Management and outcomes of postoperative endophthalmitis since the endophthalmitis vitrectomy study: the Endophthalmitis Population Study of Western Australia (EPSWA)'s fifth report.

PURPOSE: To examine if changes in the diagnosis and management of postoperative endophthalmitis have occurred since 1995, and to identify factors that might predict final visual outcome.

DESIGN: Retrospective, population-based, noncomparative, consecutive case series.

PARTICIPANTS: Patients with clinically diagnosed endophthalmitis after cataract surgery and lens-related surgery in Western Australia from 1980 to 2000.

METHODS: Endophthalmitis cases were identified using record linkage and cross-referencing with the surgical logbooks of vitreoretinal surgeons before validation by medical record review.

MAIN OUTCOME MEASURES: Microbiological data (microorganisms isolated and antibiotic susceptibilities), diagnostic interventions, surgical procedures, therapeutic interventions, and visual acuity (VA).

RESULTS: During the 21-year period, 213 episodes of endophthalmitis occurred after cataract surgery. Since 1995, both anterior chamber sampling and vitreous sampling have increased significantly. The overall use of vitrectomy has also increased, but we did not observe a difference according to presenting VA. Intravitreal antibiotic use increased significantly, whereas the use of both subconjunctival and IV antibiotics decreased. In one third of patients, the VA at least 6 months after admission for endophthalmitis was worse than 6/18. This was associated with treatment that did not include the use of oral antibiotics (odds ratio [OR], 3.86; 95% confidence interval [CI], 1.21-12.39; P = 0.02), growth from intraocular samples of organisms other than coagulase-negative staphylococci (OR, 9.84; 95% CI, 2.84-34.09; P<0.001), and a discharge VA worse than 6/18 (OR, 6.10; 95% CI, 1.63-22.89; P = 0.01).

CONCLUSIONS: Although we observed noticeable changes in the diagnosis and management of endophthalmitis since 1995, visual outcomes have not improved and remain poor. Our finding that treatment with oral antibiotics may be associated with a better visual outcome warrants further investigation.