Describing and Predicting Trajectories of Healthcare Utilization Among Older Adults Presenting to an Emergency Department Using the interRAI Emergency Department Screener

Abstract

Introduction: Although older adults visit emergency departments (EDs) more than any other age group, the trajectories of healthcare utilization older adults experience post-ED are not well described. Further, whether rapid ED assessment tools can predict trajectories and discharge destinations remains unclear. Methods: Older adults (≥65 years) who presented to an ED at a large Canadian urban academic hospital were recruited (January 2018–April 2019). The interRAI ED Screener (EDS) was completed on presentation. Patients were categorized by EDS risk score (1/2=low, 3/4=moderate, 5/6=high) and had their discharge destinations tracked. Patients admitted to hospital were tracked until their final discharge destination. Crude and age/sex-adjusted odds ratios and c-statistics were obtained to examine associations between EDS scores and discharge destinations. Results: Of 751 patients (mean/SD age 77.68/8.43; 41.3% male), 200/26.6% had a high-risk EDS score. 58.3% were discharged home, 39.7% were admitted to hospital, and 2.0% were discharged to rehabilitation/long-term care (LTC) settings directly from the ED. The high-risk group had lower odds of home discharge (aOR=0.47, 95%CI 0.31–0.71, p<0.001) and therefore greater odds of hospital admission (aOR=1.84, 1.23–2.76, p<0.001). Of those admitted, 75.4% were discharged home, 16.4% were discharged to a rehabilitation/LTC setting, 5.6% transferred institutions (psychiatric, oncology, etc.), and 2.7% died in-hospital. The high-risk group were more likely to stay ≥1 week on inpatient units (aOR=2.11, 1.04–4.31, p=0.038) and have a geriatrician consulted (aOR=3.72, 1.17–11.86, p=0.026). The EDS had poor prediction of post-ED hospitalization (C-statistic= 0.58, 95%CI 0.54–0.62), but reasonable prediction of post-ED LTC home/rehabilitation centre admission (0.75, 0.63–0.87), albeit the number of these outcomes were small (n=15). Conclusion: We describe a range of healthcare trajectories older adults experience following ED presentation. Stratification by EDS risk groups could help to proactively identify the need for geriatric consultation earlier and resource utilization trajectories after an index ED visit, which could better enable the planning and organization of acute healthcare services for older adults.