Evolution of Health and Medication Use in a Cohort of Nursing Home Residents: A Data Linkage Study

Describe the evolution of chronic medication use, potentially inappropriate medications (PIMs), and health outcomes in nursing home residents, and explore associations between changes in medication use and health outcomes.

A retrospective longitudinal data linkage study.

Nursing home residents in Flanders, insured in Belgium, aged ≥65 years, with the Belgian version of the interRAI (BelRAI) long-term care facility assessment between 2016 and 2022.

Linked data from the BelRAI long-term care facility and national health care claims (InterMutualist Agency/Agence InterMutualiste). Residents were assessed at baseline (first BelRAI after admission) and 12 months later (±1 month) window. PIM selection was based on a published umbrella review. Health outcomes were measured using validated interRAI scales.

At baseline, 8713 residents were assessed, of whom 1691 (19.4%) had a 1-year reassessment. At baseline, 94.6% used ≥1 chronic medication, with polypharmacy in 48.4% and excessive polypharmacy in 15%. Over 1 year, the mean chronic medications increased from 5.87 to 6.1, whereas PIM use declined from 72.7% to 56.0%. Polypharmacy transitions involved 13% of residents moving in and 8% moving out, associated with changes in activities of daily living (ADL) performance and cognitive function. For PIMs, 48.3% continued use, 24.5% discontinued, and 7.8% started, associated with ADL, aggressive behavior, and comorbidity. Across all health domains, 70% to 88% of residents maintained baseline status. When changes occurred, deterioration was more frequent than improvement, with the highest declines in health instability (23.6%), mood (18.3%), and cognition (13.6%), and lowest in pain (9.7%) and ADL (9.2%).

Over 1 year, chronic medication use increased, whereas PIMs decreased. Most health outcomes remained stable, though deterioration was more frequent than improvement. Changes in medication use were associated with changes in health outcomes, and some residents improved, suggesting targeted interventions may help. This innovative data linkage demonstrates strong potential to generate evidence on the causal effects of individual PIMs on health outcomes.