|
齐丽伟.老年急性肺栓塞预后不良预测模型的构建及其价值验证[J].内科急危重症杂志,2026,32(3):254-259
|
| DOI：10.11768/nkjwzzzz20260310 |
| 中文关键词: 老年急性肺栓塞 预后不良 影响因素 预测模型 |
| 英文关键词:Clinical features Elderly acute pulmonary embolism Poor prognosis Influencing factors Prediction model |
| 基金项目:河北省医学科学研究重点课题计划项目（20211674） |
|
| 摘要点击次数: 117 |
| 全文下载次数: 77 |
| 中文摘要: |
| 摘要 目的：探究老年急性肺栓塞预后不良的影响因素，构建预测模型并验证。方法：回顾性选择急性肺栓塞患者260例，按7:3的比例分为建模组（n=200）和验证组（n=60），根据随访3个月的临床转归将建模组分为预后不良组150例和预后良好组50例。收集所有患者一般资料、临床特征及实验室指标，Logistic回归分析独立危险因素；R语言软件4.0“rms”包构建预测模型，采用受试者工作特征（ROC）曲线分析模型预测价值。结果：预后不良组与预后良好组在年龄、急性心肌梗死、窦性心动过速、ST段改变、SⅠQⅢTⅢ、肺动脉高压、单纯低分子肝素(LMWH)+华法林、LMWH+非维生素κ拮抗剂口服抗凝药（NOACs）占比、白细胞计数、血红蛋白、中性粒细胞与淋巴细胞比值、红细胞分布宽度（RDW）、B型脑钠肽、心肌肌钙蛋白-I、降钙素原水平方面存在差异（P均＜0.05）。二分类Logistic分析显示， LMWH+华法林治疗、LMWH+NOACs治疗为老年急性肺栓塞预后的独立保护因素，RDW、SⅠQⅢTⅢ、右心室功能不全为独立危险因素（P均＜0.05）。建模组与验证组的C-index分别为0.785（95%CI：0.674~0.863）、0.801（95%CI：0.702～0.830），ROC分析的曲线下面积分别为0.868（95%CI:0.786~0.950）、0.860（95%CI：0.765～0.955），表明预测模型的准确度良好，校准曲线显示预测模型校准度优良，DCA显示预测模型具有极高的临床效益，适用性较强。结论：基于临床特征和实验室检测指标构建的老年急性肺栓塞预后不良预测模型具有较好的预测准确性和临床实用性，有望为临床提供决策支持，提高疗效。 |
| 英文摘要: |
| Abstract Objective: To explore the influencing factors of poor prognosis in elderly patients with acute pulmonary embolism (APE), and to construct and validate a corresponding predictive model. Methods: A retrospective analysis was conducted on the case data of APE patients. The patients were divided into a modeling group (n= 200) and a validation group (n= 60) in a 7:3 ratio. Based on the clinical outcomes at a 3-month follow-up, the modeling group was further divided into a poor prognosis subgroup (n= 150) and a good prognosis subgroup (n= 50). General information, clinical characteristics, and laboratory indicators of all patients were collected. Independent risk factors were analyzed using logistic regression. A predictive model was constructed using the "rms" package in R software (version 4.0), and its predictive value was analyzed using the receiver operating characteristic (ROC) curve.Results: Significant differences (all P< 0.05) were observed between the two groups in terms of age, acute myocardial infarction, sinus tachycardia, ST-segment changes, SⅠQⅢTⅢ pattern, pulmonary hypertension, the proportion of patients receiving LMWH + warfarin or LMWH + NOACs treatment, white blood cell count, hemoglobin, neutrophil-to-lymphocyte ratio, red cell distribution width (RDW), B-type natriuretic peptide, cardiac troponin-I, and procalcitonin levels. Binary logistic analysis showed that treatment with LMWH + warfarin and LMWH + NOACs were independent protective factors against poor prognosis in elderly APE patients, while RDW, SⅠQⅢTⅢ pattern, and right ventricular dysfunction were independent risk factors (all P< 0.05) . The C-index values for the modeling group and validation group were 0.785 (95% CI: 0.674-0. 863) and 0.801 (95% CI: 0.702-0.830), respectively. The areas under the ROC curve (AUC) were 0.868 (95% CI: 0.786-0.950) and 0.860 (95% CI: 0.765-0.955), respectively, indicating good accuracy of the prediction model. The calibration curve showed excellent calibration, and the decision curve analysis (DCA) demonstrated that the model had high clinical benefit and strong applicability. Conclusion: The predictive model for poor prognosis in elderly APE patients, constructed based on clinical characteristics and laboratory indicators, demonstrates good predictive accuracy and clinical utility. It is expected to provide effective decision support for clinicians and improve treatment outcomes for elderly APE patients. |
|
|
|
|
