Subconjunctival antibiotics in the treatment of endophthalmitis managed without vitrectomy

PURPOSE: To determine if the treatment outcomes for endophthalmitis are influenced by subconjunctival antibiotics. METHODS: A retrospective, nonrandomized consecutive series of patients with clinically diagnosed bacterial endophthalmitis confirmed with positive cultures who presented between December 1, 1995, and February 28, 2002, was studied. Patients with cataract surgery, glaucoma filtering blebs, or trauma who presented with visual acuity of hand motions or better were included. All patients received intravitreal and topical antibiotics. Management by pars plana vitrectomy or vitreous tap and use or nonuse of subconjunctival antibiotics were at the discretion of the treating physician. RESULTS: There were 59 patients identified; 54 met the follow-up criteria. These patients were divided into two groups based on whether subconjunctival antibiotics were used (group ABX; n = 21) or not used (group noABX; n = 33). The median pretreatment visual acuity was hand motions in both groups. The median age in both groups was 74 years. Etiology, duration of symptoms, vitreous culture organisms, percentage of cases with wound complications such as leaks or vitreous incarceration, and intraocular lens type were similar in the two study groups. Intravitreal and topical antibiotics and corticosteroids used were not significantly different in the two groups, except that topical ceftazidime was used less frequently in group ABX than in group noABX (43% vs. 82%, respectively; P = 0.007). The median follow-up was 13 months in both groups (range: 3-87 months for group ABX and 3-63 months for group noABX). Final visual acuity in groups ABX and noABX was at least 20/50 (33% vs. 39%, respectively), 20/60 to 5/200 (29% vs. 39%, respectively), 4/200 to better than hand motions (0 vs. 3%, respectively), or hand motions or worse (38% vs. 18%, respectively). These differences were not significant (P = 0.37). Reinjection rates (14% vs. 15%, respectively) were also similar in groups ABX and noABX. The additional procedures rate was significantly higher in group ABX than in group noABX (P = 0.024), with cumulative rates of 33% and 3%, respectively, at the 12-month follow-up. CONCLUSIONS: These data suggest that subconjunctival antibiotics may not be necessary to treat infectious endophthalmitis managed with otherwise standard tap and injection techniques and topical antibiotics.