基于急诊临床资料预测急性缺血性轻型卒中患者短期预后的列线图分析

A Nomogram Analysis of Emergency Clinical Data for Predicting Short-term Prognosis of Patients with Acute Ischemic Stroke

目的:建立预测轻型卒中患者发病30d内不良预后的列线图风险模型。方法:回顾性分析2018年10月—2020年8月期间200例发病72h内的轻型缺血性卒中住院患者的急诊临床资料,随访时间为30d,筛选出对预后有意义的危险因素,并构建列线图风险模型。结果:既往有糖尿病史、入院时NIHSS评分、舒张压≥110mmHg、超敏C反应蛋白≥3mg/L是轻型卒中患者30d内不良预后的危险因素,构成的列线图模型的ROC曲线下面积AUC值为0.789(95%CI:0.700~0.877),模型校准图预测曲线与标准曲线拟合较好,差异无统计学意义。结论:基于急性缺血性轻型卒中患者急诊临床资料(糖尿病史、NIHSS评分、舒张压、超敏C反应蛋白)建立的列线图模型具有良好的分区度和校准度,对早期识别轻型卒中不良预后具有一定临床应用价值。

Objective:To establish a nomogram risk model for predicting the poor prognosis of patients with minor stroke within 30 days of onset.Methods:The emergency clinical data of 200 hospitalized patients with minor ischemic stroke within 72 hours of onset from October 2018 to August 2020 were retrospectively analyzed.The follow-up time was 30 days.The meaningful risk factors to the prognosis were screened out and constructed the nomogram risk model.Results:Diabetes,NIHSS score at admission,diastolic blood pressure≥110mmHg,and high-sensitivity C-reactive protein≥3mg/L were risk factors for poor prognosis in patients with minor stroke within 30 days,and the AUC value of the nomogram model was 0.789(95%CI:0.700~0.877),the prediction curve of the model calibration chart fit well with the standard curve,and the difference was not statistically significant.Conclusion:The nomogram model established based on the emergency clinical data(diabetes history,NIHSS score,diastolic blood pressure,high-sensitivity C-reactive protein)of patients with acute ischemic stroke has good discrimination and calibration.It has certain clinical application value for early identification of poor prognosis of patients with minor stroke.