Correlating Ophthalmology Exposure in Medical School to Applying and Matching Into an Ophthalmology Residency

Abstract

Purpose 

To correlate ophthalmology curricular exposure in medical school to the number of students who applied and matched into ophthalmology residency programs. Given the high curricular burden placed on medical schools, the authors sought to better characterize existing ophthalmology curricula and to delineate which offerings are closely related to high numbers of students applying and matching into ophthalmology residencies.

Method 

The authors reviewed the extent of ophthalmology curricula between 2007 and 2017 via a survey administered in 2018 to all U.S. Association of American Medical Colleges (AAMC)-affiliated medical schools. They obtained residency application and match data with permission from the Association of University Professors of Ophthalmology. The authors compared metrics of ophthalmology exposure with the number of students who applied and matched into ophthalmology during the corresponding year using mixed-effects Poisson regression analysis.

Results 

This study includes 49 U.S. AAMC-affiliated medical schools. When adjusted for the number of applicants per year, the following were significantly (P < .05) associated with matching into an ophthalmology residency: the presence of an ophthalmology department, an ophthalmology residency program, an ophthalmology interest group, ophthalmologists on faculty, ophthalmology content in the preclinical curriculum, preclinical lectures taught by ophthalmologists, and the availability of an optional fourth-year ophthalmology elective. Multivariable analysis indicated both that the presence of an ophthalmology residency program was the only independent predictor of matching into an ophthalmology residency and that the presence of an ophthalmology residency program, ophthalmology content in the preclinical curriculum, and preclinical lectures taught by ophthalmologists are independent predictors for applying.

Conclusions 

A foundation in ophthalmology is crucial for all physicians, especially those who may encounter patients with eye problems in emergency or primary care settings. However, for those students considering a career in ophthalmology, choosing a medical school with an ophthalmology department and residency program is particularly important.

The role of ophthalmology education in medical school has been a long-standing question.^1–7 Though medical education became standardized in the early 1900s and was heavily influenced by the Flexner Report, the Flexner Report itself did not detail requirements for specialty education.^6,7 Maintaining a robust ophthalmic clinical experience is becoming increasingly difficult. Challenges include the lack of Liaison Committee on Medical Education guidelines regarding ophthalmology education in medical school, increased curricular demands placed on medical schools, and Medicare funding of medical education that is more amenable to inpatient hospital settings.^4,6

Ophthalmology exposure during medical school is important for many reasons: to inspire the next generation of ophthalmologists, to engage the ophthalmology department with the medical school,^4,5 and to address emergency and primary care chief complaints (2%–19% of which are ophthalmic).⁴ Despite considerable efforts to delineate core knowledge competencies for medical students in all specialties and the persistent presence of ophthalmology content on United States Medical Licensing Examination (USMLE) exams, ophthalmology teaching has continually decreased in undergraduate medical education. The prevalence of required ophthalmology rotations in medical school has gradually declined since 2000.^3–5,8,9 However, the current trend toward small-group, interactive, technologically enhanced, and standardized patient–based education presents an opportunity to incorporate more ophthalmology into future curricula.^3,7

We therefore examined the ophthalmology experience in U.S. medical schools. Specifically, given the high curricular burden placed on medical schools, we aimed to better delineate which offerings are most closely related to high numbers of students applying and matching into ophthalmology residency programs.

Method

The Institutional Review Board at Columbia University Irving Medical Center ruled that approval was not required for this study. This research adheres to the tenants of the Declaration of Helsinki. We did not collect identifying information on individual students during this study.

In 2018, we administered a survey to medical education offices and ophthalmology departments of U.S. Association of American Medical Colleges (AAMC)-affiliated medical schools via email and telephone to collect retrospective data on the extent of ophthalmology exposure. The complete survey is available as Supplemental Digital Appendix 1 (https://links.lww.com/ACADMED/A940). We asked a representative from each school (e.g., a dean of medical education, an ophthalmology department representative) to comment on the presence of and any changes to the ophthalmology content in the preclinical and clinical curricula (including the quantity of lecture, clinical, and workshop-based instruction hours). We also asked representatives to note whether their schools had an ophthalmology student interest group, affiliated ophthalmology faculty, an ophthalmology department, and/or an ophthalmology residency program from 2007–2017. Schools that were established in the midst of the study period were excluded.

With permission from the Association of University Professors of Ophthalmology (AUPO), we obtained the number of medical students who applied and matched into ophthalmology residencies between 2010 and 2017 from each medical school. Application and match data before 2010 were not readily available due to a change in the data system, and medical schools could not reliably supply this information.

We conducted a nonresponse bias analysis to assess the rates of students applying and matching into ophthalmology residencies from all eligible institutions that either did not respond to the survey or declined to participate. We compared the rates of applicants and matched applicants from those schools with the rates from the institutions that we included in the analysis.

Then, using univariable, bivariable, and multivariable mixed-effects Poisson regression, we compared our metrics of ophthalmology exposure, detailed in Supplemental Digital Appendix 2 (https://links.lww.com/ACADMED/A940), with the number of students who applied and matched into ophthalmology residency programs each year. This mixed-effects model takes into account random effects associated with different intercepts, therefore allowing for analyses of correlated data within units of observation (i.e. medical schools at different points in time). We considered metrics that affected application statistics in future years, such that preclinical and clinical metrics from a student’s first, second, and third years were appropriately aligned with application/match data from the fourth year. That is, if a student’s match year was 2010–2011, the preclinical data are from 2007–2009 while the third-year data are from 2009–2010. We aligned fourth-year metrics with fourth-year application and match data. We tested predictors for the number of applicants in univariable models. We tested predictors for the number of matched students, first, in univariable models and, then, in bivariable models. In the bivariable model, we adjusted for the number of applicants per year since a higher number of applicants would lead to a greater likelihood of matching (described below, in Results). We entered the variables that had P values less than .1 in the bivariable model into the multivariable model. Additionally, we ran the multivariable model examining solely factors that influence the number of applicants, irrespective of the number of students who matched.

We conducted our computerized statistical analyses with STATA (version 14.2, StataCorp, Inc., College Station, Texas). We set Type-1 error at 5% to define statistical significance.

Results

We distributed surveys to 148 U.S. AAMC-affiliated medical schools. Of these, 26 (18%) were established after 2007, and were thus ineligible for inclusion. Of the 122 remaining eligible institutions, 54 (44%) responded. Forty-nine chose to participate; thus, our final response rate (RR), using definition RR3 from the American Association for Public Opinion Research, was 40% (of 122).¹⁰ See also Supplemental Digital Appendix 3 (https://links.lww.com/ACADMED/A940).

Our nonresponse bias analysis demonstrated no statistically significant differences in the number of applicants between the schools included in and excluded from this study (3.6 [standard deviation or SD 2.8] vs 3.9 [SD 2.6], P = .118). Further, we detected no significant difference in the number of matched students from the schools included in and excluded from this study (3.1 [SD 2.5] vs 3.3 [SD 2.3], P = .283).

Table 1 shows the number and percentage of the responding schools that had a department of ophthalmology, an ophthalmology residency program, ophthalmology faculty members, an ophthalmology interest group, preclinical ophthalmology exposure, lectures taught by ophthalmologists, a required (and/or elective) third-year ophthalmology clinical experience, and one or more optional fourth-year ophthalmology electives during the years 2007–2017.

Table 1: The Percentage of U.S. Association of American Medical Colleges (AAMC)-Affiliated Institutions With Various Ophthalmology-Related Curricular Offerings, by Year, 2007–2017

The results of our univariable analysis, which investigated the factors that influence the number of applicants into ophthalmology (see Table 2), indicated that the following variables were significantly associated with the number of students who applied into ophthalmology (P < .05):

Table 2: Univariable Analysis Showing the Relationship Between Metrics of Ophthalmology Exposure and the Number of Students Applying to an Ophthalmology Residency^a

• the presence of an ophthalmology interest group,
• preclinical lectures taught by ophthalmologists (2 and 3 years before the match year),
• ophthalmology exposure through the presence of an optional third-year curriculum (as well as the number of students per year, total days, total hours, number of lecture hours, and number of clinical hours of the third-year curriculum),
• the presence of an optional fourth-year ophthalmology elective,
• the number of available fourth-year ophthalmology electives,
• the presence of an ophthalmology department, and
• the presence of an ophthalmology residency program.

Our analysis indicated that the relationship between applicants and matched applicants was highly significant (beta = 0.19; 95% confidence interval [CI]: 0.17, 0.21; P < .00; see Supplemental Digital Appendix 4 at https://links.lww.com/ACADMED/A940); therefore, in subsequent analyses, we adjusted for the number of applicants in the bivariable and multivariable models. Notably, approximately 90% of the seniors who applied each year matched into an ophthalmology residency, with an average of 99.4% of ophthalmology residency spots filled on match day between 2010 and 2017.¹¹

The results of our bivariable analysis, as shown in Table 3, indicate that the following variables were significantly associated with the number of students who matched into an ophthalmology residency when adjusted for the number of applicants (P < .05):

Table 3: Bivariable Analysis Showing the Relationship Between Metrics of Ophthalmology Exposure and the Number of Matched Applicants When Adjusted for the Total Number of Applicants^a

• the presence of an ophthalmology interest group,
• exposure to ophthalmology content in the preclinical curriculum (2 and 3 years before the match year),
• preclinical lectures taught by ophthalmologists (2 and 3 years before the match year),
• the presence of an optional fourth-year ophthalmology elective,
• the presence of an ophthalmology department,
• the presence of ophthalmology associated faculty, and
• the presence of an ophthalmology residency program.

To test for independent variable significance on the number of students who apply into an ophthalmology residency, we ran a multivariable model with variables that were significant in the bivariable model (P < .1). These results, available in Table 4, indicate that the presence of an ophthalmology residency program (beta = 0.92; 95% CI: 0.34, 1.51; P = .002), exposure to ophthalmology content in the preclinical curriculum (beta = −0.78; 95% CI: −1.23, −0.03; P = .001), and preclinical lectures taught by ophthalmologists (beta = 0.67; 95% CI −1.09, −0.27; P = .014) were independent predictors of applying into an ophthalmology residency.

Table 4: Multivariable Analysis Showing the Relationship Between the Number of Students Who Applied to an Ophthalmology Residency and Outcomes From the Bivariable Analysis for Which P < .1

To test for independent variable significance on the number of matched students, we ran a multivariable model that included the variables from the bivariable model whose P value was less than .1. These results, shown in Table 5, indicate that the presence of an ophthalmology residency program is the only independent predictor of matching into an ophthalmology residency (beta = 0.55; 95% CI: 0.16, 0.94; P = .005).

Table 5: Multivariable Analysis Showing the Relationship Between the Number of Students Who Matched Into an Ophthalmology Residency and Outcomes From the Bivariable Analysis for Which P < .1

Discussion

Over the past decades, changes in the structure of the ophthalmic curriculum have been heavily influenced by the increasing curricular demands placed on medical schools. Our current study is, to the best of our knowledge, the first to gather information about ophthalmology curricula over time to elucidate which offerings are most closely related to higher numbers of students applying and matching into ophthalmology residency programs. Our results show that required ophthalmology rotations continue to decline. Previous literature has reported that 68% of medical schools required a formal ophthalmology clinical rotation in 2000, 30% in 2004,⁵ and 18% in 2013.⁴ This study reveals that only 7% of medical schools required a clinical rotation in 2017.

When adjusted for the number of applicants in a bivariable analysis, the following were significantly associated with matching into an ophthalmology residency: the presence of an ophthalmology department, faculty who are ophthalmologists, and an ophthalmology interest group. In addition, the presence of preclinical ophthalmology instruction alone and, specifically, the presence of preclinical lectures taught by ophthalmologists were significantly associated with matching into an ophthalmology residency. The quantity of lecture hours was, however, not statistically significant. Additionally, the presence of both required and optional third-year ophthalmology clinical curricula was not statistically significantly associated with matching into an ophthalmology residency, nor was the length of total instruction, nor the number of lecture and clinical hours. These findings align with a previous study of medical schools in the United Kingdom, which showed that the duration of ophthalmology teaching was not associated with the percentage of students entering residency.¹²

Importantly, the aforementioned significant variables lost significance in the multivariable analysis. The only independent predictor of matching into an ophthalmology residency in the multivariable analysis was the presence of an ophthalmology residency program associated with the medical school. Notably, the presence of a residency program requires the presence of an ophthalmology department, which facilitates ophthalmology research and robust mentorship for students. Medical schools with an ophthalmology department likely have known faculty who can write letters of recommendation for their students. Further, these schools may offer a potential home-institution advantage for matching. Additionally, the presence of an ophthalmology residency program may increase the chances that ophthalmologists will deliver preclinical lectures and may enhance the resources available for preclinical workshops—both of which were significantly associated with matching in the bivariable analysis (adjusted for the number of applicants).

Our examination of the decision to apply to an ophthalmology residency showed that, in addition to the presence of an ophthalmology residency program, the presence of preclinical ophthalmology instruction and preclinical lectures given by ophthalmologists was significant. This finding is of particular interest as it indicates that early exposure to ophthalmology through the preclinical curriculum, despite not being a significant predictor of matching, is crucial to stimulating interest in ophthalmology. Therefore, institutions should not discount the value of ophthalmic resource allocation to their preclinical offerings.

This study is limited by the fact that the curricular data we collected are retrospective; however, match data were obtained through an AUPO computerized database. As this study relied on survey responses, we acknowledge a possible response and/or selection bias. Although our analysis of nonresponse bias indicated no significant differences in the number of applicants or matched students between the schools included in and excluded from the study, we do not know if the schools that responded are representative of all medical schools. Additionally, for the purposes of the study, we assumed that students took no time off during medical school to undertake research, to pursue another degree, or for any other cause. Lastly, we collected no information about the quality or mastery of ophthalmic education.

Aside from the curricular factors included in the analysis, other factors may influence application into ophthalmology. The average USMLE Step 1 score for those who matched into ophthalmology during this study period was approximately 240.¹¹ As ophthalmology is known by reputation to be a competitive field, there is likely a self-selection bias of students who elect to pursue it as a career.

In conclusion, matching into an ophthalmology residency program is most associated with the presence of an established ophthalmology residency program affiliated with the student’s medical school. The decision to apply into an ophthalmology residency program is additionally associated with exposure to ophthalmology in the preclinical curriculum and preclinical lectures taught by ophthalmologists. The presence of an ophthalmology department and ophthalmology interest group also plays a meaningful role.

A foundation in ophthalmology is crucial for all physicians, especially those who may encounter patients with eye problems in emergency or primary care settings. Particularly for those students considering a career in ophthalmology, choosing a medical school with a robust ophthalmology department and residency program is of great importance.