Comparison of different cardiac risk scores for coronary artery disease in symptomatic women: do female-specific risk factors matter?

BACKGROUND

Cardiovascular diseases remain the leading cause of death in women and there is a need for more accurate risk assessment scores. The aims of our study were to compare the accuracy of several widely used cardiac risk assessment scores in predicting the likelihood of obstructive coronary artery disease (CAD) on CT coronary angiography (CTCA) in symptomatic women and to explore which female-specific risk factors were independent predictors of obstructive CAD on CTCA and whether adding these risk factors to pre-test probability scores would improve their predictive value.

METHODS AND RESULTS

Data were obtained from a cohort of 228 consecutively included symptomatic women undergoing evaluation for CAD and referred for CTCA. Obstructive CAD was defined as ≥50% luminal stenosis on CTCA. Pre-test probability for CAD was calculated according to the Diamond and Forrester score, New score, Duke clinical score, and an updated Diamond and Forrester score. Female-specific factors were obtained by a written questionnaire. Pre-test probability scores were compared with ROC analysis and showed that only the New score and the updated Diamond and Forrester score were significant predictive scores for obstructive CAD on CTCA (area under the curve, AUC, 0.67, p < 0.01; AUC 0.61, p = 0.04, respectively). Multivariable logistic regression analysis identified that gestational diabetes mellitus (GDM) and oestrogen status were independent predictors of obstructive CAD when adjusted for the pre-test probability scores. The updated Diamond and Forrester score was used for net reclassification improvement (NRI) analysis, since the New score already accounts for oestrogen status. Adding GDM and oestrogen status to the updated Diamond and Forrester score resulted in a significant NRI (p = 0.04).

CONCLUSIONS

There is a large variability in prediction of obstructive CAD using different pre-test probability risk scores in symptomatic women. Logistic regression analysis revealed that oestrogen status and GDM were independently associated with the occurrence of obstructive stenosis on CTCA. The predictive ability of cardiac pre-test probability scores improved significantly with the addition of oestrogen status and GDM.