Is there a difference in the diagnostic accuracy of computed tomography coronary angiography between women and men?
OBJECTIVE: To assess the influence of sex on the diagnostic performance of computed tomography coronary angiography (CTCA). METHODS: A total of 916 symptomatic patients (30.5% women) without earlier history of coronary artery intervention underwent both CTCA and invasive coronary angiography. Descriptive diagnostic parameters, to detect obstructive coronary artery disease (CAD; ≥50% lumen diameter narrowing) on CTCA, were compared between women and men on a per-patient, per-vessel, and per-segment level. Adjusted values were calculated for clustered segments and differences in sex variables using logistic multivariate regression models in general estimated equations. RESULTS: Women were older, had less typical chest complaints, and had a lower prevalence, extent, and severity of CAD compared with men. Multivariate analysis on a per-patient level revealed no difference in sensitivity (98 vs. 99%, P=0.15), specificity (78 vs. 82%, P=0.65), positive predictive value (PPV; 87 vs. 95%, P=0.10), negative predictive value (NPV; 97 vs. 98%, P=0.63), and diagnostic odds ratio (DOR; 198 vs. 721, P=0.07). No difference was found on per-vessel level analysis (sensitivity 95 vs. 97%, P=0.14; specificity 89 vs. 87%, P=0.93; PPV 73 vs. 79%, P=0.06; NPV 98 vs. 98%, P=0.72; and DOR 143 vs. 240, P=0.08). Per-segment analysis revealed a lower sensitivity (88 vs. 94%, P<0.001) and DOR (163 vs. 302, P=0.002) in women compared with men, without a difference in specificity (96 vs. 95%, P=0.19), PPV (64 vs. 69%, P=0.07), and NPV (99 vs. 99%, P=0.08). CONCLUSION: CTCA can accurately rule out obstructive CAD in both women and men. CTCA is less accurate in women to detect individual obstructive disease.
- diagnostic accuracy
- coronary artery disease
- multislice computed tomography
- multivariate analyses