The reliability and validity of the Disabilities of Arm, Shoulder, and Hand, EuroQol-5D, Health Utilities Index, and Short Form-6D outcome instruments in patients with proximal humeral fractures

Abstract

Hypothesis: The Disabilities of Arm, Shoulder, Hand (DASH), EuroQol-5D (EQ-5D), Health Utilities Index Mark 3 (HUI3), and Short Form (SF)-6D questionnaires are reliable and valid measures of functional outcome and health state values in patients with proximal humeral fractures.

Materials and methods: Patients aged 55 and older treated for a proximal humeral fracture during a 5-year period completed the DASH, EQ-5D, HUI3, and SF-12 questionnaires. Test-retest reliability was quantified using intraclass correlation (ICC 2,1) and Bland-Altman agreement statistics during a second administration of the questionnaires. Correlations between the 4 study instruments, the SF-12, and a subjective global assessment of shoulder function were used to test construct validity. Ceiling/floor effects were quantified for each questionnaire.

Results: Sixty-one individuals (mean age, 69+/-10 years) participated. ICC showed the reliability (95% confidence interval) was 0.926 (0.860-0.963) for the DASH, 0.783 (0.604-0.875) for the EQ-5D, 0.794 (0.634-0.889) for the SF-6D, and 0.469 (0.184-0.686) for the HUI3. The Bland-Altman limits of agreement, however, highlighted limitations for repeated measurements with all 4 instruments at the individual patient level. Moderate construct validity was confirmed for all instruments. A significant ceiling effect was observed with the EQ-5D: 30% of participants reported "perfect health," compared with less than 7% with DASH, HUI3, or SF-6D questionnaires.

Discussion: The DASH and SF-6D questionnaires demonstrated the best psychometric properties among the study instruments. These results support their use as appropriate measures of functional outcome and health state values in patients with proximal humeral fractures.