Endophthalmitis After Pars Plana Vitrectomy: Clinical Features, Risk Factors, and Management Outcomes

Purpose: To report the clinical presentations, risk factors, and outcomes of endophthalmitis after pars plana vitrectomy at 4 tertiary eye care centers of an institute in South India. Design: A retrospective case series. Methods: The records of 38,591 patients undergoing vitrectomy were reviewed using the coding assigned by the medical records department. Consecutive cases diagnosed as endophthalmitis after pars plana vitrectomy (PPV) between 1990 and 2014 for various indications were analyzed. Results: The clinical incidence of postvitrectomy endophthalmitis was 0.052%, and culture-positive incidence of postvitrectomy endophthalmitis was 0.031%. Twelve cases (60%) were culture positive. Mean presenting vision was 2.16 +/- 1.51 logMAR (Snellen equivalent 20/2890). Seventeen eyes had received sutureless vitreous surgery (15 cases 23G, 2 cases 25G) and 3 eyes had received 20G suture-assisted vitreous surgery (P < 0.0001). The odds of developing endophthalmitis in sutureless versus sutured vitrectomy were 25.14 [95% confidence interval (CI), 7.37-85.84] (P < 0.0001) and those of developing endophthalmitis in sutureless surgery versus sutured with final tamponade of Ringer lactate (RL) were 19.53 (95% CI, 5.37-71.03) (P < 0.0001). In sutureless surgeries, the odds of developing endophthalmitis in RL tamponaded eyes versus non-RL ones was 4.39 (95% CI, 1.67-11.56) (P = 0.002). Mean interval between vitreous surgery and endophthalmitis was 4 +/- 6.89 days; median, 1.5 days. Mean postoperative vision was 1.7 +/- 1.36 logMAR (Snellen equivalent 20/1002) (P = 0.31). Conclusions: Endophthalmitis after vitrectomy is an acute presentation. Sutureless surgery, especially with aqueous tamponade, has a higher risk. The visual outcome is relatively poor.