Point/Counterpoint: Arthroscopic vs Open Bankart Surgery 

Arthroscopic Reconstruction for Shoulder Instability

By COL Patrick St. Pierre, MD

Is it time to recommend arthroscopic shoulder reconstruction for every patient? Arthroscopic techniques have improved significantly and gained in popularity over the past few years, but does the literature support abandoning open reconstruction?

Open shoulder reconstruction has been the accepted surgical intervention for recurrent anterior instability for over a century. Many procedures have been advocated, including labral repair (Bankart), capsular shift (Neer), capsular and subscapularis shortening (Magnuson-Stack and Putti-Platt), and bone-block tendon transfer (Bristow). Nonanatomic reconstructions initially were thought to produce results similar to the anatomic Bankart repair but were easier to perform. These have fallen out of favor because of higher recurrence rates and complications such as the development of glenohumeral arthritis. Today, most surgeons advocate direct labral repair, retensioning of the lengthened capsular ligaments, and closure of the rotator interval if indicated. This approach has generally been accepted as the "gold standard" for which to compare other techniques. Recurrence rates following open Bankart reconstructions have been thought to be very low, with Bankart reporting 0%,1 Morrey 11%,2 and Rowe 3.5%.3 However, these reports had variable follow-up and did not include recurrent subluxation as failures.

Recent studies have questioned whether the results of open reconstruction are really as good as touted, especially if maintaining external rotation and preventing subluxations are required for success. Chapnikoff and associates4 recently reported a redislocation rate of 9.5% after 16 years, with a 76% follow-up after open reconstruction. Uhorchak and colleagues5 reported a 23% redislocation and resubluxation rate in West Point cadets who must return to a very high level of physical activity. Most recently, Magnusson and colleagues6 reported a 17% recurrence rate at a 4- to 9-year follow-up of 47 of 54 patients.

Early arthroscopic techniques were met with high failure rates and surgeon dependant results. Transglenoid techniques and the arthroscopic use of staples also had high complication rates. This led to the development of bioabsorbable tacks and the use of suture anchors to allow a more anatomic repair and reduce the complications caused by exposed metal in the joint. The higher failure rate of these early arthroscopic reconstructions has been attributed to the failure to restore the labrum to the edge of the glenoid. With both the transglenoid technique and the staple, the point of fixation on the glenoid was medial to the edge, failing to recreate the bumper effect of the labrum and failing to properly tension the inferior glenohumeral ligament.

The most recent technique for arthroscopic Bankart repair is to use suture anchors to repair the labrum to the rim of the glenoid. Early results are mixed, with Koss and associates7 reporting a 30% recurrence rate. Seven of the 8 patients with recurrence had a traumatic event and there was a higher rate in patients with more than 5 presurgical dislocations. Hoffman and colleagues8 reported a recurrence of 13%, and Bacilla and associates9 reported a 7% redislocation rate in athletes and laborers. Cole and associates10 found no difference between arthroscopic and open reconstructions if the pathology determined the type of treatment. Patients with Bankart lesions and unidirectional instability underwent arthroscopic reconstruction, while those with multidirectional instability underwent open reconstruction.

Current arthroscopic reconstruction techniques advocate the placement of the suture anchor on the articular surface of the glenoid, thus restoring the bumper effect of the labrum and allowing proper retensioning of the ligament. While the labrum does not heal to the articular surface and may remodel with time, its placement there allows the ligament to heal properly along the exposed bone of the freshened medial glenoid. If the ligaments are still lax after Bankart repair, the capsule should be addressed with plication or thermal shrinkage to restore the proper tension of the ligament. This is commonly done with open surgery by adding a capsular repair following the repair of the labrum. Therefore, the goal in either open or arthroscopic reconstruction is to repair the torn labrum and inferior glenohumeral ligament complex to the glenoid to create the most anatomic repair possible.

This review of the literature indicates that we may not have been as good as once thought with open Bankart reconstructions, and we are getting better with our arthroscopic reconstructions. We have to ask why one has been better than the other. Higher arthroscopic recurrence rates initially were likely due to the labrum not being anatomically repaired, and I think we now can achieve that. High recurrence rates, even with anatomic labral repair, may have been due to failure to adequately address increased capsular volume and/or the rotator interval. If the anatomic repair of labrum to the glenoid is the same and we can tension the capsule equally with both methods, then the results should be the same. The key to arthroscopic instability surgery is to do the repair as well, if not better than open surgery. Studies reveal that patients with multidirectional instability, with more than 5 dislocations, and patients returning to contact sports are at higher risk for failure with any reconstruction. This needs to be taken into consideration and an arthroscopic repair must be perfect in these patients. Capsular laxity must be addressed in addition to the Bankart repair.

Therefore, I plan an arthroscopic reconstruction for every instability patient I take to the operating room. I consent every patient to convert to an open repair if necessary. My goal is to restore the labrum and global capsular tension to its original state. If I am not able to do so, especially with a patient returning to a contact sport, I will convert to an open reconstruction. Every surgeon needs to tailor this to his or her own experience and ability. The decision to reconstruct the shoulder arthroscopically or open is dependent on pathology present and the surgeon’s ability to address that pathology by the best method possible.


1. Bankart AB. The pathology and treatment of recurrent dislocation of the shoulder joint. Br J Surg. 1938;26:23-29.

2. Morrey BF, Janes JM. Recurrent anterior dislocation of the shoulder. Long-term follow-up of the Putti-Platt and Bankart procedures. J Bone Joint Surg Am. 1976;58A:252-256.

3. Rowe CR, et al. The Bankart procedure: A long-term end-result study. J Bone Joint Surg Am. 1978;60A:1-16.

4. Chapnikoff D, et al. Bankart procedure: Clinical and radiological long-term outcome [in French]. Rev Chir Orthop Reparatrice Appar Mot. 2000;86:558-565.

5. Uhorchak JM, et al. Recurrent shoulder instability after open reconstruction in athletes involved in collision and contact sports. Am J Sports Med. 2000;28:794-799.

6. Magnusson L, et al. Revisiting the open Bankart experience: A four- to nine-year follow-up. Am J Sports Med. 2002;30: 778-782.

7. Koss S, et al. Two to five-year followup of arthroscopic Bankart reconstruction using a suture anchor technique. Am J Sports Med. 1997;25:809-812.

8. Hoffmann F, Reif G. Arthroscopic shoulder stabilization using Mitek anchors. Knee Surg Sports Traumatol Arthrosc. 1995;3:50-54.

9. Bacilla P, Field LD, Savoie FH 3rd. Arthroscopic Bankart repair in a high demand patient population. Arthroscopy. 1997;13:51-60.

10. Cole BJ, et al. Prospectively determined arthroscopic versus open shoulder stabilization: A 2-6 year follow-up. J Shoulder Elbow Surg. 1998;7:3-13.