Long-term, Prospective, Multicenter Study of Isolated Bankart Repair for a Patient Selection Method Based on the Instability Severity Index Score

Abstract

Background: An isolated arthroscopic Bankart repair carries a high mid- and long-term risk of recurring instability. Preoperative patient selection based on the Instability Severity Index Score should improve outcomes. Purpose: To report the overall long-term recurrence rate for isolated Bankart repair, investigate the predictive factors for recurrence, analyze time to recurrence, and determine a quantitative cutoff point for recurrence in terms of Instability Severity Index Score. Study Design: Cohort study; Level of evidence, 2. Methods: This was a prospective multicenter study. Inclusion criteria were recurring anterior instability and an Instability Severity Index Score of 4 or less. Of the 125 patients included, 20 patients had a score of 0, 31 patients scored 1, 29 patients scored 2, 34 patients scored 3, and 11 patients scored 4. All centers used the same arthroscopic technique and rehabilitation protocol. Follow-up data were collected at 3, 6, 12, and 24 months and 3 and 9 years. The primary endpoint was recurrence of instability (total or partial dislocation). The statistical analysis was performed by use of the software package SAS 9.4. Results: We initially identified 328 patients, of whom 125 patients were prospectively included. The main reason for excluding the 202 patients was the presence of bony lesions, which carry 2 points each in the Instability Severity Index Score (humeral head notch and/or glenoid lesion visible on standard radiographs). Of the 125 eligible patients, 73% were athletes and 22.5% competitors; 16% were lost at the last follow-up. At the endpoint, 23% had experienced a recurrence after a mean interval of 35 months (range, 5.5-103 months). No statistical differences were found between patients with and without bony lesions in the overall group of 125 patients or in the subgroup with an Instability Severity Index Score of 3 or 4 points ( P = .4). According to univariate analysis, the only predictive factor for recurrence was age less than 20 years at the time of surgery, with a 42% rate of recurrence in this group ( P = .03). Multivariate analysis showed that the Instability Severity Index Score was the only predictive factor with a quantitative cutoff point (namely, a score of ≤2 points) that was statistically associated with a decreased long term recurrence rate ( P = .02). The recurrence rate was 10% for a preoperative Instability Severity Index Score of 2 or less compared with 35.6% for a score of 3 or 4. The survival curves demonstrated no new dislocations after year 4 for patients with an Instability Severity Index Score of up to 2 points. Conclusion: In a preselected population, mainly without bony lesions, the Instability Severity Index Score cutoff value that provides an acceptable recurrence rate at 9 years after isolated Bankart repair is 2 out of 10.