National Health Literacy Mapping to Inform Healthcare Policy

Our study objective is to provide valid estimates of health literacy for every census tract in the U.S. and demonstrate the value of such a health literacy measure for informing public health and healthcare policy and practice. Health literacy, or one’s ability to obtain, process and understand information needed to make health decisions, is considered not only a reflection of an individual’s skills and abilities, but also how well health systems provide information and services. Individuals who are older, less educated, lower income or racial/ethnic minorities are at greatest risk for low health literacy. This concept therefore serves as an indicator of the quality, equity, and safety of public health and healthcare. Unlike many other indicators, however, health literacy is considered modifiable, as interventions designed around health literacy ‘best practices’ have been shown to lead to improved health outcomes. Despite its promise, health literacy has rarely been included in public health and healthcare policymaking. We seek to address this by demonstrating how health literacy can be used to better inform current policies relating to hospital readmission. Health literacy is one of the most important, independent risk factors for hospital readmission. Strategies that incorporate health literacy principles have successfully reduced readmissions. This is particularly significant as the Centers for Medicaid & Medicare Services (CMS) recently began penalizing hospitals with adjusted readmission rates above the national average for acute myocardial infarction (AMI), heart failure (HF) and pneumonia. To date, the adjusted rates used by CMS include only medical comorbidity and age. While the purpose of the penalties is to promote quality of care and reduce costs, the lack of adjustment for health literacy places an undue burden on hospitals that serve patients who are at greatest risk for low health literacy and are more likely to struggle with self-care activities. Adjusting for health literacy would not only ensure that penalties are fair, but would provide hospitals with more explicit targets for quality improvement efforts (i.e. patient education and health literacy). Through this study, we will demonstrate that a derived health literacy measure can be estimated for each of the 73,057 census tracts in the U.S. using the RAND predictive model. By linking health literacy estimates to a national cohort of ~300,000 Medicare beneficiaries with hypertension and new-onset diabetes, we will demonstrate the predictive strength of the health literacy estimate. We will then link the health literacy estimates to a national cohort of ~176,000 Medicare patients with AMI to examine the impact of health literacy estimates on hospital readmissions for AMI. Finally, we will we will use the health literacy estimates to augment the CMS analytical models (clinical profiling shrinkage estimation and risk-adjustment) of adjusted readmission rates for AMI. Our specific aims are to: 1a) Estimate and map the average health literacy score of each census tract in the United States, 1b) Examine the relationship between health literacy estimates and a composite of clinical endpoint outcomes (mortality, HF, AMI, stroke, end-stage renal disease) among ~300,000 Medicare beneficiaries, 2) Investigate the impact of health literacy estimates on 30-day readmission among a national cohort of ~176,000 Medicare patients with AMI and 3) Compare the risk standardized readmission rates with and without adjustment for the derived health literacy score among the national Medic