Mapping FACT-P to EQ-5D in a large cross-sectional study of metastatic castration-resistant prostate cancer patients

PURPOSE: To construct a model to predict preference-adjusted EuroQol 5D (EQ-5D) health utilities for patients with metastatic castrate-resistant prostate cancer (mCRPC) using the disease-specific health-related quality of life (HRQoL) measure, functional assessment of cancer therapy-prostate (FACT-P). METHODS: HRQoL data were collected from patients with mCRPC who were enrolled in an observational study conducted in 47 centers across six European Union countries. Utility values were generated using a UK-specific EQ-5D value set. The predictive validity of the five FACT-P subscales, patient demographics, comorbidities and prior chemotherapy was tested using ordinary least squares (OLS), median, Gamma and Tobit multivariate regression models. RESULTS: FACT-P and EQ-5D questionnaires were completed by 602 (86 %) patients. Mean age [standard deviation (SD)] was 72.1 (7.9) years, mean time from diagnosis (SD) was 5.4 (4.4) years, and mean time since failure of androgen deprivation therapy (SD) was 1.0 (1.6) years. At study inclusion, 39 % of patients were chemotherapy-naïve, 37 % were undergoing chemotherapy, and 24 % were post-chemotherapy. Mean FACT-P and EQ-5D utility values were 104 and 0.66, respectively. OLS regression was the best-performing model, explaining 61.2 % of the observed EQ-5D variation. All FACT-P subscales were significantly predictive; the physical and functional well-being subscales had the highest explanatory value (coefficient 0.023 and 0.001, respectively, p < 0.0001). The other variables did not add additional explanatory value. CONCLUSIONS: The algorithm developed enables translation of cancer-specific HRQoL measures to preference-adjusted health status in patients with mCRPC. The function may be useful in calculating EQ-5D scores when EQ-5D data have not been gathered directly.