Residency Selection Preferences and Orthopaedic Career Perceptions: A Notable Mismatch

Abstract

Background: Despite near-equal enrollment of males and females in medical schools, orthopaedic surgery continues to have one of the lowest percentage of female orthopaedic residents. This suggests there may be factors that specifically influence females to select other specialties. Some of these possible reasons have been explored in other studies; however, in this study, we sought to identify latent or unobserved variables that may be influencing this difference by conducting an explanatory factor analysis of male and female residency preferences.

Purposes/questions: In this study, by surveying a cohort of medical students at a single institution, we asked, is there a difference between males and females (1) in their perception of orthopaedic surgery and (2) in their preferences for residency and practice? We further asked, if there are differences, (3) is there a correlation between perception and preferences for residency and practice?

Methods: A 46-question survey was sent to all current medical students (n = 628) at a major urban university with near-equal enrollment of males (55%, 345 of 628) and females (45%, 283 of 628) from September 2017 to November 2017. The survey consisted of two main parts: (1) desired attributes of a residency program and (2) perceptions of orthopaedic residency and practice. The design of the survey instruments closely followed The Checklist for Reporting Results of Internet E-Surveys (CHERRIES) checklist and went through several variations and pilot studies before release. In all, 33% (205 of 628) total medical students responded to the email survey, 55% (112 of 205) were male and 45% (93 of 205) were female. The proportion of male and female respondents matched the gender distribution of the total population surveyed, which at the time of the survey was 55% male and 45% female.The data analysis was performed using a Mann-Whitney U test and an explanatory factor analysis. The explanatory factor analysis was used to identify the correlation between survey variables among male and female students. An alpha less than 0.05 was considered significant for the Mann-Whitney U test and a factor greater than 0.5 was considered significant for the factor analysis.

Results: Both male and female students ranked "work-life balance" and "variety in specialty" among the top three most important preferences. Females ranked "range of practice options," higher than males (72% females versus 60% males, r = 0.18; p = 0.009), and males ranked "previous exposure to the specialty" higher than females (65% females versus 71% males, r = 0.03; p = 0.70). Both male and female students had similar overall perceptions of orthopaedic surgery. Both males and females indicated that orthopedics is "male dominated," has "competitive entrance requirements," and requires "long residency work hours." They differed in their perception of "requires physical strength" (60% females versus 38% males, r = 0.28; p < 0.0005), and by how much orthopaedics is "male dominated" (95% females versus 77% males, r = 0.26; p < 0.0005). The factor analysis recognized that, although male and female students do have some similar residency preferences, the influence or weighing of those preferences is different for male and female students. In a manner similar to a personality assessment, the factor analysis produced four latent factors that can help explain variation seen in responses and helped identify influential factors that were not directly tested by the survey. The first such latent factor for females consisted of "work-life balance," "residency length," 'residency work hours," and "family-friendly specialty." Although the first latent factor for males consisted of "prestige," "income potential," "grade or step scores," and "competitiveness of residency program." The three subsequent latent factors also displayed variation in the make-up of the latent factors between males and female students.

Conclusions: This study of medical students at an urban medical school found that male and female students shared many preferences for residency specialties and held many of the same perceptions regarding orthopaedic surgery. The explanatory factor analysis indicated that male and female students weight preferences differently when selecting a specialty; this difference may account for the large differences in proportion between males and females in orthopaedic residency.

Clinical relevance: Attracting talented residents and attending physicians is important for the success of any medical department. Although orthopaedics attracts some of the most talented students, these students are predominantly male. By identifying the multifactorial areas that may be inadvertently discouraging females from applying, orthopaedic residency programs may be able to better address those issues and attract the best talent of both genders.

Fig. 1

This chart shows the distribution of female students’ responses to questions on how they perceive orthopaedics across 14 metrics or variables. We used a five-point Likert scale from 5 = extremely important to 1 = not important. The responses were consolidated into three groups consisting of the two highest, the middle, and the two lowest responses.

Fig. 2

This chart shows the distribution of male students’ responses to questions on how they perceive orthopaedics across 14 metrics or variables. We used a five-point Likert scale from 5 = extremely important to 1 = not important. The responses were consolidated into three groups consisting of the two highest, the middle, and the two lowest responses.

Fig. 3

This chart shows the distribution of the 18 variables female students use or consider important when selecting a specialty. We used a five-point Likert scale from 5 = always present to 1 = not present at all. The responses were consolidated into three groups consisting of the two highest, the middle, and the two lowest responses.

Fig. 4

This chart shows the distribution of the 18 variables male students use or consider important when selecting a specialty. We used a five-point Likert scale from 5 = always present to 1 = not present at all. The responses were consolidated into three groups consisting of the two highest, the middle, and the two lowest responses.

Fig. 5

This is a graphical representation of the explanatory factor analysis. The surveyed variables are grouped into latent factors 1-4. The variance explained by each factor along with the total variance explained is annotated in the lower corner of each latent factor.