Cohort Validation of the Endophthalmitis Vitrectomy Study for Generalized Etiologies

To evaluate longitudinal visual acuity (VA) outcomes in patients with endophthalmitis of any etiology at a university-affiliated hospital, using a modified Endophthalmitis Vitrectomy Study (EVS) framework comparing immediate pars plana vitrectomy (PPV), tap and injection (TAP), and TAP followed by PPV.

The study included 125 patients diagnosed with endophthalmitis between 2011 and 2023. Patients were categorized into three treatment groups: TAP, PPV, and TAP followed by PPV. Data included demographics, etiology, causative organisms, adjunctive treatments, and VA at presentation, postoperative month 3 (POM3), and final follow-up. Statistical tests included t-tests, ANOVA, Kaplan-Meier analysis, and Mann-Whitney U-tests (p < 0.05).

Mean patient age was 61 years; 67.2% were male, and 45.6% were transferred from another facility. Endogenous endophthalmitis was the most common etiology (32.8%), and bacterial organisms accounted for 41.6% of cases. Mean LogMAR VA at presentation was 1.984; 76.8% had VA ≤ 20/800. Fungal cases had better initial VA than bacterial (p < 0.001), though final VA was similar. Poorer initial VA was associated with primary PPV (p < 0.001). Final VA did not differ significantly among treatment groups, though patients with light perception (LP) or worse trended toward better outcomes with PPV. Systemic antimicrobials and dexamethasone injections were not associated with improved outcomes.

Initial VA strongly influenced treatment. While overall outcomes were similar across groups, patients with LP or worse may benefit from primary PPV. EVS principles may apply broadly across endophthalmitis etiologies.