Incidence of endophthalmitis after 20- and 25-gauge vitrectomy causes and prevention

Abstract

Purpose: To compare endophthalmitis incidence after inpatient 20-gauge (20-G) and 25-G vitrectomies, and to examine the causes and prevention of postvitrectomy endophthalmitis.

Design: Retrospective, interventional, comparative cohort study.

Participants: Six thousand nine hundred thirty-five consecutive patients undergoing pars plana vitrectomy.

Methods: We compared the incidence of endophthalmitis in 3592 consecutive eyes that underwent 20-G vitrectomy between January 2000 and September 2004, and 3343 consecutive eyes that underwent 25-G vitrectomy between April 2004 and December 2007. For 25-G vitrectomy, 542 eyes with sclerotomies produced by straight incision and 2801 eyes with angled incisions were also compared. From 85 eyes that underwent 20-G vitrectomy and 128 eyes that underwent 25-G vitrectomy, ocular surface irrigation fluid and vitreous samples were collected at the end of surgery for bacterial culture.

Main outcome measures: Incidence of postvitrectomy endophthalmitis.

Results: The incidence of postoperative endophthalmitis was 0.0278% (1 of 3592 eyes) for 20-G vitrectomies and 0.0299% (1 of 3343 eyes) for 25-G vitrectomies, with no significant difference. Two eyes developed endophthalmitis after vitrectomy, and visual acuity deteriorated to no light perception despite emergency vitreous surgery. The causative bacteria were methicillin-resistant Staphylococcus aureus and Enterococcus faecali; both were resistant to postoperative antibiotics. In 25-G vitrectomy, the endophthalmitis incidence was 0.18% (1/542 eyes) for straight incision versus 0% (0/2801 eyes) for angled incision, with no significant difference (P = 0.1621). Bacterial contamination rates in ocular surface irrigation fluid and the vitreous were 5.9% (5/85 eyes) and 1.2% (1/85 eyes), respectively, in 20-G vitrectomies, and 5.5% (7/128 eyes) and 2.3% (3/128 eyes) in 25-G vitrectomies, with no significant difference.

Conclusions: The incidence of endophthalmitis was 0.03% for both 20-G and 25-G vitrectomies. This is the first data set to demonstrate no statistically significant difference between endophthalmitis rates in 20-G and 25-G vitrectomy. At the completion of 25-G vitrectomy, the ocular surface irrigation fluid and vitreous were, on rare occasion, contaminated by antibiotic-resistant bacteria. In 25-G vitrectomy, conjunctival irrigation, ensuring sclerotomy closure, and excision of peripheral vitreous may contribute to the prevention of postvitrectomy endophthalmitis.

Financial disclosure(s): The authors have no proprietary or commercial interest in any materials discussed in this article.