Effect of a Home Health and Safety Intervention on Emergency Department Use in the Frail Elderly: A Prospective Observational Study

INTRODUCTION: Geriatric patients are often frail and may lose independence through a variety of mechanisms including cognitive decline, reduced mobility, and falls. Our goal was to measure the effect of a multidisciplinary home health program that assessed frailty and safety and then coordinated ongoing delivery of community resources on short-term, all-cause emergency department (ED) utilization across three study arms that attempted to stratify frailty by fall risk. METHODS: Subjects became eligible for this prospective observational study via one of three pathways: 1) by visiting the ED after a fall (2,757 patients); 2) by self-identifying as at risk for falling (2,787); or 3) by calling 9-1-1 for a “lift assist” after falling and being unable to get up (121). The intervention consisted of sequential home visits by a research paramedic who used standardized assessments of frailty and risk of falling (including providing home safety guidance), and a home health nurse who aligned resources to address the conditions found. Outcomes of interest were all-cause ED utilization at 30, 60, and 90 days post-intervention compared with subjects who enrolled via the same study pathway but declined the study intervention (controls). RESULTS: Subjects in the fall-related ED visit arm were significantly less likely to have one or more subsequent ED encounters post-intervention than controls at 30 days (18.2% vs 29.2%, P<0.001); 60 days (27.5% vs 39.8%, P<0.001); and 90 days (34.6% vs 46.2%, P<0.001). In contrast, participants in the self-referral arm had no difference in ED encounters post-intervention compared to controls at 30, 60, or 90 days (P=0.30, 0.84, and 0.23, respectively). The size of the 9-1-1 call arm limited statistical power for analysis. CONCLUSION: A history of a fall requiring ED evaluation appeared to be a useful marker of frailty. Subjects recruited through this pathway experienced less all-cause ED utilization over subsequent months after a coordinated community intervention than without it. The participants who only self-identified as at risk for falling had lower rates of subsequent ED utilization than those recruited in the ED after a fall and did not significantly benefit from the intervention.