• BACKGROUND
    • Tendon repair and physiotherapy are frequently used treatment methods for small and medium-sized rotator cuff tears. In 2 previous publications of the 1 and 5-year results of this study, we reported significant but small between-group differences in favor of tendon repair. Long-term results are needed to assess whether the results in both groups remain stable over time.
  • METHODS
    • In this study, 103 patients with a rotator cuff tear not exceeding 3 cm were randomly assigned to primary tendon repair or physiotherapy with optional secondary repair. Blinded follow-up was performed after 6 months and 1, 2, 5, and 10 years. Outcome measures included the Constant score; the self-report section of the American Shoulder and Elbow Surgeons score; the measurement of shoulder pain, motion, and strength; and patient satisfaction. Magnetic resonance imaging (MRI) was performed on surgically treated shoulders after 1 year, and ultrasound was performed on all shoulders after 5 and 10 years. The main analysis was by 1-way analysis of covariance and by intention to treat.
  • RESULTS
    • Ninety-one of 103 patients attended the last follow-up. After 10 years, the results were better for primary tendon repair, by 9.6 points on the Constant score (p = 0.002), 15.7 points on the American Shoulder and Elbow Surgeons score (p < 0.001), 1.8 cm on a 10-cm visual analog scale for pain (p < 0.001), 19.6° for pain-free abduction (p = 0.007), and 14.3° for pain-free flexion (p = 0.01). Fourteen patients had crossed over from physiotherapy to secondary surgery and had an outcome on the Constant score that was 10.0 points inferior compared with that of the primary tendon repair group (p = 0.03).
  • CONCLUSIONS
    • At 10 years, the differences in outcome between primary tendon repair and physiotherapy for small and medium-sized rotator cuff tears had increased, with better results for primary tendon repair.
  • LEVEL OF EVIDENCE
    • Therapeutic Level I. See Instructions for Authors for a complete description of levels of evidence.