FULL PDF
Abstract
This study assessed the measurement invariance of the Kidney Disease Quality of Life 36-item (KDQOL TM-36) Burden of Kidney Disease (BKD), Symptoms and Problems with Kidney Disease (SPKD), and Effects of Kidney Disease (EKD) scales between Black (n = 18,404) and White (n = 21,439) US dialysis patients surveyed in 2015 and 2016, and calculated normative scores for each scale for the US dialysis population. Measurement invariance was analyzed using multiple group confirmatory factor analysis at three levels: Configural Model (invariant factor structure), Metric Model (invariant factor loadings), and Scalar Model (invariant intercepts). Criteria for invariance included non-significant chi-square tests, a difference in comparative fit index (DCFI) below 0.002, and root mean squared error of approximation and standardized root mean square residual differences below 0.015. Differential item functioning (DIF) was examined by freeing loadings and thresholds item-by-item. Normative scores were calculated by applying survey weights to match the sample to key clinical characteristics of the US dialysis population (n = 443,947). The results indicated significant chi-square differences between models, with a DCFI of 0.006 between Metric and Scalar Models, reduced to 0.001 after freeing intercepts for BKD and SPKD. Standardized factor means for the Black group compared to the White group were 0.218, 0.061, and 0.161 for BKD, SPKD, and EKD, respectively. Sequentially released loadings and thresholds resulted in factor mean differences ranging from 0.001 to 0.048. Normative mean (95% CI) scores for the overall sample were 52.8 (52.6–53.1) for BKD, 79.0 (78.9–79.2) for SPKD, and 74.1 (74.0–74.3) for EKD. Despite evidence of DIF, its impact on kidney-targeted HRQOL estimates with the KDQOL TM-36 was minimal, and normative scores can serve as reference values for research and clinical assessments.
View more >>