Adults with diagnosed and untreated diabetes: Who are they? How can we reach them?

Policy Relevance

Diabetes is a serious chronic condition that can lead to complications such as cardiovascular or kidney disease. Appropriate treatment of diabetes can lower the risk of complications and healthcare costs. Information about the factors associated with untreated diabetes could help physicians and health plans target and refine their efforts to promote appropriate treatment of this condition.

Data and Sample

The Medical Expenditure Panel Survey (MEPS) for 2000 – 2006 is the primary data source. The MEPS is a household-based survey that can be used to produce nationally representative estimates of healthcare use, insurance coverage, health status, and demographic and socioeconomic characteristics for the non-institutionalized, civilian population in the U.S. The study population includes 10,746 person-year observations for adults (age 18 and older) who reported that they had been diagnosed with diabetes. Diabetes treatment is defined to include use of insulin, oral diabetes medications, or diet modification; these treatments are components of the Standards of Medical Care in Diabetes published by the American Diabetes Association. By this definition, approximately 2.6 percent of the adults with diagnosed diabetes in our sample were untreated. Additional data sources include the Census Bureau Zip Business Pattern files (diet and weight reduction centers per capita within each zip code) and TNS Media Intelligence (direct-to-consumer advertising for diabetes medications within market areas).

Methods

We use logit models to investigate whether demographic and socioeconomic characteristics, health status, insurance coverage, diabetes severity (Medical Outcomes Study scale), communication from providers, patient attitudes, direct-to-consumer advertising for diabetes medications, and diet centers per capita are associated with diagnosed but untreated diabetes. We use multinomial logit models to investigate the relationship between these same factors and four treatment categories: any insulin, oral medications only, diet only, and no treatment.

Preliminary Findings

Estimates from a preliminary logistic model indicate that Asian-Pacific Islanders and persons who lack insurance or a usual source of care are more likely than others to have untreated diabetes (p values < 0.05). By contrast, persons who are married or who have medical providers who usually explain things in a manner they understand are less likely than others to have untreated diabetes. Future work includes investigating the association of drug advertisement, diet centers, and other provider and market characteristics with diabetes treatment. These variables are plausibly related to information dissemination through which policy, programs, or practitioners might be able to change patient behavior.