ÁP DỤNG BẢNG PHÂN LOẠI CAD-RADS TRONG ĐÁNH GIÁ BỆNH ĐỘNG MẠCH VÀNH MẠN TÍNH TRÊN CẮT LỚP VI TÍNH ĐA DÃY

SUMMARY

Purpose: To apply the CAD-RADS classification in the assessment of chronic coronary artery disease on multislice computed tomography.

Material and methods: Cross-sectional description of 179 patients undergoing coronary computed tomography angiography, diagnosed according to CAD-RADS classification by two physicians independently, the intra-class correlation was used to test the inter-reviewer agreement (IRA), and comparing with results invasive coronary angiography.

Results: There was an excellent IRA between the two for CADRADS (κ=0.904), by each CAD-RADS classification (κ=0.827-1.00), by coronary artery stenosis (κ=0.878-0.931). There is an excellent IRA for modifiers G (κ=1), and S (κ=1), moderate IRA for V (κ=0.574). The best cutoff value for predicting significant CAD was ≥ CAD-RADS 3. The diagnostic value of CAD-RADS classification according to the common segment: sensitivity 77.54%, specificity 87.23%, positive predictive value 89.92%, negative predictive value 72.56%. Conclusion: There is an excellent inter-reviewer agreement when applying the CAD-RADS classification to clinical practice, with high sensitivity, specificity, and accuracy when compared with invasive coronary angiography. Keywords: coronary artery, CAD-RADS, CT angiography, graft, stent