有创冠状动脉造影和CT冠状动脉成像对诊断冠心病的可重复性研究

Reproducibility of invasive coronary angiography and CT coronary angiography in the diagnosis of coronary heart disease

目的探讨不同医师及同一医师前后两次采用有创冠状动脉造影（ICA）和CT冠状动脉成像（CTCA）诊断冠心病（CAD）的可重复性。方法选取2011年11月-2013年1月在北京天坛医院放射科同时完成ICA及CTCA检查，且符合入选标准的患者100例。由2名经验丰富的影像科医师按照盲法原则独立分析这100例患者的ICA影像资料，30d后，其中1名按照盲法原则重复分析；另2名CTCA医师按照同样的方法分析CTCA影像资料；最后分别核对2名ICA及CTCA医师的诊断结果并协商形成一致的看法作为最后诊断结果。比较ICA及CTCA医师间、同一医师两种方法诊断CAD的Kappa值，若K值=0．65。0．75为一致性好，K值〉0．75为一致性极好。结果ICA及CTCA医师间诊断CAD的K值分别为0．89和0．79，95％CI分别为0．85-0．93和0．73—0．85，χ2值分别为504．21和404．13；同一医师诊断CAD的K值分别为0．90和0．81，95％CI分别为0．86～0．94和0．75～0．87，χ2值分别为505．42和431．25；K值均〉0．75。ICA同一医师诊断冠状动脉狭窄程度一致的冠状动脉节段数占总节段数的比率为86．86％，高于医师间的80．81％，两者比较，差异有统计学意义（χ=8．286，P〈0．05）；诊断CAD一致的冠状动脉节段数占总节段数的比率为97．32％，高于医师间的96．49％，但两者比较，差异无统计学意义（χ2=0．719，P〉0．05）。CTCA同一医师诊断冠状动脉狭窄程度一致的冠状动脉节段数占总节段数的比率为81．89％，高于医师间的75．15％，两者差异有统计学意义（χ2=8．375，P〈0．05）；诊断CAD一致的冠状动脉节段数占总节段数的比率为95．29％，高于医师间的94．34％，但两者差异无统计学意义（χ2=0．587，P〉0．05）。结论ICA和CTCA诊断CAD的可重复性极好，ICA法优于CTCA法，不同医师和同一医师前后两次的诊断会出现一定的误差。

Objective To explore the intra-and inter-observer repeatability of the diagnosis of coronary artery disease （CAD） based on invasive coronary angiography （ICA） and CT coronary angiography （CTCA）. Methods 100 patients were selected from November 2011 to January 2013 in Beijing Tiantan Hospital Affiliated to Capital Medical University, who accepted both ICA and CTCA check, and met the inclusion criteria. The ICA image data of 100 patients with a blind principle were inde- pendently analyzed by two readers with comparable experience, after 30 days, one of them reviewed the same patients again, the CTCA image datas from the same 100 patients with the same method were by another two CTCA physicians, finally, the two ICA and CTCA physician diagnosis results and consulted to the formation of the same views as the final diagnostic resuhs were checked. The inter-reader and intra-reader repeatability of ICA and CTCA were analyzed by performing Kappa teat, if the value of K = 0.65-0.75 for consistency, K values 〉 0.75 for consistency excellent. Results ICA and CTCA inter-reader K values of CAD diagnosis was 0.89 and 0.79, 95%CI was 0.85-0.93 and 0.73-0.85, χ2 values was 504.21 and 404.13, respec- tively; the intra-reader K values of CAD diagnosis were 0.90 and 0.81, 95%CI was 0.86-0.94 and 0.75-0.87, χ2 values was 505.42 and 431.25, respectively; K value 〉 0.75. The percentage of the segments with agreement within the same readers on the degree of stenosis was 86.86% in ICA, which was higher than 80.81% between readers, the difference was significant （χ2= 8.286, P 〈 0.05）; the percentage of the segments with agreement within the same readers on the CAD was 97.32%, which was higher than 96.49% between readers, but the difference was not significant （χ2=0.719, P 〉 0.05）. The percentage of the seg- ments with agreement within the same readers on the degree of stenosis was 81.89% in CTCA, which was higher than 75.15% between readers, the difference was significant （χ2=8.375, P 〈 0.05）; the percentage of the segments with agreement within the same readers on the CAD was 95.29%, which was higher than 94.34% between readers, but the difference was not significant （χ2=0.587, P 〉 0.05）. Conclusion Both ICA and CTCA have excellent repeatability of CAD diagnosis, ICA method is superior to the CTCA method, a certain discrepancy exists in two readings from the different readers or the same reader.