368 Determining the Combined Impact of Geriatric-Specific Predictors of Hospitalization Among Geriatric Emergency Department Patients

Abstract

Study Objective: To address the persistently increasing volume and complexity of older emergency department (ED) patients > 65 years in the U.S., several geriatric-specific screening tools have been reported to predict hospitalization among these older patients. However, few EDs have the resources to implement multiple tools, but data on which tool performs best in comparison with one another is lacking. This study’s objective was to determine the comparative performance of common published geriatric screening tools in a multi-ethnic ED patient population.Methods: This is a prospective observational study using data collected for monitoring as part of Geriatric Emergency Department Accreditation from September 2019 to April 2022. The primary outcome was hospitalization during the index visit. Independent variables included age, means of arrival, chief complaint classification, prior ED visits within the past 30 days, prior inpatient hospitalizations within the past year, and the following ED screens: clinical frailty scale (CFS), Morse fall scale, a nurse medication screen which was positive if either the patient was on a 2015 Beers list medication or identified having difficulty with his or her medications at home, Katz Activities of Daily Living (ADLs), the Short Portable Mental Status Questionnaire (SPMSQ), the brief Confusion Assessment Method (bCAM), and the Identification of Seniors at Risk (ISAR) score. For patients with multiple visits, only the index ED visit was included. Multivariable models adjusting for demographic and clinical characteristics associated with hospitalization were constructed to compare the strength of association of each screening tool relative to others.Results: In multivariable analysis of 1,183 unique ED patients, prior ED visit in the past 30 days (OR=0.6, p<0.05), and arrival by private vehicle or walk in were associated with a decreased odds of hospitalization on the index ED visit. Prior hospitalization in past 1 year (OR=1.5, p<0.001) and age (in years) were significantly associated with hospitalization OR for each year above 65 = 1.008, p<0.01. Among the screening tools, a CFS score >4 (OR 1.3; p<0.05), a Morse fall scale score >25 (OR 1.9; p<0.051), medication risk screen (OR 1.9; p<0.01) and a bCAM (OR 3.3; p<0.05) were independently associated with hospitalization. ISAR and ADLs were significantly associated with hospitalization in multivariable models without other geriatric ED screening included, but had no significant association with hospitalization when included in the final model: ISAR = OR 1.3; p=0.2 and the Activities of Daily Living = OR 1.3, p=0.2. SPMSQ was not significantly associated with hospitalization in any model (OR 1.18, p=0.1).Conclusion: When multiple published geriatric screening tools for hospitalization among older ED patients were examined together, two readily available and easily implementable tools (CFS and bCAM) were strongly predictive, with the bCAM being the strongest predictor of hospitalization. These results suggest that of the geriatric ED screens used in this study, the CFS and bCAM have the most utility in predicting hospitalization during an index ED visit for older adults.No, authors do not have interests to disclose Study Objective: To address the persistently increasing volume and complexity of older emergency department (ED) patients > 65 years in the U.S., several geriatric-specific screening tools have been reported to predict hospitalization among these older patients. However, few EDs have the resources to implement multiple tools, but data on which tool performs best in comparison with one another is lacking. This study’s objective was to determine the comparative performance of common published geriatric screening tools in a multi-ethnic ED patient population. Methods: This is a prospective observational study using data collected for monitoring as part of Geriatric Emergency Department Accreditation from September 2019 to April 2022. The primary outcome was hospitalization during the index visit. Independent variables included age, means of arrival, chief complaint classification, prior ED visits within the past 30 days, prior inpatient hospitalizations within the past year, and the following ED screens: clinical frailty scale (CFS), Morse fall scale, a nurse medication screen which was positive if either the patient was on a 2015 Beers list medication or identified having difficulty with his or her medications at home, Katz Activities of Daily Living (ADLs), the Short Portable Mental Status Questionnaire (SPMSQ), the brief Confusion Assessment Method (bCAM), and the Identification of Seniors at Risk (ISAR) score. For patients with multiple visits, only the index ED visit was included. Multivariable models adjusting for demographic and clinical characteristics associated with hospitalization were constructed to compare the strength of association of each screening tool relative to others. Results: In multivariable analysis of 1,183 unique ED patients, prior ED visit in the past 30 days (OR=0.6, p<0.05), and arrival by private vehicle or walk in were associated with a decreased odds of hospitalization on the index ED visit. Prior hospitalization in past 1 year (OR=1.5, p<0.001) and age (in years) were significantly associated with hospitalization OR for each year above 65 = 1.008, p<0.01. Among the screening tools, a CFS score >4 (OR 1.3; p<0.05), a Morse fall scale score >25 (OR 1.9; p<0.051), medication risk screen (OR 1.9; p<0.01) and a bCAM (OR 3.3; p<0.05) were independently associated with hospitalization. ISAR and ADLs were significantly associated with hospitalization in multivariable models without other geriatric ED screening included, but had no significant association with hospitalization when included in the final model: ISAR = OR 1.3; p=0.2 and the Activities of Daily Living = OR 1.3, p=0.2. SPMSQ was not significantly associated with hospitalization in any model (OR 1.18, p=0.1). Conclusion: When multiple published geriatric screening tools for hospitalization among older ED patients were examined together, two readily available and easily implementable tools (CFS and bCAM) were strongly predictive, with the bCAM being the strongest predictor of hospitalization. These results suggest that of the geriatric ED screens used in this study, the CFS and bCAM have the most utility in predicting hospitalization during an index ED visit for older adults. No, authors do not have interests to disclose