基于临床资料和血清指标构建脑出血合并卒中相关性肺炎的预测模型构建及验证

Construction of a predictive nomogram for stroke-associated pneumonia in patients with intracerebral hemorrhage using clinical data and serum markers and its validation

目的研究脑出血(ICH)患者合并卒中相关性肺炎(SAP)的危险因素,并构建Nomogram预测模型,预测ICH患者合并SAP的风险。方法选取2017年12月至2022年6月321例ICH患者合并SAP的情况进行调查,根据是否发生SAP分为SAP组和对照组,调查2组临床资料及中性粒细胞/淋巴细胞比值(NLR)、血小板/淋巴细胞比值(PLR)水平,以是否合并SAP为因变量,采用Lasso回归和Logistic回归筛选ICH患者合并SAP的独立危险因素,使用R语言构建Nomogram预测模型,并对构建的模型进行评估。结果321例ICH患者中57例发生SAP,Lasso回归和Logistic回归筛选出:年龄、肺部基础疾病、糖尿病、吞咽困难、NLR、PLR和气管插管是ICH患者合并SAP的危险因素(P<0.05),基于上述因素构建Nomogram预测模型,对于构建Nomogram预测模型进行评估显示,C-index值为0.867,95%CL(0.856~0.945),模型区分度良好:AUC为0.815,95%CL(0.854~0.945),模型准确性良好,校准曲线证明模型预测能力尚可,临床决策曲线显示:当MCI概率阈值>11%以及<85%使用此模型可以获得较高净获益。结论本研究构建的ICH患者合并SAP的Nomogram预测模型,具有良好区分度和精准度。

Objective To investigate the risk factors for stroke-associated pneumonia(SAP)in patients with intracerebral hemorrhage(ICH)and to develop a nomogram to predict them.Methods A survey was conducted involving 321 cases of ICH patients with SAP.They were categorized into the SAP group and the control group based on the presence or absence of SAP.Clinical data,neutrophil-to-lymphocyte ratio(NLR)and platelet-to-lymphocyte ratio(PLR)levels were collected.With SAP occurrence as the dependent variable,independent risk factors for SAP in ICH patients were identified using Lasso regression and Logistic regression.A nomogram was constructed using the R package,and validated for its performance.Results Among the 321 ICH patients,57 developed SAP.Lasso regression and Logistic regression identified that age,underlying lung disease,diabetes,dysphagia,NLR,PLR and tracheal intubation were independent risk factors for SAP in ICH patients.A nomogram was constructed based on these factors.The C-index value of the nomogram was 0.867(95%CI:0.856-0.945),demonstrating an excellent discriminative ability.The area under the curve(AUC)was 0.815(95%CI:0.854-0.945),indicating a high level of model accuracy.The calibration curve demonstrated that the nomogram possessed a reasonably reliable predictive capability.Clinical decision curve analysis yielded a higher net benefit of the nomogram when the MCI probability threshold was above 11%and below 85%.Conclusion The nomogram for predicting SAP in ICH patients exhibits outstanding discrimination and precision.