Cancer Health Literacy Test-30-Spanish (CHLT-30-DKspa), a New Spanish-Language Version of the Cancer Health Literacy Test (CHLT-30) for Spanish-Speaking Latinos

Abstract

This article describes the adaptation and initial validation of the Cancer Health Literacy Test (CHLT) for Spanish speakers. A cross-sectional field test of the Spanish version of the CHLT (CHLT-30-DKspa) was conducted among healthy Latinos in Louisiana. Diagonally weighted least squares was used to confirm the factor structure. Item response analysis using 2-parameter logistic estimates was used to identify questions that may require modification to avoid bias. Cronbach’s alpha coefficients estimated scale internal consistency reliability. Analysis of variance was used to test for significant differences in CHLT-30-DKspa scores by gender, origin, age and education. The mean CHLT-30-DKspa score (N = 400) was 17.13 (range = 0–30, SD = 6.65). Results confirmed a unidimensional structure, χ2(405) = 461.55, p = .027, comparative fit index = .993, Tucker–Lewis index = .992, root mean square error of approximation = .0180. Cronbach’s alpha was .88. Items Q1-High Calorie and Q15-Tumor Spread had the lowest item–scale correlations (.148 and .288, respectively) and standardized factor loadings (.152 and .302, respectively). Items Q19-Smoking Risk, Q8-Palliative Care, and Q1-High Calorie had the highest item difficulty parameters (difficulty = 1.12, 1.21, and 2.40, respectively). Results generally support the applicability of the CHLT-30-DKspa for healthy Spanish-speaking populations, with the exception of 4 items that need to be deleted or revised and further studied: Q1, Q8, Q15, and Q19.