The objectives of this study were to assess
patient activation levels, to assess association between sociodemographic
characteristics and patient activation, to assess association between health
status characteristics and patient activation, and to assess association
between patient activation and medication adherence among Medicare
beneficiaries with type 2 diabetes. A
retrospective cohort study was conducted using data from the 2009 through 2013 Medicare
Current Beneficiary Survey (MCBS).
Patient activation was measured with the Patient Activation Supplement
in the MCBS and was categorized as low, moderate, and high levels based on
activation scores. Medication adherence was assessed with
proportion of days covered (PDC) using Medicare Part D administrative records
from the MCBS within a period of six months after measurement of patient
activation. The
sample included Medicare beneficiaries who completed the MCBS Patient Activation
questionnaire, who were diagnosed with type 2 diabetes, and who were 18 or
older. Beneficiaries were excluded if
they responded “Not ascertained,” “Not Applicable,” “Don’t know” or “Refused”
to more than 50 percent of the Patient Activation questions, did not have
continuous Medicare Part A and Part D coverage throughout the assessment
period, had less than two Medicare Part D claims for an antidiabetic medication
throughout the assessment period, used insulin during the assessment period,
resided in long-term care facilities, or had Alzheimer’s disease, dementia,
mental retardation or mental disorder. All analyses were conducted in SAS 9.4 for Unix environment. An a
priori alpha level of 0.05 was used to determine significance. Bivariate and multivariable
weighted ordinal logistic regression were applied for assessing associations. A
total of 571 individuals met sample selection criteria. The mean age was 72.4
years. Of the 571 persons in the sample,
27.5 percent were at low activation level, 38.7 percent were at moderate
activation level, and 33.7 percent were at high activation level. Approximately three-fourths of the sample persons were adherent to antidiabetic medications. Low activation was more
likely to be found in males, less educated patients, and patients without
arrhythmia. Ex-smokers as compared to
non-smokers and overweight patients as compared to those with healthy weight
were less likely to report low activation.
In multivariable logistic analysis adjusting for race, gender,
osteoporosis, Charlson Comorbidity Index score, and number of prescribed
medications, patient activation level was not significantly associated with
medication adherence. Non-Whites
and patients with a Charlson Comorbidity Index score of 1 as
compared to those with a score of 0 were more likely to be non-adherent. A lower number of prescribed medications was
associated with higher odds of non-adherence.