(99) Group-Based Trajectory Modeling of Medication Refill Patterns of Novel Oral Anticoagulant (NOACs): An Analysis of The Phollow Cohort

Sunday, August 28, 2022

8:00 AM – 1:45 PM CEST

Location: Hall C

Publication Number: 1144

Presenting Author(s)

Antonio Teixeira Rodrigues, PharmD, PhD
– Invited Assistant Professor, School of Medicine, University of Minho, Braga, Braga, Portugal

Co-Author(s)

José Guerreiro
– Data analyst, Centre for Health Evaluation and Research, National Association of Pharmacies (CEFAR/Infosaude-ANF), Lisboa, Lisboa, Portugal
Carla Torre
– Faculty of Pharmacy University of Lisbon, Lisboa, Lisboa, Portugal

Background: Adherence to medication is one of the most important factors of disease management for chronic conditions. Group-based trajectory modeling (GBTM) is being widely used as a method to characterize the dynamic pattern of real-world medication adherence over time and can provide a better understanding of patient journey throughout their treatment course.

Objectives: To explore the use of GBTM in the identification of clusters or groups of patients with similar patterns of medication refill behavior among novel oral anticoagulant (NOACs) users.

Methods: Retrospective, multicenter cohort study (Phollow cohort) of adult patients taking NOACs (ATC: B01AA03, B01AA07, B01AE07, B01AF01, B01AF02 and B01AF03). Patients were recruited from Portuguese community pharmacies (CP). Medication refill data was retrieved through the dispense pharmacy records for 3 years prior to the recruitment date. Adherence was defined as the proportion of days covered (PDC) and evaluated for a period of 2 years. GBTM was used to identify groups of distinct patterns of adherence.

Results: About the sample, 66.6% were male and 79.4% were aged 65 years or older. Overall, mean PDC rate at 1-year was 70.8.% and at 2-year was 73.6%. Patients were categorized in four different trajectories: the majority were adherent (61.7%), followed by gradual decline (19.3%), gaps in adherence (13.3%) and rapid decline (5.8%). Average probability of group membership was higher or equal to 93%.

Conclusions: Clinically relevant adherence trajectories (patterns) were identified among NOAC users. The identification of different trajectories on the adherence pattern can help to improve strategies for interventions in therapeutic management leading better outcome results on adherence, especially in patients with low-adherence trajectories. More work is needed to study factors that can be potentially associated with (non)adherent trajectories.