ABSTRACT
Objective:To investigate whether stress hyperglycemia (SH) is an independent risk factor for the occurrence and mortality of sepsis-associated encephalopathy (SAE).Methods:From August 2016 to October 2021, sepsis patients admitted to the ICU of Guangdong Provincial People's Hospital were selected as the study subjects. According to whether they developed to SH (RBG>11.1 mmol/L) within 7 days of enrollment, the pat ients were divided into the SH group and the non-SH group for analysis. Logistic regression was used to analyze whether SH was an independent risk factor for SAE occurrence, and ROC curve was used to analyze the predictive value of SH to SAE. Kaplan-Meier curve was used to compare the 90-day survival of SAE patients with or without SH. Cox regression analysis was used to analyze the risk factors of 28-day and 90-day death in SAE patients.Results:A total of 183 sepsis patients were included, including 62 patients in the SH group and 121 in the non-SH group. Logistic regression analysis demonstrated that SH was an independent risk factor for SAE ( OR=4.452, 95% CI: 2.021-9.808, P <0.001). ROC curve demonstrated that SH could accurately predict SAE (AUC=0.831; Sensitivity=78.4%; Specificity=76.8%; and Yoden index=0.553). Kaplan-Meier curve demonstrated that the 90-day survival of SAE patients with SH significantly declined (log-rank test: P<0.01). Cox regression analysis suggested that SH was a risk factor for death at day 28 and day 90 in SAE patients (28 d, HR=2.272, 95% CI: 1.212-4.260, P=0.010; 90 d, HR=2.456, 95% CI: 1.400-4.306, P<0.01). Conclusions:SH is an independent risk factor for SAE and can predict SAE occurrence. SH significantly reduces 90-day survival and increase mortality at 28 and 90 days in SAE patients.