2003 年 27 巻 3 号 p. 435-438
The JOA shoulder score has been widely used for the evaluation of the function of the shoulder. However, no studies have been published yet concerning its interobserver reproducibility. In this study, we attempted to clarify this point. Fifty shoulders of forty-six patients (average age, fifty-two), who were newly admitted to the shoulder clinic in the Tohoku University Hospital, were assessed by 3 orthopaedic doctors on the same day. The JOA score was evaluated independently using the manual for Evaluations of the Shoulder. Pearson's correlation coefficient (r2) was calculated between each observer for the total score. As for the scores of each criterion including pain, function, active daily life (ADL), range of motion (ROM), radiologic finding and stability, Spearman's rank correlation coefficient was calculated. In the total score, the r2 value between each observer was more than 0.78. It seemed that the total score strongly correlated between each observer. In the analysis for each criterion, the ρ-value between each observer was high enough in function, ADL and ROM Especially, the score for ADL showed the highest ρ-value (0.88-O.93). Hewever, the scores for pain, radiologic findings and stability represented the lower ρ-value than those for other categories. The total score of the JOA shoulder score represented satisfactory interobserver reproducibility. To obtatin better interobserver reproducibility in each criterion, one has to double-check the manual during the assessment of patients.