BACKGROUND: Disodium cromoglycate (DSCG) is included in the BTS guidelines on the treatment of asthma for use in children, but is now used only infrequently. We have identified and interpreted the findings of all published randomised, placebo controlled trials of DSCG in the prophylactic treatment of children with asthma.
METHODS: Several databases were searched to identify trials. Studies were included if they investigated subjects with asthma aged 0-18 years old, addressed maintenance treatment with inhaled DSCG, and were published in English. The methodological quality of the studies was assessed independently by three reviewers. The 95% confidence intervals (CI) of differences in the treatment effect for cough and wheeze between placebo and treatment with DSCG were computed. The estimates were pooled and tested for homogeneity and, to assess possible publication bias, a funnel plot was made and tested for symmetry.
RESULTS: Of the 24 randomised, placebo controlled trials identified, the methodological scores varied widely. The null hypothesis of homogeneity was rejected. Under the assumption of heterogeneity the overall CI for wheeze was 0.11 to 0.26 and for cough was 0.13 to 0.27. The overall tolerance intervals (-0.11 to 0. 48 and -0.04 to 0.43 for wheeze and cough, respectively) both included zero, so it cannot be concluded that future studies will show an effect of DSCG compared with placebo. Older studies were more often in favour of DSCG. The funnel plots suggest publication bias; small studies with negative or equal outcomes are lacking.
CONCLUSION: Given the apparent publication bias, the small overall treatment effect, and the tolerance interval including zero, there is insufficient evidence that DSCG has a beneficial effect as maintenance treatment in children with asthma.
|Number of pages||8|
|Publication status||Published - Nov 2000|
- Administration, Inhalation
- Anti-Asthmatic Agents/administration & dosage
- Asthma/drug therapy
- Child, Preschool
- Cromolyn Sodium/administration & dosage
- Data Interpretation, Statistical
- Infant, Newborn
- Meta-Analysis as Topic
- Randomized Controlled Trials as Topic