Association Between Emergency Department-to-ICU Transfer Time and Hospital Mortality Across ICU Admission Diagnoses: A Post Hoc Subgroup Analysis.

Abstract

To assess the association between emergency department (ED)-to-ICU transfer time and hospital mortality across common ICU diagnoses. Secondary analysis of a Dutch retrospective cohort (2009-2020). Twelve hospitals (four academic and eight nonacademic teaching [NACT]) provided ED arrival and ICU admission times. Adult patients directly admitted from the ED to the ICU were included. Seven diagnostic groups (> 1500 patients each) were analyzed: out-of-hospital cardiac arrest (OHCA), nonoperative trauma, overdose, sepsis, pneumonia, respiratory failure (excluding pneumonia), and intracranial hemorrhage (ICH). Logistic regression assessed associations between ED-to-ICU time quintiles and hospital mortality, adjusting for hospital of admission, and Acute Physiology and Chronic Health Evaluation III score. Analyses were stratified by hospital type. None. Among 18,798 patients, median ED-to-ICU time was 1.9 hours (interquartile range, 1.2-3.1 hr). In OHCA (n = 3,818), associations differed by hospital type. In academic hospitals, prolonged ED-to-ICU time was associated with higher mortality (odds ratio [OR], 1.48; 95% CI, 1.08-2.02 at 1.1-1.6 hr; OR, 2.94; 95% CI, 1.80-4.78 at > 3.4 hr; Wald χ2p < 0.001). In NACT hospitals, prolonged ED-to-ICU time was negatively associated with mortality, with ORs less than 1.0 across quintiles (Wald χ2p < 0.001). In nonoperative trauma, a positive association between ED-to-ICU time and hospital mortality was observed in the overall cohort (OR, 1.90; 95% CI, 1.12-3.21 at > 3.4 hr; Wald χ2p = 0.05) but was not reproduced after stratification for hospital. No associations were observed in overdose, sepsis, pneumonia, respiratory failure, or ICH. The association between ED-to-ICU time and hospital mortality varied across diagnostic groups and hospital types. In OHCA, opposing associations were observed in academic and NACT hospitals. In nonoperative trauma, a positive association was observed only in the overall cohort. Prospective studies in homogeneous, risk-defined subgroups with detailed process-of-care data are needed to identify modifiable delays and define transfer-time thresholds.