Work

The triglyceride glucose index and delirium risk in sepsis patients: A causal inference study

Year: 2025

Type: article

Abstract: BACKGROUND: Sepsis, a grave systemic infection, presents substantial health challenges. While insulin resistance frequently occurs in sepsis conditions, its relationship with sepsis-associated delirium remains insufficiently explored. AIM: This study aimed to explore the causal effect between the triglyceride glucose (TyG) index and its risk of delirium in patients with sepsis through the use of causal inference. STUDY DESIGN: A cohort of 5461 sepsis patients admitted to the intensive care unit (ICU) was selected from the Medical Information Mart for Intensive Care IV database. Patients were grouped into high TyG (≥9.48) and low TyG (<9.48) categories. Propensity score matching was applied to control for confounders, and the average treatment effect on the treated was calculated. RESULTS: Of the 5461 patients, 59.6% experienced delirium. The incidence of delirium was higher in the high TyG group (1751 patients; 66.6%) than in the low TyG group (56.3%) (p < .001). The results of the logistic regression analysis indicated that the risk of delirium was significantly higher in the high TyG group (adjusted odds ratio 1.34, 95% confidence interval: 1.16-1.53). Following matching, the delirium risk increased by 6.9% in the high TyG group (T = 3.29), a finding that was confirmed by a Rosenbaum sensitivity analysis. CONCLUSIONS: The TyG index represents a straightforward and efficacious instrument for nursing staff to ascertain the likelihood of delirium in patients with sepsis during the routine monitoring of their condition. The ability to make causal inferences in observational studies provides a novel approach to research. RELEVANCE TO CLINICAL PRACTICE: The TyG index represents a readily applicable instrument for ICU nurses to identify the risk of delirium in sepsis patients. This enables the possibility of early intervention in high-risk individuals and the optimization of care outcomes.

Source: Nursing in Critical Care

DOI: 10.1111/nicc.13287