Range of motion at 6 weeks is an independent risk factor for non-healing at 6 months after arthroscopic supraspinatus repair.

Abstract

PURPOSE: This study aimed to analyze the effect of the range of motion (ROM) progression pattern on healing after supraspinatus (SSN) repair. HYPOTHESIS: We hypothesized that a faster ROM recovery in the immediate postoperative period would be detrimental for tendon healing. Methods: This was a retrospective study of prospective collected data. All primary arthroscopic SSN repairs for full-thickness tears, with retraction Patte grade < 2, fatty infiltration Goutallier grade < 3 and no glenohumeral osteoarthritis were eligible. Patients who did not attend all follow-ups were excluded. One surgeon performed all arthroscopic repairs using a double-row technique. All patients followed the same rehabilitation protocol with an abduction pillow and passive ROM for 6 weeks. One observer assessed all patients preoperatively and at 1.5, 3 and 6 months postoperatively. Collected outcome measures included: ROM, Visual Analogue Scale (VAS) for pain and Constant Score (CS). Repair integrity was evaluated at 6 months by ultrasound. 2 groups were established based on repair integrity (healed/non-healed). A multivariable analysis was performed to assess potential confounding factors and ROC curve analysis was used to identify predictive thresholds for non-healing. Results: Of 1397 consecutive SSN repairs, 169 were excluded and 1207 were finally included. The overall healing rate was 87.3%. All preoperative outcome measures were similar between groups. Non-healed repairs at 6 months were slightly younger (57.8±7.9 vs 61.6±8.8, p < 0.001) and had a faster postoperative ROM recovery pattern. This difference was most distinct at 6 weeks in passive anterior elevation (AE) (143±22 vs 128±26, p < 0.001) and external rotation (ER) (32±17 vs 23±17,p < 0.001) but was fully compensated by 6 months. VAS pain and CS at each follow-up were similar between groups. Multivariable analysis confirmed 6-weeks passive EA (OR:1.36) and ER (OR:1.16) as independent risk factors for non-healing. The non-healing rate increased when 6-weeks passive EA >140° (20.4 vs 7.9%) and ER > 30° (18.4 vs 8.8%). Conclusions: In a large cohort of isolated arthroscopic SSN repairs treated with the same surgical technique and rehabilitation protocol, a faster postoperative ROM progression pattern, reflected by higher 6-weeks AE and ER, was associated with an increased rate of non-healing. This study illustrates the benefit of a conservative rehabilitation protocol on tendon healing without affecting ROM at 6 months.