Assistant Professor BWH Div of Pharmacoepidemiology and Pharmacoeconomics Boston, United States
Background: Poor medication adherence remains highly prevalent and adversely affects health outcomes. Patients frequently describe properties of the pills themselves, like size and shape, as barriers to use but has not been evaluated objectively using real-world data.
Objectives: We sought to determine the extent to which oral medication properties thought to be influential on patient medication-taking translate into lower objectively-measured adherence.
Methods: We identified commercially-insured US beneficiaries newly initiating hypertension, diabetes, or hyperlipidemia treatment with one of five first-line classes in a nationwide claims database. As exposures, we measured pill size, shape, color, and flavoring, number of pills, and fixed-dose combination status as properties. Outcomes included short-term non-adherence (>7 days between first and second fill), discontinuation after first fill, and long-term non-adherence (1-year proportion of days covered < 0.80). We estimated associations between each property and each outcome, by first-line class (e.g., statins), with multivariable logistic regression adjusting for numerous baseline patient demographic and clinical characteristics.
Results: Across 604,323 patients meeting inclusion criteria, short-term non-adherence was 43.3%, 14.6% of patients discontinued after filling once, and 46.0% of patients were non-adherent over 1-year follow-up. Greater pill burden was associated with a higher risk of non-adherence and non-persistence across all classes (e.g., statin short-term non-adherence adjusted OR: 1.57,95%CI:1.34-1.83 for ≥2 pills/day). Larger sizes were also consistently associated with short-term non-adherence (e.g., metformin aOR: 1.02,95%CI:1.01-1.03) and long-term non-adherence except for thiazides. Color, shape, and flavoring had mixed associations across classes. Initiating combination pills was associated with short-term non-adherence but was inconsistent on long-term adherence.
Conclusions: Pill burden is a key property affecting adherence to medication; other factors, like pill size and combination, could modestly affect short-term medication-taking. Researchers, clinicians and others evaluating adherence and designing interventions to improve adherence should consider these factors.
