Evaluating the Diagnostic Accuracy of Ophthalmology Residents: Balancing Resident Autonomy with Patient Safety

Background: Ophthalmology residency programs aim to cultivate competent, independent-thinking ophthalmologists. However, achieving a balance between resident autonomy and patient safety presents a difficult challenge. Purpose: To evaluate the diagnostic accuracy and safety of a resident-driven on-call model, wherein consulting with a senior resident or attending physician is at the discretion of a postgraduate year 2 or 3 resident. Methods: This was a retrospective study evaluating emergency department ophthalmology consultations at Erie County Medical Center, a level one trauma center in Western New York, between 1 January 2019 and 31 December 2021. Resident diagnoses and critical examination findings were categorized as delineated in The Wills Eye Manual, 7th edition, and compared with attending ophthalmologist documentation at a follow-up visit at Ross Eye Institute. Each consultation was graded as accurate, partially accurate, or inaccurate. Partially accurate and inaccurate diagnoses were individually reviewed to assess patient safety. Results: A total of 1,021 consultations were analyzed. Resident diagnoses were accurate in 95.9% of cases, partially accurate in 1.6% of cases, and inaccurate in 2.5% of cases. Only 2 cases (0.2%) required emergent treatment and workup after an inaccurate resident diagnosis that was identified at the follow-up visit. Trauma diagnoses and correct examination findings during the consultation were associated with greater diagnostic accuracy. Orbital diagnoses were associated with a lower rate of diagnostic accuracy. Conclusions: A resident-driven on-call model, utilized by many ophthalmology residency programs, enables residents to hone their independent decision-making skills while maintaining a high level of patient safety.