Risk of Infectious Endophthalmitis From a Resident-Performed Intravitreal Injection

Abstract

This work assesses the safety of resident-performed intravitreal injections by comparing the incidence of postinjection endophthalmitis associated with residents vs surgical retina attendings. In this retrospective noninferiority cohort study, the Current Procedural Technology code for an intravitreal injection was used to identify patients from July 1, 2015, to October 31, 2018, at a single-center Veterans Affairs hospital where patients receive intravitreal injections from attending vitreoretinal surgeons and residents. The medical record for each injection was reviewed for the training level of the proceduralist, indication, medication, and laterality. The main outcome was the incidence of postinjection endophthalmitis occurring within 30 days of an intravitreal injection, which was confirmed by at least 2 medical-record reviewers. In this cohort of 593 patients, most were male (97.0%), and the average age was 70.3 years (±10.8 years). Of the 6934 injections given, 3877 (55.9%) were performed by residents and 3057 (44.1%) by attendings. Of the 5 cases of endophthalmitis identified, 3 were associated with a resident-performed injection in his or her fourth year of postgraduate training. The difference (0.012%; 95% CI, -0.168% to 0.169%) in the incidence of postinjection endophthalmitis between residents (0.077%) and attendings (0.065%) did not cross the predetermined noninferiority limit (0.200%). At our teaching clinic, resident-performed injections were not associated with an elevated risk of endophthalmitis when compared with attending-performed injections. Under varying levels of supervision, residents appeared to demonstrate appropriate sterile injection technique.