OBJECTIVES: The Patient-Oriented Prostate Utility Scale (PORPUS-U) is the only econometric instrument specifically designed for prostate cancer patients. The Expanded Prostate cancer Index Composite (EPIC) and the SF-36 are the most used instruments in these patients. This study aims to estimate the PORPUS-U index by EPIC and SF-36 scores to allow cost-utility analysis.

METHODS: The Spanish Multicentric Localized Prostate Cancer Study in new therapeutic modalities included patients diagnosed between 2014 and 2017 (n=469). PORPUS-U, EPIC and SF-36 were administered by telephone. The predictive model was constructed with the evaluation 3 months after treatment, and validated with evaluations at 6, 12, 24 and 36 months. A linear regression model was constructed with the logarithmic transformation of the PORPUS-U index as the dependent variable, and EPIC scores and the SF-36 physical and mental summary components as explanatory variables. All possible first level interactions were tested. R and Root Mean Square Error (RMSE) of the model were considered. Agreement between the estimated index and the observed one was calculated with the Intraclass Correlation Coefficient (ICC).

RESULTS: The model chosen included six EPIC scores (urinary incontinence and irritative-obstructive, sexual function and bother, bowel and hormonal) and the two summary components of the SF-36: R=0.793, RMSE=0.036 and ICC=0.914. The validation models showed RMSE ranging 0.031-0.045 and ICC ranging 0.838-0.922. The means (SD) of the PORPUS-U index observed and predicted by the model were: 0.97 (0.04) vs. 0.96 (0.05) at 6 months, 0.97 (0.05) vs. 0.96 (0.06) at 12 months, 0.96 (0.05) vs. 0.96 (0.07) at 24 months and 0.97 (0.06) vs. 0.95 (0.06) at 36 months.

CONCLUSIONS: The model constructed with EPIC scores and the two SF-36 summary components presented a good predictive capacity to estimate the PORPUS-U utility index, which could be useful to allow cost-utility analysis in studies without an econometric instrument.