Applicant Factors for Matching Into an Orthopaedic Foot and Ankle Fellowship

Abstract

Background:

Over the past 2 decades, various initiatives have aimed to enhance diversity in orthopaedic surgery, promoting greater racial, ethnic, and gender equity. Building on this progress, demographic data on orthopaedic fellowship matches has been collected over the last 3 years. This study seeks to analyze trends in applicants to foot and ankle fellowships, characterize the applicant pool, and identify traits associated with successful matches.

Methods:

All applicant information from a fellowship application service site was obtained for match years 2022-2024. Applicants were selected only if they applied to the specific subspecialty fellowship. Gender, race, and ethnicity were recorded. Applicant factors such as Alpha Omega Alpha (AOA) status, Gold Humanitarian status, United States Medical Licensing Examination (USMLE) Step 2 score, number of applications, and number of interview invitations were used. Applicant medical school status, including allopathic, osteopathic, Canadian, and foreign medical graduate (FMG) were analyzed. χ² test was performed between US and FMG applicants. Univariate and multivariate binomial logistic regression was performed for FMGs.

Results:

There were 286 applicants, 82.8% males, 16.5% females, 133 US- and Canadian-trained graduates, and 153 FMGs. The match rate for US- and Canadian-trained graduates was 99.2% compared with FMGs, which was 43.7% and associated with lower matching rates (P < .00001). When performing analysis in US and FMG groups independently because of multicollinearity, no factors could be associated with matching. Only when the applicant had FMG status, then the number of interview invitations were associated with matching. When the number of interviews approached 6, the likelihood of matching was >95%.

All except 1 US F&A applicant matched into an F&A fellowship. During the match period, US-trained applicants were 28%-32% female, 4%-8% Black/African American, 8%-17% Asian, 65%-73% White, 2%-4% American Indian, and 2%-8% Hispanic, with no Native Hawaiian and Pacific Islanders applying. Female applicants were above representation compared to Accreditation Council for Graduate Medical Education (ACGME) numbers, but the remaining race and ethnicity applicants were within the range of current ACGME standings, which is still lower than US Census results.

Conclusion:

Nearly all US-trained foot and ankle applicants matched, whereas FMG applicants matched 43% of the time.

Visual Abstract.

This is a visual representation of the abstract.

Introduction

Orthopaedic surgery has seen an increase in specialization primarily through postresidency fellowship training. In recent years, acquiring a fellowship has become more frequent, with approximately 90% of all orthopaedic residents pursuing a fellowship position. ⁴ The number of foot and ankle applicants has been slowly increasing over the last 3 years as well.

Similarly, orthopaedic surgery has been undergoing changes within the field, strides to become a more inclusive and diverse field, supported by pipeline initiatives and efforts to address barriers to entry and advancement for underrepresented groups. Since the recent reporting of orthopaedic fellowship applicant demographic data, we sought to analyze trends in applicants to foot and ankle fellowships, characterize the applicant pool, and identify traits associated with successful matches. Our hypothesis was that based on these findings, residents interested in a foot and ankle fellowship can gain insights into criteria and attributes that may increase their likelihood of matching into these programs.

Methods

Orthopaedic fellowship applicant data from 2022 to 2024 were obtained from the San Francisco Match (San Francisco, CA), a US-based residency and fellowship matching service currently used by all orthopaedic fellowships (except hand fellowship). The subspecialties included were shoulder and elbow, trauma, adult reconstruction, pediatrics, foot and ankle, sports, spine, and oncology. Applicants who did not submit a rank list were excluded from the data analysis. If an applicant applied to more than 1 subspecialty, they were considered an applicant for all subspecialty matches applied to. SF Match did not collect data on race and ethnicity prior to 2021 as this was not a requirement. The American Academy of Orthopaedic Surgeons (AAOS) Board of Specialty Societies was petitioned to start collection and reporting of these data beginning in the 2022 Fellowship Match.

Demographics such as gender, race, and ethnicity were reported. Gender options reported were male, female or do not wish to disclose. Race options reported were African American/Black, Asian, Caucasian, American Indian/Alaskan Native, Native Hawaiian/Pacific Islander, and ethnicity options were Hispanic or non-Hispanic with options to not wish to disclose for both race and ethnicity. This led to 10 unique race and ethnicity categories. We grouped all Hispanic groups together to reduce low sample size errors.

Applicant variables recorded included Alpha Omega Alpha status, Gold Humanitarian status, applicant medical school status, United States Medical Licensing Examination (USMLE) Step 2 score, number of fellowship applications applied to, number of interview invitations, and matched or unmatched status. USMLE Step 1 score was excluded because of missing data from more than 70% of the applicants.

Statistical Analysis

Data were analyzed on all applicants who applied to the selected fellowship(s). The outcome variable “successful match” was defined as binary, where 0 = not a successful match and 1 = successful match. Binary logistic regression was performed to evaluate the relationship between the outcome variable and all predictors based on matched versus unmatched applicants. Multivariate logistic regression was then conducted using statistically significant predictors identified in the univariate analysis. The model fit was assessed demonstrated significant multicollinearity because of 99.2% of US-trained applicants (US allopathic, US osteopathic, Canadian) matching and 43.7% of foreign medical graduates (FMGs) matching. Thus, we performed χ² test of association. FMG binomial logistic regression was performed for matching and nonmatching candidates.

The probability of achieving a successful match was plotted as a function of the number of interview invitations, along with 95% CIs. During interviewing, many applicants use the number of interviews they receive to rate their likelihood of matching into a fellowship. We sought to plot the probability of a successful matching with the number of interview invitations. Significance was set at P <.05. Statistical analysis was performed using Jamovi v2.3.21.0 (Syndey, Australia).

Results

There were 286 applicants, 82.8% male, 16.5% female, 133 US- and Canadian-trained graduates, and 153 FMGs (Table 1). The match rate for US- and Canadian-trained graduates was 99.2% compared with FMGs, which was 43.7%, and associated with lower matching rates (P < .00001). When performing analysis in US and FMG groups independently because of multicollinearity, no factors could be associated with matching. Only when the applicant had FMG status, then the number of interview invitations were associated with matching. When the number of interviews approached 6, the likelihood of matching was >95% (Figure 1 and Table 2).

Table 1.

Applicant Gender, Race, and Ethnicity Demographics of the 2022-2024 Orthopaedic Foot and Ankle Fellowship Match.

	2022-2024 Match
Category	Total	% total applicants	US-trained, n (%)	FMG, n (%)
Male	237	82.8	93 (70)	146 (95.4)
Female	47	16.4	40 (30)	7 (4.5)
Black / African American	11	3.8	8 (6)	3 (2)
Asian	62	21.6	18 (13.5)	44 (33)
White	158	55.2	92 (69)	66 (50)
American Indian / Alaskan Native (non-Hispanic)	0	0	0	0
Native Hawaiian/Pacific Islander	0	0	0	0
Multiracial (non-Hispanic)	11	3.8	4 (3)	7 (5)
Hispanic, Black / African American	1	0.3	7 (5.2)	7 (5)
Hispanic, Asian	2	0.3
Hispanic, White	7	2.4
Hispanic, American Indian / Alaskan Native	0	0
Hispanic, Native Hawaiian / Pacific Islander	0	0
Hispanic, multiracial	5	1.7
Did not disclose	28	9.7
Allopathic	113	39.5
Osteopathic	18	6.3
Canadian	2	0.6
FMG	153	53.4
Total	286

Abbreviation: FMG, foreign medical graduate.

Figure 1.

Binomial logistic regression of matched versus unmatched foreign medical graduate applicants. Number of interviews was the only predictor of likelihood of matching with odds ratio (OR) 1.85 [95% CI 1.3-2.6]. The likelihood of matching approached 95% with 6 interview invitations.

Table 2.

Binomial Logistic Regression of Matched vs Unmatched Foreign Medical Graduate Applicants.

Model Fit Measures a
Model		Deviance					AIC	R 2 McF
1		121					127	0.131
Omnibus Likelihood Ratio
Predictor				χ2			df	P
Step 2				2.59			1	.107
No. of interviews				16.29			1	<.001
Model Coefficients: Match Status b
		95% CI						95% CI
Predictor	Estimate	Lower	Upper	SE	Z	P	Odds ratio	Lower	Upper
Intercept	4.2202	−2.5189	10.95932	3.4384	1.23	.220	68.048	0.0805	57487.24
Step 2	−0.0236	−0.0530	0.00577	0.0150	−1.58	.115	0.977	0.9483	1.01
No. of interviews	0.6175	0.2624	0.97251	0.1812	3.41	<.001	1.854	1.3000	2.64

^aModels estimated using sample size of N = 103.

^bEstimates represent the log odds of “Match status = Matched” vs “Match status = No Match”.

All except 1 US applicant matched into a foot and ankle fellowship. During the match period, US-trained applicants were 28%-32% female, 4%-8% Black/African American, 8%-17% Asian, 65%-73% White, 2%-4% American Indian, and 2%-8% Hispanic, with no Native Hawaiian and Pacific Islanders applying. Female applicants were above representation compared to Accreditation Council for Graduate Medical Education (ACGME) numbers, but the remaining race/ethnicity applicants were within the range of current ACGME standings, which is still lower than US Census results.

Analysis was additionally performed on those who applied to a foot and ankle fellowship and did not match. The average number of FMG applications who matched was higher, 67.2 vs 49.2 (P = .013); however, the average USMLE Step 2 score was no different, 232 vs 228 (P = .167). Of the 67 FMGs who matched, 20 matched into foot and ankle fellowships whereas others matched into different subspecialties. For US-trained applicants, they had a lower number of applications compared with FMGs, 25.4 vs 57.1 applications (P < .00001).

Discussion

The results of this study show that no significant association could be made between US-trained foot and ankle fellowship applicants matching into a fellowship of their choice because of the high rate of matching. This suggests that the foot and ankle fellowship application cycle has remained relatively stable over the past 7 years, which is shown by the recent match statistics. ²

Despite the high match rate, the demographics of US applicants to foot and ankle fellowships still have underrepresented groups when compared to US census data. However, diversification in the field of orthopaedics through programs like Nth Dimension, The Perry Initiative, the Ruth Jackson Orthopaedic Society, and The J. Robert Gladden Orthopaedic Society have made strides in increasing the exposure and recruitment of women and minorities. As more orthopaedic residents from diverse backgrounds enter the field, it is likely that there will be increased diversity in fellowship applicants as well. As such, there was a moderate positive Pearson correlation coefficient found for both the number of women and Asian applicants into foot and ankle fellowships over the years 2011 to 2022 using ACGME data. ⁶ Future studies could use recent SF Match data to see if there are upward trends of other ethnicities.

There has been a general increase in the number of female orthopaedic fellowship applications from 2010 to 2021. ⁷ However, the percentage of female applicants has remained relatively stable the past few years. Our data align with findings by Naclerio et al, ⁷ which showed that from 2011 to 2021 the percentage of female applicants in foot and ankle were between 10% and 26% based on data from the SF match. This consistency from our data suggests that this trend has not changed significantly in the most recent years.

Joshi et al ⁶ conducted a comparison of diversity in orthopaedic surgery residents vs foot and ankle fellows from 2007 to 2022. Their findings showed that while the diversity in residents is on an upward trend, foot and ankle fellows still lagged behind. Several proposed mechanisms were due to the lag from a long residency or the fact that foot and ankle is relatively new compared with other orthopaedic fellowships. Their analysis used the ACGME Data Resource Book from 2007 to 2022 as the means to gather data. Overall, their percentages of race and ethnicity for foot and ankle fellows differed from our results as they had a larger percentage of unknown race and ethnicity reported. They pointed out that Black/African American and Hispanics were only present in 2 of the years they analyzed. Because our study used recent SF Match data that include non-ACGME programs, that may have increased our reporting of gender, race, and ethnicity statistics. Future studies should use SF Match data to accurately report candidate demographics in foot and ankle fellowships.

When comparing the most recent ACGME Data Resource Book for 2023-2024 data, there were no Black/African American, Hispanic, American Indian or Alaskan Native, nor native Hawaiian/Pacific Islander applicants, whereas SF match reported low but present numbers. ¹ One reason for this discrepancy could be that foot and ankle fellowships have the least amount of ACGME accredited programs.^5,8 This could lead to differences in the data when looking at the SF match and ACGME Data Resource Book. This could suggest SF match has more accuracy regarding gender, race, and ethnicity, because of a large amount of foot and ankle fellowships that are non-ACGME accredited.

The number of FMG applications has gradually increased from 46% to 54% of foot and ankle applications over the past 3 years. However, less than half of the FMGs matched into a foot and ankle fellowship. This may be due to inability of other fellowships to accept international graduates, because of requirements of medical state licensure or citizenship status. Although we were unable to determine which specific programs FMGs applied to, it may be because only certain programs may be able to accept FMGs because of local hospital rules or state medical licensing laws. Some foot and ankle programs require an H-1B visa status or green card, whereas other programs may evaluate on a case-by-case basis. There are also programs that only accept applicants who graduated from a US residency. ³ These factors could contribute to why FMGs are not matching into foot and ankle despite making up the majority of applicants. Future research could focus on other factors like funding, program preferences, and even visa issues to see why FMGs are being turned away.

Overall, our study found that there were no significant factors that played a role in successfully matching into a foot and ankle fellowship for US applicants. Comparing our data to other studies using ACGME suggests that SF Match data may be more accurate in regard to ethnicity and race. For FMGs, the number of total interviews played a role in successfully matching. Other factors like published research, presentations at national conferences, and visa status could be investigated to see how much they affect matching for FMGs into foot and ankle.