RESUMO
BACKGROUND: Shoulder pain is the third leading musculoskeletal complaint seen by general practitioners. Physical therapy is often the first line of intervention in this population; however, there is limited description of what constitutes effective physical therapy treatment. No study has examined the effectiveness of therapeutic exercise across all painful shoulder conditions. Our purpose was to examine the effectiveness of therapeutic exercise as an intervention across all pathoanatomic mechanisms of shoulder pain in terms of range of motion (ROM), pain, and function. METHODS: Medline via Ovid, CINAHL (Cumulative Index to Nursing and Allied Health Literature), and the Cochrane Central Register of Controlled Trials were searched from 1997 through March 2011. Randomized controlled trials comparing physical therapist-prescribed exercises against any other type of intervention were included. Articles were qualitatively evaluated by use of the Physiotherapy Evidence Database scale by 5 separate reviewers. Data from included studies were extracted and synthesized with respect to the primary outcomes of ROM, pain, and function. Individual effect sizes were calculated with a standard formula, and overall effect was calculated by use of random- and fixed-effects models. RESULTS: We qualitatively reviewed 19 articles; 17 achieved the criterion of 6 or better on the Physiotherapy Evidence Database scale. Significant heterogeneity in reporting among included studies limited quantitative assessment. Overall, therapeutic exercise has a positive effect on pain and function above all other interventions. The findings for ROM were inconclusive. CONCLUSION: Therapeutic exercise is an effective intervention for the treatment of painful shoulder conditions; however, subsequent research is necessary for translation into clinical practice.