在院前环境下对曼彻斯特急性冠状动脉综合征心电图风险模型进行外部验证。
External validation of the Manchester Acute Coronary Syndromes ECG risk model within a pre-hospital setting.
机构信息
Division of Cardiovascular Sciences, The University of Manchester, Manchester, UK
College of Applied Medical Science, King Saud bin Abdulaziz University for Health Sciences, Riyadh, Saudi Arabia.
出版信息
Emerg Med J. 2023 Jun;40(6):431-436. doi: 10.1136/emermed-2022-212872. Epub 2023 Apr 17.
OBJECTIVES
The Manchester Acute Coronary Syndromes ECG (MACS-ECG) prediction model calculates a score based on objective ECG measurements to give the probability of a non-ST elevation myocardial infarction (NSTEMI). The model showed good performance in the emergency department (ED), but its accuracy in the pre-hospital setting is unknown. We aimed to externally validate MACS-ECG in the pre-hospital environment.
METHODS
We undertook a secondary analysis from the Pre-hospital Evaluation of Sensitive Troponin (PRESTO) study, a multi-centre prospective study to validate decision aids in the pre-hospital setting (26 February 2019 to 23 March 2020). Patients with chest pain where the treating paramedic suspected acute coronary syndrome were included. Paramedics collected demographic and historical data and interpreted ECGs contemporaneously (as 'normal' or 'abnormal'). After completing recruitment, we analysed ECGs to calculate the MACS-ECG score, using both a pre-defined threshold and a novel threshold that optimises sensitivity to differentiate AMI from non-AMI. This was compared with subjective ECG interpretation by paramedics. The diagnosis of AMI was adjudicated by two investigators based on serial troponin testing in hospital.
RESULTS
Of 691 participants, 87 had type 1 AMI and 687 had complete data for paramedic ECG interpretation. The MACS-ECG model had a C-index of 0.68 (95% CI: 0.61 to 0.75). At the pre-determined cut-off, MACS-ECG had 2.3% (95% CI: 0.3% to 8.1%) sensitivity, 99.5% (95% CI: 98.6% to 99.9%) specificity, 40.0% (95% CI: 10.2% to 79.3%) positive predictive value (PPV) and 87.6% (87.3% to 88.0%) negative predictive value (NPV). At the optimal threshold for sensitivity, MACS-ECG had 50.6% sensitivity (39.6% to 61.5%), 83.1% specificity (79.9% to 86.0%), 30.1% PPV (24.7% to 36.2%) and 92.1% NPV (90.4% to 93.5%). In comparison, paramedics had a sensitivity of 71.3% (95% CI: 60.8% to 80.5%) with 53.8% (95% CI: 53.8% to 61.8%) specificity, 19.7% (17.2% to 22.45%) PPV and 93.3% (90.8% to 95.1%) NPV.
CONCLUSION
Neither MACS-ECG nor paramedic ECG interpretation had a sufficiently high PPV or NPV to 'rule in' or 'rule out' NSTEMI alone.
目的
曼彻斯特急性冠状动脉综合征心电图（MACS-ECG）预测模型基于客观心电图测量值计算得分，以给出非 ST 段抬高型心肌梗死（NSTEMI）的概率。该模型在急诊科（ED）表现出良好的性能，但在院前环境中的准确性尚不清楚。我们旨在对 MACS-ECG 在院前环境中进行外部验证。
方法
我们对多中心前瞻性研究 Pre-hospital Evaluation of Sensitive Troponin（PRESTO）进行了二次分析，该研究旨在验证院前环境中决策辅助工具的准确性（2019 年 2 月 26 日至 2020 年 3 月 23 日）。研究纳入胸痛且急救护理人员怀疑急性冠状动脉综合征的患者。急救护理人员同时收集人口统计学和病史数据并解读心电图（“正常”或“异常”）。完成招募后，我们分析了心电图以计算 MACS-ECG 评分，使用了预先定义的阈值和优化以区分 AMI 与非 AMI 的灵敏度的新阈值。将其与急救护理人员的主观心电图解读进行比较。AMI 的诊断由两名研究人员根据入院后连续进行的肌钙蛋白检测进行裁定。
结果
在 691 名参与者中，有 87 名患有 1 型 AMI，687 名有完整的急救护理人员心电图解读数据。MACS-ECG 模型的 C 指数为 0.68（95%CI：0.61 至 0.75）。在预定的截止值下，MACS-ECG 的灵敏度为 2.3%（95%CI：0.3%至 8.1%），特异性为 99.5%（95%CI：98.6%至 99.9%），阳性预测值（PPV）为 40.0%（95%CI：10.2%至 79.3%），阴性预测值（NPV）为 87.6%（87.3%至 88.0%）。在灵敏度最佳的阈值下，MACS-ECG 的灵敏度为 50.6%（39.6%至 61.5%），特异性为 83.1%（79.9%至 86.0%），PPV 为 30.1%（24.7%至 36.2%），NPV 为 92.1%（90.4%至 93.5%）。相比之下，急救护理人员的灵敏度为 71.3%（95%CI：60.8%至 80.5%），特异性为 53.8%（95%CI：53.8%至 61.8%），PPV 为 19.7%（17.2%至 22.45%），NPV 为 93.3%（90.8%至 95.1%）。
结论
MACS-ECG 或急救护理人员心电图解读均未达到足够高的 PPV 或 NPV，无法单独“排除”或“诊断”NSTEMI。
Zotero 插件