Outcomes of 72-hour emergency department return visits requiring hospital admission in older adults: a nationally representative analysis

BACKGROUND: Older adult Emergency department disposition decisions are uniquely challenging. Providers must balance the risk of admission (delirium, deconditioning, iatrogenic complications, cost) with the risk of discharging evolving illness leading to potentially worse outcomes. National-level data on return visit admissions among older adults (RVAs)-hospitalizations occurring within 72 hours of an emergency department (ED) treat-and-release visit- and their outcomes in this population remain limited. METHODS: We conducted a retrospective cohort study using the National Hospital Ambulatory Medical Care Survey (NHAMCS), analyzing data from 1/1/2007-12/31/2022. We included ED visits for adults aged ≥65 resulting in hospital admission and excluded direct admissions to procedural or intensive care units. RVAs were defined as hospital admissions preceded by a treat-and-release ED visit at the same facility within 72 hours. Comparator group comprised index ED admissions to general inpatient floor or ward units, excluding procedural and critical care destinations. The primary outcome was in-hospital mortality. Secondary outcomes included hospital length of stay and nursing home placement. We performed multivariable logistic regression, adjusting for age, sex, race, Emergency Severity Index (ESI), and comorbid conditions. Adjusted odds ratio (OR) and their 95% confidence intervals (95% CI) were reported. RESULTS: Among the sampled visits representing 3,375,062 national admissions, 206,211 (6.1%) were RVAs. In-hospital mortality was 4.0% for RVAs vs 2.3% for index visit admissions (p=0.082). After adjustment, RVA was an independent predictor of in-hospital death (aOR 2.72; p=0.028). No significant differences were seen in hospital length of stay or new nursing home placement. The five most common diagnoses in RVAs were infectious: sepsis, pneumonia, COVID-19, cellulitis, and urinary tract infection. CONCLUSIONS: In a nationally representative sample of older adult patients, 72-hour RVA were associated with higher in-hospital mortality compared to index visit admissions to the floor. Although this difference was not statistically significant in unadjusted analyses, multivariable regression revealed a significant independent association. These findings should be interpreted as a potential risk signal rather than evidence of causation, suggesting that short delays in inpatient-level care may have meaningful consequences for older adults, particularly with infectious conditions. Future work to delineate subgroups and causal pathways will be essential to improving ED risk stratification, disposition decision-making, and post-ED care transitions.