Initial treatment of complete rotator cuff tear and transition to surgical treatment: systematic review of the evidence

Background: rotator cuff tear affects many people. Natural history, and evidence for non-operative treatment remains limited. Our objective is to assess evidence available for the efficacy and morbidity of commonly used systemic medications, physiotherapy, and injections alongside evaluating any negative long-term effects. Methods: a systematic search was performed of PubMed, Cochrane, EMBASE and CINAHL dates (1 January 1960 -1 December 2014), search terms: 'rotator cuff tear', 'natural history', 'atraumatic', 'injection', 'physiotherapy'or 'physical therapy', 'injection', 'corticosteroid', 'PRP', 'MSC', risk of conservative treatment', and 'surgical indication'. Results: eleven studies were included. The mean Coleman Methodology Score modified for conservative therapy is 69.21 (range 88-44) (SD 12.31). This included 2 RCTs, 7 prospective, and 2 retrospective studies. Evidence suggests it is safe to monitor symptomatic rotator cuff tears, as tear size and symptoms are not correlated with pain, function, and/or ultimate outcome. Conclusions: complete rotator cuff tears may be effectively treated with injections, exercise in the short and intermediate terms respectively. Negative effect of corticosteroids on rotator cuff tissue has not been demonstrated. Timing to end conservative treatment is unknown, but likely indicated when a patient demonstrates increased weakness and loss of function not recoverable by physiotherapy.