Hypertonic saline (HS) for acute bronchiolitis: Systematic review and meta-analysis

Chin Maguire, Hannah Cantrill, Daniel Hind, Mike Bradburn, Mark L. Everard

Acute bronchiolitis is the commonest cause of hospitalisation in infancy. Currently management consists of supportive care and oxygen. A Cochrane review concluded that, "nebulised 3 % saline may significantly reduce the length of hospital stay". We conducted a systematic review of controlled trials of nebulised hypertonic saline (HS) for infants hospitalised with primary acute bronchiolitis. Searches to January 2015 involved: Cochrane Central Register of Controlled Trials; Ovid MEDLINE; Embase; Google Scholar; Web of Science; and, a variety of trials registers. We hand searched Chest, Paediatrics and Journal of Paediatrics on 14 January 2015. Reference lists of eligible trial publications were checked. Randomised or quasi-randomised trials which compared HS versus either normal saline (+/- adjunct treatment) or no treatment were included. Eligible studies involved children less than 2 years old hospitalised due to the first episode of acute bronchiolitis. Two reviewers extracted data to calculate mean differences (MD) and 95 % Confidence Intervals (CIs) for length of hospital stay (LoS-primary outcome), Clinical Severity Score (CSS) and Serious Adverse Events (SAEs). Meta-analysis was undertaken using a fixed effect model, supplemented with additional sensitivity analyses. We investigated statistical heterogeneity using I(2). Risk of bias, within and between studies, was assessed using the Cochrane tool, an outcome reporting bias checklist and a funnel plot. Fifteen trials were included in the systematic review (n = 1922), HS reduced mean LoS by 0.36, (95 % CI 0.50 to 0.22) days, but with considerable heterogeneity (I(2) = 78 %) and sensitivity to alternative analysis methods. A reduction in CSS was observed where assessed [n = 516; MD -1.36, CI -1.52, -1.20]. One trial reported one possible intervention related SAE, no other studies described intervention related SAEs. There is disparity between the overall combined effect on LoS as compared with the negative results from the largest and most precise trials. Together with high levels of heterogeneity, this means that neither individual trials nor pooled estimates provide a firm evidence-base for routine use of HS in inpatient acute bronchiolitis.