Background: Clinical trials and observational studies lacking measures of health-related quality of life (QoL) are often inapplicable when conducting cost-effectiveness analyses using quality-adjusted life-years (QALYs). The only solution is to map QoL ex post from additionally collected clinical outcomes and generic QoL instruments. Nonetheless, mapping studies are absent in psoriatic arthritis (PsA).
Methods: In this 2-year, prospective, multicentre, non-interventional study of PsA patients, EQ-5D and key clinical parameters such as Disease Activity in PsA (DAPsA), clinical DAPsA (cDAPsA; DAPsA without C-reactive protein [CRP]), and Health Assessment Questionnaire disability index (HAQ) were collected. We employed a linear mixed-effect regression model (ME) of the longitudinal dataset to explore the best predictors of QoL.
Results: A total of 228 patients were followed over 873 appointments/observations. DAPsA, cDAPsA and HAQ were stable and highly significant predictors of EQ-5D utilities in both cross-sectional and longitudinal analyses. The best prediction was provided using a linear ME with HAQ and cDAPsA or DAPsA. A HAQ increase of 1 point represented a decrease in EQ-5D by -0.204 or -0.203 (p < 0.0001); a one-point increase in cDAPsA or DAPsA dropped EQ-5D equally by -0.005 (p < 0.0001). The ME revealed steeper and more accurate association compared with cross-sectional regressions or non-linear models/transformations.
Conclusions: This is the first mapping study conducted in PsA and we hope that our study will encourage further mapping studies in PsA. The results showed that in cases where CRP is absent, cDAPsA provides similar results to DAPsA in predicting QoL.