Measuring breast cancer-specific health-related quality of life in South Asia: psychometric properties of the Sinhala version of the EORTC QLQ-BR23

Abstract

To evaluate the psychometric properties of the Sinhala version of the breast cancer-specific health-related quality of life (HRQL) module of the European Organization for Research and Treatment of Cancer (QLQ-BR23). Psychometric testing assessed the hypothesized scale structure, internal consistency, construct validity and acceptability of the Sinhala version of the QLQ-BR23 in a consecutive series of 356 newly diagnosed breast cancer patients recruited from tertiary care oncology treatment centres in Sri Lanka. Compliance and self-completion rates were high (98% and 88%, respectively), and missing data low (0.06%). Multitrait scaling confirmed the scale structure of the QLQ-BR23 with excellent item convergence (95%), item discrimination (99%), and scaling success (99%) rates. The Cronbach's alpha coefficients of the scales for internal consistency reliability ranged from 0.68 to 0.93. Construct validity was confirmed with satisfactory results for interscale correlations and known-groups comparisons. QLQ-BR23 item-scale correlations met or exceeded the convergent validity criterion of 0.40 for all but one item. QLQ-BR23 interscale correlations met this criterion for three comparisons and for five comparisons with conceptually related QLQ-C30 scales. Correlations between QLQ-BR23 scales and QLQ-C30 functional scales were lower as expected. As expected, most dimensions of the QLQ-BR23 were able to discriminate clearly between pretreatment and current treatment patients. Overall psychometric results for the Sinhala version of the QLQ-BR23 confirmed it as a reliable and valid questionnaire for assessing breast cancer-specific HRQL in Sri Lanka.