Reducing bacterial contamination inside fluid catch bag in 25-gauge vitrectomy by use of 0.25 % povidone-iodine ocular surface irrigation.

To examine the bacterial detection rate in infusion fluid collected inside the fluid catch bag during 25-gauge (25G) vitrectomy when the ocular surface was irrigated with infusion fluid or 0.25 % povidone-iodine. Two groups using different fluids for ocular surface irrigation during 25G vitrectomy were studied. Fifty-five consecutive eyes received ocular surface irrigation with infusion fluid (IF group) and 52 consecutive eyes with 0.25 % povidone-iodine (PI group). Samples of ocular surface fluid were collected at the beginning of surgery and samples of infusion fluid inside the fluid catch bag were collected at the end of surgery for bacteriological cultures. At the beginning of surgery, the bacterial detection rates in ocular surface fluid samples were 5.8 % (3 of 52 eyes) in the IF group and 7.7 % (4 of 52 eyes) in the PI group, with no significant difference (P = 0.6955). At the end of surgery, the bacterial detection rates in infusion fluid collected inside the fluid catch bag were 23.1 % (12 of 52 eyes) in the IF group and 3.8 % (2 of 52 eyes) in the PI group, with a significant difference (P = 0.0041). No endophthalmitis occurred in either group. These results demonstrate the risk of bacterial contamination when surgical instruments fall accidentally into the fluid catch bag during conventional 25G vitrectomy. Irrigating the ocular surface with 0.25 % povidone-iodine instead of infusion fluid significantly reduces the bacterial contamination rate in the fluid catch bag.