Data do not justify study's conclusions

EDITOR—Chapman has highlighted the confusion caused by the different interpretations of the report by the International Agency for Research on Cancer on the effects of passive smoking.1 By ignoring the size and direction of the effect and focusing on the lower limit of the confidence interval the agency came to the erroneous conclusion that passive smoking does not cause lung cancer. Unfortunately, Fahey et al have fallen into the same trap in reporting the results of a systematic review of the use of antibiotics in acute cough.2

They state categorically in their discussion: “This systematic review shows that antibiotic treatment has no effect on the resolution of acute cough.” This conclusion is not justified by the data in their review. Two of the outcomes measured—the resolution of productive cough and clinical improvement—show a pooled effect that favours antibiotics but does not reach significance at the 95% level when a random effects model is used. The authors seem to have confused the significance of these findings with the size of the effect. There is around a 1 in 40 chance of this pooled result arising because of random variation rather than because of a real difference between antibiotic and placebo; this is hardly grounds to claim that the review shows that antibiotics have no effect.

The authors do not show an even handed approach when they deal with the data concerning the efficacy of antibiotic and side effects of treatment. In the case of efficacy they state that “antibiotic treatment was no better than placebo,” and in the case of side effects they state that the data showed “a non-significant increase in the risk of side effects from antibiotics.” They then proceed to exclude the only trial that showed more side effects in the placebo group than the antibiotic …