Statistics from Altmetric.com
Commentary on: OpenUrl
Standard of care for emergency department (ED) treatment of children with acute asthma exacerbation includes systemic steroids.1 Traditionally, the systemic steroid course of choice has been a 5-day burst of oral prednisone or prednisolone. Recent studies have begun to show that a shorter course of dexamethasone may have comparable efficacy.2–4 Dexamethasone is an attractive alternative to prednisone and prednisolone because of its longer duration of metabolic effect and subsequent shorter treatment duration. Others have touted the palatability of dexamethasone over that of prednisone and prednisolone, as well as potentially improved clinical outcomes and cost savings due to superior compliance with the shorter treatment course dexamethasone requires.5
This was a review of randomised controlled trials that compared dexamethasone to prednisone/prednisolone among children with acute asthma exacerbation in the ED. The primary outcomes analysed were: relapse, including unscheduled clinic visits; return visits to the ED; hospitalisation at either 5, 10–14 or 30 days after the index ED visit. The secondary outcome was vomiting, either in the ED or at home. The review clearly stated the question being asked, the literature search strategy, study selection process and grading of study quality. Data were pooled using a fixed-effects model. The authors assessed for heterogeneity and small study effects. The results were reported as relative risks (RR) and 95% CIs.
Six randomised controlled trials were included. The total number of patients was not reported. Three of the studies used single-dose intramuscular dexamethasone, one used single-dose oral dexamethasone and two used multiple oral doses. The dosing varied greatly between studies, from 0.3 to 1.7 mg/kg; the max single dose ranged from 15 to 36 mg. Prednisone/prednisolone dosing also varied considerably.
The risk of relapse when comparing dexamethasone to prednisone/prednisolone was not statistically significant for any of the time periods studied (5 days: RR=0.90, 95% CI 0.46 to 1.78; 10–14 days: RR=1.14, 95% CI 0.77 to 1.67; 30 days: RR=1.20, 95% CI 0.03 to 56.93). Dexamethasone was associated with less frequent vomiting in the ED (RR=0.29, 95% CI 0.12 to 0.69) as well as at home (RR=0.32, 95% CI 0.14 to 0.74).
This systematic review and meta-analysis attempts to answer the following question: ‘Can I use one or two doses of dexamethasone instead of 5 days of prednisone/prednisolone to treat children with acute asthma exacerbation?’ Unfortunately, a meta-analysis may not be the best way to build evidence for this practice change. The heterogeneity of the available studies makes it difficult to draw conclusions from them when combined. The inclusion of studies that used either oral or intramuscular dosing may have severely affected the internal validity. The wide range of steroid dosing is also of concern. This makes it difficult to determine the effects attributable to the steroid itself, rather than the effects of the dosing or route used in the individual study.
Although the authors suggest that the finding of ‘no difference’ between the risks of dexamethasone and prednisone/prednisolone suggests non-inferiority, all it actually does is show that the study was not powered to find a difference between the two treatments. Because the individual and total sample size of this study is unknown, power to find a difference is hard to judge. The casual reader may misinterpret the conclusion of this meta-analysis as evidence of non-inferiority; ‘no difference’ is certainly not the same as non-inferiority.
Owing to the limited number of trials from which these authors were able to draw data, they were unable to assess the comparative effects of 1-day and 2-day dexamethasone dosing. The author's conclusions regarding vomiting are also limited by the fact that half of the included studies used intramuscular rather than oral dexamethasone. To properly answer this question, oral prednisone and prednisolone must be compared to oral dexamethasone.
Limitations aside, the question the authors raise is an important and timely one, and they do their best to draw conclusions from the existing literature in an attempt to offer an evidence-based treatment recommendation. Their article provides a good summary of the existing evidence on the subject. However, in this case the only way to overcome the problems of combining individual studies with significant heterogeneity is to conduct a randomised controlled trial powered for non-inferiority.
Competing interests None.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.