Predictors of Patient-Reported and Pharmacy Refill Measures of Maintenance Inhaler Adherence in Veterans with COPD

Rationale: Suboptimal adherence to inhaled medications in patients with chronic obstructive pulmonary disease (COPD) remains a challenge. Objectives: Examine the sociodemographic, clinical characteristics, and medication beliefs associated with adherence measured by self-report and pharmacy data. Methods: A cross-sectional analysis of data from a prospective observational cohort study of patients with COPD was completed. Participants underwent spirometry and completed questionnaires regarding sociodemographic data, inhaler use, dyspnea, social support, psychological and medical comorbidities, and medication beliefs (Beliefs about Medicines Questionnaire, BMQ). Self-reported adherence with inhaled medications was measured with the Adherence to Refills and Medications Scale (ARMS) and pharmacy-based adherence was calculated from administrative data using the ReComp score. Multivariable linear regression was used to examine the sociodemographic, clinical, and medication belief factors associated with both adherence measures. Results: Among 269 participants with both ARMS and ReComp data, adherence was the same for each measure (38.3%), however only 18% of participants were adherent by both measures. In multivariable adjusted analysis, a 10-year increase in age (β 0.54, 0.14-0.94 95% confidence interval) and number of maintenance inhalers (β 0.53, 0.04-1.02) were associated with increased adherence by self-report. Improved ReComp adherence was associated with chronic prednisone use (β 0.18, 0.04-0.31) and number of maintenance inhalers (β 0.11, 0.05-0.17). In adjusted analyses examining patient beliefs on medications, an increase in the BMQ general Harm score (β -0.15, -0.26 to –0.04) and COPD-specific Concerns score (-0.12, -0.20 to –0.05) was associated with reduced self-reported adherence. No significant associations between ReComp adherence and BMQ were found in adjusted analyses. Conclusions: Adherence to COPD inhaled medications was poor as measured by either self-report or pharmacy refill data. There were notable differences in factors associated with adherence based on the method of adherence measurement. Older age, chronic prednisone use, number of prescribed maintenance inhalers, and patient beliefs about medication safety were associated with adherence. Overall, fewer variables were associated with adherence by pharmacy refill. Pharmacy refill and self-report adherence may measure distinct aspects of adherence and be affected by different factors. These results also underscore the importance of addressing patient beliefs when developing interventions to improve medication adherence.