重症脑出血患者肠内营养支持发生误吸风险预测模型的构建及验证

Construction and validation of risk prediction model for aspiration in patients with severe intracerebral hemorrhage receiving enteral nutrition support

目的:分析重症脑出血患者肠内营养支持发生误吸的危险因素,初步构建并验证其风险预测列线图模型。方法:采用回顾性队列研究法,于2018年1月至2021年3月抽取某三级甲等医院行肠内营养支持的重症脑出血患者为研究对象,依据是否发生误吸分为非误吸组(n=221)与误吸组(n=147),采用Logistic回归分析误吸的危险因素,应用R软件构建风险预测列线图模型。于2021年4月—6月,选取同一医院重症脑出血且行肠内营养支持的53例患者对模型施以验证。结果:Logistic回归分析筛选出鼻饲管管径(OR=0.080)、胃残留量(OR=1.710)、误吸史(OR=0.356)、美国国立卫生研究脑卒中量表(NIHSS)评分(OR=1.257)与洼田饮水试验等级(OR=2.499)为误吸的独立危险因素(P均<0.05)。内部验证显示,ROC曲线下面积为0.955,约登指数最大值所对应的风险预测值为0.234,且校准曲线斜率接近于1。外部验证显示,ROC曲线下面积为0.891,校准曲线在理想曲线附近。结论:本研究所构建的列线图模型区分度与准确度较高,可为临床个体化预测行肠内营养支持的重症脑出血患者误吸发生风险提供参考依据。

Objective:To analyze the risk factors of aspiration in patients with severe intracerebral hemorrhage receiving enteral nutrition support,and preliminary develop and verify its risk prediction graph model.Methods:Retrospective cohort study method was used to systematically select patients with severe cerebral hemorrhage who received enteral nutrition support from a tertiary grade A hospital in Liaoning province from January 2018 to March 2021.According to whether aspiration occurred or not,they were divided into non-aspiration group(n=221)and aspiration group(n=147).Logistic regression was used to analyze the risk factors of aspiration,and R software was used to build a risk prediction nomogram model.From April 2021 to June 2021,53 patients with severe cerebral hemorrhage who received enteral nutrition support in the same hospital were selected to validate the model.Results:Logistic regression analysis screened the nasogastric feeding tube diameter(OR=0.080),gastric residual volume(OR=1.710),history of aspiration(OR=0.356),National Institute of Stroke Scale score(OR=1.257)and low-water drinking test(OR=2.499)were the risk factors for aspiration(P<0.05).Internal validation showed that the area under the ROC curve was 0.955,the risk prediction value corresponding to the maximum value of the Youden index was 0.234,and the slope of the calibration curve was close to 1.External validation showed that the area under ROC curve was 0.891,calibration curve was near ideal curve.Conclusion:The differentiation and accuracy of the constructed linear graph model are high,which can provide a reference for predicting the risk of aspiration in patients with severe intracerebral hemorrhage receiving enteral nutrition support.