Outcomes of initial nonoperative treatment of traumatic full-thickness rotator cuff tears

Nonsurgical management of chronic, degenerative rotator cuff tears (RCTs) can be an effective treatment strategy, but there is limited evidence to support conservative treatment of acute, traumatic RCTs. The objective of this study was to assess clinical outcomes and predictors of treatment success in patients with traumatic RCTs who elected for initial nonoperative treatment.

Patients from a single institution were retrospectively identified using diagnostic codes for traumatic RCTs followed by confirmed initial treatment with ≥2 months of physical therapy. The exclusion criteria included surgery within 2 months of injury and greater than grade I fatty infiltration on magnetic resonance imaging. At minimum 2-year follow-up, patients were contacted by telephone to collect interval surgical history and standardized patient-reported outcomes. Physical therapy was considered to have failed in all those who underwent surgical treatment and those with satisfaction ratings of "moderately dissatisfied" or "very dissatisfied."

Follow-up outcomes were obtained in 40 of 49 patients (82%), with an average follow-up time of 4.2 years. Of the RCTs, 9 (22%) were small (<1 cm), 22 (54%) were medium (>1 cm to <3 cm), and 9 (22%) were large (>3 cm to <5 cm). Grade I fatty infiltration was seen on 25% of magnetic resonance imaging scans (n = 10). Rotator cuff repair was performed in 18 patients (45%) following an average of 6 months of nonoperative treatment (range, 3-12 months). Nonoperative treatment was determined to have failed in 23 of 40 patients (58%) in total. Conservative management was more likely to fail in patients with multiple tendons torn (P = .014). Tear size and retraction were not significantly different between patients who underwent surgery and those who did not. Patients who underwent surgical management had an 83% satisfaction rate at final follow-up compared with a 55% satisfaction rate for patients who did not undergo surgery (P = .054). There was no statistically significant difference in the American Shoulder and Elbow Surgeons score or visual analog scale score between these groups. Although patients who underwent surgery had a higher mean Single Alpha Numeric Evaluation score (86.3 vs. 75.1, P = .041), this difference was below the previously established minimal clinically important difference.

Nonoperative treatment remains a viable option for certain patients with traumatic RCTs; however, the results of our study demonstrate a considerable early failure rate. This study further supports historical literature demonstrating reliably successful outcomes with surgical treatment of acute, traumatic RCTs.