血糖波动及其控制与老年内科危重症短期死亡率的关系

The association of variability of blood glucose concentration and its control with short-term mortality in aged patients with critical medical disease

目的研究血糖波动及其控制对老年内科危重症患者短期生存/死亡率的影响,评价其预测老年内科危重症短期死亡的价值。方法回顾性研究分析152例老年内科危重症所致全身炎症反应综合征（SIRS）患者的病例资料,计算诊断SIRS当日（d0）的血糖标准差及其控制时点和APACHEⅡ评分值,按d0血糖标准差分成3组,统计各组31d生存率,用Kaplan-Meier法绘制生存曲线,用Log-rank检验比较生存率,用COX回归模型进行多因素分析。结果 d0血糖标准差0-2 mmol/L、2-4mmol/L及〉4mmol/L共3层,31d生存率分别为75.0%、43.3%及11.5%（P〈0.05）。d0血糖标准差、血糖标准差控制时点和APACHEⅡ评分值进入COX风险比例模型。其相对危险度（RR）分别为d0血糖标准差1.214（P=0.001,95%CI：1.082-1.363）、血糖标准差控制时点1.369（P=0.021,95%CI：1.049-1.787）及APACHEⅡ评分值1.243（P=0.000,95%CI：1.183-1.305）。死亡风险函数方程H（t）=h0（t）exp（0.194d0血糖标准差＋0.314血糖标准差控制时点＋0.217 APACHEⅡ评分）。结论血糖指标预测老年内科危重症短期死亡的价值大小依次为：血糖标准差控制时点〉APACHEⅡ评分值〉诊断SIRS当日血糖标准差;血糖波动越大,控制越晚,短期死亡风险越大。建议将血糖标准差4mmol/L和7d内未控制血糖作为预测老年内科危重症短期死亡的临界值。老年危重症中多点监测血糖以评估血糖波动,并早期控制血糖对于减少其短期死亡风险至关重要。

【Objective】 To explore the association of average blood glucose variability（standard deviation） of blood glucose concentration（Glu SD） and its control with short-term mortality in aged patients with critical medical disease,and to evaluate the value of these parameters to predict the short term death.【Methods】 A restrospective analysis was performed on 152 cases of systematic inflammatory response syndrome（SIRS） caused by medical critical diseases in the elderly patients,calculate the data of the Glu SD on the first day of diagnosing SIRS（d0） to determine their standard deviation and the controlling time（d3,d7,d15）of Glu SD,and the APACHEⅡscore.The patients were divided into three groups according to the blood glucose data.Survial curve and survival rate of 152 elderly SIRS patients were estimated with Kaplan-Meier method and Log-rank test respectively.Univariate and multivariate analysis of prognostic factors including blood glucose parameters were carried out by COX＇s regression model.【Results】 Different levels of the Glu SD on the first day of diagnosing SIRS had different survival rates.The higher level of the Glu SD had lower survival rate： the survival rates at 31 days of the patients with Glu SD 0-2 mmol/L,2-4 mmol/L and 4 mmol/L were 75.0%,43.3% and 11.5% respectively.The Glu SD,controlling time of Glu SD,and APACHEⅡscore entered the COX regression model.The RR were： Glu SD 1.214（P =0.001,95%CI： 1.082-1.363）,controlling time of Glu SD 1.369（P =0.021,95%CI： 1.049-1.787）,APACHEⅡscore 1.243（P =0.000,95%CI： 1.183-1.305）.H（t）=h0（t）exp（0.194d0 Glu SD ＋ 0.314 controlling time of Glu SD ＋0.217 APACHEⅡscore）.【Conclusion】 The value of these prediction factors about blood glucose one by one in order is： controlling time of Glu SD APACHEⅡscore Glu SD.The bigger the variability of blood glucose level,the later the time of blood glucose level on control,the greater is the danger of short-term death.Critical levels of blood glucose parameters are recommended： standard deviation of blood glucose concentration on the first day of diagnosing SIRS 4 mmol/L and blood glucose still uncontrolled within 7 days.So monitoring multi-point blood glucose levels to evaluate its variability,and controlling blood glucose and at the early stage of SIRS are essential for reducing the danger of short-term death in aged patients with critical medical disease.