Abstract
BACKGROUND: The best method for nonsurgical treatment after primary dislocation of the shoulder is not clear. The efficacy of immobilization with the arm in external rotation (ER) compared with internal rotation (IR) remains controversial. PURPOSE: To determine the efficacy of ER immobilization versus IR immobilization on recurrence rate after primary dislocation of the shoulder from the evidence of randomized controlled trials. STUDY DESIGN: Systematic review; Level of evidence, 2. METHODS: Electronic databases (MEDLINE [Ovid SP], PubMed, Web of Science, EBM reviews, and CINAHL) and available proceedings according to the abstracts of major international meetings related to or including shoulder injuries and trauma were used to search for randomized controlled trials. Two independent investigators determined eligibility and carried out data extraction from the selected studies. RESULTS: A total of 9 studies (817 patients) were selected for this meta-analysis. They included 668 male and 149 female patients, with a mean age ranging from 20.3 to 37.5 years. In the 9 pooled studies, the recurrence rate of shoulder dislocation was 21.5% (84/390) in the ER group versus 34.9% (130/373) in the IR group. ER immobilization significantly reduced the recurrence rate compared with IR immobilization (risk ratio, 0.56; P = .007). In the subgroup analysis of those immobilized full-time, ER immobilization was significantly more effective than IR immobilization in reducing the recurrence rate (risk ratio, 0.57; P = .01). In the subgroup analysis of age, ER immobilization was significantly more effective than IR immobilization in those aged 20 to 40 years but not in those younger than 20 years. CONCLUSION: This meta-analysis demonstrates that ER immobilization reduces the recurrence rate after primary shoulder dislocation compared with IR immobilization in patients older than 20 years. When treating a patient with primary shoulder dislocation, the clinician should provide this information to the patient before a treatment method is selected.