- Email: [email protected]
National Orthopedic Residency Attrition: Who Is At Risk?
ORIGINAL REPORTS
National Orthopedic Residency Attrition: Who Is At Risk? Jennifer M. Bauer, MD, and Ginger E. Holt, MD Vanderbilt University Medical Center, Nashville, Tennessee BACKGROUND: National U.S. orthopedic resident attri-
CONCLUSIONS: Orthopedic residents who are female,
tion rates have been historically low, but no literature exists as to the characteristics of those who leave nor the circumstance of the departure. We aimed to determine factors that may place a resident at higher risk for attrition. Additionally, we planned to determine whether the 2003-work hour restriction affected attrition rate.
single, or without children are statistically more likely to undergo attrition. Consideration could be given to targeted C mentoring of these resident groups. ( J Surg Ed ]:]]]-]]]. J 2016 Association of Program Directors in Surgery. Published by Elsevier Inc. All rights reserved.)
MATERIALS AND METHODS: All orthopedic surgery
orthopedic residency
residency program directors in the United States were surveyed on demographic data for their current resident class, the number of residents who left the program, as well as demographic description for each of the residents who left their program from 1998 to 2013. Exclusion criteria included military programs and those younger than 3 years. All data were deidentified and compared to the Accreditation Council for Graduate Medical Education Data Resource book to protect against sample error in respondents. RESULTS: Of 146 programs included, the overall response rate was 54% of residency directors, representing 51% of orthopedic residents. The respondent demographic makeup of 13.7% female, and average program size of 22.3 residents, compared similarly to the Accreditation Council for Graduate Medical Education national average of 13% female and 23-resident program size. Compared to all respondents, residents who left their program were more likely to be female (27%, p ¼ 0.0018), single (51%, p ¼ 0.0028), and without children (80%, p ¼ 0.0018). There was no statistical difference based on minority status or 2003-instituted work hour restriction. Of those who left, 45% transferred to another specialty, 34% were dismissed, 14% voluntarily withdrew or cited personal reasons, and 6% transferred to another orthopedic program. The most common specialties to transfer into were radiology (30%), emergency medicine (25%), and anesthesia (18%).
Correspondence: Inquiries to Jennifer M. Bauer, MD, Vanderbilt University Medical Center, 1215 21st Ave S, Medical Center East, South Tower, Suite 4200, Nashville, TN 37232-8774. fax: +615 875 1915; e-mail: [email protected]
KEY WORDS: resident attrition, resident education, COMPETENCIES:
Professionalism, Communication Skills
Interpersonal
and
INTRODUCTION Losing residents from U.S. orthopedic surgery residencies is not common, with an average annual loss in the last 7 years of 0.96% of residents, as reported by the Accreditation Council for Graduate Medical Education (ACGME).1-7 Of those who do not complete residency each year, 16% do so for reasons of dismissal or lack of success, whereas 84% are due to transfer and withdrawal that can be considered voluntary attrition. Although these numbers are low because the average residency class size is 4.5 residents and much time and diligence is spent by programs to pick and train a residency class, the loss of one resident can weigh heavily on a class and a program. Although there is currently little information on who leaves an orthopedic program, other specialties have been investigated. Abundant literature exists examining the annual rates and reasons for general surgery attrition, cited as high as 20% by some,8,9 with at least one study noting an increase in the rate despite 80-hour work restrictions.10 Other specialties’ annual loss of residents includes 14% per year in neurosurgery, with a statistically significant increased rate for women11-12; 4.2% in obstetrics and gynecology, with a higher rate in men13; 1.1% in ophthalmology that was equal between both the sexes14; and 1.2% in otolaryngology.15 This data from peer reviewed literature does not always match that reported by the ACGME. There are no recent studies that examine the rate and characteristics of orthopedic resident attrition. The purpose
Journal of Surgical Education & 2016 Association of Program Directors in Surgery. Published by 1931-7204/$30.00 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jsurg.2016.03.010
1
TABLE 1. Survey Questions Basic Questions for All Programs How How How How
many many many many
Questions for Each Resident Who Left
residents are currently in your program? current residents are female? current residents identify as non-Caucasian? current residents are single? (or unknown)
How many current residents have children? (or unknown) How many current residents left your program between July 1998 and 2003? How many current residents left your program between July 2003 and 2013?
of our study was to determine if statistically significant characteristics are shared among orthopedic surgery residents who leave their programs, the reason residents leave, and if the rate of attrition has been affected by the 2003 80-hour work restriction.
MATERIALS AND METHODS We compiled a comprehensive list of U.S. orthopedic surgery residency programs from the American Medical Association’s online data system. Using this, individual program websites, and Council of Orthopaedic Residency Directors directories, we gathered e-mail addresses for each resident program director. There are 157 individual orthopedic programs currently active. Programs were excluded from the data collection if they were noncivilian programs, because of the complexity of commitment to military programs. Those started within the last 3 years were also excluded. We included a total of 146 programs. An electronic survey was created on Research Electronic Data Capture (REDCap) with 7 basic questions and a set
Residency class size for resident's year at departure? Resident's sex? What race/ethnicity did the resident identify as? Was the resident married/have a partner/comparable significant other? Did the resident have children? At what point in the resident's training did he/she leave? To your understanding, why did he/she leave? Dismissal Left medicine completely Transfer to different specialty—if so which Health/Hardship Unknown Other—please explain further
of further detailed questions for each resident who left (Table 1). Residency directors were asked about the basic demographics of their current residency class; the number of residents who left in the previous 5 years of 2003-work hour restrictions and in the 10 years after; and specifics about those who left, including demographics, reasons for leaving, and what the resident pursued after leaving. There was an option for additional questions or comments at the end. To collect participants, the REDCap survey that collects deidentified data was sent twice to nonrespondents, followed by 4 separate rounds of individualized emails with both the REDCap link and an option to fill a spreadsheet with the same questions, with that data then deidentified on receipt. To ensure the data we captured did not represent significant sampling error, the demographics of our sample were compared against published ACGME data averaged from the last 7 years.1-7 Data was analyzed using 2-tailed Fisher exact test for categorical variables, or Student t-test for continuous variable means with 95% confidence intervals included. Responses that were left blank or answered as “unknown” were factored out of the calculations for averages and percentages.
TABLE 2. Demographics, U.S. Orthopaedic Surgery Residents Total Survey Respondents Class of 2013 to 2014
ACGME Class of 2013 to Significance 2014
Total residents Average program size
1763 22.3
3566 23
Percent female Percent Ethnic or racial minority Percent without significant other Percent without children
13.7 20.3
13 35.3
40.0
—
30.8
—
2
p ¼ 0.5 95% CI: 2.65 to 1.29 p ¼ 0.5 p o 0.001
Journal of Surgical Education Volume ]/Number ] ] 2016
TABLE 3. Characteristics of Residents Who Left Orthopaedic Residency Study Cohort Who Left Residency
Total Survey Respondents Class of 2013 to 2014
Average program size
23.0
22.3
Percent female Percent ethnic or racial minority Percent without significant other Percent without children
27.0 22.6
13.7 20.3
p ¼ 0.5 95% CI: 2.82 to 1.37 p o 0.005 p ¼ 0.65
51.3
40.0
p o 0.005
80.0
69.2
p o 0.005
Significance
RESULTS
CONCLUSIONS
We received 79 survey responses from 146 orthopedic residency directors, accounting for 54% of programs and 51% of orthopedic residents. The data for all responding programs showed an average class size of 22.3 residents, and a demographic make-up of 13.7% female, 20.3% ethnic minority, 40% without a significant other, and 30.8% without children. Compared to the ACGME demographic categories available (Table 2), our sample was an appropriate representation of class size and sex distribution but significantly different with regard to ethnicity. Of the 79 programs who responded, 57 programs had lost at least 1 resident since 1998. There was a 0.58% rate of loss of residents per year for 1998 to 2003, and 0.59% for 2003 to 2013; there was no demonstrated difference from before the 80-hour work week restrictions to after. These rates of loss reflect a statistically significant difference from the ACGME reported average of 0.96% over the last 7 years (p o 0.0015, 95% CI: 0.21-0.55). The demographic characteristics of those who left compared to the overall sample group is shown in Table 3. There was a statistically significant increase seen for females, those without a significant other, and those without children. There was no difference in rate of attrition based on ethnicity or class size that the resident came from. The most common year of training to leave from was second year (44.3%), followed by intern year (25.4%) and third year (23%). Only 30% of those who left did so after second year. Overall, 54% were offered remediation prior to departure. Those who left did so for a variety of reasons (Fig. 1), including 34% due to involuntary dismissal. There was no significant difference between males and females with respect to involuntary departure. The remainder left for the voluntary reasons of changing specialties (44%), withdrawal/personal reasons (14%), or transferring orthopedic programs (5%). Those who left orthopedics for a different specialty entered 1 of 10 various fields (Fig. 2), with radiology, emergency medicine, and anesthesia capturing 72% overall.
This is the most comprehensive study to survey orthopedic residency directors on the characteristics of and reasons leading to resident attrition. An earlier survey drawing on the classes of 1980 to 1987 did show a significantly higher rate of females withdrawing, but did not include further characteristics besides sex nor detailed causes for those who left.16 This study again finds significantly more women leaving. Several studies point out the low rate of women and minorities in orthopedic residency. Because no sex bias has been shown in initial review of residency applications,17 it might be inferred that the disparity is indicative of a low number of female applicants. Both attracting females and retaining them may be tied to the importance of having faculty role models and mentors.18,19 Lack of sex difference in our results for dismissal, as well as studies showing residency performance correlating to no clear characteristics besides the OITE and ABOS score associations,20-22 point away from female performance as a culprit. Although several laud the increased percentage of women in both residency and academic faculty positions,23,24 many are quick to point out that this raw growth mirrors only that of women into medical school itself and therefore represents no actual proportional growth of diversity.25-28 Moreover, what growth that does exist may be concentrated within only
Journal of Surgical Education Volume ]/Number ] ] 2016
3
Reasons Given for Leaving Residency 5%
4%
7% 43%
7%
Transferred Specialty Dismissal Withdrew Health/personal Transferred Ortho Unknown
34% FIGURE 1. The breakdown of reasons for leaving orthopaedic residency, from the survey cohort.
Other Residency Types Pursued 4%
2% 2% 2% Radiology
5%
30%
7%
Emergency Anesthesia
7%
General surgery Family med. PM&R
18%
Neurosurgery
25%
Pscyhiatry Internal med. Pathology
FIGURE 2. The breakdown of different specialties that those who left orthopaedic residency entered, from the surveyed cohort.
a minority of programs.29 In addition, increase of women in orthopedic surgery is significantly less than that in all other specialties studied except for neurosurgery and cardiovascular.25 Racial and ethnic diversity is also noted to have experienced slower growth in both orthopedic residents and faculty positions compared to growth in nearly all other specialties, as well as medical school classes.27-28 The importance of educating more minority surgeons is illustrated not just by nondiverse mentorship affecting resident loss, but also by studies showing higher self-reported satisfaction in minority patients when treated by minority physicians,29-31 and others showing minority physicians are up to 4 times more likely to practice in underserved areas.30,32 Our study did not show a statistical significance in the attrition rate of ethnic minorities; however, our sample was not representative of the total population of minority orthopedic residents, as a larger percentage was reported in the ACGME data than in our data set. Our results demonstrated a varied distribution of reasons for residency departure, similar to those seen in other specialties. The majority of orthopedic residents who pursue different specialties do so in nonsurgical fields, as do those who leave other surgical residencies.10,15,16 Several studies of general surgery attrition place lifestyle or an undesired amount of work as the top reason for attrition10,33,34; despite this, we did not find a difference in orthopedic attrition after the 80-hour work week. In seemingly conflicting data, however, having a significant other or children was found to be protective,33,35 possibly because of both the support network as well as this population’s statistically significant higher likelihood of looking forward to work.35 We found these characteristics protective in orthopedic residents too. The biggest weakness of the study is the overall response rate of 54% and bias that could present. However, because our sample included an average program size and sex breakdown nearly identical to the national residency population, we feel the data can be considered in the 4
national context. This response rate is also comparable to other published residency survey studies.15,24 Besides response rate, another weakness is potential recall bias from the surveyed program directors, which could account for the lower rate of attrition in our sample compared to complete national data. We felt that the rarity of losing a resident, given the low attrition rate, was likely to be well remembered by programs, but with an average tenure of 7.3 years for orthopedic residency directors,1 all losses may not have been known. Recall bias may have also affected the reporting of ethnicity, particularly as this is often determined by self-identification and our data was completely from program directors’ responses. Because of the anonymous nature of the data, it is also possible that 1 individual left more than 1 program during residency and was thus counted twice. In addition, the program directors may have overreported the category of voluntary departure, as some residents may have left under technically “voluntary” terms when faced with an alternative of dismissal or additional years of training. In our study population, the orthopedic surgery residents who left training were more likely to be women, single, or without children compared to the overall residency population, and in the first half of their residency training. These groups, and certainly residents early in their training, may benefit from more targeted mentoring. Further research direction might include capturing more specific data directly from those individuals who left about expectations and conditions leading to their departures.
REFERENCES 1. Accreditation Council for Graduate Medical Educa-
tion. Data Resource Book: Academic Year 2013-2014. Chicago: ACGME; 2014. 2. Accreditation Council for Graduate Medical Educa-
tion. Data Resource Book: Academic Year 2012-2013. Chicago: ACGME; 2013. 3. Accreditation Council for Graduate Medical Educa-
tion. Data Resource Book: Academic Year 2011-2012. Chicago: ACGME; 2012. 4. Accreditation Council for Graduate Medical Educa-
tion. Data Resource Book: Academic Year 2010-2011. Chicago: ACGME; 2011. 5. Accreditation Council for Graduate Medical Educa-
tion. Data Resource Book: Academic Year 2009-2010. Chicago: ACGME; 2010. 6. Accreditation Council for Graduate Medical Educa-
tion. Data Resource Book: Academic Year 2008-2009. Chicago: ACGME; 2009. Journal of Surgical Education Volume ]/Number ] ] 2016
7. Accreditation Council for Graduate Medical Educa-
tion. Data Resource Book: Academic Year 2007-2008. Chicago: ACGME; 2008. 8. Yeo H, Bucholz E, Ann Sosa J, et al. A national study of
are doing and why? 2007;459:255-259.
Clin
Orthop
Relat
Res.
21. Dirschl DR, Campion ER, Gilliam K. Resident
attrition in general surgery training: which residents leave and where do they go? Ann Surg. 2010;252(3):529-534.
selection and predictors of performance: can we be evidence based? Clin Orthop Relat Res. 2006;449: 44-49.
9. Burkhart RA, Tholey RM, Guinto D, Yeo CJ,
22. Dirschl DR, Dahners LE, Adams GL, Crouch JH,
Chojnacki KA. Grit: a marker of residents at risk for attrition? Surgery. 2014;155(6):1014-1022.
Wilson FC. Correlating selection criteria with subsequent performance as residents. Clin Orthop Relat Res. 2002;399:265-271.
10. Everett CB, Helmer SD, Osland JS, Smith RS.
General surgery resident attrition and the 80-hour workweek. Am J Surg. 2007;194(6):751-756. 11. Lynch G, Nieto K, Puthenveettil S, et al. Attrition
rates in neurosurgery residency: analysis of 1361 consecutive residents matched from 1990 to 1999. J Neurosurg. 2015;122(2):240-249. 12. Renfrow JJ, Rodriguez A, Liu A, et al. Positive trends
in neurosurgery enrollment and attrition: analysis of the 2000-2009 female neurosurgery resident cohort. J Neurosurg. 2016;124(3):834-839. 13. Kennedy
23. Meals RA, Bassewitz HL, Dorey FJ. Academic lon-
gevity and attrition of full-time orthopaedic faculty members. J Bone Joint Surg Am. 2000;82-A(7). 24. Daniels EW, French K, Murphy LA, Grant RE. Has
diversity increased in orthopaedic residency programs since 1995? Clin Orthop Relat Res. 2012;470(8): 2319-2324. 25. Blakemore LC, Hall JM, Biermann JS. Women in
surgical residency training programs. J Bone Joint Surg Am. 2003;85-A(12):2477-2480.
KA, Brennan MC, Rayburn WF, Brotherton SE. Attrition rates between residents in obstetrics and gynecology and other clinical specialties, 2000-2009. J Grad Med Educ. 2013;5(2):267-271.
26. Biermann JS. Women in orthopaedic surgery residen-
14. Hatton MP, Loewenstein J. Attrition from ophthal-
race, ethnicity, and sex between academic orthopaedic surgery and other specialties: a comparative study. J Bone Joint Surg Am. 2010;92 (13):2328-2335.
mology residency programs. Am J Opthalmol. 2004; 138(5):863-864. 15. Prager JD, Myer CM, Myer CM 3rd. Attrition in
otolaryngology residency. Otolaryngol Head Neck Surg. 2011;145(5):753-754. 16. Walker JL, Janssen H, Hubbard D. Gender differences
in attrition from orthopaedic surgery residency. J Am Med Womens Assoc. 1993;48(6):182-184 [193]. 17. Scherl SA, Lively N, Simon MA. Initial review of
electronic residency application service charts by orthopaedic residency faculty members. Does applicant gender matter? J Bone Joint Surg Am. 2001;83-A (1):65-70. 18. McCord JH, McDonald R, Sippel RS, Leverson G,
Mahvi DM, Weber SM. Surgical career choices: the vital impact of mentoring. J Surg Res. 2009;155(2):136-141. 19. Flint JH, Jahangir AA, Browner BD, Mehta S. The
value of mentorship in orthopaedic surgery resident education: the residents’ perspective. J Bone Joint Surg Am. 2009;91(4):1017-1022.
cies in the United States. Acad Med. 1998;73(6): 708-709. 27. Day CS, Lage DE, Ahn CS. Diversity based on
28. Templeton K, Wood VJ, Haynes R. Women and
minorities in orthopaedic residency programs. J Am Acad Orthop Surg. 2007;15(suppl 1):S37-S41. 29. VanHeest AE, Agel J. The uneven distribution of
women in orthopaedic surgery resident training programs in the United States. J Bone Joint Surg Am. 2012;94(2):e9. 30. Cohen JJ, Gabriel BA, Terrell C. The case for diversity
in the health care workforce. Health Aff (Millwood). 2002;21(5):90-102. 31. Reeded JY. A recurring theme: the need for minority
physicians. Health Aff (Millwood). 2003;22(4):91-93. 32. White AA. Resident selection: are we putting the cart
before the horse? Clin orthop Relat Res. 2002;399: 255-259. 33. Gifford E, Galante J, Kaji AH, et al. Factors associated
Schnall SB, Patzakis MJ. Resident selection: how we
with general surgery residents’ desire to leave residency programs: a multi-institutional study. J Am Med Assoc Surg. 2014;149(9):948-953.
Journal of Surgical Education Volume ]/Number ] ] 2016
5
20. Thordarson DB, Ebramzadeh E, Sangiorgio SN,
34. Morris JB, Leibrandt TJ, Rhodes RS. Voluntary
35. Sullivan MC, Yeo H, Roman SA, Bell RH Jr., Sosa JA.
changes in surgery career paths: a survey of the program directors in surgery. J Am Coll Surg. 2003; 196(4):611-616.
Striving for work-life balance: effect of marriage and children on the experience of 4402 US general surgery residents. Ann Surg. 2013;257(3):571-576.
6
Journal of Surgical Education Volume ]/Number ] ] 2016
National Orthopedic Residency Attrition: Who Is At Risk? Jennifer M. Bauer, MD, and Ginger E. Holt, MD Vanderbilt University Medical Center, Nashville, Tennessee BACKGROUND: National U.S. orthopedic resident attri-
CONCLUSIONS: Orthopedic residents who are female,
tion rates have been historically low, but no literature exists as to the characteristics of those who leave nor the circumstance of the departure. We aimed to determine factors that may place a resident at higher risk for attrition. Additionally, we planned to determine whether the 2003-work hour restriction affected attrition rate.
single, or without children are statistically more likely to undergo attrition. Consideration could be given to targeted C mentoring of these resident groups. ( J Surg Ed ]:]]]-]]]. J 2016 Association of Program Directors in Surgery. Published by Elsevier Inc. All rights reserved.)
MATERIALS AND METHODS: All orthopedic surgery
orthopedic residency
residency program directors in the United States were surveyed on demographic data for their current resident class, the number of residents who left the program, as well as demographic description for each of the residents who left their program from 1998 to 2013. Exclusion criteria included military programs and those younger than 3 years. All data were deidentified and compared to the Accreditation Council for Graduate Medical Education Data Resource book to protect against sample error in respondents. RESULTS: Of 146 programs included, the overall response rate was 54% of residency directors, representing 51% of orthopedic residents. The respondent demographic makeup of 13.7% female, and average program size of 22.3 residents, compared similarly to the Accreditation Council for Graduate Medical Education national average of 13% female and 23-resident program size. Compared to all respondents, residents who left their program were more likely to be female (27%, p ¼ 0.0018), single (51%, p ¼ 0.0028), and without children (80%, p ¼ 0.0018). There was no statistical difference based on minority status or 2003-instituted work hour restriction. Of those who left, 45% transferred to another specialty, 34% were dismissed, 14% voluntarily withdrew or cited personal reasons, and 6% transferred to another orthopedic program. The most common specialties to transfer into were radiology (30%), emergency medicine (25%), and anesthesia (18%).
Correspondence: Inquiries to Jennifer M. Bauer, MD, Vanderbilt University Medical Center, 1215 21st Ave S, Medical Center East, South Tower, Suite 4200, Nashville, TN 37232-8774. fax: +615 875 1915; e-mail: [email protected]
KEY WORDS: resident attrition, resident education, COMPETENCIES:
Professionalism, Communication Skills
Interpersonal
and
INTRODUCTION Losing residents from U.S. orthopedic surgery residencies is not common, with an average annual loss in the last 7 years of 0.96% of residents, as reported by the Accreditation Council for Graduate Medical Education (ACGME).1-7 Of those who do not complete residency each year, 16% do so for reasons of dismissal or lack of success, whereas 84% are due to transfer and withdrawal that can be considered voluntary attrition. Although these numbers are low because the average residency class size is 4.5 residents and much time and diligence is spent by programs to pick and train a residency class, the loss of one resident can weigh heavily on a class and a program. Although there is currently little information on who leaves an orthopedic program, other specialties have been investigated. Abundant literature exists examining the annual rates and reasons for general surgery attrition, cited as high as 20% by some,8,9 with at least one study noting an increase in the rate despite 80-hour work restrictions.10 Other specialties’ annual loss of residents includes 14% per year in neurosurgery, with a statistically significant increased rate for women11-12; 4.2% in obstetrics and gynecology, with a higher rate in men13; 1.1% in ophthalmology that was equal between both the sexes14; and 1.2% in otolaryngology.15 This data from peer reviewed literature does not always match that reported by the ACGME. There are no recent studies that examine the rate and characteristics of orthopedic resident attrition. The purpose
Journal of Surgical Education & 2016 Association of Program Directors in Surgery. Published by 1931-7204/$30.00 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jsurg.2016.03.010
1
TABLE 1. Survey Questions Basic Questions for All Programs How How How How
many many many many
Questions for Each Resident Who Left
residents are currently in your program? current residents are female? current residents identify as non-Caucasian? current residents are single? (or unknown)
How many current residents have children? (or unknown) How many current residents left your program between July 1998 and 2003? How many current residents left your program between July 2003 and 2013?
of our study was to determine if statistically significant characteristics are shared among orthopedic surgery residents who leave their programs, the reason residents leave, and if the rate of attrition has been affected by the 2003 80-hour work restriction.
MATERIALS AND METHODS We compiled a comprehensive list of U.S. orthopedic surgery residency programs from the American Medical Association’s online data system. Using this, individual program websites, and Council of Orthopaedic Residency Directors directories, we gathered e-mail addresses for each resident program director. There are 157 individual orthopedic programs currently active. Programs were excluded from the data collection if they were noncivilian programs, because of the complexity of commitment to military programs. Those started within the last 3 years were also excluded. We included a total of 146 programs. An electronic survey was created on Research Electronic Data Capture (REDCap) with 7 basic questions and a set
Residency class size for resident's year at departure? Resident's sex? What race/ethnicity did the resident identify as? Was the resident married/have a partner/comparable significant other? Did the resident have children? At what point in the resident's training did he/she leave? To your understanding, why did he/she leave? Dismissal Left medicine completely Transfer to different specialty—if so which Health/Hardship Unknown Other—please explain further
of further detailed questions for each resident who left (Table 1). Residency directors were asked about the basic demographics of their current residency class; the number of residents who left in the previous 5 years of 2003-work hour restrictions and in the 10 years after; and specifics about those who left, including demographics, reasons for leaving, and what the resident pursued after leaving. There was an option for additional questions or comments at the end. To collect participants, the REDCap survey that collects deidentified data was sent twice to nonrespondents, followed by 4 separate rounds of individualized emails with both the REDCap link and an option to fill a spreadsheet with the same questions, with that data then deidentified on receipt. To ensure the data we captured did not represent significant sampling error, the demographics of our sample were compared against published ACGME data averaged from the last 7 years.1-7 Data was analyzed using 2-tailed Fisher exact test for categorical variables, or Student t-test for continuous variable means with 95% confidence intervals included. Responses that were left blank or answered as “unknown” were factored out of the calculations for averages and percentages.
TABLE 2. Demographics, U.S. Orthopaedic Surgery Residents Total Survey Respondents Class of 2013 to 2014
ACGME Class of 2013 to Significance 2014
Total residents Average program size
1763 22.3
3566 23
Percent female Percent Ethnic or racial minority Percent without significant other Percent without children
13.7 20.3
13 35.3
40.0
—
30.8
—
2
p ¼ 0.5 95% CI: 2.65 to 1.29 p ¼ 0.5 p o 0.001
Journal of Surgical Education Volume ]/Number ] ] 2016
TABLE 3. Characteristics of Residents Who Left Orthopaedic Residency Study Cohort Who Left Residency
Total Survey Respondents Class of 2013 to 2014
Average program size
23.0
22.3
Percent female Percent ethnic or racial minority Percent without significant other Percent without children
27.0 22.6
13.7 20.3
p ¼ 0.5 95% CI: 2.82 to 1.37 p o 0.005 p ¼ 0.65
51.3
40.0
p o 0.005
80.0
69.2
p o 0.005
Significance
RESULTS
CONCLUSIONS
We received 79 survey responses from 146 orthopedic residency directors, accounting for 54% of programs and 51% of orthopedic residents. The data for all responding programs showed an average class size of 22.3 residents, and a demographic make-up of 13.7% female, 20.3% ethnic minority, 40% without a significant other, and 30.8% without children. Compared to the ACGME demographic categories available (Table 2), our sample was an appropriate representation of class size and sex distribution but significantly different with regard to ethnicity. Of the 79 programs who responded, 57 programs had lost at least 1 resident since 1998. There was a 0.58% rate of loss of residents per year for 1998 to 2003, and 0.59% for 2003 to 2013; there was no demonstrated difference from before the 80-hour work week restrictions to after. These rates of loss reflect a statistically significant difference from the ACGME reported average of 0.96% over the last 7 years (p o 0.0015, 95% CI: 0.21-0.55). The demographic characteristics of those who left compared to the overall sample group is shown in Table 3. There was a statistically significant increase seen for females, those without a significant other, and those without children. There was no difference in rate of attrition based on ethnicity or class size that the resident came from. The most common year of training to leave from was second year (44.3%), followed by intern year (25.4%) and third year (23%). Only 30% of those who left did so after second year. Overall, 54% were offered remediation prior to departure. Those who left did so for a variety of reasons (Fig. 1), including 34% due to involuntary dismissal. There was no significant difference between males and females with respect to involuntary departure. The remainder left for the voluntary reasons of changing specialties (44%), withdrawal/personal reasons (14%), or transferring orthopedic programs (5%). Those who left orthopedics for a different specialty entered 1 of 10 various fields (Fig. 2), with radiology, emergency medicine, and anesthesia capturing 72% overall.
This is the most comprehensive study to survey orthopedic residency directors on the characteristics of and reasons leading to resident attrition. An earlier survey drawing on the classes of 1980 to 1987 did show a significantly higher rate of females withdrawing, but did not include further characteristics besides sex nor detailed causes for those who left.16 This study again finds significantly more women leaving. Several studies point out the low rate of women and minorities in orthopedic residency. Because no sex bias has been shown in initial review of residency applications,17 it might be inferred that the disparity is indicative of a low number of female applicants. Both attracting females and retaining them may be tied to the importance of having faculty role models and mentors.18,19 Lack of sex difference in our results for dismissal, as well as studies showing residency performance correlating to no clear characteristics besides the OITE and ABOS score associations,20-22 point away from female performance as a culprit. Although several laud the increased percentage of women in both residency and academic faculty positions,23,24 many are quick to point out that this raw growth mirrors only that of women into medical school itself and therefore represents no actual proportional growth of diversity.25-28 Moreover, what growth that does exist may be concentrated within only
Journal of Surgical Education Volume ]/Number ] ] 2016
3
Reasons Given for Leaving Residency 5%
4%
7% 43%
7%
Transferred Specialty Dismissal Withdrew Health/personal Transferred Ortho Unknown
34% FIGURE 1. The breakdown of reasons for leaving orthopaedic residency, from the survey cohort.
Other Residency Types Pursued 4%
2% 2% 2% Radiology
5%
30%
7%
Emergency Anesthesia
7%
General surgery Family med. PM&R
18%
Neurosurgery
25%
Pscyhiatry Internal med. Pathology
FIGURE 2. The breakdown of different specialties that those who left orthopaedic residency entered, from the surveyed cohort.
a minority of programs.29 In addition, increase of women in orthopedic surgery is significantly less than that in all other specialties studied except for neurosurgery and cardiovascular.25 Racial and ethnic diversity is also noted to have experienced slower growth in both orthopedic residents and faculty positions compared to growth in nearly all other specialties, as well as medical school classes.27-28 The importance of educating more minority surgeons is illustrated not just by nondiverse mentorship affecting resident loss, but also by studies showing higher self-reported satisfaction in minority patients when treated by minority physicians,29-31 and others showing minority physicians are up to 4 times more likely to practice in underserved areas.30,32 Our study did not show a statistical significance in the attrition rate of ethnic minorities; however, our sample was not representative of the total population of minority orthopedic residents, as a larger percentage was reported in the ACGME data than in our data set. Our results demonstrated a varied distribution of reasons for residency departure, similar to those seen in other specialties. The majority of orthopedic residents who pursue different specialties do so in nonsurgical fields, as do those who leave other surgical residencies.10,15,16 Several studies of general surgery attrition place lifestyle or an undesired amount of work as the top reason for attrition10,33,34; despite this, we did not find a difference in orthopedic attrition after the 80-hour work week. In seemingly conflicting data, however, having a significant other or children was found to be protective,33,35 possibly because of both the support network as well as this population’s statistically significant higher likelihood of looking forward to work.35 We found these characteristics protective in orthopedic residents too. The biggest weakness of the study is the overall response rate of 54% and bias that could present. However, because our sample included an average program size and sex breakdown nearly identical to the national residency population, we feel the data can be considered in the 4
national context. This response rate is also comparable to other published residency survey studies.15,24 Besides response rate, another weakness is potential recall bias from the surveyed program directors, which could account for the lower rate of attrition in our sample compared to complete national data. We felt that the rarity of losing a resident, given the low attrition rate, was likely to be well remembered by programs, but with an average tenure of 7.3 years for orthopedic residency directors,1 all losses may not have been known. Recall bias may have also affected the reporting of ethnicity, particularly as this is often determined by self-identification and our data was completely from program directors’ responses. Because of the anonymous nature of the data, it is also possible that 1 individual left more than 1 program during residency and was thus counted twice. In addition, the program directors may have overreported the category of voluntary departure, as some residents may have left under technically “voluntary” terms when faced with an alternative of dismissal or additional years of training. In our study population, the orthopedic surgery residents who left training were more likely to be women, single, or without children compared to the overall residency population, and in the first half of their residency training. These groups, and certainly residents early in their training, may benefit from more targeted mentoring. Further research direction might include capturing more specific data directly from those individuals who left about expectations and conditions leading to their departures.
REFERENCES 1. Accreditation Council for Graduate Medical Educa-
tion. Data Resource Book: Academic Year 2013-2014. Chicago: ACGME; 2014. 2. Accreditation Council for Graduate Medical Educa-
tion. Data Resource Book: Academic Year 2012-2013. Chicago: ACGME; 2013. 3. Accreditation Council for Graduate Medical Educa-
tion. Data Resource Book: Academic Year 2011-2012. Chicago: ACGME; 2012. 4. Accreditation Council for Graduate Medical Educa-
tion. Data Resource Book: Academic Year 2010-2011. Chicago: ACGME; 2011. 5. Accreditation Council for Graduate Medical Educa-
tion. Data Resource Book: Academic Year 2009-2010. Chicago: ACGME; 2010. 6. Accreditation Council for Graduate Medical Educa-
tion. Data Resource Book: Academic Year 2008-2009. Chicago: ACGME; 2009. Journal of Surgical Education Volume ]/Number ] ] 2016
7. Accreditation Council for Graduate Medical Educa-
tion. Data Resource Book: Academic Year 2007-2008. Chicago: ACGME; 2008. 8. Yeo H, Bucholz E, Ann Sosa J, et al. A national study of
are doing and why? 2007;459:255-259.
Clin
Orthop
Relat
Res.
21. Dirschl DR, Campion ER, Gilliam K. Resident
attrition in general surgery training: which residents leave and where do they go? Ann Surg. 2010;252(3):529-534.
selection and predictors of performance: can we be evidence based? Clin Orthop Relat Res. 2006;449: 44-49.
9. Burkhart RA, Tholey RM, Guinto D, Yeo CJ,
22. Dirschl DR, Dahners LE, Adams GL, Crouch JH,
Chojnacki KA. Grit: a marker of residents at risk for attrition? Surgery. 2014;155(6):1014-1022.
Wilson FC. Correlating selection criteria with subsequent performance as residents. Clin Orthop Relat Res. 2002;399:265-271.
10. Everett CB, Helmer SD, Osland JS, Smith RS.
General surgery resident attrition and the 80-hour workweek. Am J Surg. 2007;194(6):751-756. 11. Lynch G, Nieto K, Puthenveettil S, et al. Attrition
rates in neurosurgery residency: analysis of 1361 consecutive residents matched from 1990 to 1999. J Neurosurg. 2015;122(2):240-249. 12. Renfrow JJ, Rodriguez A, Liu A, et al. Positive trends
in neurosurgery enrollment and attrition: analysis of the 2000-2009 female neurosurgery resident cohort. J Neurosurg. 2016;124(3):834-839. 13. Kennedy
23. Meals RA, Bassewitz HL, Dorey FJ. Academic lon-
gevity and attrition of full-time orthopaedic faculty members. J Bone Joint Surg Am. 2000;82-A(7). 24. Daniels EW, French K, Murphy LA, Grant RE. Has
diversity increased in orthopaedic residency programs since 1995? Clin Orthop Relat Res. 2012;470(8): 2319-2324. 25. Blakemore LC, Hall JM, Biermann JS. Women in
surgical residency training programs. J Bone Joint Surg Am. 2003;85-A(12):2477-2480.
KA, Brennan MC, Rayburn WF, Brotherton SE. Attrition rates between residents in obstetrics and gynecology and other clinical specialties, 2000-2009. J Grad Med Educ. 2013;5(2):267-271.
26. Biermann JS. Women in orthopaedic surgery residen-
14. Hatton MP, Loewenstein J. Attrition from ophthal-
race, ethnicity, and sex between academic orthopaedic surgery and other specialties: a comparative study. J Bone Joint Surg Am. 2010;92 (13):2328-2335.
mology residency programs. Am J Opthalmol. 2004; 138(5):863-864. 15. Prager JD, Myer CM, Myer CM 3rd. Attrition in
otolaryngology residency. Otolaryngol Head Neck Surg. 2011;145(5):753-754. 16. Walker JL, Janssen H, Hubbard D. Gender differences
in attrition from orthopaedic surgery residency. J Am Med Womens Assoc. 1993;48(6):182-184 [193]. 17. Scherl SA, Lively N, Simon MA. Initial review of
electronic residency application service charts by orthopaedic residency faculty members. Does applicant gender matter? J Bone Joint Surg Am. 2001;83-A (1):65-70. 18. McCord JH, McDonald R, Sippel RS, Leverson G,
Mahvi DM, Weber SM. Surgical career choices: the vital impact of mentoring. J Surg Res. 2009;155(2):136-141. 19. Flint JH, Jahangir AA, Browner BD, Mehta S. The
value of mentorship in orthopaedic surgery resident education: the residents’ perspective. J Bone Joint Surg Am. 2009;91(4):1017-1022.
cies in the United States. Acad Med. 1998;73(6): 708-709. 27. Day CS, Lage DE, Ahn CS. Diversity based on
28. Templeton K, Wood VJ, Haynes R. Women and
minorities in orthopaedic residency programs. J Am Acad Orthop Surg. 2007;15(suppl 1):S37-S41. 29. VanHeest AE, Agel J. The uneven distribution of
women in orthopaedic surgery resident training programs in the United States. J Bone Joint Surg Am. 2012;94(2):e9. 30. Cohen JJ, Gabriel BA, Terrell C. The case for diversity
in the health care workforce. Health Aff (Millwood). 2002;21(5):90-102. 31. Reeded JY. A recurring theme: the need for minority
physicians. Health Aff (Millwood). 2003;22(4):91-93. 32. White AA. Resident selection: are we putting the cart
before the horse? Clin orthop Relat Res. 2002;399: 255-259. 33. Gifford E, Galante J, Kaji AH, et al. Factors associated
Schnall SB, Patzakis MJ. Resident selection: how we
with general surgery residents’ desire to leave residency programs: a multi-institutional study. J Am Med Assoc Surg. 2014;149(9):948-953.
Journal of Surgical Education Volume ]/Number ] ] 2016
5
20. Thordarson DB, Ebramzadeh E, Sangiorgio SN,
34. Morris JB, Leibrandt TJ, Rhodes RS. Voluntary
35. Sullivan MC, Yeo H, Roman SA, Bell RH Jr., Sosa JA.
changes in surgery career paths: a survey of the program directors in surgery. J Am Coll Surg. 2003; 196(4):611-616.
Striving for work-life balance: effect of marriage and children on the experience of 4402 US general surgery residents. Ann Surg. 2013;257(3):571-576.
6
Journal of Surgical Education Volume ]/Number ] ] 2016