What works for whom? Individual participant data network meta-analysis of exercise for persistent low back pain

Low back pain is a leading cause of disability in Canada and globally. Almost every Canadian suffers from low back pain at some point in their life. Most of the cost associated with low back pain is generated by a small percentage of individuals with persistent symptoms. Exercise therapy is a commonly recommended treatment strategy. However, there is no evidence to support the use of one type of exercise over another and very limited evidence about individual patient characteristics associated with effectiveness of specific treatments. There are two specific directions of research evidence that are needed to answer, 'what works for whom?' for patients with persistent low back pain. Neither of these directions can be pursued with more randomized controlled trials or traditional evidence synthesis projects: 1.Determine what types of exercise therapy, design or delivery characteristics improve its effectiveness for patients with persistent low back pain, and 2.Determine the persistent low back pain patient characteristics and patient subgroups that have better or worse outcomes with exercise therapy. To identify different types and characteristics of exercise programs that may be more effective we will use an approach called network meta-analyses. To identify patient characteristics and subgroups of individual patients who may benefit more from exercise therapy compared to no treatment, non-exercise usual care, and other conservative treatments, we will conduct an analysis using individual patient data from the best available trials. There are currently more than 350 randomized controlled trials of exercise therapy for persistent low back pain. Individually, these trials provide limited evidence and have not advanced decision-making for back pain patients. This team will advance the low back pain field by combining and re-analyzing trial data using best-practice statistical methods.