PICC相关性上肢静脉血栓风险评估模型的初步构建

发布时间：2018-06-25 00:17

  本文选题：PICC + 血栓　； 参考：《山西医科大学》2017年硕士论文

【摘要】：目的:1.应用Meta分析法进一步明确PICC相关性上肢静脉血栓的危险因素。2.在确定血栓危险因素的基础上,初步建立PICC相关性上肢静脉血栓风险评估模型,为识别置管的高危患者和预防血栓的发生提供理论依据。方法:1.计算机检索国内外数据库,主要为CNKI、Wan Fang、Pub Med、Embase、Science Direct和Cochrane Library中公开发表的关于PICC相关性上肢静脉血栓形成影响因素的文献。用Cochrane协作网提供的Meta分析软件即Revman5.3对提取到的纳入文献的数据资料进行定量分析。分类资料的合并效应量用比值比(odds ratio,OR)及95%可信区间(confidence interval,CI)表示。2.回顾性研究分析2015年7月至2016年2月在山西省某三级甲等综合医院就诊并住院的146例PICC置管患者的病例资料,包括PICC相关性血栓组12例,非血栓组134例。以是否发生PICC相关性上肢静脉血栓作为因变量,以Meta分析后有意义的危险因素即肥胖(BMI≥30)、癌症、血栓史、PICC或CVC置管史、穿刺次数≥2次、置管侧肢体肿胀、卧床72h为自变量,建立Logistic回归模型。结果:1.按照文献筛选的流程和标准,最终纳入10篇文献,包括5篇队列研究,5篇病例对照研究。研究总病例数为3471例,血栓组为662例,非血栓组为2809例。Meta分析的合并结果中具有统计学意义的有7项因素,即肥胖(BMI≥30):OR=2.20,95%CI(1.41,3.44);癌症:OR=1.68,95%CI(1.12,2.53);血栓史:OR=1.77,95%CI(1.36,2.32);PICC或CVC置管史:OR=3.83,95%CI(1.51,9.66);穿刺次数≥2次:OR=1.77,95%CI(1.09,2.87);置管侧肢体肿胀:OR=15.44,95%CI(6.44,37.05);卧床72h:OR=9.44,95%CI(3.87,23.02)。结果表明这些因素是影响血栓发生的危险因素。2.依据Logistic回归分析结果建立回归模型,得到回归模型为Logit(P)=-1.681+2.184×肥胖(BMI≥30)+2.283×癌症+2.086×血栓史+2.177×PICC或CVC置管史+3.518×穿刺次数≥2次+3.027×置管侧肢体肿胀+1.153×卧床72h。该模型的预测正确率为95.9%,ROC曲线下面积为0.908,统计学认为对PICC相关性上肢静脉血栓的发生有较好的预测能力。并且得到最佳临界值为4.632,相应的灵敏度和特异度是81.2%和75.7%。结论:本研究初步构建了PICC相关性上肢静脉血栓的风险评估模型,方便实用,可以帮助护理人员对PICC置管患者进行个体化评估,明确高危患者,从而及早预防。
[Abstract]:Purpose 1. Meta analysis was used to further determine the risk factors of PICC associated upper limb venous thrombosis. 2. 2. On the basis of determining the risk factors of thrombus, a PICC related assessment model of venous thrombus risk in upper extremity was established to provide theoretical basis for identifying high risk patients with catheterization and preventing thrombosis. Method 1: 1. To search the domestic and foreign databases by computer, the literatures published in CNKI Wan Fangli Pub Medbase Science Direct and Cochrane Library on the influencing factors of PICC associated upper limb venous thrombosis were studied. The Meta-analysis software provided by Cochrane collaboration, Revman 5.3, was used to quantitatively analyze the data collected from the literature. The combined effect of classified data was expressed as odds ratio (OR) and 95% confidence interval (confidence CI). From July 2015 to February 2016, we retrospectively analyzed the data of 146 patients who were admitted to a general hospital in Shanxi province, including 12 patients with PICC-associated thrombus and 134 patients with non-thrombus. Whether or not PICC was associated with venous thrombosis in upper extremity was taken as dependent variable, obesity (BMI 鈮，

本文编号：2063634