Modifiable barriers to adherence to inhaled steroids among adults with asthma: It's not just black and white

Background: Sociobehavioral factors influence adherence to inhaled corticosteroid (ICS) in adults with asthma and warrant exploration as explanations of apparent racial disparities in adherence. Objective: The purposes of this study were to identify barriers to adherence, potentially modifiable by healthcare providers, in a group of African Americans and non-African Americans and to test modifiable barriers as explanations of racial-ethnic differences in adherence. Methods: We conducted a cohort study of 85 adults (mean age, 47 +/- 15 years; 61 [72%] female; 55 [65%] African American) with moderate or severe persistent asthma to determine modifiable sociobehavioral predictors of adherence. These were knowledge of the function of ICS, patient-perceived adequacy of communication with the provider, social support, attitude (perception of risks/benefits of ICS), depression, and self-efficacy. Adherence was calculated from electronic monitoring data as the mean of the number of doses recorded per 12 hours divided by the number prescribed, truncated at 100%. Past adherence, baseline severity of symptoms, and sociodemographics were treated as fixed confounders in ordinal logistic modeling. Results: Adherence was 60% +/- 30%. In bivariate analyses, favorable attitude to ICS (P = .01) was associated with better adherence. Of immutable predictors, African American race-ethnicity (P = .001), lower educational achievement (P = .01), lower household income (P = .002), and more baseline symptoms (P = .003) were associated with poorer adherence. In multivariable analysis, controlling for immutable predictors, favorable attitude was associated with adherence. Favorable attitude was associated with greater adherence in African Americans and non-African Americans. Controlling for immutable factors, the race-adherence relationship was not mediated by the mutable factors, but economic factors (income and insurance) were mediators. Conclusion: Attitude is strongly related to adherence but does not mediate the effect of race-ethnicity.