interval, which one is inclined to use given the other shortcomings of the study (relatively small sample size with recurrences, not all participants were still blinded at the time of the questionnaire, the diagnosis of AOM was not verified, and the questionnaire response rates were less than ideal), the findings do not seem clinically significant (ie, recurrences 1.1 times more common in the children who had received antibiotics). In summary, we feel that modeling a higher rate of recurrences in children receiving antibiotics, as the letter suggests, would be contrary to the bulk of the available evidence on the topic.