Emergency risk stratification using the TyG index: a multi-center cohort study on nonlinear association with 28-day mortality among critically ill patients transferred from the ED to the ICU

Background In the emergency department (ED), rapid risk stratification of critically ill patients is essential for timely intervention. The triglyceride-glucose (TyG) index, a simple marker of insulin resistance, may aid in early mortality prediction, but its utility in ED-to-ICU patients remains unexplored. Methods Using data from the eICU Collaborative Research Database, we conducted a retrospective multicenter cohort study of 11,593 ED-to-ICU critically ill patients. The TyG index was calculated at ED presentation. The primary outcome was 28-day all-cause mortality. Multivariable Cox regression, restricted cubic splines, and sensitivity analyses were performed to assess associations. Results Among patients (mean age 63.6 ± 15.7 years, 57.3% male), 28-day mortality was 6.96%. The relationship between the TyG index and mortality was nonlinear, featuring a critical threshold at a TyG index value of 9.84. Below this cutoff, each unit increase in TyG index significantly elevated mortality risk (HR 1.47, 95% CI 1.20–1.69, p < 0.0001), while above it, the risk plateaued (HR 1.04, 95% CI 1.03–1.05, p = 0.097). The association remained robust after adjustment for confounders (adjusted HR 1.19, 95% CI 1.04–1.35, p = 0.0089) and across sensitivity analyses. Conclusion The TyG index, readily obtainable at ED presentation, provides emergency clinicians with a practical tool for early mortality risk stratification in critically ill patients. Its nonlinear association with 28-day mortality suggests a saturation effect, enabling rapid identification of high-risk patients who may benefit from intensified monitoring and intervention.