Purpose: To assess the effect of topical antibiotic prophylaxis on the rate of post-operative endophthalmitis after intravitreal injection (IVI).
Methods: We conducted a systematic review of studies comparing the rates of endophthalmitis in eyes receiving IVI of different drugs with and without topical antibiotic prophylaxis, by searching MEDLINE and EMBASE up to June 2016. The methodological quality of the included studies was assessed using the Cochrane Risk of Bias tool and the Risk Of Bias in Non-randomized Studies of Interventions (ROBINS-I) for randomized clinical trials (RCTs) and non-randomized studies, respectively. We used a random-effects meta-analysis to compute the odds ratio (OR) of endophthalmitis with antibiotic prophylaxis compared with no prophylaxis and conducted subgroup analyses to compare the efficacy of different regimens and classes of antibiotics on endophthalmitis rates.
Results: We identified 1 randomized and 12 non-randomized studies that reported 74 cases of endophthalmitis in 147,203 IVIs using antibiotic prophylaxis compared with 55 cases in 211,418 IVIs with no prophylaxis. The overall OR of endophthalmitis for antibiotic prophylaxis vs. no prophylaxis was 1.33 (95% CI 0.75-2.38). Leave-one-out sensitivity analyses showed that the exclusion of the only study with a serious risk of bias significantly increased the risk of endophthalmitis in the antibiotic prophylaxis group compared with control (OR: 1.62, 95% CI: 1.17, 2.34). There was no difference in the endophthalmitis rate associated with any other factor analyzed, including type of antibiotic, type of drug injected, or antibiotic prophylaxis regimen.
Conclusions: Antibiotic prophylaxis does not reduce the rate of endophthalmitis following IVI and might potentially be associated with an increased risk of post-operative infection.