Background

Scarce data exist for a validated risk model to predict concomitant coronary artery disease (CAD) in patients undergoing valvular heart surgery.

Objectives

We aimed to identify risk factors and develop a model to predict coexistent CAD in these patients.

Materials and Methods

This retrospective cohort study included patients who underwent preoperative coronary angiography (CAG) before valvular heart surgery from January 2015 to December 2018 at two tertiary care hospitals. Data from one hospital were analyzed to develop the predictive risk score, which was validated using data from the other hospital. A receiver-operating-characteristic curve was constructed to evaluate the score diagnostic ability.

Results

Of 690 patients, the prevalence of coexistent CAD was 11.4%. According to multivariate analysis, risk factors significantly associated with CAD were age (OR 1.04; 95%CI: 1.00-1.08), typical angina pain (OR 2.67; 95%CI: 1.24-5.73), family history of premature CAD (OR 5.51; 95%CI: 1.30-23.20), dyslipidemia (OR 2.11; 95%CI: 1.14-3.91), and diabetes (OR 2.98; 95%CI: 1.49-5.95). Factors significantly lowering the CAD risk were rheumatic heart disease (OR 0.21; 95%CI: 0.10-0.41) and aortic valve lesions (OR 0.32; 95%CI: 0.13-0.75). The predictive score created from these variables yielded a c-statistic of 0.84 (95%CI: 0.79-0.88) in the development and 0.64 (95%CI: 0.46-0.82) in the validation cohorts. Furthermore, at a 61.75 cutoff, this simplified predictive score exhibited 70.9% sensitivity, 80.7% specificity, 32.2% positive predictive value, and 95.5% negative predictive value.

Conclusions

In patients undergoing valvular heart surgery, traditional CAD risk factors were associated with concomitant CAD, whereas aortic valve lesions and rheumatic heart disease were the protective factors. Our predictive risk score can identify low-risk patients with concomitant CAD, which may avoid unnecessary CAG.

Keywords: valvular heart disease, coronary artery disease, coronary angiography, predictive risk score.