• The difference in the mean numbers of bacterial colony forming units from pre-irrigation to post-irrigation with PI.
  

• Iodine excretion after surgery, assessed as urine iodine concentration per creatinine clearance.
  
• Postoperative erosion of the cornea and corneal oedema.
  

Background: Endophthalmitis after strabismus surgery in young children leads to blindness and loss of the affected eye. It is caused by conjunctival bacteria. PI solutions between 1% and 5% reduce the number of bacteria on the conjunctiva. The concentration used varies widely among clinics, from 1% to 5%. In vitro studies have shown that PI is paradoxically more effective at lower concentration, but in cataract surgery in elderly, 1% PI has been shown to be less effective than 5% PI. Dilution by tear fluid or binding of PI to proteins in tear fluid may lower its effectiveness. Since endophthalmitis after strabismus surgery especially affects young children and the bacterial flora of the conjunctiva in children is different from that in adults, the cataract PI study should be repeated in young children operated for strabismus.

Objective: To compare the efficacy of 5 % vs. 1.25 % povidone-iodine (PI) as preoperative antiseptic prior to strabismus surgery in children as a prophylaxis of endophthalmitis. Given the low rate of endophthalmitis the conjunctival bacterial flora rate is used as surrogate marker to determine the effectiveness of topical PI in reducing or eliminating bacteria from the ocular surface at the time of the surgery.

Design: The study is a multi-centre, prospective, randomized-controlled, parallel-groups, assessor-blind (microbiological assessments), investigator-initiated trial.

Study population: All children under 6 years of age attending the 15 participating clinics for routine strabismus surgery will be eligible for the study at the point that a strabismus operation is planned. The clinics (5 Dutch, 10 German) will each recruit approximately 20 patients. The minimum sample size is 2 x 100 patients.

Intervention: Diluted PI, 1.25% or 5%, will be prepared in a sterile fashion, and distributed in single-use dispensers. These will be coded for randomization. Before initiation of surgery, children randomized to the 5% PI group will have their conjunctival fornices irrigated with 5 ml PI 5%. Children randomized to the 1.25% PI group will have their conjunctival fornices irrigated with 5 ml PI 1.25%. Conjunctiva cultures for aerobic and anaerobic bacteria will be obtained (1) after general anesthesia has been established, (2) 5-10 min after PI irrigation, (3) after reattachment of the eye muscles and (4) after closing the conjunctiva with sutures.

Primary outcome: The difference in the mean numbers of bacterial colony forming units (CFUs) from pre-irrigation (l) to post-irrigation with PI (2-4).

Secondary outcome: Iodine excretion after surgery, assessed as urine iodine concentration per creatinine clearance.

Postoperative erosion of the cornea and corneal oedema. Both of these have been described as side-effects of PI use.

Nature and extent of the burden and risks associated with participation:

Risks are limited to the act of taking the four bacterial cultures, as both 1.25% PI and 5% PI are approved preoperative antiseptic applications of PI and both are used, rather indiscriminately, by the university departments of ophthalmology participating in the study.