Using group-based trajectory modeling to characterize the association of past ACEIs/ARBs adherence with subsequent statin adherence patterns among new statin users.

Background: Despite well-documented benefits, statin adherence remains suboptimal. Studies have suggested that previous adherence to other chronic medications is a strong predictor of future adherence to newly initiated statins. Group-based trajectory modeling (GBTM) has been applied as a method to longitudinally depict the dynamic nature of adherence.

Objectives: This study aimed to examine the association between patients' adherence patterns to newly initiated statins and previous adherence trajectories of angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin receptor blockers (ARBs) using GBTM.

Methods: This retrospective cohort study was conducted among continuously enrolled statin initiators using claims data. Patients were included if they had ACEI/ARB use within 1 year before statin initiation (preindex period). Monthly adherence to ACEIs/ARBs was calculated during the preindex period and monthly adherence to statins was assessed 1 year after statin initiation using proportion of days covered (PDC). The monthly PDCs were modeled as a longitudinal response in a logistic GBTM to provide distinct patterns of adherence for ACEIs/ARBs and statins, separately. A multinomial logistic regression was conducted to determine an association between ACEI/ARB adherence trajectories and future statin trajectories, controlling for patient characteristics.

Results: A total of 1078 patients were categorized into 4 distinct statin adherence trajectories: adherent (40.8%), gradual decline (37.4%), gaps in adherence (13.9%), and rapid decline (7.9%). Patients were further categorized into 4 groups on the basis of their distinct past ACEIs/ARBs trajectories: adherent (43%), gaps in adherence (29%), delayed nonadherence (15.2%), and gradual decline (12.8%). In the multinomial logistic regression, patients in the gaps in adherence or gradual decline groups were more likely to follow similar trajectories for future statin use than the adherent trajectory.

Conclusion: Previous adherence trajectories of ACEIs/ARBs may predict future adherence patterns for newly initiated statins. Knowledge of past medication-taking behavior could provide valuable information for developing tailored interventions to improve adherence.