[Qualitative and quantitative diagnostic performance of 320-slice computed tomography for detecting coronary artery disease with respect to atherosclerotic plaque characteristics]

Zhonghua Yi Xue Za Zhi. 2014 Oct 28;94(39):3071-4.
[Article in Chinese]

Abstract

Objective: To investigate qualitatively and quantitatively the diagnostic performance of 320-slice CT for detection of coronary artery disease with respect to different atherosclerotic plaque characteristics.

Methods: A retrospective search was performed for inpatients underwent both coronary CT and further coronary angiography (CAG) from December 1, 2008 to December 31, 2012. The diagnostic performance of 320-slice CTA for detecting significant stenosis ( ≥ 50% diameter) with respect to atherosclerotic plaque characteristics were analyzed by calculating sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), accuracy, kappa index (κ), and area under the receiver operating characteristic curve (AUC). Chi-square test was used to evaluate whether there were significant differences of the true-case frequency (true positive + true negative) and false-case frequency (false positive + false negative) among groups. Bland-Altman analysis was used to determine limits of agreement between CTA and CAG.

Results: A total of 454 patients and 6 779 segments were analyzed. Diagnostic accuracy was higher in non-calcified segments; whereas they decreased in the presence of both mild-moderately and heavily calcified plaques. Excellent agreement (κ = 0.810) between CT and CAG was observed for non-calcified segments, while good agreement was observed for both mild-moderately (κ = 0.701) and heavily calcified segments (κ = 0.750). Both mild-moderate (P = 0.000) and heavy (P = 0.000) calcification decreased the true-case frequency and increased the false-case frequency when compared to non-calcification. There were no significant underestimation or overestimation for non-calcified (P = 0.087) and mild-moderately calcified (P = 0.704) segments, while there was significant overestimation for heavily calcified segments (P = 0.001).

Conclusions: Great qualitative and quantitative diagnostic performances of 320-slice CT were observed in non-calcified coronary segments. However, qualitative diagnostic performance decreased in both mild-moderately and heavily calcified segments, and quantitative overestimation were observed in heavily calcified segments.

MeSH terms

  • Calcinosis
  • Coronary Angiography
  • Coronary Artery Disease*
  • Humans
  • Plaque, Atherosclerotic*
  • ROC Curve
  • Retrospective Studies
  • Tomography, X-Ray Computed