Health Policy and Health Services Research
Background: Return visits to the emergency department (RVED) within days reflects the quality of emergency care. However, not many studies evaluated the clinical outcomes of patients who were hospitalized after RVED. We wonder whether patients who were hospitalized after RVED would cause delay treatment and thereby worsen the outcomes comparing with if they were hospitalized at first ED visits?
Methods: This is a retrospective cohort study using the National Health Insurance Research Database in Taiwan. We identified eligible adult indexed emergency department (ED) visits between 2000 and 2013. Patients who were hospitalized after RVED within 3 days were compared to those who were hospitalized after their initial ED visits using propensity-score matching. Outcomes included mortality, the risk of ICU admission, and length of stay. We used logistic regression for binary outcomes and log-linear model for skewed continuous outcomes. The generalized estimating equation method was used after matching. A hospitalization after RVED within 7 days was done as sensitivity analysis.
Results: Among 2,027,019 eligible indexed ED visits, 1,675,425 (83%) were discharged from the ED, and 351,594 (17%) were hospitalized. Four percent (n = 65,541) of discharged patients had a RVED within 3 days. Patients initially discharged from the ED were younger and had fewer comorbidities. After propensity-score matching, patients who were hospitalized after RVED had a significantly lower overall mortality (6.1% vs 7.4%, odds ratio [OR] 0.84, 95% confidence interval [CI] 0.75-0.93), fewer ICU admission (10.3 % vs 13.8%, adjusted OR: 0.73, 95% CI 0.67-0.79), and borderline shorter length of stay (10.9 vs 11.6, incidence rate ratio 0.95, 95% CI 0.92-0.99) compared with patients who were hospitalized after their first ED visits. Similar outcomes were observed in patients returning to ED within 7 days.
Conclusion: Patients who were hospitalized after RVED within 3 days showed lower mortality, fewer ICU admission, and similar length of stay with patients initially admitted in ED visit. We proposed that RVED within short time period may have limited roles in assessing hospital performance.