Association between timing of medical intensive care unit admission and outcome of emergency department patients: a retrospective cohort study


Intensive Care Units, Critical care, Emergency Medicine, Patient Transfer, Treatment Outcome

Published online: May 29 2024

M. Quisquater1, W. Swinnen2,4, W. Van Paesschen3, A. Koch4, E. Pannier2,4, K. Mignolet2,4, W. Temmerman4, T. Sarens4, J. Pauwels2

1 Department of Anesthesiology, Ghent University, Ghent University Hospital, Ghent, Belgium
2 Department of Anesthesiology, az Sint-Blasius, Dendermonde, Belgium
3 Emergency Department, az Sint-Blasius, Dendermonde, Belgium
4 Intensive Care Unit, az Sint-Blasius, Dendermonde, Belgium


Background: Critically ill Emergency Department (ED) patients may benefit from timely triage to the Intensive Care Unit (ICU), as there is a “window of critical opportunity.” Several authors have investigated the relationship between delayed ED-to-ICU transfer and poor outcome. However, covariates often obscured this relationship.

Objectives: To examine the impact of direct (DICU-P) versus indirect (IDICU-P) ED-to-ICU admission on patient outcomes and assess whether delay in critical care provision is a contributing factor. To compare survival for up to 12 months.

Design and Setting: Single-center retrospective cohort study.

Methods: Unplanned medical ED-to-ICU admissions between 2015 and 2019 were classified as DICU-P or IDICU-P (hospital ward stay < 48 hours). Groups were divided according to Length Of Stay (LOS) as ICU-LOS < 48h or ICU-LOS ≥ 48h. A timeline analysis was conducted. Propensity Score Matching (PSM) was used to account for bias (age, gender, SAPS II, APACHE IV admission diagnosis) and achieve pseudo-randomization.

Main outcomes: LOS and mortality, both for ICU and in-hospital, and 1 year mortality.

Results: IDICU-P patients had higher mortality rates (ICU, p = 0.006; post-ICU, p = 0.0005; hospital, p < 0.0001), longer LOS (hospital, p = 0.007), but were older (p <0.0001) and sicker (SAPS II, p = 0.0002). After PSM, a trend for higher mortality rates (hospital, p = 0.030; early in ICU (LOS-ICU < 48h), p = 0.034) and longer LOS (hospital, p = 0.030) persisted, with elderly patients being responsible for this disparity. ICU mortality was equal after 48 hours, while post-ICU and long-term mortality up to 30 days and 12 months were higher in IDICU-P (both p < 0.0001; after PSM, p = 0.018 and p = 0.009, respectively). COPD exacerbations, pneumonia, and congestive heart failure showed higher hospital mortality in IDICU-P.

Conclusion: Indirect ICU admission of ED patients in need of critical care was associated with higher mortality and longer LOS but also with higher age and severity of illness. Mortality was consistently higher for up to 12 months after ICU admission and showed no catch-up mortality. After correcting for biases with PSM, the significance often diminished; however, a general trend was confirmed. This finding highlights the importance of correct triage in the ED.