Relative effectiveness of different forms of exercises for treatment of chronic low back pain: protocol for a systematic review incorporating Bayesian network meta-analysis

Introduction Exercise is considered as an effective intervention in the management of patients with chronic low back pain (cLBP). However, the relative effectiveness as well as the hierarchy of exercise interventions have not been well established, although various exercise options are available. Therefore, the present protocol proposes to conduct a network meta-analysis (NMA) aiming to evaluate the effectiveness of different forms of exercise for treatment of cLBP. Methods and analysis Medline, Embase, PsycINFO, the Cumulative Index to Nursing and Allied Health Literature, the Cochrane Central Register of Controlled Trials, and the Physiotherapy Evidence Database will be searched to identify all randomised controlled trials that evaluate the effectiveness of exercise in the treatment of cLBP. There will be no restrictions on date or language. Two authors will screen the literature and extract data independently based on predesigned rules, and evaluate the risk of bias of included studies using the Cochrane Risk of Bias Tool. Disagreements will be resolved through discussion or consultation with a senior reviewer. The primary outcomes of this study will be pain relief and improvement in function or disability for all interventions. Traditional pairwise meta-analysis and Bayesian NMA will be conducted to compare the effectiveness of different exercise interventions. The ranking probabilities for all interventions will be estimated and the hierarchy of each intervention will be summarised as surface under the cumulative ranking curve. The quality of evidence will be assessed using the Grading of Recommendations Assessment, Development and Evaluation instrument. Ethics and dissemination Ethical approval and informed consent are not required since this is a protocol for a meta-analysis with no confidential personal data to be collected. The results of this NMA will be submitted to a peer-reviewed journal for publication. PROSPERO registration number CRD42018090576.