| Summary: | AbstractWe aimed to derive and test abbreviated versions of the Shortened Assessment of Health Literacy-Spanish and English (SAHL-S&E) that accurately identify English- or Spanish-speaking lower health literacy adult emergency department (ED) patients. Recursive partitioning of the SAHL-S&E was used to derive four abbreviated versions of the SAHL-S&E by mode of administration (self-administered or staff-administered) and language (English or Spanish). Test performance characteristics of the four abbreviated versions of the SAHL-S&E in distinguishing persons with lower health literacy from those with higher health literacy were assessed against the original full version of the SAHL-S&E. The test performance characteristics of the self-administered English abbreviated SAHL-S&E were: AUC 0.84 (0.79, 0.89), sensitivity 0.84 (0.76, 0.91), and specificity 0.68 (0.61, 0.75); and for the self-administered Spanish version were: AUC 0.88 (0.85, 0.92), sensitivity 0.88 (0.82, 0.93), and specificity 0.78 (0.73, 0.83). For the staff-administered English version, the performance characteristics were: AUC 0.94 (0.91, 0.96), sensitivity 0.98 (0.95, 1.00), and specificity 0.74 (0.69, 0.80), and for the staff-administered Spanish version were AUC 0.89 (0.85, 0.92), sensitivity 0.89 (0.84, 0.94), and specificity 0.80 (0.75, 0.85). Although the four abbreviated versions of the SAHL-S&E performed well they differed by content, length, language and how they are administered, which could add complexity in their routine administration in emergency medicine practice.
|
|---|