For the past decade, there have been studies appearing on the use of hypertonic saline for treatment of bronchiolitis. Some of these studies were favorable, and some less so—prompting systematic reviews and meta-analyses to be published that seemed to initially suggest there was merit in trying hypertonic saline in a baby with acute viral bronchiolitis. Unfortunately, meta-analyses can be deceiving depending on the characteristics of the included studies. A statistical technique called trial sequential analysis (TSA) can account for potential sources of error in conventional meta-analysis. TSA is just what Harrison et al. (10.1542/peds.2018-1144) have done in an article being early released this week and suggested findings from meta-analyses of a benefit in reducing the likelihood of admission or length of stay for hospitalized children may represent type 1 errors (i.e., falsely claiming a difference). So what do we do now?
We asked two hospitalists (Drs. Katherine Auger and Michelle Parker) and a biostatistician (Dr. Bin Huang) from Cincinnati Children’s Hospital to weigh in with an accompanying commentary (10.1542/peds.2018-1868). The authors further explain TSA and how it can be applied to a meta-analysis. They note that it is still possible certain studies were not included in the meta-analyses used by Harrison et al. and that the TSA method may not be correct for small differences. Before you turn the light completely out on hypertonic saline, Auger et al. call for more definitive large sample studies. Until then, the decision to use hypertonic saline is still one that eludes us—although papers like the one from Harrison et al. certainly do not suggest that this particular treatment for bronchiolitis is the panacea we’ve been looking for. Are you still using it? Do you think it works? Share with us your thoughts on this article and commentary by offering a comment to this blog, to the article on our website, or on our Facebook or Twitter pages.
