Group-Based Trajectory Models to Evaluate the Association of Lipid Testing and Statin Adherence.

Abstract

Background and objective: Performing lipid testing after statin initiation is recommended to monitor response. Inadequate response may indicate non-adherence, which is associated with an increased risk of cardiovascular events and increased costs. Group-based trajectory modeling is an approach to establish probabilistic developmental trajectories of adherence, differentiating individuals by their distinct longitudinal medication-taking behaviors. We examined whether lipid testing is associated with distinct trajectories of statin adherence among individuals enrolled in a Medicare fee-for-service plan in the USA.

Methods: A retrospective cohort study was conducted using the Centers for Medicare & Medicaid Chronic Condition Warehouse 5% sample of Medicare fee-for-service data between 2006 and 2015. Statin use and lipid testing were identified using claims data. The proportion of days covered was calculated for each 30 days after the index date, which was used to estimate the probability of belonging to each potential adherence trajectory.

Results: In a cohort of 138,101 statin initiators, four statin adherence trajectory groups were identified. The four groups were differentiated as "rapid decline" (21.53%), "gradual decline" (10.25%), "decline first then improve later" (26.47%), and "high adherence" (41.75%). Compared with "high adherence," initiators who had lipid tests within 360 days after statin initiation were less likely to fall into "rapid decline" (adjusted odds ratio: 0.661; 95% confidence interval 0.641-0.683), "gradual decline" (adjusted odds ratio: 0.834; 95% confidence interval 0.801-0.868), and "decline first then improve later" groups (adjusted odds ratio: 0.936; 95% confidence interval 0.910-0.962).

Conclusions: Lipid testing is positively associated with greater use of statin medication across different adherence trajectories in the present study.