Factors contributing to the ceiling effect of the EQ-5D-5L: an analysis of patients with prostate cancer judged "no-problems".

ABSTRACT

PURPOSE: The goal of the present study was to determine factors related to a ceiling effect (CE) on the EQ-5D-5L among Japanese patients with prostate cancer (PC). METHODS: An existent cross-sectional observational study dataset was used. Patients were ≥ 20 years of age and diagnosed with PC. For CE determinants on the EQ-5D-5L, we excluded possible "full-health" patients flagged by the EQ-VAS (score = 100) and/or FACT-P (score = 156) instruments. We then divided them into binary variables A CE group (EQ-5D-5L score = 1) and others (< 1). The associations between CE, sociodemographic and medical characteristics, and FACT-P subscale scores were examined using a multivariate LASSO selection followed by a binomial logistic regression analysis performed to calculate odds ratios (ORs) and 95% confidence intervals (CIs). RESULTS: A total of 362 patients were analyzed. The LASSO selection variables, including all obtained variables, were as follows age, palliative treatment, FACT-P physical well-being, and PC subscale score. Statistically significant variables predicting CE were palliative treatment (OR 0.23; 95% CI 0.09-0.60), physical well-being (OR 1.54; 95% CI 1.34-1.76), and PC subscale (OR 1.08; 95% CI 1.03-1.14). CONCLUSIONS: This study revealed that palliative treatment and two FACT-P physical well-being and PC subscale scores were positively related to CE on the EQ-5D-5L. To our knowledge, this is the first study to examine predictors of CE on the EQ-5D-5L. The present results may be helpful for facilitating the consideration of "bolt-on" studies from the standpoint of PC patients.