|
| 开发并验证慢性阻塞性肺疾病急性加重病毒感染风险预测模型 |
| Development and Validation of a Viral Infection Risk Prediction Model in Patients with Acute Exacerbation of Chronic Obstructive Pulmonary Disease |
| 投稿时间：2025-10-15 修订日期：2026-06-02 |
| DOI： |
| 中文关键词: 慢性阻塞性肺疾病急性加重，病原体靶向测序，病毒感染，预测模型，在线APP |
| 英文关键词: Acute exacerbation of chronic obstructive pulmonary disease (AECOPD) Pathogen-targeted sequencing Viral infection Predictive model Web application |
| 基金项目:湖南省卫健委科研课题(D202314019409) |
|
| 摘要点击次数: 74 |
| 全文下载次数: 0 |
| 中文摘要: |
| 目的：基于支气管肺泡灌洗液病原体宏基因组二代测序(Next Generation Sequencing, NGS)检测结果,建立慢性阻塞性肺疾病急性加重期(acute exacerbation of chronic obstructive pulmonary disease, AECOPD)患者病毒感染风险预测模型。方法：回顾性收集2024年9月1日-2025年2月28日本院呼吸与危重症医学科接受支气管镜检查并行BALF-NGS检测的AECOPD患者临床资料。通过随机抽样将数据按8:2比例分为训练集与验证集。采用R 4.4.0软件进行多因素Logistic回归分析确定病毒感染独立危险因素,构建列线图预测模型。通过混淆矩阵及相关指标评估了预测模型的性能,并开发基于Shiny框架的在线预测工具。结果：共纳入86例符合标准患者。多因素分析显示：慢性肾功能不全(OR=7.515, 95%CI:0.599-94.278, p=0.118)、肺部影像提示磨玻璃影(OR=6.028, 95%CI:1.528-23.775, p=0.010)、长期糖皮质激素使用(OR=5.916, 95%CI:1.612-21.717, p=0.007)及淋巴细胞比例升高(OR=1.105, 95%CI:1.022-1.196, p=0.012)与病毒感染显著相关。模型在训练集与验证集的曲线下面积分别为0.872(0.783-0.960)和0.827 (0.618-1.000)。基于研究成果开发的在线预测工具已部署于：https://polymyxinaki.shinyapps.io/dynnomapp/。结论：本研究建立的AECOPD病毒感染预测模型展现出良好的区分效能(AUC>0.8)与临床适用性,为早期识别病毒感染相关AECOPD提供了可靠的量化工具。 |
| 英文摘要: |
| Objective: To develop a viral infection risk prediction model for patients with acute exacerbation of chronic obstructive pulmonary disease (AECOPD) based on metagenomic next-generation sequencing (NGS) of bronchoalveolar lavage fluid (BALF). Methods: A retrospective cohort study was conducted on AECOPD patients who underwent bronchoscopy with BALF-NGS testing in the Department of Respiratory and Critical Care Medicine at our institution from September 1, 2024, to February 28, 2025. The cohort (n=86) was randomly divided into a training set (n=68) and a validation set (n=18) at an 8:2 ratio. Multivariable logistic regression analysis using R 4.4.0 identified independent predictors, followed by nomogram construction. Model performance was evaluated through confusion matrices and receiver operating characteristic (ROC) analysis. A Shiny-based web application was developed for clinical implementation. Results: Chronic kidney disease (OR=7.515, 95% CI:0.599-94.278, p=0.118), ground-glass opacities on imaging (OR=6.028, 95% CI:1.528-23.775, p=0.010), long-term glucocorticoid use (OR=5.916, 95% CI:1.612-21.717, p=0.007), and elevated lymphocyte percentage (OR=1.105, 95% CI:1.022-1.196, p=0.012) were independently associated with viral infections. The model demonstrated robust discrimination in both training (AUC=0.872, 95% CI:0.783-0.960) and validation cohorts (AUC=0.827, 95% CI:0.618-1.000). The online tool is publicly accessible at: https://polymyxinaki.shinyapps.io/dynnomapp/. Conclusion: This validated prediction model exhibits excellent discriminatory capacity (AUC>0.8) and clinical utility for early identification of viral-associated AECOPD. External validation is warranted to confirm generalizability. |
|
View Fulltext
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
