Potentially inappropriate medications in nursing homes and the community older adults using the French health insurance databases.

Abstract

Potentially inappropriate medications (PIMs) have become a major issue in improving prescribing practices and reducing the risk of adverse drug events in older people. However, very few studies have compared exposition to PIMs controlling for differences in demographic and health between nursing home residents (NHRs) and community-dwelling older adults (CDOAs). This study aimed to assess the prescribing pattern of PIMs between NHRs and CDOAs. We conducted a cross-sectional study over three months in 2019 using the French Health Insurance databases. The study population included 274 971 NHRs and 4 893 721 CDOAs aged 75 years or over. The prevalence ratio (PR) between NHRs and CDOAs of 17 PIM indicators, based on the Beers and STOPP criteria lists, was assessed using multivariable robust Poisson regression adjusted for age, sex, diseases, and polypharmacy. During the study period, 54% of NHRs and 29% of CDOAs received at least one PIM. After adjustment, the prevalence of PIMs was 33% higher among NHRs compared to CDOAs (aPR=1.33; 95% CI [1.33-1.34]). NHRs received PIMs related to benzodiazepines (aPR=1.43; 95% CI [1.42-1.43]), anticholinergic drugs (aPR=1.29; 95% CI [1.27-1.31]), and at least three central nervous system-active drugs (aPR=1.94; 95% CI [1.92-1.96]) more frequently. Prevalence of PIMs related to non-steroidal anti-inflammatory drugs (aPR=0.50; 95% CI [0.48-0.52]) and long-acting benzodiazepines (aPR=0.84; 95% CI [0.82-0.85]) was lower among NHRs. The NHRs were at greater risk for PIM than CDOAs, although differences exist according to the category of PIMs. As the population is aging, it is essential to promote and evaluate interventions in NHs and the community to enhance medication optimization.