Washington University School of Medicine

Digital Commons@Becker
Open Access Publications

2018

Probability of success in the ophthalmology

residency match: Three-year outcomes analysis of
San Francisco Matching Program data
R. Michael Siatkowski
University of Oklahoma

Shahzad I. Mian
University of Michigan-Ann Arbor

Susan M. Culican
Washington University School of Medicine in St. Louis

Laura K. Green
Sinai Hospital of Baltimore

Grace Sun
Weill Cornell Medical College

See next page for additional authors

Follow this and additional works at: https://digitalcommons.wustl.edu/open_access_pubs

Recommended Citation
Siatkowski, R. Michael; Mian, Shahzad I.; Culican, Susan M.; Green, Laura K.; Sun, Grace; Waxman, Evan L.; Wayman, Laura L.;
Stoner, Julie; Chen, Xi; and Feldon, Steven, ,"Probability of success in the ophthalmology residency match: Three-year outcomes
analysis of San Francisco Matching Program data." Journal of Academic Ophthalmology.10,01. e150-e157. (2018).
https://digitalcommons.wustl.edu/open_access_pubs/7701

This Open Access Publication is brought to you for free and open access by Digital Commons@Becker. It has been accepted for inclusion in Open
Access Publications by an authorized administrator of Digital Commons@Becker. For more information, please contact engeszer@wustl.edu.
Authors
R. Michael Siatkowski, Shahzad I. Mian, Susan M. Culican, Laura K. Green, Grace Sun, Evan L. Waxman,
Laura L. Wayman, Julie Stoner, Xi Chen, and Steven Feldon

This open access publication is available at Digital Commons@Becker: https://digitalcommons.wustl.edu/open_access_pubs/7701

 THIEME
e150 Research Article

Probability of Success in the Ophthalmology

Residency Match: Three-Year Outcomes Analysis
of San Francisco Matching Program Data
R. Michael Siatkowski, MD1 Shahzad I. Mian, MD2 Susan M. Cullican, MD, PhD3 Laura K. Green, MD4
Grace Sun, MD5 Evan L. Waxman, MD, PhD6 Laura L. Wayman, MD7 Julie Stoner, PhD8
Xi Chen, MD, MS8 Steven Feldon, MD, MBA9 for the Association of University Professors of Ophthalmology

1 Department of Ophthalmology, Dean McGee Eye Institute, Address for correspondence R. Michael Siatkowski, MD, Department
University of Oklahoma, Oklahoma City, Oklahoma of Ophthalmology, Dean McGee Eye Institute, University of
2 Department of Ophthalmology and Visual Sciences, University of Oklahoma, 608 Stanton L. Young Blvd., OKC, Oklahoma, OK 73104
Michigan Kellogg Eye Center, Ann Arbor, Michigan (e-mail: Rmichael-siatkowski@dmei.org).
3 Department of Ophthalmology and Visual Sciences, Washington
University School of Medicine, St. Louis, Missouri
4 Krieger Eye Institute, Sinai Hospital of Baltimore, Baltimore, Maryland
5 Department of Ophthalmology, Weill Cornell Medical College, New York,
New York
6 Eye Center, University of Pittsburgh, Pittsburgh, Pennsylvania
7 Department of Ophthalmology, Vanderbilt University Medical
Center, Nashville, Tennessee
8 Department of Biostatistics and Epidemiology, University of
Oklahoma Health Sciences Center, Oklahoma City, Oklahoma
9 Department of Ophthalmology, Flaum Eye Institute, University of
Rochester School of Medicine and Dentistry, Rochester, New York

J Acad Ophthalmol 2018;10:e150–e157.

Abstract Objective To develop a probability model of matching into a US ophthalmology

residency program using San Francisco Matching Program (SF Match) data.
Design Retrospective data analysis of de-identified application and matching data.
Participants Registrants for the 2013, 2014, and 2015 ophthalmology residency
matches conducted by the SF Match.
Methods Descriptive statistics of candidates, comparison of continuous and catego-
rical variables between matched and nonmatched candidates, and linear regression
modeling were performed. A recursive partitioning method was used to create a
probability of matching algorithm.
Main Outcome Measures Probability of successfully matching based on quantifiable
candidate characteristics.
Results Over the 3-year period, 1,959 individuals submitted an average of 64 applications
Keywords and received a mean of nine interview invitations. The overall match rate was 71%, with 78%
► ophthalmology matching at one of their top five choices. Successful matches were more likely to occur in US
residency match medical school graduates (78% vs 20%, p < 0.001) and applicants on their first attempt
► success rate (76% vs 29%, p < 0.001). The association between matching and number of programs
► outcomes analysis applied became negative with > 48 applications. Probability of matching was “high”

received DOI https://doi.org/ Copyright © 2018 by Thieme Medical

August 17, 2018 10.1055/s-0038-1673675. Publishers, Inc., 333 Seventh Avenue,
accepted after revision ISSN 2475-4757. New York, NY 10001, USA.
September 3, 2018 Tel: +1(212) 584-4662.
 Probability of Success in the Ophthalmology Residency Match Siatkowski et al. e151

(> 80%) among US graduates with a step 1 United States Medical Licensing Examination
(USMLE) score >243 (regardless of number of programs applied to), a step 1 USMLE score of
231 to 243 who applied to at least 30 programs, and first-time applicants with a step 1 score
>232. No international medical graduates or repeat applicants had a “high” probability of
matching.
Conclusions Although advice must be individualized for each candidate, applicants
for ophthalmology residency who fall into a “high” probability of matching group are
likely to be successful with applications to 45 or fewer programs. Applying to 80 or
more programs should be considered for international medical graduates and/or
applicants who are previously unmatched. Modification of the match application data
form may allow more detailed analysis of variables such as Alpha Omega Alpha or Gold
Humanism Honor Society membership, research activity, and composite evaluation on
a standardized letter of recommendation.

Candidates for ophthalmology residency apply through the with diminishing chances of matching for students who
ophthalmology residency matching program, administered applied to more than 39 programs; similar data for USMLE
by the San Francisco Residency and Fellowship Matching score 237–249 and <237 are 80% and 66%, and 43 and 44
Services (SF Match). A universal application is submitted to programs, respectively. Providing similar data to candidates for
SF Match, which in turn distributes it to the programs ophthalmology residency programs will assist candidates in
specified by the applicant. After completion of the interview effectively allocating effort and investment in the application
process, the programs and the applicants each submit rank process, while at the same time maximizing the successful
lists to SF Match, which uses an algorithm based on prefer- match rate among qualified applicants.
ences submitted to place candidates in a training program. The goal of this study was to develop a model to predict
Ophthalmology continues to be perceived as one of the the probability of matching to an ophthalmology residency
most competitive medical specialties. This is demonstrated program based on applicant characteristics and number of
by the steady increase in the mean number of applications application submissions by analyzing SF Match data.
submitted per applicant over the past 10 years, from 48 in
2008, to 68 in 2017 based on SF Match tracking data. Despite
Methods
this trend, the overall match rate has remained relatively
stable (mean 74%, range 70–78%, for the period 2008–2017). Data Source: Deidentified application and ophthalmology
This continued rise in the number of applications submitted residency matching data for the 2013, 2014 and 2015 match
per applicant places a significant financial burden on stu- cycles collected by the Association of University Professors of
dents as well as a tremendous administrative load on resi- Ophthalmology were used in the study. Independent vari-
dency programs. The cost structure consists of a $100 ables include the number of program applications, the
registration fee and a $60 base fee for application to 10 number of invitations for interview, USMLE step 1 score,
programs, with incremental cost increases of $ 100 to 350 for international medical graduate (IMG) status, Alpha Omega
each additional 10 programs. Thus, the total fee for applying Alpha membership, presence of research activity as stated by
to 40 programs is $610, which increases to $2,010 for 80 the applicant, number of times that the applicant applied for
programs and $2,710 for 100 programs. an ophthalmology residency, number of programs ranked by
Due to the perceived difficulty of the ophthalmology match, the applicant, the total number of programs that ranked the
applicants may feel the need to apply to more programs to gain applicant, the matched position on the program’s rank list,
more interview invitations and increase the likelihood of and the applicant’s rank of the matched program. Data on
matching successfully. However, this assumption may be individual interview invitations offered and interviews com-
based on incomplete or inaccurate information from peers pleted is provided by programs and applicants, respectively,
or advisors, given that the match is a complex process based on through a self-reporting process, which may result in inad-
both quantitative and qualitative characteristics. vertent errors, and is discussed in more detail later.
Although the National Residency Matching Program Statistical Analysis Methods: Descriptive statistics were
(NRMP) provides an annual report detailing match outcomes estimated to summarize the characteristics of the applicants
for other specialties, this information is not available for across the three application cycles. Summaries were pre-
ophthalmology. For example, NRMP data for otolaryngology sented after stratifying by program matching success, first
(found at https://students-residents.aamc.org/applying-resi- versus repeat attempt, and IMG status. An independent
dency/article/apply-smart-data-consider/) demonstrate that sample t-test was used to compare the means of continuous
candidates with United States Medical Licensing Examination variables, and a chi-square test was used to compare
(USMLE) step 1 scores > 250 have a mean match rate of 88% categorical variables between independent groups. The

Journal of Academic Ophthalmology Vol. 10 No. 1/2018

e152 Probability of Success in the Ophthalmology Residency Match Siatkowski et al.

Wilcoxon rank sum test and Fisher’s exact test were used in [SD] ¼ 27, 95% confidence interval [CI]: 62.9–65.3), received
the case of skewed distributions for continuous variables or 9 invitations for interview (SD ¼ 6, 95% CI: 9.0–9.6), and had
small expected frequencies for categorical variables, respec- a score of 238 in USMLE step 1 (SD ¼ 16, 95% CI: 237.6–
tively. Linear and segmented regression modeling1 was used 239.0). Among applicants, 13% were IMGs (95% CI: 11–14%),
to quantify the association between number of applications 88% were first-time submissions (95% CI: 87–90%), and 71%
submitted and the number of interview invitations. SAS 9.4 accomplished a successful match (95% CI: 69–73%). For
was used to perform the descriptive summary and regression applicants who submitted multiple applications (n ¼ 226,
modeling (SAS Institute, Cary, NC). 12%), the most recent attempt was retained in the study. The
To provide an intuitive model to predict the probability of most recent attempt was selected to reflect the overall status
matching, a recursive partitioning method was used in the of each applicant during the study time period (2013–2015)
multivariable analysis.2 The recursive partitioning analysis and was necessary to ensure independent applicants, avoid-
used USMLE score and number of program applications as ing correlated measures, in the analysis set. The approach
the two predictors of matching status. The analysis was first results in a slightly increased probability of matching (71%),
stratified by IMG status and then by first-attempt status as a compared with analysis of first attempts only (68%).
separate stratification factor. Models were estimated sepa- ►Table 2 includes a summary of characteristics according to
rately within each resulting subgroup. The minimum terminal matching status (matched or did not match). Data analyzed
node size was set to 20. A fivefold cross-validation method was included number of interviews offered, number of programs
used to identify a best-fit model. A fivefold R2 value was ranked by the candidate, number of programs that ranked the
presented, which was the proportion of the variability in the applicant, and rank position of program by candidate and
response that was explained by the model. The resulting candidate by program. Applicants who successfully matched
models were presented as decision trees. Subgroups with submitted 66 applications (SD 24) and ranked 12 programs
similar probabilities of matching in the decision trees were (SD 7) on average, whereas those who failed to match submitted
combined. JMP software was used to fit the decision trees (JMP 60 applications (SD 33) (p ¼ 0.0004) and ranked 2 programs
version 11.2.0, 2013 SAS Institute).3 (SD 4) on average (p < 0.0001). In addition, matched applicants
performed better on USMLE step 1, with a mean score of 243
(SD 13), compared with a mean score of 226 (SD 18) for
Results
unmatched applicants (p < 0.0001). Furthermore, 95% of
Across the 2013, 2014, and 2015 application cycles, 1,959 matched applicants were applying on their first attempt,
unique individuals were identified. ►Table 1 includes a whereas only 72% of unmatched applicants were on their first
summary of characteristics for all candidates. On average, attempt (p < 0.0001). IMGs were significantly less likely to be in
applicants submitted 64 applications (standard deviation the matched group (4% vs 35%, p < 0.0001).

Table 1 Descriptive summary for all applicants (n ¼ 1,959)

Variable Mean SD Min Max Median Q1 Q3 IQR

No. of program applications 64.06 26.91 1 113 63 45 83 38
No. of invites for interview 9.29 6.3 0 29 9 4 14 10
USMLE step 1 scorea 238.3 16.42 182 275 241 229 250 21
No. of attempts 1.16 0.49 1 5 1 1 1 0
b
No. of institutions ranked by the applicant 10.45 5.04 1 113 11 8 13 5
Total no. of programs that ranked the applicantb 9.85 4.19 1 14 10 1 4 3
Matched position on the program’s rank listb 12.49 9.52 1 57 11 5 18 13
b
Applicant’s rank of the matched program 2.92 2.41 1 14 2 1 1 0
N %
IMG 247 13
Successful match 1391 71
a
AOA member 422 44
Published research 1948 99
Invited for  1 interview 1802 92
First attempt 1733 88

Abbreviations: AOA, Alpha Omega Alpha Honor Medical Society; IMG, international medical graduate; IQR, interquartile range (75th percentile–
25th percentile); Max, maximum; Min, minimum; Q1, 25th percentile; Q3, 75th percentile; SD, standard deviation; USMLE, United States Medical
Licensing Examination.
a
Number of missing observations for USMLE ¼ 46, AOA status ¼ 992.
b
For these four variables, summaries are restricted to applicants who matched (n ¼ 1,391).

Journal of Academic Ophthalmology Vol. 10 No. 1/2018

 Probability of Success in the Ophthalmology Residency Match Siatkowski et al. e153

Table 2 Descriptive summary of characteristics according to matching status

Variable Matched Did not match p-Valuea

(n ¼ 1,391) (n ¼ 568)
Mean SD Mean SD
No. of program applications 65.63 23.64 60.21 33.31 0.0004
No. of invites for interviewb 12 7 2 4 <0.0001b
USMLE step 1 scorec 242.99 13.28 226.23 17.54 <0.0001
No. of attempts 1.06 0.28 1.4 0.74 NPd
Count % Count % p-Valuee
IMG 49 4 198 35 <0.0001
c
AOA member 391 55 31 12 <0.0001
Published research 1390 99 558 98 <0.0001
Invited for  1 interview 1391 100 411 72 NP f
First attempt 1325 95 408 72 <0.0001

Abbreviations: AOA, Alpha Omega Alpha Honor Medical Society; IMG, international medical graduate; IQR, interquartile range (75th
percentile–25th percentile); Max, maximum; Min, minimum; NP, not performed; Q1, 25th percentile; Q3, 75th percentile; SD, standard deviation;
USMLE, United States Medical Licensing Examination.
a
p-Values are based on t-test.
b
Descriptive summary based on median and interquartile range. p-Value based on Wilcoxon rank sum test.
c
Number of missing observations for USMLE (matched) ¼ 13, AOA status (matched) ¼ 679, USMLE (not matched) ¼33, AOA status (not matched)
¼313.
d
Given the skewed nature of the variable indicating the number of application attempts, hypothesis testing is based on the dichotomous variables
indicating the first application attempt.
e
p-Value for published research is based on Fisher’s exact test, other p-values are based on chi-square test.
f
Hypothesis testing was not performed given that by definition, those who match to a program had at least one interview.

►Figs. 1 and 2 illustrate the flow of applicants through the and indicates the change point for the slope of the best-fit
application, interview, and matching process after stratifying line. Based on the estimated segmented regression line, the
by IMG status and the number of attempts. The overall match- number of invitations for interviews will generally increase
ing rate was higher among non-IMGs compared with IMGs among IMGs. In contrast, for non-IMGs, the change point of
(78% vs 20%, p < 0.0001) and among those on a first attempt the association between number of applications and number
compared with a repeat attempt (76% vs 29%, p < 0.0001). of interviews is 39. This indicates that the association was
The association between the number of applications positive for individuals who submitted up to 39 applications,
submitted and the number of interviews offered at programs after which point, the association became negative
ranked is summarized in ►Fig. 3 after stratifying by IMG (p < 0.0001). It is important to note that among the 1,712
status. The red line corresponds to the best-fit simple linear non-IMGs, 1,484 (87%) submitted more than 39 applications.
association. The blue line allows for a nonlinear association When all applicants are considered, regardless of IMG status,

IMG 247 applicants Not IMG 1,712 applicants

146 (59%) invited for 1 interview 1,656 (97%) invited for 1 interview

97 (66%) were ranked by 1 program 1,616 (97%) were ranked by 1 program

49 (51%, overall 20 %) successfully 1,342 (83%, overall 78%) successfully

matched matched

Fig. 1 Flow of applicants (stratified by international medical graduate [IMG] status) through the application, interview, and matching process.
Note: percentage was calculated based on the number of applicants in the previous level.

Journal of Academic Ophthalmology Vol. 10 No. 1/2018

e154 Probability of Success in the Ophthalmology Residency Match Siatkowski et al.

First attempt 1,733 applicants Second/repeat attempt 226 applicants

1,634 (94%) invited for 1 interview 168 (74%) invited for 1 interview

1,577 (97%) were ranked by 1 program 136 (81%) were ranked by 1 program

1,325 (84%, overall 76%) successfully 66 (49%, overall 29%) successfully

matched matched

Fig. 2 Flow of applicants (stratified by first/repeat attempts) through the application, interview, and matching process. Note: percentage was
calculated based on the number of applicants in the previous level.

Fig. 3 Plot number of interviews (y-axis) versus number of applications (x-axis) for (A) international medical graduates (IMGs) (n ¼ 247) and (B)
non-IMGs (n ¼ 1,712).

the change point of the association between number of (probability > 0.8). The analysis used USMLE score and num-
applications and number of interviews is 48, indicating ber of program applications as the two predictors because
that when the number of applications is less than 48, the these were nonmissing for a large percentage of applications,
association is positive; when the number of applications is with 1,913 applicants who had nonmissing observations for
greater than 48, the association is negative. Among the 1,959 these variables. The analysis was first stratified by IMG status
applicants, 1,398 (71%) submitted more than 48 applications. and then by first-attempt status as a separate stratification
►Fig 4 demonstrates that the probability of matching factor. Models were estimated separately within each resulting
reaches the peak of 81% with the group who submitted 41 to subgroup. ►Figs. 5 and 6 present the prediction models based
60 applications and decreases thereafter. Additionally, 538 on stratification factors. In summary, among the non-IMG
(39%) applicants matched with their top-ranked program, applicants, those with a “high” probability of matching were
while 268 (19%) applicants matched at their second-ranked those with USMLE  244 or those with a USMLE between 231
program. A majority (87%) of the applicants who matched and 243 who submitted at least 30 applications. Similarly,
did so within their top five choices. among those at their first attempt, individuals with a USMLE
After univariate data analysis, a multivariable, recursive  233 were predicted to have a “high” probability of matching.
partitioning algorithm was created to identify three different None of the IMG applicants and none of those at a second
groups based on their predicted probability of matching, attempt were categorically predicted to have a “high” prob-
resulting in those with a low probability of matching (prob- ability of matching. Each model explained no more than 18% of
ability < 0.4), a moderate probability of matching (probability the variability in the probability of matching (fivefold cross-
between 0.4 and 0.8), and a high probability of matching validation R2  0.18).

Journal of Academic Ophthalmology Vol. 10 No. 1/2018

 Probability of Success in the Ophthalmology Residency Match Siatkowski et al. e155

Fig. 4 Summary of matching characteristics. (A) Probability of matching by categories of number of applications (n ¼ 1,959) and
(B) cumulative percentage of applicants’ rank of the matching program for applicants who matched (n ¼ 1,391).

Fig. 5 Classification tree summary of factors predictive of matching for (A) non-IMG (international medical graduate) applicants (n ¼ 1,671).
The resulting decision tree explains 18% of the variability in the probability of matching (5-fold cross-validation R2 ¼ 0.18). (B) IMG applicants
(n ¼ 242). The resulting decision tree explains 7% of the variability in probability of matching (5-fold cross-validation R 2 ¼ 0.07). Green, high
probability of matching (>0.8); NPA, number of program applications; P, predicted probability of matching; Red, low probability of matching
(<0.4); USMLE, United States Medical Licensing Examination; Yellow, moderate probability of matching (0.4–0.8).

Fig. 6 Classification tree summary of factors predictive of matching for (A) first attempt applicants (n ¼ 1,690). The resulting decision tree
explains 14% of the variability in probability of matching (5-fold cross-validation R 2 ¼ 0.14). (B) second or higher attempt applicants (n ¼ 233).
The resulting decision tree explains 14% of the variability in probability of matching (5-fold cross-validation R2 ¼ 0.14). Green, high probability
of matching (>0.8); NPA, number of program applications; P, predicted probability of matching; Red, low probability of matching (<0.4);
USMLE, United States Medical Licensing Examination; Yellow, moderate probability of matching (0.4–0.8).

Journal of Academic Ophthalmology Vol. 10 No. 1/2018

e156 Probability of Success in the Ophthalmology Residency Match Siatkowski et al.

Discussion junior and senior admission, while others offer only senior
admission. Additionally, differences among schools, timing of
Roughly 70% of candidates match in ophthalmology, and 70% application submissions, and the fact that individual chapters
match at one of their top three choices. Yet a majority (53%) have nonstandard admission criteria and protocols are signifi-
apply to more than 60 programs and over one-quarter (27%) cant confounders which limit interpretation of results.
apply to more than 80. Clearly, more data are needed for Similarly, research activity was common in both matched
program directors and faculty advisors to assist them in the and unmatched groups. However, the analysis simply used the
application and matching process. presence of absence of any research activity; we did not
The recursive partitioning model that we have described is distinguish many relative criteria regarding research, for
extremely useful in this regard. Since the match rates are so example, whether the research activity was print publica-
disparate for IMG versus US graduates and first- versus nonfirst tion(s), the impact factor of any journals, presentation at local
attempts, these are logical stratification points. The group with versus regional versus national meetings, or whether the
the highest match rate (93%) is US graduates, regardless of first activity was merely participation of some sort in an uncom-
or repeated attempt, with a USMLE step 1 score of 244 or pleted project. Finally, in recent years the Gold Humanism
greater. The second highest match rate (87%) was among all Honor Society has grown across the medical community and is
first-time candidates with a USMLE score of 233 or greater. For increasingly associated with many of the character traits felt to
these two groups, the number of programs applied to or be important for competent physicians. It is certainly possible
ranked was not a factor in matching success. The third highest that any of these factors could be significantly correlated with
match rate (83%) occurred among US graduates with USMLE probability of matching, and revision of the application form to
scores from 231 to 243 who applied to 30 or more programs. clarify and standardize reporting of these factors would offer
Although our data do not allow specific recommendations on further assistance in advising candidates. In a similar vein, a
exact numbers, it would seem unlikely that many candidates in standardized letter of recommendation format has been
these groups should be advised to apply to more than 40 recommended for ophthalmology candidates, although to
programs, especially as among all applicants overall, applying date it has not been widely adopted.
to more than 48 programs is not associated with an increasing A final set of limitations to this study surrounds the fact that
number of interviews; this number decreases to 39 when data on interview invitations and completions is self-reported
confined to US graduates alone. These suggested thresholds by both programs and applicants. In the match system, pro-
are well below the average number of applications submitted grams must mark an applicant as invited for an interview
by all applicants (mean 64, SD 27), US graduates (mean 65, SD before they can view candidate photos or add them to a rank
25), and IMG applicants (mean 56, SD 36). list. Similarly, applicants are asked to mark an interview as
The fourth highest success rate (66%) was among US having been completed, but in some cases may make or not
graduates with USMLE scores of 217 to 230 who applied to make this designation erroneously. As an example of the
43 or more programs, and the next highest (61%) among first- consequences of this methodology, in the 2015 application
time applicants with USMLE < 233 who applied to 47 or cycle there were 726 candidates who submitted 634 rank lists.
more programs. However, those with scores 229 to 232 who Of these, 624 applicants were ranked. Programs reported
applied to fewer than 47 programs had a similar match rate offering a total of 6,594 interview invitations with completion
(60%). These groups seem the most likely to benefit from of 5,503 of interviews. However, in the same cycle, applicants
applying to more than 40 programs. For candidates who have reported 6,655 interview invitations with 5,749 interviews
previously not matched, the single most important factor in completed. The newly formed SF Match Oversight Committee
matching was a USMLE score > 235 (55%), independent of of the AUPO intends to review this design system and consider
number of programs applied to. alternatives to minimize such discrepancies.
For candidates not in the groups described above, there Although this analysis yields the ability to provide overall
were no subgroups with match rates > 50%. US graduates and advice to various categories of candidates, it does not permit
first-time applicants with USMLE scores < 231 should be more detailed candidate-specific data that includes individual
advised to apply to more than 45 programs. Non-US graduates, portfolios and “good fits” for either candidates or programs in
regardless of USMLE score, and previously unmatched candi- creating their rank lists. The focus of this study may make
dates with USMLE score 220–235 did not achieve a moderate many candidates and advisors more comfortable with a
probability of matching until they had applied to more than smaller number of programs applied to, but does not assist
100 programs (43 and 42%, respectively). with determining which candidates should apply to which
While we did find that AOA (Alpha Omega Alpha Honor programs. In this regard, programs should consider providing
Medical Society) membership conferred an increased prob- candidates with standardized data regarding their residents
ability of matching (55% vs 12%, p < 0.0001), this finding must (e.g., program mean USMLE scores, grade point average (GPA),
be interpreted cautiously. AOA membership status was missing AOA status, research participation in medical school) to help
for 992 (51%) of applicants and, therefore, could not accurately them determine their best application strategy.
be evaluated as an independent predictive factor of matching, In summary, these data indicate that many candidates for
relative to USMLE score or number of submitted applications in ophthalmology residency (first-time applicants/US grads with
our analysis. Some US medical schools, and many non-US USMLE scores >243) need not apply to 60 or 80 programs to
schools, may not have an AOA chapter, and some offer both successfully match; for most of them, 40 to 45 applications

Journal of Academic Ophthalmology Vol. 10 No. 1/2018

 Probability of Success in the Ophthalmology Residency Match Siatkowski et al. e157

should suffice. The 45 to 60 range may be indicated for initial the NIH Vision Core Grant P30 EY 0268 (Bethesda, MD).
applicant US graduates with USMLE scores in the 217 to 243 The funding organizations had no role in the design or
range. A relatively small number of candidates, especially conduct of this research. The study was supported in part
international graduates and previously unmatched candidates by an unrestricted grant from Research to Prevent Blind-
(12% each of the entire applicant pool), should consider ness, Inc, NY, NY (RMS).
applying to more than 80 programs, particularly if their USMLE
score is <236. Conflict of Interest
Finally, we recognize that these recommendations must Dr. Feldon reports personal fees from Association of
always be individualized, as there are many additional University Professors of Ophthalmology, outside the sub-
factors that cannot be easily standardized or quantified mitted work, and is Chair of a Department of Ophthalmol-
which nevertheless play important roles in the likelihood ogy that has a residency training program that could be
of candidates matching. These include variable and qualita- affected by the information in the article. Dr. Stoner
tive grading rubrics, communication skills, strength of reports support for data analysis from the Association
letters of recommendation, extracurricular and employment of University Professors of Ophthalmology (AUPO), during
experiences, and interview performance. In the end, appli- the conduct of the study. All the other authors report no
cants for ophthalmology residency are ranked, and thus conflict of interest.
matched, by a holistic consideration of their entire candidate
portfolio.
References
Funding 1 Muggeo VM. Segmented: an R package to fit regression models
The Association of University Professors of Ophthalmol- with broken-line relationships. R News 2008;8:20–25
ogy supported the statistical analysis performed by 2 Hawkins DM, Kass GV. Automatic interaction detection. In: Haw-
kins DM, ed. Topics in Applied Multivariate Analysis. Cambridge:
Dr. Stoner and Ms. Chen. Dr. Cullican received support
Cambridge University Press; 1982:269–302
from awards to the Department of Ophthalmology and 3 Sall J, Lehman A, Stephens ML, Creighton L. Exploratory modeling.
Visual Sciences at Washington University from a Research In :JMP Start Statistics: A Guide to Statistics and Data Analysis
to Prevent Blindness, unrestricted grant (New York, NY), Using JMP. Cary, NC: SAS Institute; 2012:519–530

Journal of Academic Ophthalmology Vol. 10 No. 1/2018