Recurrent shoulder instability is frequently associated with glenohumeral bone loss. Recently there has been a surge of interest in arthroscopically performed bone block procedures. The aim of this systematic review was to determine the clinical and radiological outcomes of arthroscopic glenoid bone block stabilisation for recurrent anterior dislocation.

This systematic review was performed in accordance with PRISMA guidelines. The search strategy was applied to MEDLINE and Embase databases on 20^th July 2020. Studies reporting either clinical or radiological outcomes following arthroscopic bone block stabilisation for recurrent anterior dislocation were included. Primary outcomes were function and instability scores. Secondary outcomes included recurrent instability, graft union and resorption rates, return to activity/sports, and complications. Pooled analysis was performed when an outcome was uniformly reported by more than one study. Critical appraisal of studies was conducted using the Methodological Index for Non-Randomized Studies (MINORS) tool.

Application of the search strategy resulted in the inclusion of 15 eligible studies; 12 used iliac crest bone graft while 3 used distal tibial allograft. The overall population comprised 265 patients (mean age range, 25.5–37.5 years; 79% of participants were men). All post-operative outcome scores were significantly improved, and the overall rate of recurrent instability was low (weighted mean 6.6%, range 0–18.2%) at mean follow up of 30.4 months. The Rowe score was the most frequently reported outcome measure, improving on average by 53.9 points at final follow-up, exceeding the minimal clinically important difference (MCID) threshold. Graft union rates ranged between 92–100% in 8 out of 10 studies at mean follow up range 6–78.7 months but two reported lower rates ranging from 58.3–84% for autografts and 37.5% for allografts. Graft resorption rates averaged between 10–16% for autografts and 32% for allografts. Hardware-related complications occurred in 2% with the most frequent being screw breakage or symptomatic mechanical irritation.

Arthroscopic bone block stabilisation is associated with high rates of graft union, significant improvements in the WOSI, Rowe, Constant and SSV scores (exceeding MCID thresholds where known), and a low rate of complications, including re-dislocation in the short to mid-term. Graft union rates were high, but the long-term implications of graft resorption (which occurs more frequently with allograft) are unknown. Longer follow-up of these patients and future experimental studies are required to further examine the effects of graft type and fixation methods.

IV; systematic review.