Objective  To study the association of glucose variability and ICU delirium of patients after liver transplantation.

  Methods  This was a retrospective, single-center cohort study. Patients who admitted to ICU after liver transplantation during Aug. 2016 to Dec. 2018 were enrolled. They were divided into two groups accoding to whether they had delirium in ICU. Multivariate logistic regression analysis model was used to analyze the relationship between glucose variability and ICU delirium, and Cochran-Armitage trend test was used to analyze the linear relationship between blood glucose variability levels and the incidence of delirium.

  Results   A total of 242 patients were enrolled, among them, 36 patients had delirium. The occurrence rate of delirium was 14.9% (36/242). Results indicated that glucose variability was an independently risk factor of ICU delirium for liver transplant patients (P=0.045), and delirium was more common in patients with higher glucose variability (fourth quartile vs. first quartile, odds ratio=5.283, 95% confidence interval: 1.092～25.550, P=0.038). Results of Cochran-Armitage trend test indicated that there was a linear relationship between blood glucose variability level and ICU delirium rate, with the increase of glucose variability level, the risk of ICU delirium was increased too (P<0.001).

  Conclusion  Glucose variability was an independently risk factor of ICU delirium in liver transplantation patients.