Intravitreal injections as a leading cause of acute postoperative endophthalmitis—a regional survey in England

BACKGROUND: To evaluate the characteristics, treatment patterns and outcomes of acute postoperative endophthalmitis. METHODS: Patients presenting with acute postoperative endophthalmitis between January 2017 to December 2019 were identified from hospital records in this multicentre retrospective cohort study. Clinical records were reviewed for visual acuity (VA) at various timepoints, cause of endophthalmitis, microbiological results, treatments and complications. RESULTS: Forty-six eyes of 46 patients were included. Intravitreal injections were the leading cause of acute postoperative endophthalmitis (n = 29; 63%), followed by cataract surgery (n = 8; 17%), vitreoretinal surgery (n = 7; 15%), and secondary intraocular lens insertion (n = 2, 4%). The absolute risk of endophthalmitis was 0.024% (1:4132) for intravitreal injections, 0.016% (1:6096) for cataract surgery, and 0.072% (1:1385) for vitreoretinal surgery. The majority of patients (n = 38; 83%) had better VA at 6 months compared to presentation, although fewer (n = 13; 28%) maintained similar or better VA compared to before the precipitating surgery. Twenty-four cases yielded positive culture results, of which staphylococcus epidermidis was the most commonly isolated organism. Microbiological yield was not associated with better final visual outcomes. Patients who underwent therapeutic vitrectomy (n = 15; 33%) had poorer VA at presentation, but subsequently achieved visual outcomes comparable to those who received medical treatment alone. There was no difference in time to presentation, visual outcome and retinal detachment rates among the different causative procedures. CONCLUSION: Intravitreal injections were the most common cause of endophthalmitis in our region, primarily because of their higher frequency compared to other intraocular procedures. In this cohort, the primary procedure had no effect on presentation, management or visual outcomes.