Prognostic Value of Stress-Induced Hyperglycemia in High-Acuity Emergency Department Patients

TL;DR

High blood sugar upon arrival in the emergency department is linked to higher death rates in critically ill patients, regardless of diabetes status.

Contribution

Identifies stress-induced hyperglycemia as an independent predictor of mortality in high-acuity emergency department patients.

Findings

Stress-induced hyperglycemia (SIH) is independently associated with increased in-hospital mortality.

SIH remains a significant predictor of mortality in both diabetic and non-diabetic patients.

Admission glucose levels improve risk stratification when included in clinical models.

Abstract

Background/Objectives: Stress-induced hyperglycemia (SIH) is frequently observed in critically ill patients and has been associated with adverse outcomes in individuals both with and without known diabetes mellitus (DM). However, evidence regarding its prognostic utility for in-hospital mortality in high-acuity emergency department (ED) populations remains limited. Methods: We conducted a retrospective observational cohort study of consecutive adult ED patients classified as Emergency Severity Index (ESI) triage level 1. SIH was defined a priori as an admission serum glucose > 140 mg/dL, a pragmatic cutoff widely applied in clinical practice despite ongoing debate regarding optimal pathophysiological thresholds. Associations with in-hospital mortality were assessed using logistic regression in the overall cohort and stratified by DM status. Additional analyses assessed the prognostic performance of admission glucose as a continuous variable. Results: Of 470 included patients, 435 had complete mortality data; 247 (56.8%) died during hospitalization. SIH was present in 258/435 (59.3%)and known DM in 114/435 (26.2%). SIH was associated with higher in-hospital mortality in univariate analysis (OR 2.90, 95% CI 1.91–4.43; p < 0.001) and remained independently associated after adjustment (adjusted OR 2.22, 95% CI 1.41–3.51; p < 0.001). The association between SIH and mortality persisted in both non-DM and DM subgroups, with no significant interaction by DM status. SIH alone showed modest discrimination for mortality (AUC 0.625, 95% CI 0.572–0.669), whereas continuous admission glucose performed better. Discrimination improved in the multivariable model (AUC 0.728, 95% CI 0.677–0.779). Restricted cubic spline analysis demonstrated a strong overall association between admission glucose and mortality without evidence of nonlinearity, indicating an approximately linear risk increase across the observed glucose range. Conclusions: Regarding severely ill ED patients, classified as ESI triage 1, SIH is an independent predictor of in-hospital mortality irrespective of DM status. Admission glucose may improve early risk stratification when incorporated into clinical models.

Figures