Does the EPIC-26 measure five domains in English and Spanish speakers?

e557 Background: Despite being widely used in many countries to assess health-related quality of life for patients undergoing treatment for localized prostate cancer(PC), seldom research has been done to investigate the five domain scores generated by 26 question short form of the Expanded Prostate Cancer Index Composite Short (EPIC-26). We conducted exploratory and confirmatory factor analyses searching for evidence of five unique domains of dysfunction in men being treated for PC in the US and Spain.Methods: Using two-year post treatment information, data from 400 men who were part of the original EPIC (OE) validation cohort [T1 or T2, 87% Radical Prostatectomy (RP), 10% External Beam Radiation (EBR), 3% Brachytherapy (BR)] and 450 men who were treated for prostate cancer at Institut Hospital del Mar d'Investigacions Mèdiques in Spain [33% RP, 33% EBR, 33% BR], were subjected to exploratory (EFA) and confirmatory factor analysis (CFA).Results: EFA analyses on both the OE and Spanish cohorts provides support for three, four or five factor models (accounting for approximately 83, 91 or 97% of variance in either cohort) to describe the EPIC-26 variables and weak evidence for the hormonal domain. There is evidence of separate urinary, sexual and bowel factors, but limited evidence of a separate hormonal domain. With the exception of the questions which could be described as assessing depression (lack of energy and depression and in some models weight gain), items within the hormonal domain frequently failed to consistently load onto a single factor (orthogonal communalities ranging from .22 to .27 in the OE cohort and .10 to .21 in the Spanish for hot flashes, breast tenderness and weight gain). CFA looking for the published five domain solution in the Spanish data provides mediocre support for a five factor solution (Standardized RMR = .0624, RMSEA Estimate = .07, Bentler Comparative Fit Index = .91). Conclusions: While all questions asked on the EPIC-26 are of clinical value, there is limited evidence to continue to treat the hormonal domain as a single entity. Urologists should consider looking at the questions linked to depression separately from the other hormonal questions