【作者】 万鹏；

【导师】 苏磊；

【作者基本信息】 南方医科大学 ， 急诊医学， 2014， 博士

【摘要】 第一部分即时凝血功能监测在弥散性血管内凝血中的诊断价值目的探讨即时凝血功能监测在弥散性血管内凝血(Disseminated intravascular coagulation, DIC)患者中的诊断价值。方法采用前瞻性病例对照试验设计,将具有DIC潜在危险因素、临床表现或实验室检查疑诊为DIC患者,依据国际血栓与止血学会(ISTH)／科学标准化学会(SSC)显性DIC评分标准分为显性DIC组(DIC评分≥5分)、非显性DIC组(DIC评分<5分)。收集所有患者的年龄、性别、ICU住院时间、总住院时间、急性生理和慢性健康(APACHE Ⅱ)评分及血常规、常规凝血试验(凝血酶时间(PT)、部分活化凝血酶时间(APTT)、纤维蛋白原(FIB)、国际标准化比值(INR))检测、D-二聚体定量检测等资料,同时进行Sonoclot凝血和血小板功能分析仪量化指标(激活凝血时间(ACT)、纤维蛋白凝结速率(CR)、血小板活性(PF))检测。比较显性DIC组、非显性DIC组一般临床资料及各指标间的差异,寻找可能有助于诊断DIC的指标；同时对显性DIC亚组(早期：6小时内、中期：6-12小时、晚期：12小时后)进行分析,探讨各指标的变化趋势,为不同阶段DIC诊断及病情判断相关指标的选择提供依据。基于上述临床数据,分析Sonoclot指标与其他凝血指标之间的关系,采用诊断性试验的相关方法,绘制不同时间段各凝血功能指标的受试者工作特征曲线(ROC),计算在不同时间段各凝血指标诊断DIC的ROC曲线下面积(AUC),探讨各指标预测DIC的最佳阈值并计算其敏感度、特异度。结果共498例患者纳入本研究。患者平均年龄为52.5±19.8岁,其中男性362例(72.7%),女性136例(27.3%)。①结果提示组间哪些指标差异有统计学意义,可能用于DIC诊断。显性DIC组与非显性DIC组患者在年龄、性别之间差异无统计学意义(P分别为0.400和0.453),具有可比性。显性DIC组患者APACHE Ⅱ评分24.8±6.0分、ISTH评分5(5,8)分,显著高于非显性DIC组18.5±5.5分,3(0,4)分(P<0.01)。显性DIC组、非显性DIC组患者APTT、 PT、ACT以及INR分别为(76.6±46.2秒,55.7±40.1秒)、(21.4±9.5秒,16.1±3.3秒)、(231.1±165.4秒,134.8±71.3秒)、(2.2±1.1,1.5±0.7),显性DIC组明显延长(均P<0.001)；显性DIC组D-二聚体为5880.2±2733.5ng/L,较非显性DIC组3695.0±2510.3ng/L明显升高(P<0.01)；显性DIC组纤维蛋白原含量2.5±2.1g/L、血小板计数54.5±38.1×109/L以及PF1.1(0,4.6)均较非显性DIC组4.2±2.1g/L、154.1±90.4×109/L、2.4(0.1,6.7)下降(均P<0.001)。②结果提示,针对组间有差异的指标,初步判断哪些指标用于哪个阶段可能更有价值?利用单因素Logistic回归分析对传统的凝血指标(APTT、PT、血小板计数、纤维蛋白原、D-二聚体、INR)以及Sonoclot指标(ACT、CR、PF)与患者是否发生DIC进行了分析,结果显示无论是传统的凝血指标还是Sonoclot指标的P值均<0.001,说明凝血相关指标与DIC的发生均具有良好的相关性。利用SPSS产生0-1均匀分布的随机数,选择70%样本作为训练样本,30%样本为验证样本。将传统的凝血指标(APTT、PT、INR,纤维蛋白原、D-二聚体以及血小板计数)是否发生DIC进行多因素分析,采用强迫进入法。模型的拟合R Square值为0.551,模型方程为Log(P/1-P)=-2.586-0.226PT+0.675INR-0.233FIB+0.000D-二聚体-0.037血小板计数,ROC曲线下面积为0.841。将ACT序贯性引入该模型,新模型拟合R Square值为0.540,模型方程为Log(P/1-P)=-3.066+0.171PT+0.521INR-0.226FIB+0.000D-二聚体-0.031血小板计数+0.006ACT,曲线下面积上升为0.845；将CR引入模型,新模型拟合R Square值为0.559,模型方程为Log(P/1-P)=-3.843+0.200PT+0.487INR-0.216FIB+0.000D-二聚体+0.010ACT-0.022CR,ROC曲线下面积较前无明显变化,仍为0.845；将PF引入模型中,新模型拟合R Square值为0.454,模型方程为Log(P/1-P)=-2.827+0.OO0D-二聚体+0.009ACT+0.069PT+0.498INR-0.401FIB-0.449PF-0.002血小板计数,ROC曲线下面积上升为0.876。③Sonoclot指标与其他凝血指标之间的关系及其分别在诊断DIC中的价值比较。Sonoclot指标与传统的凝血指标相关性分析结果显示,CR与血小板呈负相关(r=-0.154,p=0.018),PF与APTT呈负相关相关(r=-0.134,p=0.039),PF与纤维蛋白原呈正相关(r=0.189,p=0.003);在各显性DIC组亚组Sonoclot指标与传统的凝血指标进行相关性分析结果显示,12小时后亚组内血小板计数与PF呈明显正相关(r=0.513,p<0.00)。以非参数法构建不同时间段的受试者工作曲线,计算各凝血指标在DIC不同的时间段诊断DIC的ROC曲线下面积,并计算不同指标的敏感度及特异度。结果显示：在早期(6小时内),传统的凝血指标中APTT、PT、D-二聚体以及INR对应的ROC曲线下面积分别为0.699、0.759、0.635、0.762(均P<0.001),Sonoclot指标中ACT曲线下面积为0.765(P<0.001),均优于随机猜测,有较好的预测价值。同时,纤维蛋白原、血小板、CR以及PF ROC曲线下面积均小于0.5(P<0.001),预测准确度较差,经转化后ROC曲线下面积分别为0.741、0.896、0.682、0.603(P<0.001)。与传统的凝血指标相比,ACT诊断早期显性DIC患者患者的敏感性度较差仅为50.6%,特异度达到95.9%。以“0”表示非显性DIC组,“1”表示显性DIC组,采用二项分类Logistic回归强行进入法建立Sonoclot联合指标的预测概率模型并计算预测概率。结果显示：ACT联合CR的模型方程为P=exp(-2.265+0.021ACT-0.046CR)/1+exp(-2.265+0.021ACT-0.046CR),模型拟合度检验卡方为7.310(P=0.504),模型拟合较好；ACT联合CR以及PF预测概率模型为：P=exp(-1.517+0.022ACT-0.041CR-0.398PF)/1+exp(-1.517+0.022ACT-0.041CR-0.398PF),模型拟合度检验卡方为6.001(P=0.647),模型拟合较好。利用ROC曲线分析方法分析ROC曲线下面积,结果提示ACT联合CR,以及ACT、CR、PF联合后ROC曲线下面积分别为达到0.814,0.828(P<0.001)。三者联合诊断的敏感度为84.9%,特异度为87.8%,诊断的综合效能较好。6-12小时内,传统凝血指标中APTT、PT、INR以及D-二聚体ROC曲线下面积分别为0.624、0.768、0.713、0.736(均P<0.001)。Sonoclot指标中CR的ROC曲线下面积为0.448,无统计学意义(P=0.269),ACT的ROC曲线下面积为0.787,优于随机猜测,有较好的预测价值(P<0.001)。同时,纤维蛋白原、血小板以及PF ROC曲线下面积均小于0.5(P<0.001),预测准确度较差,经转化后ROC曲线下面积分别为0.780、0.870、0.712(P<0.001)。ACT联合CR的Logistic回归分析提示变量CR无统计学意义(P=0.420),预测概率模型为：P=exp(-3.197+0.022ACT)/1+exp(-3.197+0.022ACT),模型拟合度检验卡方为12.538(P=0.129),模型拟合较好,ROC曲线下面积为0.792(P<0.001)；ACT联合CR、PF预测概率模型为：P=exp(-2.810+0.023ACT-0.041CR-0.804PF)/1+exp(-2.810+0.023ACT-0.041CR-0.804PF).模型拟合度检验卡方为5.903(P=0.658),模型拟合较好。联合ACT、 CR以及PF后曲线下面积达到0.846(P<0.001)。与传统的凝血指标相比,ACT诊断的敏感度仅为63.3%,低于传统的凝血指标,但特异度达到88.9%,联合ACT、CR、PF后,特异度为87.8%,敏感度上升至84.9%。与DIC早期及中期相比,后期(12小时后)APTT、PT、D-二聚体以及INR对应的ROC曲线下面积均超过0.8(P<0.001),而ACT曲线下面积仅为0.733。纤维蛋白原、血小板、CR以及PF ROC曲线下面积均小于0.5(P<0.001),预测准确度较差,经转化后ROC曲线下面积分别为0.794、0.907、0.688、0.826(P<0.001)。ACT联合CR、ACT联合CR, PF的Logistic回归分析均提示变量CR无统计学意义(P分别为0.533、0.058)。预测概率模型为：P=exp(-0.856+0.006ACT)/1+exp(-0.856+0.006ACT),模型拟合度检验卡方为5.208(P=0.735),模型拟合较好。ACT联合CR、PF预测概率模型为：P=exp(1.145+0.003ACT-0.979PF)/1+exp(1.145+0.003ACT-0.979PF),模型拟合度检验卡方为1.964(P=0.982),拟合较好。ROC结果显示当ACT联合CR及PF后,曲线面积达到0.839(P<0.001),尽管诊断的特异性(72.5%)小于传统的凝血指标,但敏感性(85.2%)达到最高,优于传统的凝血指标。结论Sonoclot凝血和血小板功能分析仪作为一种即时凝血功能监测仪器,能够对临床工作中并发DIC的患者发生的出凝血功能障碍做出判断。在显性DIC患者中,Sonoclot分析仪较传统的凝血实验更能准确的诊断启动后并进入消耗性低凝期的DIC患者。随着DIC病程的进展,联合Sonoclot指标诊断的综合效能均优于传统的凝血指标。由于Sonoclot分析仪具有快速、全面、耗材低等优点,可考虑在临床实际工作中作为DIC诊断的常规手段。第二部分即时凝血功能监测对弥散性血管内凝血预后判断价值目的探讨即时凝血功能监测对弥散性血管内凝血患者预后评估的价值。方法采用前瞻性病例调查试验设计。以ISTH评分标准确诊的显性DIC(ISTH评分≥5分)患者为研究对象,收集所有患者的性别、年龄等一般临床资料及凝血功能指标、Sonoclot各项指标,并随访患者预后,比较生存组和死亡组之间各项指标的差异。采用多因素Cox比例风险回归模型对传统的凝血指标(PT、APTT、INR、D-二聚体、纤维蛋白原以及血小板计数)、Sonoclot指标(ACT, CR、PF)、ISTH评分、APACHE Ⅱ评分、是否具有DIC临床症状与预后进行相关性分析。采用Kaplan-Meier生存曲线、绘制受试者工作特征曲线下面积(AUC)等探讨Sonoclot指标对DIC预后判断的价值。结果共237例患者纳入研究。①一般情况：平均年龄为51.7±17.6岁,其中男性176例(74.2%),女性61例(25.8%),死亡91例(38.4%),存活146例(61.6%),出现器官功能不全103例(43.5%),具有DIC临床症状的患者有13例(5.4%)。②组间比较,描述哪些指标可能对预后判断有价值：两组之间年龄、性别无统计学差异(P分别为0.971和0.088)。死亡组患者出现DIC临床表现比例、合并MODS比例、ISTH评分以及APACHE Ⅱ评分均显著高于生存组患者(均P<0.05),出现DIC临床表现比例分别为13.6%、0.1%,合并MODS比例分别59%、34.2%,ISTH评分分别为6(5,8)分、5(5,8)分,APACHE Ⅱ评分分别为25.9±6.3分、24.1±5.7分。在凝血相关指标中,两组D-二聚体含量分别为6471.5±2565.9ng/L,5511.6±2778.3ng/L,死亡组中较生存组明显升高(P=0.008)=ACT分别为289.5±200.5秒、194.6±126.8秒,PF分别为1.2±0.9、1.6±1.2,死亡组与生存组相比,ACT明显延长,PF明显缩短(均P<0.05);③相关性分析,何种因素影响预后?以“0”表示存活,“1”表示死亡,多因素Cox比例风险回归模型结果显示,拟合模型的卡方值为39.999(P=0.000),具有统计学意义。其中合并DIC临床症状、APACHE Ⅱ评分、ISTH评分、ACT以及PF是DIC患者重要的死亡风险因素,相对危险度(RR)分别为2.509、1.042、1.422、1.002、0.761(均P<0.05)。Kaplan-Meier生存曲线对异常Sonoclot指标分析,结果显示在237例患者中,7例患者ACT、CR、PF均正常,均存活(100%)。78例患者有一项指标异常,67例患者存活(85.9%)；102例患者有2项指标异常,59例存活(57.8%)；50例患者ACT、CR、PF均异常,20例存活(40%),Kaplan-Meier生存曲线分析提示仅一项指标异常的患者预后较好(均P<0.05),两项指标异常的患者与三项指标异常患者的预后差异无统计学意义(P=0.496)。进一步分析,根据ACT、CR及PF的正常范围,分为ACT延长组(ACT>195秒)及缩短组(ACT<119秒),CR延长组(CR>23Sig/min)及缩短组(CR<7Sig/min),PF缩短组(PF<1.5),分析不同变量的30天生存关系。Kaplan-Meier生存曲线结果显示,CR指标异常时的生存率与CR正常组相比,无统计学差异(均P>0.05)。与ACT正常组相比,ACT及PF异常的生存率明显低于指标正常组(均P<0.05),ACT缩短组与延长组比较,差别无统计学意义(P=0.113)。以非参数法构建不同时间段的受试者工作曲线,计算各凝血指标在不同的时间段诊断DIC的ROC曲线下面积,并计算不同指标的敏感度及特异度。结果显示：ISTH评分、APACHE Ⅱ评分、ACT的ROC曲线下面积分别为0.603,0.601,0.682,差异有统计学意义(均P<0.01)。其中ACT敏感度为69.2%,ISTH评分特异度为87.7%,但敏感度较差,仅为28.6%。APACHE Ⅱ评分特异度(80.1%)低于ISTH,但敏感度较低仅为40.7%。PF的ROC曲线下面积为0.420(P=0.039),预测死亡的准确度较差,经转化后ROC曲线下面积为0.580(P=0.039)。以“0”表示存活,“1”表示死亡,将ACT联合PF进行二项分类Logistic回归分析,结果显示该模型拟合度较差(P=0.007)。当ACT联合PF. ISTH时,结果显示PF无统计学意义(P=0.110),预测概率模型：P=exp(-3.918+0.004ACT+0.502ISTH),该拟合模型的卡方值为14.590(P=0.068),拟合度尚可。通过计算联合预测概率,进行ROC曲线分析,结果显示：ACT联合PF、ISTH评分的ROC曲线下面积达到0.717,预测的特异性(78.1%)尽管小于ISTH评分,但是敏感性上升至为64.8%。结论Sonoclot凝血和血小板功能分析仪作为一种即时凝血功能监测仪器可以有效的评估DIC的严重程度及预后,随着ACT延长,PF缩短,DIC患者死亡风险也随之增加。通过联合Sonoclot指标与ISTH评分,可以有效弥补ISTH评分在DIC预后评价中的不足,综合预测效能优于APACHE Ⅱ评分及ISTH评分。由于Sonoclot仪器量化指标易于获得,能够全面反映凝血阶段的整个过程,可以作为一种较好的预测DIC患者病情严重程度及预后的工具。第三部分即时凝血功能监测对弥散性血管内凝血患者指导治疗作用目的探讨即时凝血功能监测对弥散性血管内凝血患者个体化治疗的指导作用。方法采用随机对照试验设计,以ISTH评分标准确诊并具有替代及抗凝治疗指针的显性DIC (ISTH评分≥5分)患者为研究对象,按照随机数字原则将患者分为对照组(n=121例)与Sonoclot治疗组(n=116例),对照组根据国际血栓与止血标准委员会(ISTH)2009年DIC的治疗指南与弥散性血管内凝血诊断与治疗中国专家共识(2012)确定治疗方案；Sonoclot治疗组在以上基础上,结合Sonoclot量化指标确定治疗方案。观察传统凝血指标指导治疗与Sonoclot凝血分析仪指导治疗对患者替代治疗的种类、剂量及替代治疗时间、预后之间的差异。结果共纳入237例患者,研究过程中无剔除以及退出者。①基线比较：治疗前两组患者年龄、性别、体重、APACHE Ⅱ评分以及ISTH评分、肾脏替代患者比例均无显著差异(P>0.05)。对照组和Sonoclot组患者治疗前传统凝血指标(APTT、PT、INR、血小板计数、D-二聚体、纤维蛋白原)、Sonoclot仪器指标(ACT, CR、PF)、临床表现为出血、紫癜或肢端出血的患者比例无显著性差异(P均>0.05)。②两组之间抗凝治疗效果比较：Sonoclot组使用肝素比例为38/116(32.76%)高于对照组比较(23/121,19.01%),有统计学差异(P=0.018)。两组肝素使用的时间分别为11.0±7.2天、5.2±3.4天,肝素用量分别为65739.1±43322.5U、31263.2±20222.7U,与对照组相比,Sonoclot组患者肝素抗凝时间明显缩短,肝素用量显著减少(P均为0.001)。对照组中有10例(43.4%)患者出现局部出血的临床表现,与治疗组(5例,13.1%)相比具有统计学差异(P=0.003)。两组患者均未使用鱼精蛋白。在肾脏替代治疗中抗凝比较结果显示,纳入本部分研究的患者中,共有116例患者行CRRT治疗,其中50例患者行肝素抗凝治疗,对照组20例(36.3%),Sonoclot组30例(49.1%),差别无统计学意义(P=0.164)。两组肝素使用的时间分别为11.5±7.5天、4.9±3.6天,肝素用量分别为69000.0±45257.2U、294000.0±21393.8U,与对照组相比,Sonoclot组患者抗凝时间明显缩短,用量显著减少(P均为0.001)。对照组中有7例(35%)患者出现局部出血的临床表现,与治疗组(3例,10%)相比具有统计学差异(P=0.039)。两组患者均未使用鱼精蛋白。两组之间替代治疗比较结果显示,两组血浆、血小板及冷沉淀用量分别为：(1760±1757.1ml,1104.6±762.7m1).(44.2±50.9u,25.7±15.1u).(42.4±36.6u,29.9±24.3u),与对照组相比,Sonoclot组患者血浆、血小板及冷沉淀的用量显著减少(P均<0.05),红细胞悬液两组间无统计学差异(P=0.054)；两组替代治疗的时间差别无统计学意义(P=0.206)。患者DIC症状的改善率两组间无统计学差异(P=0.676)。与对照组相比,Sonoclot组ICU住院时间明显缩短(10(1,124)天,9(1,213)天,P=0.009)),ICU内死亡率明显下降(22.4%,12.4%,P=0.027)。Kaplan-Meier生存分析显示Sonoclot分析仪指导的替代和抗凝治疗与对照组比较,显著改善了30天生存时间(P=0.001)。结论Sonoclot凝血和血小板功能分析仪作为一种即时凝血功能监测仪器较传统的凝血功能检测能更精确对血液制品成分进行合理的选择和数量的控制,减少了替代治疗的数量和时间。同时,它还有效指导抗凝治疗,显著降低抗凝治疗致出血的风险,降低DIC患者死亡率。更多还原

【Abstract】 PART I The effect of the immediate coagulation mornioring in the predicting the occurrence of DICObjective To evaluate the diagnostic effect of the effect of the immediate coagulation mornioring in the predicting the occurrence of Disseminated intravascular coagulation(DIC).Methods Consecutive patients admitted to intensive care unit (ICU) were initially assessed.498cases of patients with suspected DIC according to the potential risk factors, clinical symptoms and experimental data were prospectively included and divided into2groups:overt DIC group with ISTH DIC score≥5and non DIC group with the International Society of Thrombosis and Haemostasis (ISTH) DIC score<5. The prothrombin time (PT), activated partial thromboplastin time (APTT), plasma fibrinogen level(FIB), international normalized ratio (INR), blood platelet count(PLT), D-dime were examined, as long as Sonoclot indexes such as the actived clotting time (ACT), the clot rate (CR) and platelet function (PF) and the common characteristics like ages, sex, Acute Physiology and Chronic Health Evaluation (APACHE Ⅱ) score and length of stay in ICU and hospital. The above parameters were then compared in the overt DIC group and non-DIC group to search possible diagnostic indexes. furtherly, the overt DIC group were divided into early DIC group with definite diagnosis made less than6hours, and medium-term DIC group in6-12hours and late DIC group more than12hours. The variation trend of all parameters were analyzed to provide evidence for diagnosis and determining severity of disease at different stages of DIC. Based on the above data, related diagnostic tests were used to analyze the Sonoclot parameters with other conventional coagulation indexes, and receiver-operating characteristic curve (ROC) was applied to explore the best threshold of above indexes and calculate the sensitivity and specificity of their prediction in diagnosis of DIC.Results498patients met the inclusion criteria were enrolled, with average age of52.5+19.8years old and362(72.7%) cases of male,136(27.3%) cases of female.1. There were no differences in the age and sex between overt DIC group and non-DIC group (P=0.400and0.453, respectively). The APACHE II scores and ISTH scores were much higher (APACHE II score:24.8+6.0vs18.5+5.5, P<0.001, ISTH score:5.0vs3.0, P<0.001, respectively), and the coagulation parameters such as APTT, PT, ACT and INR were much longer in the overt DIC group than those in the non-DIC group (APTT:76.6±46.2s vs55.7±40.1s, P<0.001; PT:21.4±9.5s vs16.1±3.3s, P<0.001; ACT:231.1±165.4s vs134.8±71.3s, P<0.001; INR:2.2±1.1vs1.5±0.7, P<0.001). The D-dime level was more significantly increased in the overt DIC group than those in the non-DIC group (5880.2±2733.5ng/L vs3695.0±2510.3ng/L, P<0.001). The fibrinogen level, blood platelet count, and platelet function were much lower in the overt DIC group than those in the non-DIC group(FIB:2.5±2.1g/L vs4.2±2.1g/L, P<0.001; PLT:54.5±38.1×109/Lvs154.1±90.4×109/L, P<0.001; PF:1.1vs2.4, P<0.001).2. The association between conventional coagulatory parameters such as APTT, PT, PLT, Fib, D-dimer and INR, sonoclot indexes like ACT, CR and PF and DIC occurrence were discussed with univariate analysis by logistic regression analysis. Both the conventional and sonoclot coagulatory parameters were closely related with DIC occurrence (p<0.001). Furthermore, after randomization of0～1by SPSS,70%of total sample were chosen as training pattern, and the rest were chosen as verification sample. Multi-variate analysis were processed to evaluate the conventional coagulatory parameters with DIC occurrence, and the model equation was Log(P/1-P)=-2.586-0.226PT+0.675INR-0.233FIB+0.000D-dimer-0.037PLT, with matching R Square of0.551and the AUC of0.841. After ACT was sequentially introduced, the model equation shifted into Log(P/1-P)=-3.066+0.171PT+0.521INR-0.226FIB+0.000D-Dimer-0.031PLT+0.006ACT, with matching R Square of0.540and the AUC of0.845. After CR was sequentially introduced, the model equation shifted into Log(P/1-P)=-3.843+0.200PT+0.487INR-0.216FIB+0.000D-Dimer+0.010ACT-0.022CR, with matching R Square of0.559and the AUC of0.845. After PF was sequentially introduced, the model equation shifted into Log(P/1-P)=-2.827+0.000D-dimer+0.009ACT+0.069PT+0.498INR-0.401FIB-0.449PF-0.002PLT, with matching R Square of0.454and the AUC of0.876.3. The CR was negatively correlated with the blood platelet count (r=-0.154,p=0.0018), PF was negatively correlated with the APTT (r=-0.134, p=0.039) and positively with FIB (r=0.189, p=0.003). We also found that the PF was more positively correlated with the blood platelet count in the late subgroup(r=0.513,p=0.000) than other two subgroup. In the early subgroup, the AUC of the conventional coagulation parameters like APTT, PT, D-dime and INR were0.699,0.759,0.635and0.762, respectively(P<0.001). Compared with conventional coagulation indexes, the AUC of ACT along was0.765. We also founde tha the AUC of PLT,FIB, CR and PF were less than0.5, with poor predicting importance. When transformed, AUC of them were0.741,0.896,0.682and0.603, respectively (P<0.001). To predict early overt DIC, the specificity was95.9%and the sensitivity was50.6%in single ACT index. We used "0" and "1" as non-DIC and overt-DIC respectively, and applied Logistic regression with enter method to established model of prediction probability of Sonoclot index. The equation of model of ACT combined with CR was P=exp(-2.265+0.021ACT-0.046CR)/1+exp(-2.265+0.021ACT-0.046CR)(x2=7.310, P=0.504), and when combined with CR and PF, The equation of model was P=exp(-1.517+0.022ACT-0.041CR-0.398PF)/l+exp(-1.517+0.022ACT-0.041CR-0.398PF)(χ2=6.001, P=0.647). The AUC of ACT combined with CR and with PF were0.814and0.828, respectively (P<0.001). The sensitivity was increased to when ACT combined with CR and PF, it increased significantly to84.9%, and the specificity was87.8%, respectively. Both more significantly useful than those of conventional parameters. In the ROC analysis of medium-term subgroup, the AUC of the conventional coagulation parameters like APTT, PT, INR and D-dime were0.624,0.768,0.713and0.736, respectively (P<0.001). The AUC of ACT was0.787. We also founde tha the AUC of FIB, PLT and PF were less than0.5, with poor predicting importance. When transformed, AUC of them were0.780,0.870and0.712, respectively (P<0.001). Logistic regression analysis of ACT combined with CR showed CR was of no significant. The equation of model of ACT combined with CR was P=exp(-3.197+0.022ACT)/1+exp(-3.197+0.022ACT)(x2=12.538,P=0.129), and when combined with CR and PF, The equation of model was P=exp(-2.810+0.023ACT-0.041CR-0.804PF)/1+exp(-2.810+0.023ACT-0.041CR-0.804PF)(x2=5.903, P=0.658). The AUC of ACT combined with CR and PF were0.846, respectively (P<0.001). To predict overt DIC, sensitivity of63.3%and specificity of88.9%in single ACT index. When combined with CR and PF, the specificity of87.8%and sensitivity of84.9%. In the ROC analysis of12hour subgroup, the AUC of the conventional coagulation parameters like APTT, PT, INR and D-dime were more than0.8(P<0.001). The AUC of ACT was0.733. We also founde tha the AUC of FIB, PLT, CR and PF were less than0.5, with poor predicting importance. When transformed, AUC of them were0.794,0.907,0.688and0.826, respectively (P<0.001). Logistic regression analysis of ACT combined with CR and with CR, PF showed CR was of no significant. The equation of model of ACT combined with CR was P=exp(-0.856+0.006ACT)/1+exp(-0.856+0.006ACT)(x2=5.208,P=0.735), and when combined with CR and PF, The equation of model was P=exp(1.145+0.003ACT-0.979PF)/1+exp(1.145+0.003ACT-0.979PF)(x2=1.964, P=0.0.982). The AUC of ACT combined with CR and PF were0.839(P<0.001). When ACT combined with CR and PF, the specificity of72.5%, but sensitivity of84.9%was higher than conventionnal parametres.Conclusions Sonoclot analysis can predict the coagulation dysfunction in the patents with DIC, and is more effective and accurate to diagnose and monitor the consumed hypocoagulable stage of overt DIC than conventional coagulation parameters. It can be widespread used as a routine diagnostic method to predict DIC occurrence due to its rapid, comprehensive, and low-consumable advantages. PART II The immediate coagulation mornioring as a potential prognostic predictor in overt DIC patientsObjective Disseminated intravascular coagulation (DIC) contributed to high mortality. The study was performed to investigate the effect of immediate coagulation mornioring as a potential prognostic predictor in overt DIC patients.Methods We performed prospective investigation test design.237cases of patients with overt DIC according to ISTH score were prospectively included in the study, and was divided into2groups:non-survival group and survival group Clinical and demographic data at admission included the age, sex, conventional coagulation parameters, Sonoclot indexes and outcome were collected and compared. For multivariate analysis of prognosis, COX regression model was used to analyze the correlation in conventional parameters(PT, APTT, INR, D-dime, FIB and blood platelet count), Sonoclot indexes, ISTH and APACHE Ⅱ scores and clinical manifestations of DIC. Kaplan-Meier analysis or the log-rank test were applied to evaluate the prognostic importance of sonoclot indexes. Based on the above data, related diagnostic tests were used to analyze the Sonoclot parameters with ISTH and APACHE Ⅱ scores, and receiver-operating characteristic curve (ROC) was applied to explore the best threshold of above indexes and calculate the sensitivity and specificity of their prediction in prognosis of DIC.Results A total of237patients met the inclusion criteria were enrolled.1. The average age was51.7+17.6years old, with176(74.2%) males,61(25.8%)femal and91(38.4%) death cases. There were103(43.5%) cases with multiple organ dysfunction (MODS) and13(5.4%) cases with clinical symptoms of DIC.2. The age and sex was no difference between two groups (P=0.971and0.088, respectively). In the non-survival group, the DIC manifestation rate (13.6%vs0.1%, P<0.001), MODS rate (59%vs34.2%,P=0.001) and ISTH score (6vs5, P=0.004), ISTH score (25.9±6.3vs24.1±5.7, P=0.022) were much higher. The significantly increased D-dime levetl (6471.5±2565.9ng/L vs5511.6±2778.3ng/L, P=0.008), delayed ACT (289.5±200.5s vs194.6±126.8s, P<0.001) and much shorter PF (1.2±0.9vs1.6±1.2, P=0.010) were also found in the non-survival group.3. COX regression analysis showed that clinical manifestations of DIC(OR=2.509, P=0.013), increased APACHE Ⅱ score (RR=1.043, P=0.04) and ISTH score (RR=1.424, P=0.026), delayed ACT (RR=1.002, P=0.001) and reduced PF (RR=0.763, P=0.017) were risk factor of mortality in DIC. Kaplan-Meier survival curve analysis of abnormal sonoclot indexes indicated that100%of7cases with normal sonoclot indexes survived (n=7),85.9%of78cases with one abnormal index survived (n=67),57.8%of102cases with two abnormal index survived (n=59),40%of50cases with all abnormal3indexes survived (n=20). The above data suggested that patients with one pathological Sonoclot findings have better outcome (P<0.05), and there was no difference between with one pathological and all abnormal3Sonoclot index. The patients were furtherly divided into ACT delaying group(ACT>195s) and ACT reducing group(ACT<119s), CR delaying group(CR>23Sig/min) and CR reducing group(CR<7Sig/min) and PF reducing group(PF<1.5), and30-day survival curve analysis were repeated. Compared with normal ACT and normal PF group, the mortality rate of abnormal ACT and abnormal PF group were significantly increased(P<0.05), with no difference in the abnormal CR group and normal CR group(P>0.05). Compared with ISTH scores and APACHE Ⅱ scores, the AUC of ACT was0.682, with statistically difference (P<0.01). To predict the prognosis of DIC, the specificity was68.5%and the sensitivity was69.2%of single ACT index. The specificity of ISTH scores and APACHE Ⅱ scores were87.7%and80.1%, respectively, however, the sensitivity were only28.6%and40.7%, respectively. We also founde tha the AUC of PF was0.420(P=0.039), with poor predicting importance. When transformed, AUC of PF were0.580(P=0.039). We used "0" and "1" as Survival and non-Survival respectively, and applied Logistic regression with enter method to established model of prediction probability of Sonoclot index. The model of ACT combined with PF was poor fitting. When ACT combined with PF and ISTH score, we found that PF was of no significent(P=0.110). The equation of model was P=exp(-3.918+0.004ACT+0.502ISTH)(x2=14.590, P=0.068). When ACT combined with PF and ISTH, the AUC of0.717was much larger than those combined with ACT and PF(P<0.01), with specificity of78.1%and sensitivity of64.8%. In conclusion, ACT with PF and ISTH scores have better prognostic importance to predict30-day survival rate in DIC.Conclusions Sonoclot analysis can effectively assess the severity and prognosis of DIC. Delayed ACT and reduced PF suggest the increased mortality risk. Combined Sonoclot parameters with ISTH scores, it can effectively make up for the inadequacy of ISTH scores and is superior to predict the30-day prognostics than APACHE Ⅱ and ISTH scores alone. Since Sonoclot analysis provide easy-obtained quantitative indexes and can fully reflect the whole process of blood coagulation course, it can be used as an effective tool to forecast the severity and prognosis of DIC. PART Ⅲ Protocol based on immediate coagulation mornioring to guide blood replacement and anticoagulant therapy during DICObjective To explore the guidance of immediate coagulation mornioring in the individualized treatment of DICMethods237cases of DIC patients with ISTH scores≥5were randomly divided into either Sonoclot group (n=121) and control groups (n=116) according to the principle of random numbers. The control group was received the conventional treatment according to2009ISTH DIC guidelines and2012China expert consensus. The sonoclot group was received conventional treatment under sonoclot guidance. The types, doseage, and duration of replacement treatment and the outcome like complication rate, ICU stay and30-day survival rate were recorded and compared.Results A total of237patients met the inclusion criteria were enrolled.1. There were no differences in the demographic data including age, sex, body weight, APACHE Ⅱ score, ISTH score and the ratio of patients with renal replacement treatment, APTT, PT, INR, platelet count, FIB,D-dime, ACT, CR and PF between two groups(P>0.05). The ratio of clinical symptoms of DIC like bleeding, purpura or limb bleeding was not statistically significant (P>0.05).2. In Sonoclot group, the ratio of heparin using were much higher (32.76%vs19.01%, P=0.018), with shorter duration (11.0±7.2d vs5.2±3.4d, P=0.001) and reduced dosage (65739.1±43322.5U vs31263.2±20222.7U, P=0.001) than those in control group. Since protamine were not given in both of two groups, the regional bleeding rate were significantly lower in Sonoclot treatment group(13.1%vs43.4%, P=0.003).116cases of patients were received RRT, with50cases of heparin usage(20cases in control group and30cases in sonoclot group, P=0.164). Compared with control group, the heparin duration (11.5±7.5d vs4.9±3.6d, P=0.001) and dosage (69000.0±45257.2U vs294000.0± 21393.8U, P=0.001) were decreased significantly in sonoclot group, with reduced regional bleeding rate(35%vs10%,P=0.039). Besides, the fresh frozen plasma (1760±1757.1ml vs1104.6±762.7ml, P<0.05), PLT (44.2±50.9u vs25.7±15. lu, P<0.05) and Cryoprecipitate replacement dosage42.4±36.6u vs29.9±24.3u, P<0.05) was much lower in the sonoclot treatment group. There was no difference in the duration of replacement therapy and DIC improvement rate in two groups (P=0.054). There was no difference in improvement of DIC between two groups(P=0.676).The shorter ICU stay (9(1,92)d vs10(1,120)d, P=0.009) were found in the sonoclot treatment group. The Sonoclot group significantly improve the prognosis by Kaplan-Meier analysis curve (P<0.05).Conclusions The Sonoclot analysis can be more accurate and rigorous to select and control the quantity and composition in the blood replacement therapy. It can also reduce the duration and doseage of blood product. It is useful and effective to guide the clinical anticoagulant therapy, to significantly reduce the bleeding risk and mortality in patients with DIC.更多还原