急诊科环境下冠状动脉CT血管成像的有限轴位解读
本研究纳入74例急诊胸痛患者的冠状动脉CT血管成像(CCTA),由两名放射科医生分别以仅轴位图像(有限解读,LI)和联合轴位与多平面重组图像(完整解读,FI)评估显著狭窄(≥50%)。结果显示LI在患者层面的准确率非劣效于FI(90.5% vs 91.9%,差值-1.4%),结论提示LI可能足以用于急诊检出显著冠状动脉疾病。
要点
- 急诊胸痛患者CCTA仅用轴位图像(LI)诊断显著狭窄的准确率非劣效于联合多平面重组(FI)。
- LI与FI的阅片者间一致性均良好(κ=0.72 vs 0.70,P=.74)。
- 患者层面LI准确率90.5% vs FI 91.9%,差值-1.4%,置信区间未包含非劣效界值-10%。
核心结论
- 在急诊环境中,仅凭轴位CCTA图像(LI)检测显著冠状动脉狭窄可能已足够。
研究设计与样本
- 研究纳入74例急诊胸痛患者的CCTA检查,由两名放射科医生(一名无专门CCTA培训,一名有基础CCTA经验)在独立 sessions 中随机顺序各评估三次(LI一次,FI两次),对19个冠状动脉节段评定是否存在显著狭窄(≥50%)。
诊断标准与关键数值
- 显著狭窄定义为冠状动脉节段狭窄≥50%;非劣效界值为-10%;患者层面准确率LI 90.5% vs FI 91.9%,差值-1.4%;阅片者间一致性κ=0.72(LI)vs 0.70(FI),P=.74。
临床应用与阅片要点
- 作者结论提示,在急诊环境下仅解读轴位CCTA图像(LI)可能足以检测显著冠状动脉疾病。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
PURPOSE: Incorporating coronary CT angiographic (CCTA) imaging into emergency department (ED) workflows has been limited by the need for 24/7 real-time postprocessing. The aim of this study was to determine whether interpretation of transaxial CCTA images alone (limited axial interpretation [LI]) is noninferior to interpretation of combined transaxial and multiplanar reformation images (full interpretation [FI]) in assessing patients with acute chest pain in the ED.
METHODS: CCTA examinations from 74 patients were evaluated by two radiologists, one without dedicated CCTA training and one with basic CCTA experience. Each examination was evaluated three times in separate sessions, once by LI and twice by FI, in random order. Nineteen coronary artery segments were rated as having significant stenoses (≥50%) or not. Interreader agreement was assessed using Cohen's κ statistic. The primary analysis was whether the accuracy of LI for detecting significant stenosis was noninferior to that of FI at the patient level (margin = -10%). Secondary analyses included similar analyses of sensitivity and specificity, at both the patient and vessel levels.
RESULTS: Interreader agreement for significant stenosis was good for both LI and FI (κ = 0.72 vs 0.70, P = .74). Average accuracy for significant stenosis at the patient level was 90.5% for LI and 91.9% for FI, with a difference of -1.4%. The accuracy of LI was noninferior to FI, because the confidence interval did not include the noninferiority margin. Noninferiority was also found for patient-level sensitivity and for accuracy, sensitivity, and specificity at the vessel level.
CONCLUSIONS: LI of the coronary arteries using transaxial CCTA images may be sufficient for the detection of significant coronary artery disease in the ED setting.
引用
- Jehangir, Hippe, Huang, et al. Limited Axial Interpretation of Coronary CT Angiography in the Emergency Department Setting. J Am Coll Radiol. 2023. PMID: 37201689. doi:10.1016/j.jacr.2023.04.005
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。