Abstract
In order to summarize the available clinical evidence for the efficacy of physiotherapy, 400 randomized clinical trials were identified from the literature. Studies were found by using bibliographic databases, citation tracking, and correspondence with researchers in the field. Focusing on disorders of the musculoskeletal system, a number of criterion based meta-analyses were performed on 180 trials in order to summarize the available evidence. For each randomized clinical trial in each meta-analysis a methodological score was calculated using a set of explicit criteria and weighting factors applied by two or three independent reviewers who were blinded as to the outcomes, the journal and the authors of the publication. In each meta-analysis the randomized clinical trials were ordered hierarchically depending on their score for methodological quality. Meta-analyses were performed for spinal manipulation, exercise therapy, traction, ultrasound, and laser therapy, and for disorders of the back, neck, shoulder and knee. In general, the methodological quality of the studies appeared to be low, and the efficacy of physiotherapy was shown to be convincing for only a few indications and treatments. On the other hand, because of the prevalence of serious methodological flaws, it cannot be concluded that physiotherapy has no effect.