A short assessment of health literacy (SAHL) in the Netherlands

BACKGROUND: An earlier attempt to adapt the REALM (Rapid Estimate of Adult Literacy in Medicine) word recognition test to Dutch was not entirely successful due to ceiling effects. In contrast to REALM, the Short Assessment of Health Literacy (SAHL) assesses both word recognition and comprehension in the health domain. The aim of this study was to design, test and validate a SAHL for Dutch patients (SAHL-D). METHODS: We pretested 95 health-related terms (n = 127) and selected 33 best performing items for validation in a quantitative survey (n = 329). For each item, a correct recognition (1 point) and comprehension (1 point) contributed to the total score (scale 0–66). Internal consistency was assessed using Cronbach’s alpha. Construct validity was examined by analyzing association patterns of SAHL-D with educational level, objective and subjective health literacy, prose literacy, and vocabulary. Receiver operating characteristic (ROC) curves, with prose literacy as the reference standard, determined optimal cut-off scores. RESULTS: Cronbach’s alpha was 0.77 for recognition, 0.79 for comprehension, and 0.86 for the total score. Scores significantly differed substantially by educational level. Association patterns mostly confirmed a priori expectations in direction and strength, thereby supporting the construct validity of the SAHL-D. The optimal cut-off scores for differentiating between adequate and low literacy lie between 52.5 and 55.5. A shorter SAHL-D version presenting 22 terms offers a comparable prediction performance. CONCLUSION: The results provide positive evidence for the reliability and validity of the SAHL-D. The SAHL-D can be applied to analyze the role of health literacy in health and healthcare, and for the development and evaluation of targeted interventions. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1186/1471-2458-14-990) contains supplementary material, which is available to authorized users.