目测和定量冠状动脉造影分析软件评估冠状动脉狭窄的比较

Severity of coronary stenosis by physician′s visual assessment vs quantitative coronary angiography

摘要:

目的 比较术者目测与定量冠状动脉造影(QCA)分析软件评估冠状动脉狭窄程度的相关性及一致性。方法 回顾性分析2016年6月至2017年1月解放军总医院心血管内科冠状动脉CT造影(CCTA)检查后至少有1处病变≥50%并同期(<1周)行冠状动脉造影(CAG)检查的101例冠心病患者资料,根据每例患者CAG报告中所提及病变部位,应用2D-QCA 软件进行分析,比较术者目测和QCA软件评估冠状动脉狭窄是否一致。应用SPSS 17.0统计软件对数据进行处理。相关分析采用Pearson线性相关分析,一致性检验采用Kappa检验及Bland-Altman分析法。结果 101例患者共164支靶病变,其中左前降支病变50.6%(83/164),左回旋支病变18.9%(31/164),右冠状动脉病变30.5%(50/164)。Pearson线性相关分析结果表明目测与QCA呈正相关(r=0.745,P＜0.01),相关关系明显。Kappa一致性检验得出Kappa值为0.453,表明目测与QCA对冠状动脉狭窄程度的评估一致性一般。Bland-Altman分析表明4.9%(8/164)评估超过了一致性界限,一致性欠佳。结论 术者目测评估冠状动脉狭窄程度与QCA软件有较好的相关性但一致性欠佳。

Abstract:

Objective To investigate the correlation and the concordance of stenosis severity in patients with coronary artery disease (CAD) by physician′s visual assessment and quantitative coronary angiography (QCA). Methods Clinical data of 101 CAD patients with at least 1 lesion of ≥50% stenosis identified by coronary computed tomographic angiography (CCTA) and undergoing coronary angiography (CAG) at same period (<1 week) in our department from June 2016 to January 2017 were collected and retrospectively analyzed in this study. According to the lesion sites by CAG results, 2D-QCA software was employed to assess the coronary stenosis, and the results were analyzed for correlation with visual assessment. SPSS statistics 17.0 was used to analyze the data. Pearson linear correlation analysis was performed for correlation analysis, and Kappa test and Bland-Altman analysis for consistency test. ResultsThere were totally 164 target lesions in the 101 patients, accounting for 50.6%(83/164) in the left anterior descending branch, 18.9%(31/164) in the left circumflex branch, and 30.5%(50/164) in the right coronary artery. Pearson linear correlation analysis indicated significant positive correlation between the visual results and the QCA results (r=0.745, P<0.01). A weighted Kappa of 0.453 was found, suggesting not good enough concordance between the 2 measurements in assessment of stenosis. Bland-Altman analysis showed 4.9%(8/164) was beyond the concordance boundary, indicating unsatisfactory concordance. Conclusion There is good correlation in the assessment of coronary artery stenosis by visual assessment and QCA, but the concordance is unsatisfactory.