Copyright © 2008 by The Journal of Bone and Joint Surgery, Inc.

Commentary & Perspective

Commentary & Perspective on
"Primary Arthroscopic Stabilization for a First-Time Anterior Dislocation of the Shoulder: A Randomized, Double-Blind Trial"
by C. Michael Robinson, BMedSci, FRCSEd(Orth), et al.

Commentary & Perspective by
Frederick A. Matsen III, MD*,
University of Washington Medical Center, Seattle, Washington

Posted April 2008

In the study, "Primary Arthroscopic Stabilization for a First-Time Anterior Dislocation of the Shoulder: A Randomized Study," Robinson et al. compared two procedures: a technically simple and less expensive arthroscopic lavage, and an arthroscopic lavage followed by a technically demanding and more costly arthroscopic Bankart repair. This article is a superior example of the rarest of entities: a randomized double-blind study comparing two surgical procedures. In its excellence, it serves as an example for those who recognize the critical need for such studies to provide Level-I evidence in orthopaedic surgery. This study met even higher standards. It included only patients with a first-time anterior glenohumeral dislocation that was initially treated in the emergency room affiliated with the authors' practice, thus ensuring that complete information regarding the nature of the injury was documented; it included only patients who met strict inclusion criteria for age, type of injury, mechanism of injury, and acuteness; and it included only patients treated by an individual surgeon and evaluated by an individual observer blinded to the procedure. Its hypothesis was clearly stated. The study included the primary outcome variable of recurrent instability, but also included functional outcome, range of movement, return to sport, return to work, and satisfaction with the result. Furthermore, and uniquely, it considered not only the cost of the initial treatment but also the cost of managing treatment failures. It used the appropriate statistics, and was based on intention to treat, rather than the per protocol analysis so commonly found in our literature. Randomization was carried out preoperatively and included so-called banding of risk. Finally it brought its message home with a number-needed-to-treat analysis. This is a careful and complete study. For all these reasons, it deserves a "best in show" recognition.

Having recognized the great virtues of its methodology, let us consider some of the lessons learned with respect to the two surgical procedures, other lessons learned, and limitations in the application of this study to our practice.

The results from this series of individuals with first-time traumatic anterior shoulder dislocation robustly demonstrated that, in comparison with arthroscopic lavage alone, arthroscopic lavage followed by arthroscopic Bankart repair yielded superior results with respect to recurrent dislocation, recurrent instability, function, return to sport, and patient satisfaction. However, all patients with stable shoulders, regardless of their treatment group, had similar shoulder function.

In addition to these conclusions related directly to the hypothesis, there are other important observations that can be made from the data presented. First, the authors were careful to include only patients with a documented first-time anterior dislocation caused by substantial force (patients whose dislocation was not caused by a substantial force were excluded; many of the excluded had generalized hyperlaxity). All of the eighty-eight shoulders with traumatic unidirectional anterior instability were found at surgery to have Bankart lesions and none had substantial other soft-tissue pathology. The size of the Bankart lesions was substantial: when a repair was carried out, three to five anchors were required. Repeat arthroscopy of all shoulders in which treatment failed revealed an unhealed Bankart lesion. These findings encourage surgeons to seek and treat Bankart lesions in individuals with traumatic instability of the shoulder and to be concerned about the mechanism of instability when a Bankart lesion is not found.

Secondly, the authors realized that the total cost of a treatment must include the cost of treating treatment failures (just as the total cost of nuclear energy must include the cost of meltdowns and waste disposal). They found that the initial amount of savings accrued in performing arthroscopic lavage alone was more than offset by the cost of managing the episodes of recurrent instability when this procedure failed.

Thirdly, the number-needed-to-treat analysis demonstrated that at least three patients needed to have a Bankart repair in addition to lavage to prevent one case of recurrent instability. In other words, two of three patients would have had the procedure unnecessarily. Thus, as the authors point out, these data cannot be used to support the general use of primary Bankart repair for patients with first-time dislocations.

Fourthly, the patients with a ≥50% predicted risk of recurrence (men who are twenty-seven years of age or younger and women who are sixteen years of age or younger) had a 46% (thirteen of twenty-eight patients) prevalence of recurrence after lavage alone, but only a 7% (two of twenty-eight patients) prevalence of recurrence after lavage and Bankart repair. This banding-of-risk analysis suggests that additional studies that include patient profiling may help identify those most likely to benefit from surgical management of a primary traumatic dislocation.

Finally, having recognized the great value of the randomized controlled trial, let us consider some of its limitations. First, these results may lack external validity because of the fact that the procedures were all performed by the same expert surgeon with essentially no complications. The benefits of arthroscopic Bankart repair in comparison to lavage alone might disappear if the surgeons performing them have more failures and more complications. The surgeon is the method. Thus, we cannot know the generalizability of these results to community practice. Second, as wonderful as they are, the requirements of the double-blinded randomized controlled study prevent it from being applied to some of the most important questions about shoulder stabilization for first-time anterior dislocation. How does primary repair compare with nonoperative management? How does open repair compare with arthroscopic treatment? Are patients who consent to randomized clinical trials the same as the majority of patients who have the same diagnosis? How can we apply these results to our practice?

In conclusion, this article provides us with a wonderful guide on how to perform a Level-I double-blinded randomized clinical trial of surgical treatment. The authors are to be congratulated on attaining this highest level of clinical research. A Bankart repair of a first-time dislocation must be considered a prophylactic procedure (to prevent recurrent dislocation). In that light, the decision of whether or not to perform a Bankart repair must take into account not only the potential benefit (minimization of risk) but also the drawbacks of the procedure itself (such as the cost, time off from work and other activities, and surgical complications). The results of this study may not be generalizable to the practice of other surgeons, who may have different rates of benefit and drawbacks. This study surely cannot be considered a green light to operate on individuals with first-time shoulder dislocations without other, patient-specific indications.

*The author did not receive any outside funding or grants in support of his research for or preparation of this work. Neither he nor a member of his immediate family received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. A commercial entity (DePuy) paid or directed in any one year, or agreed to pay or direct, benefits in excess of $10,000 to a research fund, foundation, division, center, clinical practice, or other charitable or nonprofit organization with which the author, or a member of his immediate family, is affiliated or associated.