The NIH residency program also was excluded because it is not a traditional 3-year residency program.

There were 5 factors that were deemed by the authors to be the most reflective of academic achievement among dermatology residency programs: number of faculty publications in 2008; amount of NIH funding received in 2008; number of faculty lectures given at 5 society meetings in 2008; amount of DF funding received in 2008; and the number of faculty members who were on the editorial boards of 6 dermatology journals in 2008. We wished to get a broad range of subspecialties of dermatology (eg, medical, surgical) for the journals. Further, the 6 journals selected were chosen because they had the highest impact factors at the time for general dermatology and dermatopathology.

Each residency program was assigned a score from 0 to 1 for each of these factors. The program with the highest number of faculty publications was assigned a score of 1 and the program with the lowest number of publications was assigned a score of 0. The programs in between were subsequently assigned scores from 0 to 1 based on the number of publications as a percentage of the number of publications from the program with the most publications.

A weighted ranking scheme was used to rank programs based on the relative importance of each factor. The authors decided that NIH funding, number of faculty publications, and number of lectures at society meetings were relatively more important than the other factors; thus these factors were given a weight of 1.0. The remaining factors—DF funding and number of faculty members on editorial boards of journals—were given a weight of 0.5. Values were totaled and programs were ranked based on the sum of these values.

Data were analyzed using SAS 9.2 software. This study was approved by the institutional review board at Kaiser Permanente Southern California.

Results

The overall ranking of the top 20 US dermatology residency programs in 2008 is presented in Table 1. The top 5 programs based on each of the 5 factors used in the weighted ranking algorithm are presented in Tables 2 through 6.

A separate analysis was performed to evaluate the relationship between the ranking factors and the size of the residency programs, as the data seemed to favor larger programs. Table 7 demonstrates that the amount of NIH and DF funding was positively correlated with the number of faculty members in a residency program. The remaining factors were not correlated with the number of faculty in the program.

Due to space considerations, analyses are based on data not shown in this manuscript. Data about the characteristics of each residency program are available from the authors.

Comment

There previously have been few attempts to rank US dermatology residency programs based on factors related to academic achievement. Individual faculty and programs have been ranked based on the number of publications in the literature.^5-7 Dermatology institutions/organizations (eg, departments, hospitals, medical schools) have been ranked based on amount of NIH funding received and number of journal citations.⁸

Dermatology residency programs have been ranked based on amount of NIH funding received annually.² Based on the prior Wu et al³ ranking algorithm for 2004 data, the institutions with the top 5 residency programs were the University of Pennsylvania, University of California, San Francisco; Yale University; New York University; and the University of Michigan.

The current study refined this ranking algorithm by including residency programs housed within the institution’s division of dermatology of a department of internal medicine, which were previously excluded from NIH funding data, as opposed to just the department of dermatology. The authors also did not count publications coauthored by faculty members at the same institution more than once. The annual amount of DF funding received was considered rather than number of grants, as this factor was thought to better reflect the scale of research being conducted. Additionally, relatively more weight was given to annual NIH funding, annual number of faculty publications, and number of faculty lectures at annual society meetings than to annual amount of DF funding and number of faculty on journal editorial boards. The University of Pennsylvania; the University of California, San Francisco; and Yale University were in the top 5 based on data from 2004 in the prior study³ and 2008 in the current study.

Distribution of grant funding and number of faculty publications in the literature often are used as a measure of scholarly achievement. Within the specialty of dermatology, the amount of NIH and DF funding received could be considered the most prestigious achievement for a residency program. Faculty members are encouraged to develop expertise in a specialized area of dermatology as well as to conduct research and publish articles in the area of their choosing. One of the most common ways to attain tenure and achieve academic recognition is by publishing articles in peer-reviewed journals with high impact factors. Other factors that demonstrate expertise and accomplishment in one’s field include giving lectures at national society meetings and sitting on editorial boards of prestigious journals.

The authors deemed these factors to be the most reflective of academic achievement in a dermatology residency program. It should come as no surprise that the top programs according to our algorithm generally are regarded as the most prestigious programs in the country.

The current study did have some limitations. For instance, residency programs were ranked solely based on academic achievements. Although academic achievement is an important aspect of the reputation of a dermatology residency program, it does not account for other important aspects of a program such as commitment to teaching and patient care. These aspects are difficult to measure and were not included in our ranking algorithm. For this reason, there are many excellent residency programs that may not be listed in this article as top programs but still provide outstanding clinical training and patient care. Our ranking algorithm is more indicative of a program’s commitment to research and scholarship and does not necessarily reflect how well a program trains its residents.

The factors included in our ranking algorithm also were somewhat arbitrary. The 5 factors that were chosen by the authors were considered to be most reflective of academic achievement and also will be easy to obtain in future years to replicate these rankings; however, there are other important factors that could have been used instead or in addition to the factors we chose. Some of the chosen factors were more important than others, so a decision was made to weight the factors. In-training examination scores from the American Board of Dermatology, boards passing rate, or percentage of residents who received fellowships or academic appointments were not used because this information is not publicly available. The current study also appeared to favor larger residency programs. Programs with fewer faculty members generally receive less research funding and have fewer publications, fewer faculty members on journal editorial boards, and fewer lecturers at national society meetings. This factor was not controlled for in the original analysis, as larger programs generally are thought to be more prestigious; therefore, this bias should be accounted for in the rankings.

Conclusion

Based on our weighted ranking algorithm, the top 5 dermatology residency programs in 2008 were University of California, San Francisco; Northwestern University; University of Pennsylvania; Yale University; and Stanford University.

Acknowledgements—We thank all of the program coordinators, full-time faculty members, program directors, chairs, and chiefs who were kind enough to field our questions if we had any missing data about their programs. We thank Mary H. Black, PhD, Pasadena, California, for her contribution to the analytical plan.

Many hospital-based adult specialties are routinely ranked annually by US News & World Report. Its 2014-2015 rankings included cancer; cardiology and heart surgery; diabetes and endocrinology; ear, nose, and throat; gastroenterology and gastrointestinal surgery; geriatrics; gynecology; nephrology; neurology and neurosurgery; ophthalmology; orthopedics; pulmonology; psychiatry; rehabilitation; rheumatology; and urology.¹ Only teaching hospitals that saw a high volume of patients were included in the rankings. Ophthalmology, psychiatry, rehabilitation, and rheumatology were ranked based on reputation among specialists; the others were ranked based on objective data (eg, hospital mortality rates, nursing ratios) in addition to a reputation score from a survey of physicians.¹

Dermatology has never been included in the annual US News & World Report rankings. In the past, dermatology residency programs have been ranked solely according to the amount of annual funding received from the National Institutes of Health (NIH).² Wu et al³ expanded the scope by creating an algorithm used to rank dermatology residency programs based on scholarly achievement. Included were a number of NIH grants as well as 4 other factors—publications in 2001-2004, Dermatology Foundation (DF) grants from 2001-2004, faculty lectures in 2004 delivered at national conferences, and number of full-time faculty members who were on the editorial boards of the top 3 US dermatology journals and the top 4 subspecialty journals—deemed important by the authors.³

The current study refines the prior algorithm by creating a weighted ranking algorithm and using criteria that the authors considered to be more meaningful than the original criteria used.³ Specifically, the current study considered the amount of NIH and DF funding received versus the number of grants received, and less importance was given to the number of faculty members on editorial boards and DF funding relative to other criteria. We used publicly available data from Web searches to conduct this study.

Methods

The overall ranking algorithm was designed based on the methodology used by the Institute of Higher Education, Shanghai Jiao Tong University, to rank universities in the annual Academic Ranking of World Universities, which is published annually and uses a weighted ranking algorithm that includes academic and research performance factors to evaluate universities worldwide.⁴

The names of all dermatology residency programs in the United States were obtained as of December 31, 2008, from FREIDA Online using the search term dermatology; the names of all full-time faculty members at these residency programs and number of residents also were obtained by searching the programs’ Web sites.

For another related study investigating the relationship between residency program characteristics and residents pursuing a career in dermatology, the following data were obtained: total number of full-time faculty members at the program; total number of residents; amount of NIH funding received in 2008 (http://report.nih.gov/award/index.cfm); amount of DF funding received in 2008 (http://dermatologyfoundation.org/pdf/pubs/DF_2008_Annual_Report.pdf.); number of publications by full-time faculty members in 2008 (http://www.ncbi.nlm.nih.gov/pubmed/); number of faculty lectures given at annual meetings of 5 societies in 2008 (American Academy of Dermatology, the Society for Investigative Dermatology, the American Society of Dermatopathology, the Society for Pediatric Dermatology, and the American Society for Dermatologic Surgery); the number of full-time faculty members who were on the editorial boards of 6 dermatology journals (Journal of Investigative Dermatology, Archives of Dermatology [currently known as JAMA Dermatology], Journal of the American Academy of Dermatology, Dermatologic Surgery, Journal of Cutaneous Pathology, and Pediatric Dermatology); and whether a program was housed within an institution’s department of dermatology or division of internal medicine.

The data were summed for all faculty members at a given program. To avoid duplicate faculty publications, collections for each residency program were created within PubMed (ie, if 2 authors from the same program coauthored an article, it was only counted once toward the total number of faculty publications from that program).

The dermatology residency programs that were excluded from this analysis included University of Texas at Austin, University of Texas Medical Branch, and the University of Connecticut, which were started in 2008, as well as Kaiser Permanente Southern California, which was started in 2010. The combined Boston University and Tufts University dermatology residency program was established prior to 2008 and therefore was counted as such in the analysis. The program at Harbor-UCLA Medical Center was not included in the analysis because the program was reestablished in 2004 and we could not assume that the prior program had similar attributes. Military residency programs also were excluded from the analysis, as residents are assigned to faculty positions upon graduation. The NIH residency program also was excluded because it is not a traditional 3-year residency program.

There were 5 factors that were deemed by the authors to be the most reflective of academic achievement among dermatology residency programs: number of faculty publications in 2008; amount of NIH funding received in 2008; number of faculty lectures given at 5 society meetings in 2008; amount of DF funding received in 2008; and the number of faculty members who were on the editorial boards of 6 dermatology journals in 2008. We wished to get a broad range of subspecialties of dermatology (eg, medical, surgical) for the journals. Further, the 6 journals selected were chosen because they had the highest impact factors at the time for general dermatology and dermatopathology.

Each residency program was assigned a score from 0 to 1 for each of these factors. The program with the highest number of faculty publications was assigned a score of 1 and the program with the lowest number of publications was assigned a score of 0. The programs in between were subsequently assigned scores from 0 to 1 based on the number of publications as a percentage of the number of publications from the program with the most publications.

A weighted ranking scheme was used to rank programs based on the relative importance of each factor. The authors decided that NIH funding, number of faculty publications, and number of lectures at society meetings were relatively more important than the other factors; thus these factors were given a weight of 1.0. The remaining factors—DF funding and number of faculty members on editorial boards of journals—were given a weight of 0.5. Values were totaled and programs were ranked based on the sum of these values.

Data were analyzed using SAS 9.2 software. This study was approved by the institutional review board at Kaiser Permanente Southern California.

Results

The overall ranking of the top 20 US dermatology residency programs in 2008 is presented in Table 1. The top 5 programs based on each of the 5 factors used in the weighted ranking algorithm are presented in Tables 2 through 6.

A separate analysis was performed to evaluate the relationship between the ranking factors and the size of the residency programs, as the data seemed to favor larger programs. Table 7 demonstrates that the amount of NIH and DF funding was positively correlated with the number of faculty members in a residency program. The remaining factors were not correlated with the number of faculty in the program.

Due to space considerations, analyses are based on data not shown in this manuscript. Data about the characteristics of each residency program are available from the authors.

Comment

There previously have been few attempts to rank US dermatology residency programs based on factors related to academic achievement. Individual faculty and programs have been ranked based on the number of publications in the literature.^5-7 Dermatology institutions/organizations (eg, departments, hospitals, medical schools) have been ranked based on amount of NIH funding received and number of journal citations.⁸

Dermatology residency programs have been ranked based on amount of NIH funding received annually.² Based on the prior Wu et al³ ranking algorithm for 2004 data, the institutions with the top 5 residency programs were the University of Pennsylvania, University of California, San Francisco; Yale University; New York University; and the University of Michigan.

The current study refined this ranking algorithm by including residency programs housed within the institution’s division of dermatology of a department of internal medicine, which were previously excluded from NIH funding data, as opposed to just the department of dermatology. The authors also did not count publications coauthored by faculty members at the same institution more than once. The annual amount of DF funding received was considered rather than number of grants, as this factor was thought to better reflect the scale of research being conducted. Additionally, relatively more weight was given to annual NIH funding, annual number of faculty publications, and number of faculty lectures at annual society meetings than to annual amount of DF funding and number of faculty on journal editorial boards. The University of Pennsylvania; the University of California, San Francisco; and Yale University were in the top 5 based on data from 2004 in the prior study³ and 2008 in the current study.

Distribution of grant funding and number of faculty publications in the literature often are used as a measure of scholarly achievement. Within the specialty of dermatology, the amount of NIH and DF funding received could be considered the most prestigious achievement for a residency program. Faculty members are encouraged to develop expertise in a specialized area of dermatology as well as to conduct research and publish articles in the area of their choosing. One of the most common ways to attain tenure and achieve academic recognition is by publishing articles in peer-reviewed journals with high impact factors. Other factors that demonstrate expertise and accomplishment in one’s field include giving lectures at national society meetings and sitting on editorial boards of prestigious journals.

The authors deemed these factors to be the most reflective of academic achievement in a dermatology residency program. It should come as no surprise that the top programs according to our algorithm generally are regarded as the most prestigious programs in the country.

The current study did have some limitations. For instance, residency programs were ranked solely based on academic achievements. Although academic achievement is an important aspect of the reputation of a dermatology residency program, it does not account for other important aspects of a program such as commitment to teaching and patient care. These aspects are difficult to measure and were not included in our ranking algorithm. For this reason, there are many excellent residency programs that may not be listed in this article as top programs but still provide outstanding clinical training and patient care. Our ranking algorithm is more indicative of a program’s commitment to research and scholarship and does not necessarily reflect how well a program trains its residents.

The factors included in our ranking algorithm also were somewhat arbitrary. The 5 factors that were chosen by the authors were considered to be most reflective of academic achievement and also will be easy to obtain in future years to replicate these rankings; however, there are other important factors that could have been used instead or in addition to the factors we chose. Some of the chosen factors were more important than others, so a decision was made to weight the factors. In-training examination scores from the American Board of Dermatology, boards passing rate, or percentage of residents who received fellowships or academic appointments were not used because this information is not publicly available. The current study also appeared to favor larger residency programs. Programs with fewer faculty members generally receive less research funding and have fewer publications, fewer faculty members on journal editorial boards, and fewer lecturers at national society meetings. This factor was not controlled for in the original analysis, as larger programs generally are thought to be more prestigious; therefore, this bias should be accounted for in the rankings.

Conclusion

Based on our weighted ranking algorithm, the top 5 dermatology residency programs in 2008 were University of California, San Francisco; Northwestern University; University of Pennsylvania; Yale University; and Stanford University.

Acknowledgements—We thank all of the program coordinators, full-time faculty members, program directors, chairs, and chiefs who were kind enough to field our questions if we had any missing data about their programs. We thank Mary H. Black, PhD, Pasadena, California, for her contribution to the analytical plan.

Many hospital-based adult specialties are routinely ranked annually by US News & World Report. Its 2014-2015 rankings included cancer; cardiology and heart surgery; diabetes and endocrinology; ear, nose, and throat; gastroenterology and gastrointestinal surgery; geriatrics; gynecology; nephrology; neurology and neurosurgery; ophthalmology; orthopedics; pulmonology; psychiatry; rehabilitation; rheumatology; and urology.¹ Only teaching hospitals that saw a high volume of patients were included in the rankings. Ophthalmology, psychiatry, rehabilitation, and rheumatology were ranked based on reputation among specialists; the others were ranked based on objective data (eg, hospital mortality rates, nursing ratios) in addition to a reputation score from a survey of physicians.¹

Dermatology has never been included in the annual US News & World Report rankings. In the past, dermatology residency programs have been ranked solely according to the amount of annual funding received from the National Institutes of Health (NIH).² Wu et al³ expanded the scope by creating an algorithm used to rank dermatology residency programs based on scholarly achievement. Included were a number of NIH grants as well as 4 other factors—publications in 2001-2004, Dermatology Foundation (DF) grants from 2001-2004, faculty lectures in 2004 delivered at national conferences, and number of full-time faculty members who were on the editorial boards of the top 3 US dermatology journals and the top 4 subspecialty journals—deemed important by the authors.³

The current study refines the prior algorithm by creating a weighted ranking algorithm and using criteria that the authors considered to be more meaningful than the original criteria used.³ Specifically, the current study considered the amount of NIH and DF funding received versus the number of grants received, and less importance was given to the number of faculty members on editorial boards and DF funding relative to other criteria. We used publicly available data from Web searches to conduct this study.

Methods

The overall ranking algorithm was designed based on the methodology used by the Institute of Higher Education, Shanghai Jiao Tong University, to rank universities in the annual Academic Ranking of World Universities, which is published annually and uses a weighted ranking algorithm that includes academic and research performance factors to evaluate universities worldwide.⁴

The names of all dermatology residency programs in the United States were obtained as of December 31, 2008, from FREIDA Online using the search term dermatology; the names of all full-time faculty members at these residency programs and number of residents also were obtained by searching the programs’ Web sites.

For another related study investigating the relationship between residency program characteristics and residents pursuing a career in dermatology, the following data were obtained: total number of full-time faculty members at the program; total number of residents; amount of NIH funding received in 2008 (http://report.nih.gov/award/index.cfm); amount of DF funding received in 2008 (http://dermatologyfoundation.org/pdf/pubs/DF_2008_Annual_Report.pdf.); number of publications by full-time faculty members in 2008 (http://www.ncbi.nlm.nih.gov/pubmed/); number of faculty lectures given at annual meetings of 5 societies in 2008 (American Academy of Dermatology, the Society for Investigative Dermatology, the American Society of Dermatopathology, the Society for Pediatric Dermatology, and the American Society for Dermatologic Surgery); the number of full-time faculty members who were on the editorial boards of 6 dermatology journals (Journal of Investigative Dermatology, Archives of Dermatology [currently known as JAMA Dermatology], Journal of the American Academy of Dermatology, Dermatologic Surgery, Journal of Cutaneous Pathology, and Pediatric Dermatology); and whether a program was housed within an institution’s department of dermatology or division of internal medicine.

The data were summed for all faculty members at a given program. To avoid duplicate faculty publications, collections for each residency program were created within PubMed (ie, if 2 authors from the same program coauthored an article, it was only counted once toward the total number of faculty publications from that program).

The dermatology residency programs that were excluded from this analysis included University of Texas at Austin, University of Texas Medical Branch, and the University of Connecticut, which were started in 2008, as well as Kaiser Permanente Southern California, which was started in 2010. The combined Boston University and Tufts University dermatology residency program was established prior to 2008 and therefore was counted as such in the analysis. The program at Harbor-UCLA Medical Center was not included in the analysis because the program was reestablished in 2004 and we could not assume that the prior program had similar attributes. Military residency programs also were excluded from the analysis, as residents are assigned to faculty positions upon graduation. The NIH residency program also was excluded because it is not a traditional 3-year residency program.

There were 5 factors that were deemed by the authors to be the most reflective of academic achievement among dermatology residency programs: number of faculty publications in 2008; amount of NIH funding received in 2008; number of faculty lectures given at 5 society meetings in 2008; amount of DF funding received in 2008; and the number of faculty members who were on the editorial boards of 6 dermatology journals in 2008. We wished to get a broad range of subspecialties of dermatology (eg, medical, surgical) for the journals. Further, the 6 journals selected were chosen because they had the highest impact factors at the time for general dermatology and dermatopathology.

Each residency program was assigned a score from 0 to 1 for each of these factors. The program with the highest number of faculty publications was assigned a score of 1 and the program with the lowest number of publications was assigned a score of 0. The programs in between were subsequently assigned scores from 0 to 1 based on the number of publications as a percentage of the number of publications from the program with the most publications.

A weighted ranking scheme was used to rank programs based on the relative importance of each factor. The authors decided that NIH funding, number of faculty publications, and number of lectures at society meetings were relatively more important than the other factors; thus these factors were given a weight of 1.0. The remaining factors—DF funding and number of faculty members on editorial boards of journals—were given a weight of 0.5. Values were totaled and programs were ranked based on the sum of these values.

Data were analyzed using SAS 9.2 software. This study was approved by the institutional review board at Kaiser Permanente Southern California.

Results

The overall ranking of the top 20 US dermatology residency programs in 2008 is presented in Table 1. The top 5 programs based on each of the 5 factors used in the weighted ranking algorithm are presented in Tables 2 through 6.

A separate analysis was performed to evaluate the relationship between the ranking factors and the size of the residency programs, as the data seemed to favor larger programs. Table 7 demonstrates that the amount of NIH and DF funding was positively correlated with the number of faculty members in a residency program. The remaining factors were not correlated with the number of faculty in the program.

Due to space considerations, analyses are based on data not shown in this manuscript. Data about the characteristics of each residency program are available from the authors.

Comment

There previously have been few attempts to rank US dermatology residency programs based on factors related to academic achievement. Individual faculty and programs have been ranked based on the number of publications in the literature.^5-7 Dermatology institutions/organizations (eg, departments, hospitals, medical schools) have been ranked based on amount of NIH funding received and number of journal citations.⁸

Dermatology residency programs have been ranked based on amount of NIH funding received annually.² Based on the prior Wu et al³ ranking algorithm for 2004 data, the institutions with the top 5 residency programs were the University of Pennsylvania, University of California, San Francisco; Yale University; New York University; and the University of Michigan.

The current study refined this ranking algorithm by including residency programs housed within the institution’s division of dermatology of a department of internal medicine, which were previously excluded from NIH funding data, as opposed to just the department of dermatology. The authors also did not count publications coauthored by faculty members at the same institution more than once. The annual amount of DF funding received was considered rather than number of grants, as this factor was thought to better reflect the scale of research being conducted. Additionally, relatively more weight was given to annual NIH funding, annual number of faculty publications, and number of faculty lectures at annual society meetings than to annual amount of DF funding and number of faculty on journal editorial boards. The University of Pennsylvania; the University of California, San Francisco; and Yale University were in the top 5 based on data from 2004 in the prior study³ and 2008 in the current study.

Distribution of grant funding and number of faculty publications in the literature often are used as a measure of scholarly achievement. Within the specialty of dermatology, the amount of NIH and DF funding received could be considered the most prestigious achievement for a residency program. Faculty members are encouraged to develop expertise in a specialized area of dermatology as well as to conduct research and publish articles in the area of their choosing. One of the most common ways to attain tenure and achieve academic recognition is by publishing articles in peer-reviewed journals with high impact factors. Other factors that demonstrate expertise and accomplishment in one’s field include giving lectures at national society meetings and sitting on editorial boards of prestigious journals.

The authors deemed these factors to be the most reflective of academic achievement in a dermatology residency program. It should come as no surprise that the top programs according to our algorithm generally are regarded as the most prestigious programs in the country.

The current study did have some limitations. For instance, residency programs were ranked solely based on academic achievements. Although academic achievement is an important aspect of the reputation of a dermatology residency program, it does not account for other important aspects of a program such as commitment to teaching and patient care. These aspects are difficult to measure and were not included in our ranking algorithm. For this reason, there are many excellent residency programs that may not be listed in this article as top programs but still provide outstanding clinical training and patient care. Our ranking algorithm is more indicative of a program’s commitment to research and scholarship and does not necessarily reflect how well a program trains its residents.

The factors included in our ranking algorithm also were somewhat arbitrary. The 5 factors that were chosen by the authors were considered to be most reflective of academic achievement and also will be easy to obtain in future years to replicate these rankings; however, there are other important factors that could have been used instead or in addition to the factors we chose. Some of the chosen factors were more important than others, so a decision was made to weight the factors. In-training examination scores from the American Board of Dermatology, boards passing rate, or percentage of residents who received fellowships or academic appointments were not used because this information is not publicly available. The current study also appeared to favor larger residency programs. Programs with fewer faculty members generally receive less research funding and have fewer publications, fewer faculty members on journal editorial boards, and fewer lecturers at national society meetings. This factor was not controlled for in the original analysis, as larger programs generally are thought to be more prestigious; therefore, this bias should be accounted for in the rankings.

Conclusion

Based on our weighted ranking algorithm, the top 5 dermatology residency programs in 2008 were University of California, San Francisco; Northwestern University; University of Pennsylvania; Yale University; and Stanford University.

Acknowledgements—We thank all of the program coordinators, full-time faculty members, program directors, chairs, and chiefs who were kind enough to field our questions if we had any missing data about their programs. We thank Mary H. Black, PhD, Pasadena, California, for her contribution to the analytical plan.