经皮冠状动脉介入术中活性凝血时间异常降低的预测模型构建及验证  

Construction and verification of prediction model for abnormal decrease in activated coagulation time during percutaneous coronary intervention

作　　者：胡超超 安燕 姜君 牟灵英 HU Chaochao;AN Yan;JIANG Jun;MU Lingying(School of Nursing,Weifang Medical University,Weifang 261053,Shandong,China;Affiliated Hospital of Weifang Medical College,Weifang 261031,Shandong,China)

机构地区：[1]潍坊医学院护理学院,山东潍坊261053 [2]潍坊医学院附属医院,山东潍坊261031

出　　处：《护士进修杂志》2024年第19期2066-2073,共8页Journal of Nurses Training

基　　金：山东省中医药科技项目(编号:Q-2023147)。

摘　　要：目的 构建经皮冠状动脉介入治疗(percutaneous coronary intervention, PCI)患者术中活性凝血时间(activated coagulation time, ACT)低于正常范围的预测模型,评估该模型对ACT检测值低于正常范围的预测效果。方法 通过文献检索、病例回顾及相关介入专家进行半结构式访谈确定影响因素条目池,选取2023年1-5月山东省某三级甲等医院导管室行PCI术的293例患者作为建模组,2023年6-7月90例PCI患者作为验证组对模型作进一步验证。通过病历以及医院信息系统查询收集一般资料以及相关因素临床指标。将所收集影响因素应用LASSO回归进行降维处理,初筛掉不重要因素后进行单因素分析,结合临床应用将单因素分析中P<0.05的指标纳多因素分析,应用logistic回归分析进一步探讨并建立列线图预测模型。应用校准曲线、受试者工作特征曲线和决策曲线分析对列线图进行准确度、稳定性评估。结果 患者患有糖尿病、高胆固醇、手术时间>45 min、数字疼痛评价量表评分是影响ACT低于正常范围的影响因素(P<0.05)。利用上述指标构建列线图模型,其预测PCI术中ACT检测值低于正常范围曲线下面积为0.726。决策曲线阈值为0.17~0.63时,与2条极端曲线不相交,模型净获益高。结论 本研究构建的对ACT检测值低于正常范围的列线图预测模型具有良好的区分度与校准度,可直观、简洁用于对PCI介入手术患者ACT检测值低于正常范围的预估。Objective To construct a prediction model of intraoperative activated coagulation time(ACT)below the normal range in patients undergoing percutaneous coronary intervention(PCI),and to evaluate the prediction effect of the model on ACT values below the normal range.Methods A item pool of influencing factors was identified through literature search,case review,and semi-structured interviews with relevant interventionalists,and 293 patients undergoing PCI in the catheterization laboratory of a tertiary care hospital from January to May 2023 were selected as the modeling set,and 90 patients who underwent PCI from June to July 2023 were selected as the validation set for further verification of the model.General information and clinical indicators of relevant factors were queried and collected through medical records and hospital HIS system.LASSO regression was used to reduce the dimensionality of the collected influencing factors,and unifactorial analysis was performed after the initial screening of unimportant factors,and multifactorial analysis was performed by combining the indicators with P<0.05 in unifactorial analysis with clinical application,and logistic regression analysis was used to further explore and establish the prediction model of the column line graph.The accuracy and stability of the nomogram were evaluated by calibration curve,receiver operating characteristic(ROC)curve,and decision curve analysis(DCA).Results Patients with diabetes mellitus,high cholesterol,surgery time greater than 45 minutes,and score of NRS numeric pain rating scale were influential factors affecting ACT below the normal range(P<0.05).A nomogram model was constructed using the above indicators,and its area under the curve for predicting intraoperative ACT monitoring values below the normal range during PCI was 0.726.The DCA curve thresholds of 0.17–0.63 were not intersected with the two extreme curves,and the net benefit of the model was high.Conclusion The prediction model of nomogram constructed in this study for ACT monito

关 键 词：经皮冠状动脉介入治疗 冠脉介入治疗 活性凝血时间 列线图 危险因素 预测模型 

分 类 号：R473.57[医药卫生—护理学]