Objective: To investigate the association between early blood glucose level fluctuation and prognosis of critically ill patients.

Methods: A retrospective study involving 95 critically ill patients in intensive care unit (ICU) was conducted. According to the 28-day outcome after admission to ICU, the patients were divided into nonsurvivors (43 cases) and survivors (52 cases), and the blood glucose level in them was monitored in the first 72 hours. Blood glucose concentration at admission (BGadm), mean blood glucose level (MBG), hyperglycemia index (HGI), glycemic lability index (GLI), incidence of hypoglycemia and total dosage of intravenous insulin for each patient were compared. The index as an independent risk factor of mortality was determined by multivariate logistic regression analysis and the predictor value by comparing the area under the receiver operating characteristic curve (ROC curve, AUC) of each index.

Results: The BGadm (mmol/L), MBG (mmol/L), HGI and the incidence of hypoglycemia showed no significant differences between nonsurvivors and survivors [BGadm: 9.87 ± 4.48 vs. 9.26 ± 3.07, MBG: 8.59 ± 1.23 vs. 8.47 ± 1.01, HGI(6.0): 2.45 ± 0.94 vs. 1.68 ± 1.05, HGI(8.3): 0.84 ± 0.70 vs. 0.68 ± 0.51, the incidence of hypoglycemia: 9.30% vs. 5.77%, all P > 0.05], but acute physiology and chronic health evaluation II (APACHE II ) score, GLI and the total dosage of intravenous insulin (U) were significantly higher in nonsurvivors than survivors [APACHE II score: 23 ± 6 vs. 19 ± 6, GLI: 56.96 (65.43) vs. 23.87 (41.62), the total dosage of intravenous insulin: 65.5 (130.5) vs. 12.5 (90.0), all P < 0.05]. Multivariate logistic regression analysis showed that APACHE II score and GLI were both independent risk factors [APACHE II score: odds ratio (OR) = 1.09, 95% confidence interval (95%CI) 1.01-1.17; GLI: OR = 1.03, 95%CI 1.01-1.06, both P < 0.05]. When ROC curve was plotted, the AUC of APACHE II score and GLI was respectively 0.69 and 0.71, and there was no significant difference (P > 0.05).

Conclusions: Early fluctuation of blood glucose is a significant independent risk factor of mortality in critically ill patients. Control the early fluctuation of blood glucose concentration might improve the patients' outcome.