恶性肿瘤患者多重耐药菌医院感染风险预测模型的开发及验证

Development and validation of a prediction model for the risk of hospital-acquired multidrug-resistant bacteri⁃al infections in patients with malignant tumors

目的建立并验证恶性肿瘤患者多重耐药菌(MDRO)医院感染风险的预测模型。方法回顾性分析2017年10月至2023年4月期间在温州市人民医院接受恶性肿瘤治疗的患者数据。采用最小绝对收缩和选择算子(LASSO)回归模型进行MDRO的预测变量筛选。基于这些因子,应用多因素logistic回归分析构建了列线图预测模型。使用C指数和校准曲线评估预测模型的区分度和一致性。结果LASSO回归筛选出由年龄、经皮胸膜引流管、营养风险筛查(NRS)-2002评分和抗菌药物使用天数等4个预测因子构成的列线图预测模型。该模型在建模组和验证组的C指数分别为0.74和0.77,表明该模型具有较高的区分能力。校准图显示,该模型预测的概率与实际概率之间具有良好的一致性。结论本研究建立的列线图模型是一个具有临床应用价值的个体化预测模型,可帮助医务人员早期识别恶性肿瘤MDRO医院感染的高危患者。

Objective To develop and validate a nomogram model to predict the risk of hospital-acquired infection by multidrug-resistant organisms(MDRO)in patients with malignant tumors.Methods This study retrospectively analyzed the data of patients who received malignant tumor treatment in Wenzhou People’s Hospital from October,2017 to April,2023.The least absolute shrinkage and selection operator(LASSO)regression model was used to select variables and de⁃termine the best predictive factors included in the nomogram.Based on these factors,a nomogram prediction model was constructed using multivariate logistic regression analysis.The C-index and calibration curve were used to evaluate the discrimination and calibration of the prediction model.Results The LASSO regression selected four predictive factors to construct the nomogram prediction model including age,percutaneous pleural drainage tube placement,NRS-2002 score,and days of antimicrobial drug use.The C-index of the development cohort and 0.77 in the validation cohort was 0.74 and 0.77,respctively,indicating high discriminative ability.Calibration plots demonstrated good consistency between pre⁃dicted probabilities and actual probabilities.Conclusion The nomogram model established in this study is a clinically valuable individualized prediction model,which can help to identify high-risk patients of hospital-acquired infection by MDRO in patients with malignant tumors.