Non-specific chronic low back pain is a leading cause of disability-adjusted life-years (DALYs) with a lifetime prevalence of 80%–85%. Although many modalities of treating chronic low back pain are available, pharmacotherapy remains a very important treatment modality, according to many recognised guidelines. Each of these pharmacological agents has evidence supporting its efficacy in treating chronic low back pain, and each of them is associated with side effects. At present, no study has unified these two outcomes or ranked them from the most effective to the least effective. Thus, this study aims to identify pharmacological agents that rank best by conducting multivariate bivariate network meta-analysis.

Four databases (Medline/Pubmed, Cochrane Central Register for Controlled Trials, Cochrane Database for Systematic Reviews and CINAHL) will be systematically searched from their inception to 31 July 2024 for randomised control trials that meet the inclusion criteria. Appropriate data will be extracted from the relevant studies, including the two outcomes of interest: the pain intensity score and the proportion of patients who withdrew from the study because of adverse events. A bivariate Bayesian network meta-analysis will be used to combine the two outcomes of interest to form a single probability that will rank the pharmacotherapy agents from the most effective and safe to the least. The studies will be assessed for risk of bias, and their evidence will be graded using the Grading of Recommendations Assessment, Development, and Evaluation assessment tool. Inconsistency, which is a key assumption of network meta-analysis, will be assessed using both local and global methods.

This systematic review is based on already published data, so it does not require ethical approval. The findings from the review, after completion, will be disseminated via peer-reviewed publications.

CRD42024569192.