Chronic rotator cuff tears are prevalent and can be disabling. The existing literature is unclear regarding the effectiveness of nonoperative treatment. The purposes of this study were to determine whether the outcome of nonoperative treatment can be predicted on the basis of the presenting clinical characteristics and whether the outcome achieved at three months after treatment can be maintained at two years.
The prospective cohort included ninety-three patients with a documented chronic full-thickness rotator cuff tear. Patients underwent a three-month supervised program of nonoperative treatment and were then evaluated by an orthopaedic surgeon. The treatment outcome was defined as a success if surgical treatment was no longer deemed appropriate by both patient and surgeon because the patient had improved considerably and was predominantly asymptomatic. The outcome was defined as a failure if the patient elected to have surgery after failing to improve and remaining symptomatic. The presenting clinical characteristics that were analyzed included age, sex, smoking status, hand dominance, duration of symptoms, onset (traumatic etiology or insidious onset), shoulder motion, external rotation strength, tear size as documented by ultrasonography or magnetic resonance imaging, and the Rotator Cuff Quality-of-Life Index (RC-QOL).
Seventy (75%) of the patients were successfully treated. Logistic regression analysis showed that the baseline RC-QOL score was a significant predictor of outcome (p = 0.017). Eighty-nine percent of patients maintained their three-month outcome at two years of follow-up.
The RC-QOL was predictive of the outcome of nonoperative treatment of patients with a chronic full-thickness rotator cuff tear. Patients in whom the nonoperative treatment was deemed successful at the conclusion of three months of treatment had a very high chance of ongoing success at two years after the initiation of treatment.
Level of Evidence:
Prognostic Level II. See Instructions for Authors for a complete description of levels of evidence.