Continuity of Care and Medication Adherence Among Medicare Beneficiaries

Gediwon Negash Milky, Purdue University

Abstract

The objectives for this study were to develop a continuity of care scale, to assess the mean level of continuity of care, to assess association between demographic variables and clinical variables with continuity of care, and to assess association between continuity of care and medication adherence among Medicare beneficiaries. A retrospective cohort study was conducted to achieve the objectives using data from the 2015 to 2017 Medicare Current Beneficiaries Survey (MCBS). To be included in the sample, beneficiaries had to have a hyperlipidemia diagnosis, be continuously enrolled in Medicare Part D for six months from start of medication adherence, be continuously enrolled in Medicare Part A and Part B in the preceding year, and had to have at least two prescription claims for hyperlipidemia medications. Beneficiaries were excluded if they had a proxy responder, had an Alzheimer’s disease or dementia diagnosis, were enrolled in Medicare due to end-stage renal disease or disability, or were residing in a long-term care facility. Among 2,120 beneficiaries that met sample selection criteria, 57 percent were aged 75 years or older, 57 percent were female, and 87 percent were White. An overall continuity of care scale was developed using MCBS items that asked respondents about their care experience. Exploratory factor analysis was used to determine subscales of continuity of care using a randomly selected 60 percent of the sample, which yielded three subscales of continuity of care: relational continuity (Factor 1), informational continuity (Factor 2), and management continuity (Factor 3). Confirmatory factor analysis conducted using the remaining 40 percent of the sample validated factor structure of the continuity of care scale. The mean level of overall continuity of care among Medicare beneficiaries was 3.26 out of 4. Medication adherence was assessed using proportion of days (PDC) covered for anti-hyperlipidemia medications. Beneficiaries with a PDC of 80 percent or more were considered medication adherent. Approximately, 81 percent of beneficiaries were adherent to prescribed hyperlipidemia medications. Association between demographic variables and clinical variables with overall continuity of care was assessed using multivariable logistic regression based on purposeful selection of variables method. Older age, low perceived health status, and lower number of prescribed medications were associated with low overall continuity of care. Race and marital status were found to have interaction effect on overall continuity of care. Among non-white beneficiaries, married beneficiaries reported higher overall continuity of care than not-married beneficiaries. Among married beneficiaries, whites reported lower overall continuity of care than non-whites. Association between overall continuity of care and medication adherence was assessed using multivariable logistic regression with purposeful selection of variables method. There was no association found between overall continuity of care and medication adherence.

Degree

Ph.D.

Advisors

Thomas, Purdue University.

Subject Area

Communication|Aging|Gerontology|Health care management|Health sciences|Medicine|Neurosciences|Public health

COinS