An application of computerised adaptive testing for measuring health status in patients with knee osteoarthritis

The aim of this study was to explore the potential of computerised adaptive testing (CAT) for measuring health status in knee osteoarthritis.

Three hundred and fifty-one patients with knee osteoarthritis (OA) answered 118 questions from a range of widely used questionnaires. Three dimensions relevant to knee OA were identified by expert opinion as 'pain, mental and emotional status'; 'self-care and household activities' and 'mobility and social activities'. Confirmatory factor analysis (CFA) was used to test the relationship between the items and their dimensions. After CFA, these dimensions were subjected to Rasch analysis to calibrate the items onto an interval scale. A CAT was developed for each dimension, and the results checked against simulated applications from 10,000 simulees.

After CFA and Rasch analysis, 14, 22 and 23 items remained in the first, second and third dimensions, respectively. Items were mostly free of differential item functioning for age, gender and duration of disease. Reliability exceeded 0.87 for each dimension. The health status levels generated using item banks and those obtained from the simulated CAT application were highly correlated (i.e. >0.91 for each dimension). On average, 8, 10 and 10 items were used to estimate the health status levels using the CAT for the first, second and third dimensions respectively.

Using a combination approach of CFA and Rasch analysis, this study has shown that it is possible to calibrate items onto a single metric in a way that can be used to provide the basis of a CAT application.