Objective
Determine the association between initial treatment method (tap and inject [T/I] or pars plana vitrectomy [PPV]) and microbiology findings on post-treatment visual acuity (VA) in endophthalmitis patients.
Methods
This retrospective study evaluated patients treated for endophthalmitis in Toronto, Canada. The association between treatment, microbiology, and VA at 1-, 3-, 6-, and 12-months post-treatment was analyzed using linear regression, adjusted for age, sex, time to treatment, and baseline VA.
Results
A total of 111 patients were treated with initial T/I (n = 93, 84%) or PPV (n = 18, 16%) between 2011–2023. Microbiology identified 69 (62%) culture-negative, 36 (32%) bacterial, and 6 (5%) fungal cases. Adjusted VA improvement did not differ between bacterial and culture-negative groups at all timepoints, including 1 month (-0.68 ± 0.86 vs. -0.69 ± 0.87 logMAR, p = 0.98, n = 102) and last measured follow-up (-1.01 ± 0.90 vs. -0.84 ± 0.83 logMAR, p = 0.35, n = 105). VA improvement also did not differ between T/I and PPV at all timepoints, including at 1 month (-0.66 ± 0.87 vs. -0.50 ± 0.78 logMAR, p = 0.55, n = 108) and last measured follow-up (-0.91 ± 0.86 vs. -0.60 ± 0.80 logMAR, p = 0.20, n = 111). Last measured follow-up time did not differ between T/I and PPV (6.0 vs. 3.8 months, p = 0.05) or between bacterial and culture-negative groups (6.1 vs. 5.6 months, p = 0.85). Retreatment rates were similar between T/I and PPV (35% vs. 33%, OR = 0.95, p = 0.93), but higher in bacterial vs. culture-negative cases (56% vs. 22%, OR = 4.80, p < 0.01).
Conclusions
Visual outcomes did not differ by treatment or microbiology. Bacterial cases had higher retreatment rates than culture-negative cases.