PDB74 - Association Between Insurance Type and Patient Activation Among Medicare Beneficiaries with Type 2 Diabetes

Abstract

Patient activation has been increasingly recognized as a significant predictor of health behaviors, healthcare utilization, and costs. It’s important to know whether insurance types affect care delivery in a way that influences patient activation. This study assessed patient activation levels of Medicare beneficiaries with type 2 diabetes and assessed the association between insurance type and patient activation. Data from the 2011 and 2012 Medicare Current Beneficiary Survey (MCBS) was used for analysis. Patient activation levels were classified as “low,” scores less than the mean minus half of the standard deviation; as “high,” scores greater than the mean plus half of the standard deviation; or as “moderate,” scores between the cutoff points for “low” and “high.” Effects of managed care enrollment, Medicaid eligibility, and private insurance coverage were examined in bivariate and multivariate ordinal logistic regressions. SAS 9.4 for Linux and an alpha of 0.05 were used in all analyses. The sample included 3,255 patients, 26.4% with low activation, 38.3% with moderate activation, and 35.3% with high activation. Non-white patients (Odds Ratio, O.R. = 1.4, 95% C.I.: 1.2 to 1.6, p = 0.0001) and males (O.R. = 1.4, 95% C.I.: 1.2 to 1.6, p < 0.0001) were more likely to report low activation. In bivariate analyses, Medicaid eligibility (O.R. = 1.5, 95% C.I.: 1.3 to 1.7, p < 0.0001) and no private insurance (O.R. = 1.4, 95% C.I.: 1.2 to 1.6, p < 0.0001) were associated with low activation. However, there was no association between managed care enrollment and patient activation. In multivariate models adjusting for covariates, none of the insurance variables were associated with activation level. Managed care enrollment, Medicaid eligibility, and private insurance coverage were not associated with patient activation levels when other risk factors were accounted for.