• sodium cromoglycate
• children
• asthma

We note with interest the further correspondence concerning the systematic review of inhaled sodium cromoglycate as maintenance therapy in children with asthma¹ from Professor M Silverman² and Dr G Laszlo.³

Professor Silverman asked the authors to carry out separate calculations of the size of the treatment effect between inhaled sodium cromoglycate and placebo for school children and pre-school children. The results of these calculations show that the mean difference between inhaled sodium cromoglycate and placebo is greater for school children (cough 0.26, wheeze 0.29) than for pre-school children (cough 0.12, wheeze 0.08). In both age groups these differences are significant and in favour of sodium cromoglycate. In fig 1 we illustrate these differences and the comparison with all children taken from the original review.

These results support our own conclusion⁴ that the drug is either less effective in pre-school children or the apparent lesser effect is related to the difficulties in trials in younger children. In their reply to Silverman the authors dismiss his suggestion that sodium cromoglycate is more effective in schoolchildren by stating that results from early trials in older children were likely to be unreliable by implying that the trials were flawed. Why then did they include them in their review? This hardly seems a valid argument. Tasche et al state that the finding that the 95% CI tolerance interval for cough (not wheeze as in their reply) in school children includes zero also supports the view that this conclusion should be dismissed. We have already pointed out that we believe that their interpretation of the tolerance interval is incorrect and misleading.

Systematic reviews of treatments for important diseases are likely to be used in the development of treatment guidelines for doctors. In such reviews the conclusions of the authors of the review are likely to be taken into consideration. In this case the authors concluded that “there is insufficient evidence for a beneficial effect of sodium cromoglycate as a maintenance treatment in children with asthma”. This conclusion was not supported by their original review and, in the case of school children, is certainly not supported by the new calculations as requested by Silverman. In their review of the 12 trials conducted in children aged 5–17 years, they classified 11 as positive and one as positive/equal. In the pooled placebo groups the reported mean daily symptom score was 0.8 and the size of the treatment effects, particularly in school children, is certainly beneficial and of clinical importance. In the light of both the statistical and clinical benefits seen with sodium cromoglycate, we would therefore suggest that the authors be asked to reconsider their conclusions.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

95% Confidence intervals of the mean difference between sodium cromoglycate and placebo on the symptoms of cough and wheeze.