高血压患者急性脑梗死发生风险预测模型的建立与评价  

Establishment and evaluation of risk forecast model for acute cerebral infarction in the patients with hypertension

作　　者：王秀青 周月琴[1] WANG Xiuqing;ZHOU Yueqin(Department of Clinical Laboratory,Shaoxing 7th People's Hospital in Zhejiang Province,Shaoxing312000,China)

机构地区：[1]浙江省绍兴市第七人民医院检验科,浙江绍兴312000

出　　处：《中国现代医生》2020年第32期129-132,共4页China Modern Doctor

基　　金：国家自然科学基金(11574210)。

摘　　要：目的建立高血压患者急性脑梗死发生风险的预测模型并评价。方法回顾性分析我院2016年1月~2018年12月的高血压患者224例,收集患者的临床资料、主要生化指标及急性脑梗死的发生情况,使用SPSS 25.0和R3.6.1软件及其软件包进行统计分析,建立预测模型并对模型进行评价。结果单因素分析表明,年龄、尿酸、血尿素氮、肌酐在急性脑梗死组与非急性脑梗死组比较,差异均有统计学意义(P<0.05);多因素分析表明,尿酸升高、血尿素氮升高为高血压患者急性脑梗死的独立危险因素。根据多因素Logistic回归模型筛选的变量建立风险列线图,评价分析示C-index为0.764,ROC曲线AUC为0.764,提示模型区分能力、模型准确性尚可。Calibration曲线、Decision曲线评价模型一致性及获益性尚可。结论本研究用列线图建立的高血压患者急性脑梗死发生风险预测模型,简单快速,临床应用准确性尚可,临床参考价值尚可,但本研究为单中心研究,样本量较少,需要更多多中心、大样本研究验证。Objective To establish and evaluate the forecast model for the risk of acute cerebral infarction(ACI)in the patients with hypertension.Methods From January 2016 to December 2018,a tatal of 224 patients with hypertension in our hospital were analyzed retrospectively.The clinical data and main biochemical indexes of the patients and the occurrence of ACI among them were collected.SPSS25.0 and R3.6.1 software and their software packages were used for statistical analysis,and to establish and evaluate the forecast model.Results Univariate analysis showed that the differences between the ACI group and the non-ACI group in age,uric acid,blood urea nitrogen and creatinine were statistically significant(P<0.05);multivariate analysis showed that elevated uric acid and blood urea nitrogen were independent risk factors for ACI in the patients with hypertension.According to the variables screened by multivariate Logistic regression model,the risk nomogram was established,and the evaluation analysis showed that the C-index was 0.764,and the AUC of ROC curve was 0.764,which suggested that the discrimination ability and accuracy of the model were acceptable.Calibration curve and Decision curve evaluated that the consistency and profitability of the model were acceptable.Conclusion The risk forecast model for the risk of ACI in the patients with hypertension established by the nomogram in this study,applied simply and quickly,and its clinical practice accuracy and clinical reference value are acceptable.However,this study is merely a single-center study with few samples.More multi-center studies with large samples are still necessary for further verification.

关 键 词：高血压 急性脑梗死 风险预测模型 列线图 尿酸 血尿素氮 

分 类 号：R544.1[医药卫生—心血管疾病] R743.3[医药卫生—内科学]