Potentially Inappropriate Medication Use in Older Adults with Alzheimer’s Disease and Related Dementias Living in the Community: A Cross-Sectional Analysis

BACKGROUND: With the older adult population in the USA increasing, so is the population of those with Alzheimer’s disease and related dementias (ADRD). Older adults are vulnerable to the effects of potentially inappropriate medications as established by the Beers Criteria; however, some medications continue to be prescribed against recommendations. OBJECTIVES: Our objectives were to describe potentially inappropriate medication (PIM) use linked to cognitive impairment or decline (referred to as Cog-PIM) in older adults with and without ADRD and to investigate whether the odds of Cog-PIM report differ by ADRD status in ambulatory care (i.e., outpatient care) in the USA. METHODS: A cross-sectional analysis was performed using a nationally representative sample of non-perioperative, office-based ambulatory care visits by adults aged ≥ 65 years in 2016 (n = 218,182,131). Data were collected from the National Ambulatory Medical Care Survey. Cog-PIMs were identified as defined in the 2015 Beers Criteria recommendations for medications that may be potentially inappropriate in older adults with cognitive impairment or dementia. ADRD status was determined by clinician report using free text, the ADRD flag, or the presence of a diagnosis code indicating dementia. Multivariable logistic regressions were used to estimate the odds of Cog-PIM use overall and by medication class. RESULTS: In 2016, 2.1% (n = 4,651,563) of outpatient visits were made by older adults with ADRD, 33.2% of which reported at least one Cog-PIM. Anticholinergic Cog-PIMs were noted in 20.5% of ADRD visits compared with 8.1% of non-ADRD visits. Antipsychotic PIMs were noted in 15.5% of ADRD visits compared with 0.8% of non-ADRD visits. Benzodiazepine and non-benzodiazepine receptor agonist hypnotic (Z drug) Cog-PIMs were reported in 10.9% of ADRD visits and 10.7% of non-ADRD visits. ADRD status was a significant predictor of Cog-PIM report overall (adjusted odds ratio [aOR] 2.74 [95% confidence interval {CI} 1.20–6.27]) and for anticholinergics and antipsychotics specifically (aOR 3.35 [95% CI 1.24–9.03] and aOR 22.80 [95% CI 5.80–89.50], respectively). CONCLUSION: This study demonstrated a high prevalence of Cog-PIM use and increased odds of Cog-PIM use in older adults with ADRD. Future work should investigate opportunities in the ambulatory care setting for safer prescribing and de-escalation of Cog-PIMs.