Group-Based Trajectory Modeling to Identify Longitudinal Patterns of Adherence to Statins in Older Adults on Concomitant Triple Therapy (Oral Antidiabetic, RAS Antagonists, Statins)
Author(s)
Sai S. Cheruvu, MPH, Bilqees Fatima, MPhil, PharmD, Susan Abughosh, PhD;
University Of Houston-College of Pharmacy, Pharmaceutical Health Outcomes and Policy, Houston, TX, USA
University Of Houston-College of Pharmacy, Pharmaceutical Health Outcomes and Policy, Houston, TX, USA
OBJECTIVES: Diabetes, hypertension, and hyperlipidemia frequently coexist in older adults, which significantly increases cardiovascular risk. Medication adherence is crucial to mitigating these risks but is challenging due to the complexity of triple therapy regimens. Group-based trajectory modeling (GBTM) offers a dynamic approach to understanding adherence patterns and visualizing behavior over time compared to other measures like the proportion of days covered (PDC). This study evaluated statin adherence using GBTM in elderly patients on concomitant triple therapy.
METHODS: Patients on concurrent triple therapy with continuous enrollment were identified using Texas Medicare Advantage dataset (July 2016-December 2016). Patients needed to have a minimum 30-day overlap of triple therapy (renin-angiotensin system antagonists, oral antidiabetics, and statins) according to STAR metrics. Patients were also required a second prescription for each therapy component within the identification period and a 12-month follow-up. Monthly statin adherence was assessed using PDC during the follow-up period, defining adherence as ≥80% (24 out of 30 days) coverage. A logistic GBTM was applied with monthly PDC data to identify distinct adherence trajectories, estimating 2-5 groups using a second-order polynomial function of time. Group differences in trajectories based on demographic and clinical patient characteristics will be evaluated using Chi-square for categorical variables and ANOVA for continuous variables.
RESULTS: Among 7,847 patients, 3-adherence trajectories were identified based on Bayesian information criteria, clinical relevance, and a minimum group size of 5%. The trajectories were gradual decline (19.4%), adherent (23.6%), and near-perfect adherence (57%). The model was validated using criteria recommended in Nagin et al. 2005.
CONCLUSIONS: Approximately 19.4% of patients are non-adherent to statin therapy. The identification of distinct adherence trajectories and associated patient characteristics provides insight into patient behaviors and lays the foundation for targeted interventions to enhance adherence among patients in the gradual decline trajectory.
METHODS: Patients on concurrent triple therapy with continuous enrollment were identified using Texas Medicare Advantage dataset (July 2016-December 2016). Patients needed to have a minimum 30-day overlap of triple therapy (renin-angiotensin system antagonists, oral antidiabetics, and statins) according to STAR metrics. Patients were also required a second prescription for each therapy component within the identification period and a 12-month follow-up. Monthly statin adherence was assessed using PDC during the follow-up period, defining adherence as ≥80% (24 out of 30 days) coverage. A logistic GBTM was applied with monthly PDC data to identify distinct adherence trajectories, estimating 2-5 groups using a second-order polynomial function of time. Group differences in trajectories based on demographic and clinical patient characteristics will be evaluated using Chi-square for categorical variables and ANOVA for continuous variables.
RESULTS: Among 7,847 patients, 3-adherence trajectories were identified based on Bayesian information criteria, clinical relevance, and a minimum group size of 5%. The trajectories were gradual decline (19.4%), adherent (23.6%), and near-perfect adherence (57%). The model was validated using criteria recommended in Nagin et al. 2005.
CONCLUSIONS: Approximately 19.4% of patients are non-adherent to statin therapy. The identification of distinct adherence trajectories and associated patient characteristics provides insight into patient behaviors and lays the foundation for targeted interventions to enhance adherence among patients in the gradual decline trajectory.
Conference/Value in Health Info
2025-05, ISPOR 2025, Montréal, Quebec, CA
Value in Health, Volume 28, Issue S1
Code
MSR151
Topic
Methodological & Statistical Research
Disease
SDC: Cardiovascular Disorders (including MI, Stroke, Circulatory), SDC: Diabetes/Endocrine/Metabolic Disorders (including obesity), SDC: Geriatrics