6h内血糖变异度对脓毒性休克患者短期预后的预测价值

Predictive value of glycemic variability within 6 hours on the short-term prognosis of patients with septic shock

目的探讨6 h内不同血糖变异指标对脓毒性休克患者短期预后的预测价值。方法采用回顾性研究方法,收集2014年12月至2019年12月入住南京医科大学附属南京医院重症监护病房(ICU)的133例脓毒性休克患者的临床资料。以患者入住ICU治疗期间是否死亡为观察节点,死亡患者纳入死亡组,否则纳入存活组。收集患者一般资料,包括性别、年龄、基础疾病、感染部位、机械通气时间、ICU住院时间、是否使用连续性肾脏替代治疗(CRRT);记录患者入ICU时初始血糖(GLUadm)、平均动脉压(MAP)、血肌酐(SCr)、降钙素原(PCT)及24 h内急性生理学与慢性健康状况评分Ⅱ(APACHEⅡ);同时收集患者接受6 h集束化治疗期间每2 h监测1次的血糖水平,并计算6 h内血糖差值(GLUdif)、平均血糖(GLUave)、血糖标准差(GLUsd)及血糖变异系数(GLUcv)。采用多因素Logistic回归法分析影响脓毒性休克患者短期预后的因素;绘制受试者工作特征曲线(ROC曲线),评估血糖变异指标对脓毒性休克患者短期预后的预测效能。结果133例患者ICU住院期间存活87例,死亡46例。与存活组相比,死亡组患者入ICU时SCr及24 h内APACHEⅡ评分更高〔SCr(μmol/L):208.5(143.0,286.5)比172.0(91.0,234.0),APACHEⅡ评分(分):30.28±6.67比24.03±5.90,均P<0.05〕,ICU住院时间更短〔d:4.00(2.00,10.25)比9.00(4.00,13.00),P<0.01〕;而两组患者性别、年龄、基础疾病、感染部位、机械通气时间、CRRT比例及入ICU时MAP、PCT等基线资料比较差异均无统计学意义。与存活组相比,死亡组患者6 h内GLUsd及GLUcv均明显升高〔GLUsd(mmol/L):2.33(1.95,3.14)比2.02(1.66,2.52),GLUcv:(31.00±7.06)%比(23.31±10.51)%,均P<0.05〕;而两组患者6 h内GLUadm、GLUdif及GLUave比较差异均无统计学意义。多因素Logistic回归分析显示,24 h内APACHEⅡ评分及6 h内GLUsd、GLUcv是影响脓毒性休克患者短期预后的独立危险因素〔APACHEⅡ评分:优势比(OR)=1.173,95%可信区间(95%CI)为1.095~1.256,P=0.000;GLUsd:OR=1.465,95%CI为1.038~2.067,P=0.030;GLUcv:OR=1.089,95%CI为1.043~1.138,P=0.000〕。ROC曲线分析显示,6 h内GLUsd及GLUcv均可用于脓毒性休克患者短期预后的评估,其中6 h内GLUcv的ROC曲线下面积(AUC)大于APACHEⅡ评分(0.765比0.753),6 h内GLUsd的AUC接近于APACHEⅡ评分(0.629比0.753);且6 h内GLUsd联合6 h内GLUcv的诊断价值较二者单独诊断价值更大(AUC:0.809比0.629、0.765),敏感度为97.8%,特异度为66.7%。结论6 h内GLUsd联合GLUcv可以用于预测脓毒性休克患者的短期预后。

Objective To investigate the predictive value of different glycemic variability indexes within 6 hours on the short-term prognosis of septic shock patients.Methods A retrospective study was conducted.The 133 patients with septic shock admitted to intensive care unit(ICU)of Nanjing Hospital Affiliated to Nanjing Medical University from December 2014 to December 2019 were enrolled.Patients with septic shock admitted to ICU died during hospitalization were enrolled in the death group and others in the survival group.General data of the patients including gender,age,underlying disease,site of infection,duration of mechanical ventilation,length of ICU stay,whether to use continuous renal replacement therapy(CRRT)and acute physiology and chronic health evaluationⅡ(APACHEⅡ)scores within 24 hours were collected.The blood glucose(GLUadm),mean arterial pressure(MAP),serum creatinine(SCr)and procalcitonin(PCT)were recorded at ICU admission.The patients admitted to ICU received bundle therapy within 6 hours and blood glucose was observed every 2 hours.The blood glucose difference(GLUdif),average blood glucose(GLUave),blood glucose standard deviation(GLUsd)and blood glucose variation coefficient(GLUcv)within 6 hours were calculated.Multivariate Logistic regression analysis was used to analyze the prognostic factors of short-term prognosis of patients with septic shock,and receiver operating characteristic(ROC)curve was plotted to evaluate the diagnostic efficacy of glycemic parameters for short-term prognosis of septic shock patients.Results A total of 133 patients with septic shock were admitted to ICU,among them 87 patients survived and 46 patients died during the ICU hospitalization.Compared with the survival group,the SCr at ICU admission and APACHEⅡscore within 24 hours were significantly higher in the death group[SCr(μmol/L):208.5(143.0,286.5)vs.172.0(91.0,234.0),APACHEⅡscore:30.28±6.67 vs.24.03±5.90,both P<0.05],the length of ICU stay was shorter[days:4.00(2.00,10.25)vs.9.00(4.00,13.00),P<0.01].However,there were no significant differences in the baseline data of gender,age,underlying disease,infection site,CRRT ratio,MAP or PCT at ICU admission between the two groups.Compared with the survival group,the GLUsd and GLUcv within 6 hours in the death group were higher[GLUsd(mmol/L):2.33(1.95,3.14)vs.2.02(1.66,2.52),GLUcv:(31.00±7.06)%vs.(23.31±10.51)%,both P<0.05].There were no statistically significant differences in the levels of GLUadm,GLUdif or GLUave within 6 hours between the two groups.Multivariate Logistic regression analysis showed that APACHEⅡscore within 24 hours and GLUsd and GLUcv within 6 hours were independent risk factors of the short-term prognosis of septic shock patients[APACHEⅡscore:odds ratio(OR)=1.173,95%confidence interval(95%CI)was 1.095-1.256,P=0.000;GLUsd:OR=1.465,95%CI was 1.038-2.067,P=0.030;GLUcv:OR=1.089,95%CI was 1.043-1.138,P=0.000].ROC curve analysis showed that GLUsd and GLUcv within 6 hours both had certain predictive value for the short-term prognosis of septic shock patients,the area under ROC curve(AUC)of GLUcv within 6 hours was higher than that of APACHEⅡscore(0.765 vs.0.753),and AUC of GLUsd within 6 hours was close to APACHEⅡscore(0.629 vs.0.753);and the diagnostic value of GLUsd combined with GLUcv within 6 hours was higher than the two respectively(AUC:0.809 vs.0.629,0.765),the sensitivity was 97.8%,and the specificity was 66.7%.Conclusion GLUsd combined with GLUcv within 6 hours can be used to estimate the short-term prognosis of septic shock patients.