Effects of surgical management on multidirectional instability of the shoulder : a meta-analysis

Purpose: The purpose of this study was to assess the effectiveness of arthroscopic and open surgical techniques on the treatment of shoulder multidirectional instability. Methods: Literature searches were conducted using the databases MEDLINE, Embase, ClinicalTrials.gov, the Cochrane Library, and the Cochrane Central Register of Controlled Trials. Original articles on the surgical management of multidirectional instability were retrieved against selection criteria. Data were extracted and divided into three groups by surgical technique. Proportion and mean meta-analyses were performed for comparison. Results: The available evidence was from 35 level IV and 1 level II studies. The recurrent instability rate was 9.9 % (95 % CI 7.3–12.9 %) in open capsular shift (OCS) group and 6.0 % (95 % CI 3.7–8.9 %) in arthroscopic capsular plication (ACP) group, between which no difference was observed. However, thermal capsular shrinkage (TCS) group resulted in a recurrent instability rate of 23.9 % (95 % CI 16.6–32.2 %), significantly higher than the above two groups. OCS and ACP groups revealed low reoperation rates of approximately 5.2 % (95 % CI 2.7–8.5 %) and 4.8 % (95 % CI 2.3–8.0 %), respectively, which are lower than that in TCS group of 16.9 % (95 % CI 12.4–21.8 %). OCS caused more loss of external rotation than ACP, losing 7.0 (95 % CI 3.3–10.6) degrees versus 2 (95 % CI 0.9–2.4) degrees, respectively. Conclusions: ACP and OCS techniques have similar primary outcomes, but the former causes less post-operative stiffness. It is suggestible to avoid TCS in the treatment of MDI. Level of evidence: Level IV.