Prediction of adherence to treatment with statins and anti-platelet drugs in first-year post-stroke patients: Validation of beta-regression models

by Elias Edward Tannous, Shlomo Vinker, David Stepensky, Eyal Schwarzberg

Stroke is the third most common cause of disability and the second most common cause of death worldwide. Greater levels of medication adherence after stroke or transient ischemic attack are associated with improved survival. Very few medication adherence prediction models are available and have not been validated using external data. The current study aimed to evaluate the predictive performance of previously published beta regression models for statin and antiplatelet adherence at 1 year in patients’ post-stroke or transient ischemic attack. The models use the first 90-day adherence data as a single predictor for 1-year adherence. Adherence was measured using the Proportion of Days Covered (PDC), which utilized prescription-filling data. Model performance was assessed using the following metrics: R² (proportion of variance explained), the difference between the mean observed and the mean predicted PDC, and the calibration slope, which ideally should be one. 2369 were included in the statin cohort, and 2147 patients were included in the antiplatelet cohort. R² was 0.67 and 0.56 for statin and antiplatelet models, respectively. The difference between the mean observed and the mean predicted PDC was −3.7% and −2.5% for statin and antiplatelet models, respectively. The calibration slopes were 1.06 and 0.96 for the statin and antiplatelet models, respectively. The model performed well on a new patient population comprised of post-stroke patients and may be used for early identification of patients at high risk for low 1-year adherence within 90 days post-stroke, enabling timely, targeted adherence-support interventions.