[2026-10-01] Association Between ICU Admission Delay After Vasopressor Initiation and In-Hospital Mortality in Patients With Septic Shock: A Retrospective Cohort Study

Abstract

Background: Delayed intensive care unit (ICU) transfer after the emergency department (ED) management of septic shock is associated with poor outcomes. However, most studies define delay from ED arrival, which may not accurately reflect the onset of hemodynamic instability requiring ICU-level care.

Objective: To investigate the association between the delay in ICU admission after vasopressor initiation and in-hospital mortality of patients with septic shock.

Methods: This retrospective cohort study included adults with septic shock treated at the Phrapokklao Hospital ED from January 2024 to December 2025. Index time was defined as vasopressor initiation in the ED, or ED arrival for patients already receiving vasopressors before transfer. The primary exposure was ICU admission delay, and the outcome was in-hospital mortality. Data were analyzed using multivariable models, with restricted cubic spline analysis to assess nonlinear associations.

Results: Among 317 patients with septic shock, in-hospital mortality was 29.0%. ICU admission delay was significantly associated with mortality, with risk increasing per hour of delay (adjusted odds ratio [aOR], 2.33 per hour; 95% CI, 1.73-3.15; P < .001). Patients delayed ≥ 2 hours had higher mortality than those admitted within 2 hours (45.9% vs 6.6%; aOR, 9.21; 95% CI, 3.96-21.42; P < .001). A dose-response relationship was observed across delay quartiles (P for trend < .001), with consistent findings in walk-in and referred subgroups.

Conclusions: ICU admission delay after vasopressor initiation was independently associated with increased mortality, with risk increasing progressively over time. The 2-hour cutoff represents an operational, not a physiological, threshold. Potential residual confounding warrants cautious interpretation.