Antibiotics for acute infective conjunctivitis in children.

Peter Rose and colleagues (July 2, p 37) compared the efficacy of chloramphenicol eye drops with placebo eye drops for the treatment of acute infective conjunctivitis in children. They found that “most children presenting with acute infective conjunctivitis in primary care will get better by themselves and do not need treatment with an antibiotic”. Their recommendations should represent a major sea change in management, but we have some reservations. The trial is subject to selection bias since only an estimated 29% of all children presenting in the study period were recruited. There is evidence that substantial numbers of children with more severe conjunctivitis did not enter into the randomisation stage, and these are the very children who were likely to benefit the most from antibiotic treatment. Furthermore, since chloramphenicol reduces the number of pathogenic bacteria in the eye, failure to treat could result in a higher risk of transmission. Without antibiotic treatment, children would need to be isolated from their peers for longer, causing greater inconvenience and possibly a loss of earnings for the family. Had an objective primary outcome measure been used, independent of the children’s parents, with stratification of conjunctivitis by severity, we feel the study findings would have been more robust.