High long-term failure rates after arthroscopic Bankart repair in younger patients with recurrent shoulder dislocations: A plea for early treatment.

PURPOSE To determine arthroscopic Bankart repair outcomes and recurrence risk factors at a minimum 5-year follow-up. METHODS Retrospective assessment of prospectively collected data, single-cohort study of patients who underwent arthroscopic Bankart repair with a minimum 5-year follow-up. Demographical and preoperative instability features were collected. Primary outcome was recurrent instability set as dislocation or subluxation. Secondary outcomes were revision surgery, postoperative instability degree according to Manta criteria, objective and subjective clinical and functional status, assessed by the Rowe, Western Ontario Shoulder Index (WOSI) and Subjective Shoulder Value (SSV) scores. Return to sport and postoperative sports activity at the final follow-up were also recorded. RESULTS One-hundred and seventy-two patients, 82% men, average age at surgery 29.5 ± 9.2 years, were included. At a mean follow-up of 8.3 ± 2.6 years, recurrent instability occurred in 53 of 172 patients (30.8%). Revision surgery was required in 23/53 (43.4%) of shoulder with recurrent instability. Recurrence occurred within the first 2 years postoperative in 49% of the shoulders, whereas 51% of recurrences occurred after this period. Recurrence took place after a traumatic event in 25% and 56%, respectively. Recurrence rates were higher in patients who underwent surgery after two or more dislocations (p = 0.029). Patients younger at the time of first dislocation, younger at surgery and those with a higher preoperative degree of instability also showed significantly higher rates of recurrence (p = 0.04, p = 0.02, p = 0.03). Postoperative ROWE, WOSI and SSV scores were significantly worse in patients with recurrent instability (p < 0.001). Return-to-sports rate was also lower in patients with postoperative recurrence (p < 0.001). CONCLUSION The arthroscopic Bankart repair was associated with a high long-term recurrence rate, and its effectiveness decreased over time. The lowest recurrence rates in arthroscopic Bankart repair were achieved in older patients with only one prior instability episode and a lower instability degree. LEVEL OF EVIDENCE Level IV.