2D-06 - Long-Term Trajectories and Predictors of Concurrent Psychotropic Medication Use in Residential Aged Care: A Group-Based Multi-Trajectory Modeling Approach

Background: Psychotropic medicines are commonly used in residential aged care (RAC) homes despite notable safety concerns. No prior studies have examined the longitudinal concurrent use of psychotropic medicines.

Objectives: To identify trajectories of concurrent use of three psychotropic medication classes over three years and determine predictors of trajectory group membership for residents with and without dementia.

Methods: Study design: A longitudinal cohort study utilising electronic health data (2019–2022). The study included 2837 permanent residents, including 47.4% (n=1344) with dementia.
Settings: 30 RAC homes across two aged care providers in Sydney, New South Wales, Australia.
Data source: Linked electronic data containing demographics, health, and daily medication administration.
Exposure: Weekly exposure to three classes of psychotropic medications—antidepressants, antipsychotics, and anxiolytics/hypnotics—was tracked over a three-year period.
Statistical analysis: We used group-based multi-trajectory modelling to identify concurrent trajectories of the three psychotropic medicine classes and multinomial logistic regressions to determine the predictors of trajectory groups.

Results: The concurrent use of multiple psychotropic classes occurred in 23.3% of non-dementia and 31.6% of dementia cohorts. The model identified 6-group and 4-group trajectories as the optimal fit for dementia and non-dementia cohorts respectively. The model performance metrics were excellent in both cohorts, including average posterior probability ≥ 0.98, odds of correct classification >500, and relative entropy >0.98 for all groups. Psychotropic use trajectories mostly remained stable over time in non-dementia cohorts, while dementia cohorts showed more diverse and fluctuating use. Multinomial logistic regressions identified eleven predictors of trajectory membership in dementia and nine in non-dementia cohorts.

Conclusions: Our findings reveal a concerningly high rate of concurrent psychotropic medication use in RAC homes, with distinct usage patterns between residents with and without dementia. Further research should assess the appropriateness of such use and consider strategies for improving health outcomes.