A multi-group confirmatory factor analyses of the LupusPRO between southern California and Filipino samples of patients with systemic lupus erythematosus

Abstract

Introduction Systemic lupus erythematosus (SLE) leads to a range of biopsychosocial health outcomes through an unpredictable and complex disease path. The LupusPRO is a comprehensive, self-report measure developed specifically for populations with SLE, which assesses both health-related quality of life and non-health related quality of life. Given its increasingly widespread use, additional research is needed to evaluate the psychometric integrity of the LupusPRO across diverse populations. The objectives of this study were to evaluate the performance of the LupusPRO in two divergent patient samples and the model fit between both samples. Methods Two diverse samples with SLE included 136 patients from an ethnically-diverse, urban region in southern California and 100 from an ethnically-homogenous, rural region in Manila, Philippines. All patients met the ACR classification criteria for SLE. Confirmatory factor analysis (CFAs) were conducted in each sample separately and combined to provide evidence of the factorial integrity of the 12 subscales in the LupusPRO. Results Demographic analyses indicated significant differences in age, disease activity and duration, education, income, insurance, and medication use between groups. Results of the separate CFAs indicated moderate fit to the data for the hypothesized 12-factor model for both the Manila and southern California groups, respectively [χ2 (794) = 1283.32, p < 0.001, Comparative Fit Index (CFI) = 0.793; χ2 (794) =1398.44, p < 0.001, CFI = 0.858]. When the factor structures of the LupusPRO in the southern California and Manila groups were constrained to be equal between the two groups, findings revealed that the factor structures of measured variables fit the two groups reasonably well [χ2 (1697) = 2950.413, df = 1697, p < 0.000; CFI = 0.811]. After removing seven constraints and eight correlations suggested by the Lagrange multiplier test, the model fit improved significantly [χ2 (15) = 147.165, p < 0.000]. Conclusions This research provides significant support for the subscale structure of the LupusPRO in two disparate cultural samples of SLE patients. Despite significant sociodemographic and clinical differences between the two samples, for the most part, the LupusPRO performed similarly in both samples.