No effect of real-world universal face masking on post-intravitreal injection endophthalmitis rate at a single tertiary academic center

Abstract

Purpose: To determine if universal masking during COVID-19 altered rate and outcomes of post-injection endophthalmitis (PIE). Methods: Retrospective, single-site, comparative cohort study. Eyes diagnosed with endophthalmitis within 4 weeks of intravitreal injection at the University of Michigan from 8/1/2012 to 11/15/2022 were identified. Cases were considered “masking” between 3/15/2020 and 11/15/2022. Endophthalmitis rate, visual acuity (VA), and microbial spectrum were investigated. Results: There were 20 PIE cases out of 72,194 injections (0.028%; 1 in 3571 injections) pre-masking and 10 of 38,962 with universal masking (0.026%; 1 in 3846 injections) (OR 0.9; 95% [CI]: 0.4-2.0). Referral from the community was unchanged with 32 cases referred pre-masking (0.35 cases/month) and 10 cases with masking (0.31 cases/month). Presenting mean logMAR VA with masking of all PIE cases trended worse (2.35±0.40) compared to pre-masking (2.09±0.48) (P=0.05) with light perception VA more common with masking (31.6% vs 10.9%, P=0.06). There was no delay in time from procedure to initial treatment (P=0.36), no difference in rate of initial treatment with tap and inject (T/I), and similar positive-culture rates (P=0.77) between cohorts. VA after 30 days follow-up was clinically unchanged (∼20/500 vs 20/400) (P=0.59). Conclusions: Universal masking had no effect on PIE rate nor rate of culture-positive cases. While presenting VA appeared worse with masking, this was not statistically significant and current treatment paradigms resulted in similar visual outcomes.