Background: The Hospital Readmission Reductions Program (HRRP) penalizes hospitals with excessive 30-day readmission rates. The HRRP has been credited with decreasing readmission rates, but recent literature suggests this may discourage necessary readmissions and increase mortality. Increasing numbers of patients are presenting to the emergency department each year. The objective of this study is to evaluate the incidence of ED visits and subsequent mortality within 30 days (30D) of hospital discharge among geriatric patients.
Methods: We conducted a retrospective cohort study of ED visits by patients age ≥65 who visited an ED at Mount Sinai Medical Center, Northwestern Memorial Hospital, or St. Joseph’s Regional Medical Center between 2013-2015. Medicare claims data including date of death were obtained for hospitals throughout the United States from 2006-2015. Inclusion criteria were ED visits for patients within 30D of hospital discharge. ED visits were divided into 4 periods: 2 before HRRP, one after HRRP announcement, and one during HRRP implementation. Difference in mortality between discharged and admitted patients was assessed using chi-square goodness of fit test. Trends in 30D mortality for patients discharged from the ED were evaluated using Wald test for logistic regression fitting Charlson comorbidity conditions as covariates.
Results: There were 3,077,071 ED visits between 2006-2015: 17.15% (527,854) occurred within 30D of hospital discharge. ED discharge rate was 43.97% (232,082). Overall, 2.34% (5,426) of ED discharges had 30D mortality from time of inpatient discharge versus 6.74% (19,924) for ED admissions (RR 95% CI 0.70-0.70). 27.44% (63,682) ED discharges had a subsequent 30-day readmission. Patients with subsequent readmission had a 1.94% (1,237) 30D mortality rate compared to 2.39% (4,189) 30D mortality for patients not readmitted (RR 95% CI 0.93-0.95, P<0.01). Mortality rate for ED discharges increased across all four periods from 2006-2015 (95% CI 1.12-1.19), but this effect did not persist when controlling for comorbidities (95% CI 0.99-1.05).
Conclusion: Overall, mortality is higher for patients admitted through the ED versus those discharged from the ED. For patients with ED visits within 30D after a hospital discharge, however, those discharged from the ED have higher mortality rates than those who are readmitted.