Nursing home residents with dementia carry more, and longer, drug interactions

Design

Retrospective observational study, linked Medicare claims and clinical assessments

Population

485,251 long-stay US nursing home residents ≥66 (70% with dementia)

Primary outcome

Exposure to 98 potential drug–drug interactions

Effect

58.0% vs 53.3% (PR 1.09, 95% CI 1.08–1.09); ≥3 CNS-active drugs 29.6% vs 21.2%

This retrospective study linked Medicare claims with clinical assessments for 485,251 long-stay US nursing home residents aged 66 or over between 2018 and 2020, of whom 70% had dementia, and screened for 98 potential drug–drug interactions during their stay. It was published in August.

More than half of all residents (56.6%) had at least one potential interaction. Dementia was associated with a slightly higher prevalence (58.0% vs 53.3%; prevalence ratio 1.09, 95% CI 1.08 to 1.09), but the bigger differences were in specific combinations and duration: 34 interactions lasted longer in residents with dementia, and 29.6% of them were exposed to three or more central nervous system-active drugs, against 21.2% without dementia.

Residents with dementia are less able to report drowsiness, dizziness or confusion, which is exactly when interactions need active review. The study counted potential interactions, not harms, but the CNS-active burden it describes is a known falls risk. A scheduled pharmacist-led review focused on CNS-active drugs is a practical response in care homes, including in India's growing residential care sector.

• Schedule regular medication reviews for care home residents with dementia, focused on CNS-active drugs.
• Flag three or more CNS-active drugs as a priority for deprescribing.
• Check for interactions whenever a new drug is added, rather than only at annual review.
• Ask carers about drowsiness, unsteadiness or new confusion, since residents may not report them.
• Involve a pharmacist in reviews where possible.

Why it matters

The residents least able to report side effects are the ones most exposed to brain-active drug combinations.

Don't overread it

This counted potential interactions from claims; it does not show these interactions caused falls or other harm.

The statistics, in plain English

A prevalence ratio of 1.09 means potential interactions were only about 9% more common with dementia — the larger signal is the 8-point gap in residents on three or more CNS-active drugs. These are potential interactions, not confirmed harms.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.