Clinical Question: Do cough suppressants improve the sleep of children with respiratory infections, or the sleep of their parents?

Setting: Outpatient (primary care)

Study Design: Randomized controlled trial (double-blinded)

Synopsis: Although the American Academy of Pediatrics recommends against the use of antitussives in children because of the lack of demonstrated benefit, these products fly off pharmacy shelves in the winter months. This study identified 100 children who had rhinitis and cough symptoms for one week or less and did not have asthma or allergies. The average age of the children was slightly older than four years (range: two to 16 years).

The children were randomized (allocation concealment uncertain) to receive a single dose of placebo, diphenhydramine, or dextromethorphan for the single night of the study. Using a seven-point Likert scale, parents were asked to rate the effect of treatment on the child’s cough frequency, as well as the effect on their own sleep and that of the child. Compared with ratings obtained for the night before the study, parents reported a significant decrease in cough frequency and severity (from “somewhat” to “occasional” on the descriptive scale) overall. The combined symptom score decreased from 19.8 to 8.9 (of a possible 30) with any treatment (P < .01). Parents also reported significant improvement in their sleep and their child’s sleep. However, the results were similar whether the child was treated with drug or placebo. Adverse effects were reported equally in all three study groups. The study had the power to find a one-point change in scores of the three arms, if one truly existed.

Bottom Line: In this single-dose study, placebo worked as well as dextromethorphan and diphenhydramine to decrease cough frequency and severity in children. The active drugs provided no additional benefit on parents’ report of their own or their child’s sleep. Although these drugs do not work better than placebo when their child has a cold, parents can be assured that all products work equally well. (Level of Evidence: 1b–)