The Influence Of Glucose Variability On Prognosis In Critically Ill Patients

Objective: To evaluate the influence of glycemic variability on prognosis in critically ill patients by calculating the glucose variability during the ICU stay and the first day of ICU stay.Methods: Select 80 critical ill patients stayed in Comprehensive ICU during September 2010 and January 2011, exclude patients younger than 16 years old, diabetes and only for recovery and monitoring treatment in the ICU. Record general information of patients , and the initial blood glucose (GluAdm) Transferred to ICU, glycated hemoglobin HbA1c,,the average blood glucose (Glu1Ave),blood glucose maximum (Glu1Max),blood glucose standard deviation (Glu1SD),glucose coefficient of variation (Glu1CV) within the first 24h; the average blood glucose (GluAve), blood glucose maximum (GluMax), blood glucose standard deviation (GluSD), glucose coefficient of variation(GluCV),the Occurrence of hypoglycemia[blood glucose<4mmol/L is defined as hypoglycemia; hypoglycemia ratio= the number of hypoglycemia of death group (survival group)/ the number of death group (survival group)]during ICU stay time( if ICU stay time more than 14 days, only to record blood glucose values during stay in ICU within 14 days ). Record APACHE II score during stay in ICU within 24 hours, mechanical ventilation stayed in ICU, ICU stay time and the outcome(to be transferred out of ICU were defined as survival group and died during the ICU were defined as death group ). Divided the glycemic index of the significant difference into G1,G2,G3,G4 four subgroups between the death group and survival, then analyze the variation of APACHE II score, mechanical ventilation time, ICU stay time and death percentage (number of deaths within a sub-group / total number of the same sub-group) with the variation of glucose variability, relevant indicators with statistically significant analyzed by logistic regression analysis to predict death.Results: Initial blood glucose,average blood glucose and blood glucose maximum was no significant difference between the death group and survival group, P>0.05. But the proportion of hypoglycemia in death group was significantly higher than the survival group(36.84% vs 6.56%, P <0.01),the two groups were significantly difference. Between the standard deviation of blood glucose (SD) and coefficient of variation (CV) during the first day of ICU stay and the SD,CV during ICU stay were significantly different (Glu1SD: 2.28±1.20mmol / L compared with 1.66±1.08mmol / L); Glu1CV: 0.28±0.13 compared with 0.18±0.08; GluSD: 2.63±1.08mmol / L compared with 1.89±0.96mmol / L; GluCV: 0.32±0.12 vs 0.23±0.09), P all <0.05. The SD, CV with significantly difference between the death group and survival group is divided into G1,G2,G3,G4 four subgroups according to quartile, the results showed that with the increase of glucose variability, APACHE II score and the percentage of death increased. Included the significantly difference indicators between the death group and survival group into Logistic regression model,analysis showed that APACHE II score and GluCV independently correlated with the risk of death.Conclusion: The glucose variability index GluSD, GluCV was significantly correlated with the prognosis of ICU patients, means with the increase of the degree of variation, the risk of the death increased. Therefore, it may be an important aspect that control of the blood glucose variability may improve the prognosis of critically ill patients.