A systematic review of the psychometric properties of the Constant-Murley score

HYPOTHESIS: The purpose of this study was to conduct a systematic review of the psychometric evidence relating to Constant-Murley score. MATERIALS AND METHODS: A search of 3 databases (Medline, CINAHL, and EMBASE) and a manual search yielded 35 relevant publications. Pairs of raters used structured tools to analyze these articles, through critical appraisal and data extraction. A descriptive synthesis of the psychometric evidence was then performed. RESULTS: Quality ratings of 23% of the studies reviewed reached a level of 75% or higher. Studies evaluating the content validity of the Constant-Murley score suggest that the description in the original publication is insufficient to accomplish standardization between centers and evaluators. Despite this limitation, the Constant-Murley score correlates strongly (>or= 0.70) with shoulder-specific questionnaires, reaches acceptable benchmarks (rho > 0.80) for its reliability coefficients, and is responsive (effect sizes and standardized response mean > 0.80) for detecting improvement after intervention in a variety of shoulder pathologies. DISCUSSION: This systematic review provides evidence to support the use of the Constant-Murley score for specific clinical and research applications but underscores the need for greater standardization and precaution when interpreting scores. Methods to improve standardization and measurement precision are needed. Responsiveness has been shown to be excellent, but some properties still need be evaluated, particularly those related to the absolute errors of measurement and minimal clinically important difference. CONCLUSION: Given the widespread acceptance for usage of the Constant-Murley score in clinical studies and early indications that the measure is responsive, studies defining more rigid standardization of the tools/procedures are needed. LEVEL OF EVIDENCE: Level 1.