Microbiological sampling has limited value in managing acute postoperative bacterial endophthalmitis: a multicentre study in the UK.

Aims: To determine whether microbiological testing impacts clinical management in acute postoperative endophthalmitis.

Methods: Multicentre retrospective cohort study encompassing five tertiary centres in the UK. Patients presenting with acute postoperative endophthalmitis with at least 4 weeks follow-up were included. The main outcome was the proportion with a change in management (defined as further intravitreal antibiotic therapy and/or further surgical intervention within 4 weeks of the initial treatment) and the rationale for this change.

Results: 190 eyes of 189 patients were included. Patients presented at a median of 5 days postoperatively (IQR 3-10). Sampling was predominantly obtained via vitreous tap alone (80/190, 42%) or with both vitreous and anterior chamber tap (84/190, 44%). Over half were culture-positive (107/190, 56%), and only bacterial pathogens were isolated. Culture-positive cases with available antibiotic sensitivity data demonstrated in vitro sensitivity to at least one of the empirical intravitreal antibiotics administered. Seventy-six eyes (40%) had a change in management within 4 weeks of the initial treatment. These additional procedures took place within 48 hours of initial treatment in 46% (35/76) of patients. The main reasons were a lack of clinical improvement (46/76, 61%) or clinical deterioration (18/76, 24%); none of these changes were prompted or guided by culture or sensitivity results.

Conclusions: Microbiological sampling was of limited clinical utility in this series. In patients presenting with suspected acute bacterial endophthalmitis, if microbiological sampling might pose any delays to treatment, consideration should be given to immediate intravitreal antibiotic treatment without sampling to optimise visual outcomes.