David H. Howard, Ph.D.
Past research has shown that individuals with low levels of health literacy are more likely to be hospitalized and have worse disease outcomes (Baker et al., 1998, 2002; Schillinger et al., 2002). The major obstacle to extending these results by examining the relationship between health literacy and spending is the lack of data containing measures of both. Some insight about the impact of health literacy on costs can be gleaned from studies that examine the impact of years of schooling on medical costs, but health literacy is a fundamentally different concept from educational attainment (Davis et al., 1994; Stedman and Kaestle, 1991). At least one study has examined the association between general literacy and costs (Weiss et al., 1994). It found no relationship, though the sample size was small (N = 402) and not representative of the overall U.S. population (over half the study subjects qualified for Medicaid because of disability). Furthermore, inpatient and outpatient costs were not analyzed separately and the analysis did not control for confounding patient characteristics. Another paper by the same author (Weiss) shows large differences in costs by grade reading level in a Medicaid population (Weiss, 1999), but descriptions of the methods and data are not available. This study examines the relationship between health literacy and costs using a unique dataset combining cost information from an administrative claims file and a health literacy measure from a beneficiary survey. Multivariate techniques are used to adjust for underlying differences in respondents’ characteristics.
Health literacy data were collected as part of a survey of persons enrolling in a Prudential Medicare health maintenance organization between December of 1996 and August of 1997 in one of four locations: Cleveland, Ohio; Houston, Texas; South Florida (including Fort Lauderdale, Miami, and nearby areas); and Tampa, Florida. New Prudential Medicare members were contacted three months after enrollment, and those meeting the eligibility criterion were asked to complete an in-person survey. In order to be included in the study, members had to be comfortable speaking either English or Spanish, living in the community, and possess adequate visual and cognitive function. The survey included the Short Test of Functional