| Background:Cerebral hemorrhage is a serious cerebrovascular disease with high incidence rate and high mortality rate.Although clinical treatment has made significant progress,the 1 year survival rate of patients with cerebral hemorrhage is only 46%,and the 5year survival rate is only 29%.Patients with cerebral hemorrhage often have different levels of neurological dysfunction,which requires long-term hospitalization and rehabilitation,causing a heavy economic burden to families and society.Therefore,how to improve the prognosis of patients with cerebral hemorrhage has become the focus of global research.Hyperglycemia is very common in patients with cerebral hemorrhage,and elevated blood glucose often indicates a poor prognosis.With the in-depth study of blood glucose,blood glucose variability has been widely concerned.Recent studies have found that blood glucose variability is independently related to mortality in critically ill patients.However,there are few reports about the effect of blood glucose variability on patients with severe cerebral hemorrhage at home and abroad.Objective:The purpose of this study was to investigate the correlation between blood glucose variability and prognosis of patients with severe acute cerebral hemorrhage,so as to provide references for clinical blood glucose management.Methods:This study is a retrospective cohort study.We selected 137 patients with acute cerebral hemorrhage who were hospitalized in the Intensive Care Units of the First Affiliated Hospital of Zhengzhou University from December 2018 to December 2019,and collected their blood glucose values and relevant clinical data from medical records,and also calculated their Mean Blood Glucose(MBG),Standard Deviation of Blood Glucose(SD),as well as Coefficient of Variation of Blood Glucose(CV).1. Per the 28-day prognosis,the patients were divided into survival group and death group. Comparison of the relevant clinical data of each group was made including but not limited to the Mean Blood Glucose,Standard Deviation of Blood Glucose and Coefficient of Variation of Blood Glucose.2.MBG was divided into high MBG group(MBG≥median)and low MBG group(MBG<median).SDBG was divided into high SD group(SD≥median)and low SD group(SD<median).On the basis of the above two indicators,patients were divided into four groups:low MBG+low SD group,low MBG+high SD group,high MBG+low SD group and high MBG+high SD group.The differences of mortality among the above four groups were compared.The same grouping method was adopted when MBG and CVBwere combined.3.The patients were divided into three groups based on the tertiles of Blood Glucose SD and CV,respectively,and the mortality of patients with different degrees of blood glucose variability was compared.4.The meaningful variables in single factor analysis(other than the indicators of blood glucose variability)are carried out by multivariate logistic regression analysis to construct the original model.SD and CV were added to the original model and then the SD and CV models are constructed.The subjects working characteristic curves of the original model,SD model and CV model are drawn.At the same time,the area under the curve is compared to judge the predictive power of the different models on the patients’28-day mortality.Results:1.The Standard Deviation of Blood Glucose[2.19(1.60-3.53)mmol/L vs1.17(0.81-1.74)mmol/L,P<0.01],Coefficient of Variation of Blood Glucose[23.89(18.25-32.57)%vs15.02(11.31-19.62)%,P<0.01]and Mean Blood Glucose[9.29(8.30-11.22)mmol/L vs7.70(6.74-9.14)mmol/L,P<0.01]in the death group were significantly higher than those in the survival group.There existed2. Per the medians of MBG(8.17mmol/L)and SD(1.40mmol/L),the patients were divided into four groups.There was significant statistical difference in mortality among the four groups(χ~2=33.879,P<0.001).There was significant statistical differences in 28-Day mortality between the low MBG+low SD group and the low MBG+high SD group,the high MBG+low SD group and the high MBG+high SD group(P<0.05),while there was no significant statistical difference in 28d mortality between the low MBG+low SD group and the high MBG+low SD group,as well as the low MBG+high SD group and the high MBG+high SD group(P>0.05).Similar results were obtained when patients were grouped according to the median of MBG(8.17mmol/L)and the median of CV(17.25%).3. Per the two cut-off values(tertile method)of SD 1.12 mmol/L and SD 1.92mmol/L,and the two cut-off values(tertile method)of CV 14.64%and CV 20.93%,the patients were divided into three groups,i.e.T1,T2,T3.The results showed that there was statistical significance in mortalities among the groups[SD(6.5%,28.3%,57.8%,P<0.01);CV(8.7%,32.6%,51.1%,P<0.01)].4. A multi-factor logistic regression analysis showed that the Blood Glucose Variability was still independently related to the mortality for the patients suffering from cerebral hemorrhage[SD(OR=2.728,95%CI:1.693～4.397,P<0.001);CV(OR=1.147,95%CI:1.077～1.221,P<0.001),while being adjusted for age,bleeding volume and bleeding location.The AUC in the original model was 0.817(95CI%:0.742～0.878).The AUC of the SD model was 0.899(95CI%:0.836～0.944).The AUC of the CV model was 0.903(95CI%:0.840～0.947).Both the SD model and the CV model showed higher predictive powers than the original model(SD model vs original model,Z=2.813,P=0.005;CV model vs original model,Z=2.903,P=0.004).The difference between the SD model and the CV model in predicting the mortality of patients was not statistically significant(Z=0.300,P=0.764).Conclusions:1.The index of blood glucose variation can be used as an independent predictor of poor prognosis in patients with severe acute cerebral hemorrhage,and there is no significant difference between SD and CV in predicting the prognosis of severe acute cerebral hemorrhage.2.The higher the variability of blood glucose,the higher the risk of death in patients with severe acute cerebral hemorrhage.Reducing blood glucose variability may be an important aspect of improving the prognosis of patients with severe acute cerebral hemorrhage. |
PDF Full Text Request