Overall confidence in the results of systematic reviews on exercise therapy for chronic low back pain: a cross-sectional analysis using the Assessing the Methodological Quality of Systematic Reviews (AMSTAR) 2 tool

Abstract

ObjectiveTo evaluate the overall confidence in the results of systematic reviews of exercise therapy for chronic non-specific low back pain using the AMSTAR 2 tool. MethodsPubMed, Embase, Cochrane Database of Systematic Reviews, PEDro and CINAHL was searched up to February 2017. Two independent reviewers selected systematic reviews of randomized controlled trials that investigated exercise therapy in patients with low back pain. AMSTAR 2 assessment was performed by pairs of reviewers, and the overall confidence in the results of the systematic reviews were rated as ‘High’, ‘Moderate’, ‘Low’ and ‘Critically low’. Descriptive analysis was used to summarize the characteristics of included systematic reviews. The percentage of systematic reviews achieving each item from the AMSTAR 2 and the overall confidence in the results were tabulated. ResultsThe search identified 38 systematic reviews. Most of the reviews included a median of 10 clinical trials and total sample size of 813 participants per review. Five of 38 (13%) reviews were Cochrane reviews, and 8 (21%) systematic reviews had a protocol published or registered prospectively. The overall confidence in the results of 28 reviews (74%) was rated as ‘Critically low’, 6 (16%) as ‘Low’, 1 (2%) as Moderate, while 3 of 38 reviews (8%) were rated as ‘High’. ConclusionThe results demonstrate very low confidence in the results of most systematic reviews of exercise in chronic non-specific low back pain. Clinicians are more likely to deliver the most efficacious interventions to patients by critically appraising systematic reviews using AMSTAR 2 before making their decisions.