Predictors of Length of Stay of Patients at The Emergency Departments of Level 5 Government Hospitals in Nairobi City County, Kenya

Background: Emergency department (ED) length of stay (LOS) is a key indicator of healthcare quality and operational efficiency. Prolonged EDLOS contributes to overcrowding, delayed treatment, increased healthcare costs, poor patient outcomes, and reduced patient satisfaction. This study investigated the predictors of EDLOS among patients attending two level 5 government hospitals in Nairobi City County, Kenya. Methods: The study was a hospital based cross sectional analytical study, carried out in one month (22nd April- 19th May 2025). The study considered 412 patients selected through systematic random sampling from the emergency departments of the two hospitals. Seven emergency department clinicians and nurse managers were purposively selected for qualitative interviews. Data was collected using structured questionnaires, patient treatment records, and key informant interview guides. Descriptive statistics summarized patient characteristics and ED performance indicators. Chi-square tests assessed associations between independent variables and EDLOS, while multivariable logistic regression identified independent predictors of prolonged EDLOS. Statistical significance was set at p<0.05. Results: The mean EDLOS was 334.6 minutes (5.6 hours), with only 48.1% of patients completing emergency care within the recommended four-hour benchmark. The mean time to triage was 20.0 minutes, and the average waiting time before clinical consultation was 72.6 minutes. Occupation was the only socio-demographic factor significantly associated with prolonged EDLOS (p=0.047). Clinical predictors included triage category (p<0.001) and patient disposition (p<0.001). Staffing levels, fewer doctors (p < 0.01) or three or fewer nurses (p < 0.01) were significantly associated with prolonged EDLOS. Qualitative findings identified diagnostic delays, inpatient bed shortages, high patient volumes, and staffing constraints as major contributors to prolonged ED stays. Conclusion: More than half of patients experienced prolonged EDLOS exceeding the recommended four-hour benchmark. Prolonged stays were primarily influenced by clinical acuity, patient disposition, and health system factors, particularly staffing levels and resource availability. Strengthening emergency department staffing, improving patient flow, enhancing diagnostic services, and optimizing bed management may reduce EDLOS and improve emergency care delivery in Kenya.