Background: Although critical bone loss for anterior instability is well defined, a clinically significant threshold of posterior bone loss has not been elucidated. Hypothesis: Patients with failed arthroscopic posterior shoulder capsulolabral repair will have increased posterior glenoid bone loss with a defined critical threshold. Study Design: Case control study; Level of evidence, 3. Methods: Athletes older than 18 years with unidirectional posterior instability treated with arthroscopic repair were evaluated at 2-year minimum follow-up. Failure was defined as revision surgery, American Shoulder and Elbow Surgeons (ASES) score of <60, or subjective stability score of >5. Magnetic resonance imaging (MRI) measurements from 19 patients with failed arthroscopic posterior shoulder capsulolabral repair were compared with 56 patients whose surgery was successful. MRI measures included glenoid version, labral version, glenoid width, labral width, percentage bone loss using the circle technique, labral height, percent subluxation, and recently described measures of defect slope, bone loss angle, and defect length. The P value threshold was set at .05, and a multivariable logistic regression analysis was performed for evaluation of risk of surgical failure. Results: Smaller glenoid width and greater percentage glenoid bone loss (25.5 ± 0.68 mm vs 28.8 ± 0.47 mm; P < .001; 6.8% ± 0.64% vs 4.6% ± 0.43%; P = .008) were seen in those patients with failed surgery. There was no difference in glenoid version or other measurements between the failures and nonfailures. A cutoff of 11% glenoid bone loss resulted in a 10.4 times statistically higher surgical failure rate, while a 15% bone loss resulted in a 24.4 times statistically higher failure rate. Six patients had >11% bone loss (range, 11.1 to 19.3) and 1 patient had >15% bone loss. Conclusion: Risk factors for failure of arthroscopic posterior shoulder capsulolabral repair include smaller glenoid bone width and greater percentage of glenoid bone loss. A threshold of 11% posterior glenoid bone loss implicated a 10 times higher surgical failure rate, while a threshold of 15% led to a 25 times higher surgical failure rate. Surgical failure of posterior capsulolabral repair, however, is relatively rare as it is an overall successful intervention.
