Cross-cultural comparisons of limited cancer health literacy rates

Abstract

Abstract Purpose This study aims to compare the limited cancer health literacy (LCHL) rates among White (U.S.), Black (U.S.) and Chinese (H.K.) cancer patients using the Cancer Health Literacy Test - 6 (CHLT-6) and its recent adaptation in Chinese. Rationale Arbitrary nature of determining LCHL using the cut-score method is a major obstacle to scientifically investigate racial/ethnic differences in LCHL rates. To overcome, we used the multi-group latent class analysis (LCA) to test invariant measurement properties of the CHLT-6, a necessary condition to claim that the differences in LCHL rates reflect true group differences. Results Data included 1,899 cancer patients: White (N = 806) and Black (N = 491) and Chinese (N = 602). Measurement invariance test between the unconstrained and conditional probability constrained multi-group LCA favored the latter using the Bayesian Information Criterion (BIC: 10,388 vs.10,334) indicating that the LCHL label has the same meaning between three groups. The CHLT-6 determines cancer patients with LCHL with high accuracy (>90%) in all three groups. The model-based estimates of LCHL rates varied across groups: White (4%), Black (50%), and Chinese (25%). Conclusions This study provides a strong construct validity evidence for the CHLT-6 by showing that it has invariant measurement properties between racial/ethnic groups and that it assigns cancer patients into LCHL group with high accuracy. Causal mechanisms underlying a wide range of LCHL rates need further explorations. Key messages • A substantial difference in limited cancer health literacy rates among White (U.S), Black (U.S.), and Chinese (Hong Kong) cancer patients. • The CHLT-6 accurately captures the limited cancer health literacy in diverse racial/ethnic groups.