期刊正文

Randomized controlled trials are difficult to conduct, explaining their paucity in the literature on the management of rotator cuff disease and the difficulty in establishing definitive guidelines for treatment.

This study by Moosmayer and colleagues represents one of the few randomized trials on patients with rotator cuff tears that compares surgical repair with nonoperative treatment. Although rotator cuff surgery is increasing in frequency, this study should be troubling to surgeons as the amount of improvement obtained with surgery is not great compared with a standard exercise program and, in fact, the gains in the Constant score are below the threshold for a clinically important difference.

It is also interesting to note that the failure of treatment for the two groups was similar. Of the fifty-one patients randomized to physiotherapy, nonoperative treatment was successful in 76% (thirty-nine patients) and unsuccessful in 24% (twelve patients) if failure is defined as crossing over to surgery. For sixty of the sixty-four patients treated with a rotator cuff repair, assessment of the integrity of the tendon was performed by sonography after five years, and postoperative imaging demonstrated complete healing in 75% (forty-five patients), with partial failure in 12% (seven patients) and complete failure in 13% (eight patients).

There were several limitations of this study. First, the authors report that, of the fifty-two cases in the primary tendon repair group, eighteen (35%) had a concomitant biceps tenodesis performed in addition to the rotator cuff repair. Unfortunately, this introduces considerable performance bias into the study. It is reasonable to expect that at least some of the improvement in outcomes in the primary tendon repair group may be related to the treatment of the biceps tendon and not to the rotator cuff repair.

Second, the authors did not distinguish outcomes for the three different types of rotator cuff tear mechanisms: acute tears, acute-on-chronic tears, and atraumatic rotator cuff tears in their populations of patients. Although the randomized design should stratify these patients into the surgical repair and nonoperative groups equally, when these patients are combined, we cannot know if surgery may produce better outcomes for one of the mechanisms that produced the rotator cuff tear. In fact, some lower-level evidence has suggested that rotator cuff repair produces better results than nonoperative treatment for acute tears with functional loss. Interestingly, Kukkonen et al. conducted a randomized trial in patients with atraumatic rotator cuff tears for the use of home exercises; acromioplasty and home exercises; or rotator cuff repair, acromioplasty, and home exercises. Those authors failed to detect a difference in the Constant score at the one-year follow-up, suggesting that surgical repair may not be beneficial in the short term in patients with atraumatic rotator cuff tears.