A cross-cultural validation of patient-reported outcomes measures: a study of breast cancers survivors

Abstract

Psychometric inadequacy interferes with the assessment of patient-reported health outcomes. This study examined the psychometric properties of several standardized measures in health research. Participants completed health outcomes measures including the FACT-G, SF-36, MOS Social Support Survey, and CES-D. Psychometric properties examined include reliability, and construct and concurrent validity. 320 BCS including 88 African-, 95 English-proficient Latina-, and 137 Limited English-proficient Latina-Americans participated. The findings demonstrate acceptable reliability (α > 0.70) and consistent factor structures for most measures with the variance ranging from 56 to 84%. The FACT-G physical well-being and SF-36 role limitations subscales had the best fitting structures. Concurrent validity showed the FACT-G subscales correlated with their appropriate counterparts. Despite being commonly used instruments in HRQOL (e.g., FACT-G, SF-36) and QOL (e.g., CES-D, MOS) research, few studies reported the psychometric properties of these and when applied cross-culturally. However, evaluating the psychometric properties of measures in health outcomes research should be done routinely.