Rotator cuff disease is a fascinating clinical entity. Despite decades of treating rotator cuff tears, the evidence keeps mounting that the exact etiology of symptoms and the predictors of success after treatment are still not entirely known.
The purpose of the study by Russell et al. was to review the highest level of evidence available on the correlation between structural integrity of the rotator cuff after surgical repair and patient outcome measures. A combination of meta-analysis and systematic review methodology was utilized to examine this question. This study examines how well the two most commonly reported end points in rotator cuff surgery, patient clinical outcome measures and radiographic healing, correlate with each other.
There can be great power in meta-analyses and systematic reviews. The level of evidence of these studies is determined by the level of evidence of the included research. In this study, the authors included only Level-I and Level-II studies. Hence, these studies represent high-level evidence on a large number of patients. The authors only included studies with a minimum of one year of follow-up, at least one outcome measure, and imaging evaluation of healing using magnetic resonance imaging, computed tomographic arthography, or ultrasound. Overall, fourteen studies (five Level-I studies and nine Level-II studies) met inclusion criteria with an impressive combined total of 861 patients.
Russell et al. pooled studies that used similar outcome measures. Four studies used the Constant score, five studies used the American Shoulder and Elbow Surgeons (ASES) score, four studies used the University of California Los Angeles (UCLA) shoulder score, and three studies used the visual analog scale (VAS) score. Two studies examined shoulder external rotation strength, and three studies examined forward flexion strength. Overall, 78% of the rotator cuff repairs were healed on postoperative imaging. A significant improvement was noted for the Constant, VAS, and UCLA scores in patients with intact rotator cuff repairs. The ASES score was not significantly different. However, despite achieving significance, the results did not meet known benchmarks for meeting the minimal clinically important difference. Validated outcome measures have minimal clinically important difference values that allow for the determination of whether numerical differences in scores can be deemed significant. The UCLA and Constant scores are not validated. However, the differences seen between intact repairs and recurrent tears did not meet levels that would be considered clinically important. The validated VAS and ASES scores did not meet the minimal clinically important difference benchmarks either.
The included studies do seem to support that a healed rotator cuff repair is associated with increased forward flexion strength of approximately five pounds. There is no minimal clinically important difference for strength. However, this is useful information to pass along to patients considering rotator cuff surgery, especially if regaining strength is a priority for the patient.
This study summarizes what many in orthopaedics have recognized over the years. The vast majority of patients are subjectively improved after rotator cuff repair surgery, despite evidence that a variable percentage of repairs either do not heal or do retear1. However, the current study demonstrates that the presence of an intact repair does not lead to a significantly improved clinical outcome based on patient outcome measures. Taking a step back, these results also call into question whether the rotator cuff tear itself is solely responsible for the symptoms that led a patient to undergo an operation. It is possible that the underlying cause for a patient’s pain and dissatisfaction that contribute to patient-related outcome measures is something other than the rotator cuff tear itself, given that patient outcome measures are similar whether the repair is intact or not after surgery.
Again, rotator cuff disease is fascinating. It is known that a substantial portion of the population has asymptomatic rotator cuff tears2. It has also been shown that patients with an atraumatic, full-thickness rotator cuff tear can be successfully managed with conservative treatment3. Yet failure of conservative measures, patient activity levels, and expectations often lead to surgical management to repair symptomatic patients with rotator cuff tears.
The results of the current study are extremely helpful to the surgeon performing surgery on patients with rotator cuff tears. Ideally, the rotator cuff heals after surgical intervention. However, these data now support counseling patients who, even if their repair does not completely heal, are very likely to clinically do approximately as well as if the rotator cuff did completely heal. Most surgeons are primarily interested in a good clinical outcome, with a secondary goal being an improved patient with healing on imaging studies.
Lastly, this study suggests that there may be little utility for routine imaging evaluation of rotator cuff repairs, especially if the patient is doing well. It was not terribly long ago that all anterior cruciate ligament surgeries were judged merely by the objective KT-1000 arthrometer. It has subsequently been observed that knee laxity measures do not correlate with validated patient outcome measures4. Fortunately or unfortunately, rotator cuff repair surgery and anterior cruciate ligament reconstruction surgery both demonstrate a lack of correlation between some common objective and subjective outcomes used in the research and clinical settings.
The limitations of the current study were well recognized. There was a heterogeneous group of study designs included using various surgical techniques, outcome measures, and radiographic imaging methods. It was recognized that the minimum time interval to postoperative imaging was only six months, and ideally, this would be longer. Lastly, the meta-analysis portion of the study was limited by the fact that not all of the included studies used the same outcome measures, some of which are not validated and do not have a minimal clinically important difference.
This study by Russell et al. adds more evidence to the puzzle of outcomes after rotator cuff repair. It also highlights the need for further high level research into the best indications for surgery and the predictors of clinical outcomes after surgical intervention.