Objective: Computer-adaptive tests (CAT) use individualised sets of questions to assess patient-reported health states, whereas static (conventional) questionnaires present the same questions to all patients. CAT has been shown to increase measurement precision and reduce assessment length. In our study, we investigated if patients perceive CAT questions as more appropriate than static questionnaires, a claim that is frequently associated with CAT measures.
Methods: We compared the static European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC QLQ-C30) against its corresponding CAT measures focusing on two domains: Physical Functioning (PF) and Emotional Functioning (EF). Cancer patients completed the questionnaires and participated in a cognitive interview to assess how appropriate they perceive the QLQ-C30 and the CAT questions for their current health state.
Results: Forty-four cancer patients (mean age = 54.6; 56.8% female) were assessed. For the PF domain, patients considered the CAT items more appropriate (p = 0.002) than the QLQ-C30 items (56.8% vs. 15.9%; 27.2% indifferent). For the EF domain, patients were in favour of the QLQ-C30 items (p < 0.001), with 54.5% considering the QLQ-C30, and 4.5% considering the CAT items to be more appropriate; 40.9% were indifferent. Most patients (N = 36) commented on the preference for the CAT (PF), mentioning better matching of the questions and the health state (38.6%) and better item wording (15.9%).
Conclusion: For the PF domain the CAT measure better matched the score distribution in the patient sample than the QLQ-C30 PF scale and was consequently considered more appropriate by patients. For the EF domain, the CAT measure did not show better fit than the QLQ-C30 and hence no such preference in terms of appropriateness was observed.
Keywords: Computer-adaptive tests; EORTC CAT; EORTC QLQ-C30; Mixed methods; Patient-reported outcomes.