Intra- and interobserver reproducibility of the shoulder laxity examination

The present study evaluates the inter- and intraobserver reproducibility of clinical examination of glenohumeral laxity in the unanesthetized shoulder. Forty-three asymptomatic Division I collegiate athletes underwent bilateral shoulder laxity examination initially and again after 3 months. Translation of the humeral head on the glenoid fossa in the anterior, posterior, and inferior directions was graded by four physicians who were blinded to their own previous grading and that of the other examiners. Overall intraobserver reproducibility of examination was 46%. When grades 0 and 1 were equalized, overall intraobserver reproducibility improved to 74%. For both the equalized and non-equalized reproducibility values reported by all examiners, the kappa values for intraobserver correlation were less than 0.5, which suggests that correlations were not better than those achieved by chance alone. Overall interobserver reproducibility was 47%. When grades 0 and 1 were equalized, interobserver reproducibility improved to 78%. Kappa values were greater than 0.5 only in equalized posterior and inferior laxity. The data demonstrate that the laxity examination of the unanesthetized shoulder is not easily reproducible in either intra- or interobserver comparison. Equalization of grade 0 and grade 1 laxity improves both intra- and interobserver reproducibility. We recommend caution when determining diagnosis and treatment based on this examination.