Risk Factors for Hospital Admissions Among Emergency Department Patients: From Triage to Admission

Introduction: Healthcare systems typically provide multiple channels to access acute inpatient care, with the emergency department (ED) as the main route of access. The ED faces multifaceted demand and supply challenges, which implicate resource allocation and patient flow. In this study we aimed to identify factors associated with hospital admissions among ED patients in a Singapore tertiary-care hospital. Methods: Using a retrospective cohort study of all eligible visits to a Singapore ED between January 1–December 31, 2019, we conducted a multivariable, mixed-effect logistic regression model to study the factors associated with hospital admissions. The model accounted for patients’ demographics; triage category; arrival mode; referral source; time of ED visit; discharge diagnosis; and ED occupancy levels. Results: In 2019, there were 141,719 visits to the ED, with 42,238 (30%) of these visits resulting in hospital admissions. Factors associated with increased odds of hospital admissions included increasing age, being male, ethnicity (Malay vs Chinese), higher patient acuity, non-self-referred patients (vs self-referred), patient being conveyed by ambulances (vs walk-in), and category of disease. Our model demonstrated that the highest odds of inpatient admissions were attributed to the patient’s acuity (highest vs lowest acuity: odds ratio [OR] 326, 95% confidence interval [CI] 292-363), followed by patients’ age (70 and above vs 30 and below: OR 13.8, 95% CI 12.8-14.8). The ORs for all other factors with significantly increased odds of admissions were modest, ranging from 1.12-4.18. Although the ED occupancy levels at the hour of the patient’s disposition decision, the hour of the ED visit, and the month of the ED visit were significantly associated with hospital admissions, changes in the probabilities of hospital admissions across the possible range of values of these factors were marginal. Conclusion: Our study revealed several factors significantly associated with hospital admissions, with patient acuity and age as the most important factors. Moreover, emergency physicians’ decisions to admit patients were clinically consistent and only marginally influenced by the degree of ED crowding. These findings offer invaluable insights into follow-up studies that will be crucial in shaping new policies or designing new interventions to enhance current preventive health or healthcare delivery systems to curtail the growth in inpatient-bed demand among ED patients over time.