Potentially inappropriate medications in older adults visiting a geriatric emergency department

Abstract

Potentially inappropriate medications (PIMs) are associated with adverse health outcomes, higher rates of healthcare utilization, and increased costs in older adults.1 However, previous studies have reported that using explicit criteria to classify PIMs in the emergency department (ED) had limited accuracy in predicting ED visits, adverse drug events, or hospitalization outcomes (such as length of stay).2, 3 Given these studies' limitations, such as small sample sizes and use of currently outdated criteria,2, 3 as well as the relevance of the topic, we aimed to describe the prevalence of PIMs according to updated criteria in a geriatric ED and explore their association with hospital admission and length of stay in a large sample of acutely ill older patients. We completed a retrospective cohort study of clinically stable patients aged 70 years and older who visited the geriatric ED (ProAGE) of a tertiary general hospital in Sao Paulo, Brazil, between August 2017 and August 2018 (n = 2539 visits). ProAGE geriatricians systematically documented patients' and visits' characteristics (e.g., demographics, chief complaints, medical history, medications, disabilities, delirium) using REDCap resources.4 More details on our geriatric ED can be found elsewhere.5 A clinical pharmacist classified the reported medications according to the 2019 Updated American Geriatrics Society (AGS) Beers criteria and PIMs were defined as present when listed in the guidelines as medications to be avoided.6 The primary outcomes were hospital admission and length of stay. Numerical variables were reported as means and standard deviations, or medians and interquartile ranges (IQR) as appropriate. Occasionally, variables were also stratified into categories to simplify their clinical interpretation. Categorical variables were reported in absolute counts and proportions. We used multivariable generalized linear models (adjusted for age, sex, Charlson Comorbidity Index, and frailty) to explore the association between PIMs and hospital admissions and length of stay. The study protocol was reviewed and approved by our Institutional Ethics Review Board. Access to de-identified retrospective data from the geriatric ED clinical database was approved with an exemption of informed consent. Overall, the mean age was 80 years, and 1276 (57%) were female. Frailty was present in 765 (30%) patients and delirium in 266 (10%). A total of 1125 (44%) patients used six or more medications regularly, and only 14 (1%) reported not using any daily medications. Amazingly, one participant reported using 31 medications on a daily basis. Two in three patients used at least one PIM; 1260 (50%) used one or two PIMs, while 408 (16%) used three or more. The most prevalent PIMs in our population were proton pump inhibitors (37%), benzodiazepines (18%), aspirin (15%), antipsychotics (10%), antidepressants (9%), antiarrhythmics (8%), anticholinergics (7%), nonbenzodiazepine receptor agonists (6%), anticonvulsants (6%), and antibiotics (2%) (Figure 1). Patients using more PIMs were older and had a higher prevalence of heart failure, coronary heart disease, metastatic cancer, depression, dementia, frailty, hospitalization in the previous 6 months, recent falls, and delirium (Table 1). Admissions were twice as common in patients using ≥3 PIMs as in those using none (53% vs 24%; p < 0.001). In adjusted analyses, using 1–2 PIMs and ≥3 PIMs were independently associated with a higher risk of hospital admission, with respective relative ratios of 1.25 (95% confidence interval [95% CI] = 1.08–1.45) and 1.52 (95% CI = 1.30–1.79). Moreover, among patients who were hospitalized, those using ≥3 PIMs had longer lengths of stay (median 6 [interquartile range = 3, 12] vs median 4 [interquartile change = 2, 7] days; p < 0.001). However, the association between PIMs and length of hospital stay was nonsignificant in our multivariable analyses. We verified a high prevalence of PIMs, as defined by the 2019 Updated AGS Beers criteria, in a large sample of older adults visiting a geriatric ED. Patients who used more PIMs were older and had a higher prevalence of multimorbidity, frailty, and health services utilization. Previous studies have also observed that PIMs are common in older adults in the ED.2, 7 However, Harrison et al. reported in a retrospective cohort of 400 older adults that PIMs were not associated with their patients' medical complaints or ED repeated visits, leading the authors to question the use of the Beers criteria in the ED.2 Conversely, we observed that patients using PIMs were more likely to be hospitalized and stayed longer in the hospital. Possible explanations for these disparities include the prior study's smaller sample size, which may have limited its power, and the use of different versions of the AGS Beers Criteria, the most recent of which was employed in the current investigation. Therefore, our findings underline the importance of routine and structured evaluation of acute patients' prescriptions, as recommended by American and European Geriatric Emergency Department Guidelines.8, 9 More than an additional burden to their practice, emergency providers should perceive medication reviews as an opportunity to improve prescriptions and patient care. Further work is needed to develop and implement safe and effective pathways to modify prescriptions and interrupt PIMs in older adults in acute care settings. The authors thank those who supported this study, particularly the ProAGE team, who cared for our patients and helped implement our Geriatric ED program. There are no financial or personal conflicts to declare. All investigators who contributed significantly to this work have been included in the authors' list. Concept and design: Isadora D. Martini, Pedro Ramberger Castelo, Luiz A. Gil-Junior, Thiago J. Avelino-Silva. Acquisition of data: Fabiane G. Correia, Pedro Ramberger Castelo, Luiz A. Gil-Junior, Pedro K. Curiati, Thiago J. Avelino-Silva. Analysis and interpretation of data: Isadora D. Martini, Marlon J. R. Aliberti, Pedro K. Curiati, Thiago J. Avelino-Silva. Preparation of the manuscript: Isadora D. Martini, Pedro K. Curiati. Critical revision of the manuscript for important intellectual content: Fabiane G. Correia, Christian Valle Morinaga, Luiz A. Gil-Junior, Marlon J. R. Aliberti, Thiago J. Avelino-Silva. Study supervision: Christian Valle Morinaga, Pedro K. Curiati, Thiago J. Avelino-Silva. No sponsorship was received for this study.

Topics & Concepts