LASSO logistic regression and cluster analysis in predicting adherence and drug patterns among new users of monotherapy for antihypertensive drugs.

OBJECTIVE

Real-world long-term adherence and drug utilization patterns are essential for hypertension management. However, the evidence was unclear. We aimed to construct adherence and drug patterns risk prediction models.

METHODS

We designed a retrospective inception cohort study using pharmacy records from the University of Groningen IADB.nl dispensing database. Exposures were antihypertensive drug monotherapy including angiotensin-converting enzyme inhibitors, angiotensin II receptor blockers, beta-blockers, calcium channel blockers, and thiazides. Primary outcomes included adherence rates and drug utilization patterns. Cluster analysis was performed to categorize patients with similar risk factors. Least Absolute Shrinkage and Selection Operator logistic regression was then employed to construct prediction models.

RESULTS

The adherence rate in the 1 st year to the 10 th year increased from 86.4% to 92.9%. Most monotherapies, middle and older age, with initial diabetes, asthma/ chronic-obstructive-pulmonary-disease drug, and without psycholeptics drug, first exposure prescription after 2000 promoted high adherence. Thiazides and being male were harmful to continuation and helpful to discontinuation, switching and adding on. Middle and older age promoted switching and adding on but impeded discontinuation. High adherence contributed to continuation, switching and adding on but impeded discontinuation. Thiazides were better for the short term use to achieve high adherence while calcium channel blockers in both short term and long term were better to achieve high adherence than beta-blockers.

CONCLUSIONS

The results highlighted key factors influencing medication adherence and treatment changes, emphasizing the need for personalized approaches to optimize patient care. Enhancing adherence and provide specific monotherapy for short term and long term plan were beneficial to better personalized treatment.