We recently demonstrated that medical students after their first-year ophthalmology training not only examined the ocular fundus more accurately with photographs than with direct ophthalmoscopy, but also preferred using photographs. One year later, we completed a follow-up study to test our hypothesis that these differences would persist over time.

Second-year medical students were reevaluated on their ability to examine the ocular fundus after randomization to either fundus photographs or direct ophthalmoscopy on eye simulators. All students recorded answers to a questionnaire assessing their ability to visualize features of the ocular fundus, positive and negative affect, preferences, and their clinical experiences during the previous year with ocular fundus examination. The study was institutional review board–approved, in accordance with Health Insurance Portability and Accountability Act regulations, and informed consent was obtained from subjects. Of 119 students, 107 (90%) who participated in the original study completed this 1-year follow-up study. Of 48 questions, students answered 34.5 (72%) correctly using photographs vs 31.4 (65%) using direct ophthalmoscopy ( P = .004). Two series of questions used the same images for both the simulator and photographs, and students performed better using photographs (mean: 16.6 vs 14.3 of 24 correct; P = .0008). Both photography and ophthalmoscopy groups correctly answered 5 fewer questions on average than 1 year prior ( P < .001). Students rated photographs as “easier than ophthalmoscopy” (mean: 7.9/10 vs 5.9/10, respectively; P < .001). Students’ positive affect scores were higher in the photograph group (26.5) than in the ophthalmoscopy group (23.2; P = .03). Students tested on simulators reported lower positive affect than 1 year ago (−6.4 points, P < .001).

Students’ self-reported median frequency of fundus examination over the preceding year was <10% (interquartile range: 0%-20%). The students’ primary reasons for not examining the ocular fundus during a general physical examination were discomfort with the examination (41/107; 38%), discouragement by preceptors (21/107; 20%), and insufficient time (16/107; 15%). Of 107 students, 85 (79%) felt uncomfortable with ophthalmoscopy, 47 (44%) stated they would not perform ophthalmoscopy during general physical examinations, and 81 (76%) stated they would prefer using photographs over ophthalmoscopy for fundus examination.

One year after training, students continued to be more accurate using photographs than ophthalmoscopy and still preferred photographs for examining the ocular fundus. Although both groups performed significantly worse in identifying relevant fundus features than 1 year prior, the difference was equal in the 2 groups and likely related to a lack of fundus examination skill reinforcement in the interim. Most students felt uncomfortable with ophthalmoscopy, which may cause avoidance of ocular fundus examination in clinically appropriate situations. Of concern, 20% of students cited discouragement by their clinical preceptor as their primary reason for not performing ophthalmoscopy, which suggests that postgraduate education may be needed to create a long-term change in the use and performance of fundus examination, particularly with direct ophthalmoscopy. However, the students’ continued enthusiasm for the clinical use of ocular fundus photography, combined with other research, again suggests that more widespread availability of nonmydriatic fundus photography could allow for more frequent and accurate examination of the ocular fundus by medical students, residents, and faculty in a variety of clinical settings.