SW-2C - Dynamic Modeling of Medication Adherence: A Hands-on Exploration of Trajectory Analysis Techniques

Background: Medication adherence significantly influences treatment effectiveness and health outcomes, yet its dynamic and complex patterns present challenges for analysis and intervention design. While traditional methods often oversimplify adherence behaviors, trajectory analysis provides a more sophisticated approach to capturing adherence trajectories over time. Adherence researchers are increasingly interested in applying these techniques. Building on the success of the Adherence SIG-sponsored symposium at ICPE 2024, titled “Applications of Trajectory Methods in Medication Adherence Research: A Decade in Review and a Look Ahead,” which garnered high participation and interest, this workshop will provide a hands-on exploration of these advanced techniques. The workshop will also serve as a platform to solicit input and gain insight into common questions regarding implementing trajectory analysis techniques, helping to refine the development of the ISPE-funded manuscript on “Guideline and Best Practice for Classifying Medication Adherence using Latent Variable Mixture Modeling”. The interactive format ensures attendees gain practical skills and experience in dynamic modeling for medication adherence trajectories.

Objectives: This workshop will provide attendees with the necessary knowledge and hands-on experience to implement medication adherence trajectory techniques (SAS macro and R package). Participants will learn key considerations for applying adherence trajectory modeling, follow real-world examples, and understand the advantages of dynamic measures over static adherence assessments. Challenges and solutions in implementation will be discussed. All ISPE members interested in learning dynamic modeling of medication adherence using trajectory analysis techniques will benefit from this workshop.

Description: This hands-on workshop will allow attendees to develop skills in modeling medication adherence trajectories. Drs. Julie Lauffenburger and Chao Cai will introduce adherence trajectory modeling, discussing limitations of static measures, advantages of dynamic approaches, and key literature. They will further demonstrate steps in data preparation and coding for implementing group-based trajectory modeling (GBTM) in SAS and latent class mixed models (LCMM) in R. Dr. Marissa Seamans will present case-studies and discuss common challenges and solutions in adherence trajectory modeling. The final 30 minutes will be dedicated to interactive Q&A, and an engaging discussion with the audience. Interactive discussion will encourage participants to share their own experiences and insights.