Objective: To assess the validity of the e-OAKHQOL questionnaire and analyze whether the answers were affected by the form of administration (electronic vs. paper).
Methods: Two samples of patients with knee osteoarthritis were constituted. The first was recruited by general practitioners. Patients could choose to respond to the electronic or paper version. The second included subjects who responded to the paper version and were matched with respondents to the electronic version in the first sample. The OAKHQOL questionnaire measures health-related quality of life in five dimensions (43 items): physical activity, mental health, pain, social functioning, and social support. Validity was assessed by the classical test theory (CTT) and a Rasch measurement model (partial credit model).
Results: The electronic form was preferred by 471 (89.7%) patients: 345 were matched to respondents of the paper version. The percentage of missing responses was lower with the electronic than paper form (1.6 vs. 2.0%, p = .01). Rasch analysis revealed four items with underfitting. Internal consistency was excellent for physical activity (PSI = 0.96) and mental health (PSI = 0.93) but was slightly < 0.85 for the other dimensions. The top-down purification highlighted the significance of DIF by gender in the pain dimension and by form of questionnaire in the mental health dimension.
Conclusion: CTT and Rasch analysis demonstrated acceptable measurement properties for the five dimensions of the e-OAKHQOL, so it may be a valuable alternative to the paper form for measuring HRQoL.
Keywords: Knee; Osteoarthritis; Psychometrics; Quality of life.