Background: The period following hospital discharge is a vulnerable time for patients when increased self-care requirements are common. Low levels of health literacy and social support are thought to contribute to poor post-discharge outcomes. However, little research has specifically examined the role of health literacy alone or in combination with social support in predicting readmission rates.
Methods: The Vanderbilt Inpatient Cohort Study (VICS) is a prospective longitudinal study of adults hospitalized with acute coronary syndromes (ACS) and/or acute decompensated heart failure (ADHF). During hospitalization we assessed demographic characteristics (age, gender, race, education, income), health literacy/numeracy (Brief Health Literacy Screen, short Test of Functional Health Literacy in Adults (sTOFHLA), Subjective Numeracy Scale), and social support (ENRICHD Social Support Inventory, living alone). The combined outcome of interest was all-cause readmission to any acute care hospital or emergency department (ED) visit, which were determined at 30 and 90 days after discharge. We constructed logistic regression models predicting readmission or ED visits using health literacy/numeracy and social support as predictors, adjusted for demographics and diagnosis.
Results: The sample of 2594 patients included 1620 (62.5%) with ACS, 803 (31.0%) with ADHF, and 171 (6.6%) with both conditions. The 30-day incidence of readmissions or ED visits was 23.7% (N=615), and the 90-day incidence was 39.5% (N=1024). In unadjusted analyses, lower levels of health literacy/numeracy and social support were each associated with unplanned health care utilization. In multivariable models, neither health literacy nor social support remained significantly associated with 30-day outcomes. Health literacy measured by the sTOFHLA was uniquely associated with 90-day outcomes (p=0.015) and explained about 3.3% of the variation in readmissions/ED visits at 90 days.
Conclusions: Health literacy was independently associated with 90-day readmissions/ED visits, after taking other patient characteristics into account. Social support was not independently associated with readmissions at 30 or 90 days in this sample.