Trajectory Patterns of Adherence to Oral Anticoagulants and Associated Factors Among Chinese Patients With Atrial Fibrillation: A Latent Growth Mixture Model

BACKGROUND: Patients with atrial fibrillation (AF) benefit from long-term adherence to oral anticoagulants (OACs) for stroke prevention. Adherence to OACs undergoes continuous variation over time, therefore, exploring the trajectory and latent patterns of adherence to OACs among patients with AF in China is critical. OBJECTIVES: To identify the trajectory and latent patterns of adherence to OACs among Chinese patients with AF, and determine predictors associated with trajectory patterns of adherence to OACs. METHODS: A cross-sectional study was conducted in Guangzhou, China, from July 2023 to January 2024. Data on medication adherence, beliefs about medicines, and anticoagulation knowledge were collected before hospital discharge and at 1 and 3 months after discharge. Sociodemographic and clinical data were also collected. Generalized estimation equation was used to investigate the overall trajectory of medication adherence. Latent growth mixture modeling was used to explore adherence trajectory patterns, and then binary logistic regression was used to explore predictors associated with each pattern. RESULTS: A total of 240 participants (age: 61.59 ± 10.91, 68.3% male) were included. Two adherence trajectory patterns were identified: "medium adherence-slight ascent followed by a downward trajectory" (72.9%) and "non-adherence-sharp ascent followed by a downward trajectory" (27.1%). Anticoagulation knowledge, beliefs about the necessity of medication, and anticoagulation knowledge were associated with different patterns ( P < .05) of OACs medication adherence in patients with AF. CONCLUSIONS: Identification of longitudinal patterns of nonadherence to OACs can alert clinicians to the need to monitor adherence longitudinally.