Social support and medication adherence among African-Americans and whites with heart failure

Background: Proper adherence to medications may decrease hospital use among heart failure patients. African-American race is generally thought to be related to lower medication adherence. Yet, less is known about the influence of social support on medication adherence among all heart failure patients. Therefore, we assessed the relationship between social support and medication adherence in a diverse population of heart failure patients.

Methods: We surveyed 322 African-American and white heart failure patients at 2 hospitals in Birmingham, AL. Our primary outcome was medication adherence measured by the Morisky Scale. Each of the 4 Morisky items was coded as perfect (0 points) or imperfect (1 point) adherence. Our final outcome was an ordinal scale (range “0” for perfect adherence to “4” worst adherence) based on a summation of the 4 items. Our primary independent variable was perceived social support measured by the Medical Outcomes Study Social Support Scale (range 0-76, higher score=higher social support). We dichotomized our population into categories of “low” and “high” social support based on a split at the 25th percentile (48). Other independent variables included race, gender, education, income, marital status, number of family members and friends available in times of need, type of doctor (primary care vs. cardiologist), number of comorbidities, knowledge of heart failure, and number of years with heart failure. We conducted bivariable analyses to identify which factors were significantly related (p< 0.05) to medication adherence. Those factors identified were used to build a multivariable ordinal logistic regression model to assess independent relationships with medication adherence.

Results: The study population was 72% African-American and 26% white. The mean age was 60, and 47% were male and 53% female. Seventy-five percent of participants had heart failure more than one year. Only 10% reported perfect medication adherence. In the multivariable analysis, low perceived social support (vs. high) decreased the odds of better medication adherence by 50% (OR 0.5, 95% CI 0.3-0.8). Having at least one friend (vs. no friends) available in times of need also doubled the odds of better medication adherence (OR 2.0, 1.2-3.4). There were no racial differences in medication adherence.

Conclusions: Inadequate social support may partially explain poor medication adherence among African-American and white heart failure patients. The availability of friends may also be a marker of a larger social network positively impacting medication adherence. Improving social support may be a viable means of increasing medication adherence among all patients with heart failure.