To Örebro University

Background: Subacromial pain is a common and disabling condition with multifactorial etiology. Increasing evidence supports exercises as first-line treatment and need of surgery is debated. Long-term follow-ups after surgical-and nonsurgical treatment are scarce. The primary aim of the present study was to investigate the 10-year outcomes after a study comparing specific and nonspecific exercise treatment and the need for surgery. Secondarily we compared patients who had undergone exercise treatment, to those having surgery and explored the importance of rotator cuff status in relation to given treatment and outcomes.

Methods: At enrollment 2009-2010, 97 patients with long-standing subacromial pain were on the waiting list for arthroscopic subacromial decompression (ASD). They were randomized to specific exercises focusing on strengthening of the rotator cuff and scapula stabilizers (n = 51) or control: unloaded range of motion exercises (n = 46). ASD was optional during the entire observation time. Eligible patients from either group constitute the present 10-year cohort with nonoperated (n = 42) or operated (n = 41) patients. The primary outcome was shoulder function and pain after 10 years assessed by the Constant-Murley score (CMS). Secondary outcomes were proportion of patients choosing surgery and rotator cuff status related to treatment and primary outcome.

Results: At the 10-year follow-up, 83 of 97 patients (86%) participated. All patients significantly improved in CMS from baseline to 10-year follow-up, mean improvement of 37 (95% confidence interval 33-41, P <.0001). Nonoperated patients had significantly better CMS compared to operated with mean difference 11(95% confidence interval 4-18) (P = .003). Significantly more patients in the control exercise group 65% (26 of 40) had chosen surgery compared to 35% (15 of 43) in the specific exercise group up until 10 years (P = .006). At 10 years, 55% of the patients had a partial or full-thickness rotator cuff tear compared to 28% at baseline. There was no difference in tear progression between operated and nonoperated patients (P = .494). In the contralateral shoulder, 51% of the patients (39 of 76) had a cuff tear compared to 3% at baseline.

Conclusions: Specific exercise treatment for patients with subacromial pain was effective and reduced the need for surgery with maintained results after 10 years. Responders to exercise treatment had the best long-term outcomes and ASD yielded satisfying outcome in nonresponders. A rotator cuff tear was equally common in patients having undergone surgery as in those treated nonoperatively and in the contralateral shoulder.