摘要
目的:国内首次对较大样本敌草快中毒进行系统临床调查,探究患者死亡的危险因素,为进一步提高敌草快中毒的诊治水平提供科学依据。方法:收集我院急诊科2016—2018年间收治的40例敌草快中毒患者的临床资料,包括年龄、性别、中毒剂量、中毒至来院时间、入室生命体征、氧合指数、炎症指标、凝血功能、脏器损伤、治疗时间及最终预后等指标,对患者的临床特征进行分析,比较生存组和死亡组中毒剂量、临床体征、氧合指数、炎症指标和脏器功能等单因素的差异,多因素Logistic回归分析中毒患者的死亡危险因素,并构建预测模型。结果:①入组40例敌草快中毒患者,占229例农药中毒的17.5%,男25例(62.5%),年龄(36.9±14.5)岁;中毒剂量5～800 ml,平均(112.3±58.9) ml,治疗4(3.0～7.5) d,最终21例(52.5%)救治成功。②单因素分析提示心率、呼吸、首次检验白细胞计数、白介素-6、SIRS评分和中毒剂量等指标增高,显著增加患者的死亡风险(均P<0.05)。死亡组患者入室AST、ALT最大值、AST最大值、胆红素最大值和肌酐最大值高于生存组患者数据,且差异有统计学意义(均P<0.05)。③纳入服毒剂量、白细胞计数和AST最大值的Logistic回归模型正确率为92.5%,敏感度为94.7%,特异度为90.5%;3个变量进行联合诊断绘制ROC曲线图,曲线下面积为0.96,95%可信区间为90%～100%。结论:单因素和Logistic回归分析较大样本的敌草快中毒临床数据,发现基于服毒剂量、白细胞计数和AST最大值3个变量构建的预测模型,可帮助临床医生较准确地预测中毒患者预后。近年敌草快中毒患者呈快速增多趋势,临床特征不同于百草枯,提示必须加强科学研究,尽早制定出敌草快中毒的临床诊治指南。
Objective:A systematic clinical investigation was carried out to explore the risk factors of patients’death,so as to provide scientific basis for further improving the diagnosis and treatment level of diquat poisoning for the first time in China.Method:It was a retrospective study,the clinical data of 40 patients with diquat poisoning who were admitted to our emergency department from 2016 to 2018,were collected including age,sex,toxic dose,poisoning to hospital time,vital signs,oxygenation index,indicators of inflammation,coagulation,organ damage,treatment time and final prognosis.We analyzed the clinical characteristics of patients,compared the differences of toxic dose,clinical signs,oxygenation index,inflammation index and organ function between survival group and death group,and then we analyzed the risk factors of death in poisoned patients by multivariate Logistic regression,and constructed a prediction model.Result:①Forty patients with diquat poisoning were enrolled,accounting for 17.5%of 229 cases of pesticide poisoning,of whom 25(62.5%)were males and the average age was 36.9 years old.The poisoning dose ranged from 5 to 800 ml,with an average of 112.3 ml.The treatment lasted for 4(3.0~7.5)days,and 21 cases(52.5%)were cured successfully.②Univariate analysis showed that heart rate,respiration,white blood cell count,interleukin-6,SIRS score and poisoning dose increased significantly,which significantly increased the risk of death(all P<0.05).The data of AST,ALT,AST,bilirubin and creatinine in the death group were higher than those in the survival group,and there were significant differences(all P<0.05).③Logistic regression model incorporating dose,white blood cell count and AST maximum was 92.5%correct,94.7%sensitive and 90.5%specific.The ROC curve was drawn by joint diagnosis of the three variables.The area under the curve was 0.96,and the 95%confidence interval was 90%~100%.Conclusion:Univariate and Logistic regression analysis showed that the prediction model based on dose,white blood cell count
作者
梁晓丽
虎晓岷
龚阳
王仙琦
刘善收
LIANG Xiaoli;HU Xiaomin;GONG Yang;WANG Xianqi;LIU Shanshou(Department of Emergency,the First Affiliated Hospital of PLA Air Force Medical University,Xi,an,710032,China)
出处
《临床急诊杂志》
CAS
2019年第5期366-370,375,共6页
Journal of Clinical Emergency
基金
西京医院助推计划-临床探索类课题(No:XJZT18ML53)
