Development and validation of an elbow score

Abstract

Objectives. Few of the questionnaires available for evaluating the function and clinical state of the elbow have been validated. An ideal score would be consistent, sensitive, reliable and elbow-specific, incorporating both patient perception and clinician assessment. This was our aim. Methods. Items were generated using 25 patients and expert opinion, and reduced using 25 new patients to yield a nine-item patient questionnaire and a six-item clinical evaluation (of strength, motion and ulnar nerve involvement). This was validated using 63 new patients (of whom 28 were studied twice without therapy and 18 were studied again after appropriate surgery). Results. The test-retest reliability coefficient of determination (R2 =0.93) and internal consistency (Cronbach's alpha = 0.98) were both good. Convergent validity was attested by good correlations with other scores, the Disabilities of Arm, Shoulder and Hand Questionnaire (DASH) and the Nottingham Health Profile (NHP) (physical) (R2 = 0.62 and 0.29, P < 0.0005). Sensitivity to change was demonstrated by correlating preoperative-postoperative changes to those in DASH and NHP (physical) (R2 = 0.50 and 0.27, P < 0.04). Conclusion. This is a reliable, internally consistent score, correlating well with other, non-elbow specific scores and sensitive to change on treatment. © British Society for Rheumatology 2004; all rights reserved.