A risk predictive model for determining the severity of coronary artery lesions in older postmenopausal women with coronary heart disease.

Abstract

Objective: To determine the risk factors affecting the severity of coronary artery disease (CAD) in older postmenopausal women with coronary heart disease (CHD) and to construct a personalized risk predictive model.

Methods: In this cohort study, clinical records of 527 female patients aged ≥60 with CHD who were hospitalized in the First Affiliated Hospital of the University of Science and Technology of China from March 2018 to February 2019 were analyzed retrospectively. The severity of CAD was determined using the Gensini scores that are based on coronary angiography findings. Patients with Gensini scores ≥40 and <40 were divided into high-risk (n=277) and non-high-risk groups (n=250), respectively. Logistic regression analysis was used to assess independent predictors of CAD severity. The nomogram prediction model of CAD severity was plotted by the R software. The area under the receiver operating characteristic (ROC) and calibration curves were used to evaluate the predictive efficiency of the nomogram model, and the decision curve analysis (DCA) was used to assess the clinical applicability of the nomogram model.

Results: Multivariate analysis showed that high-sensitivity C-reactive protein, RBC count, WBC count, BMI, and diabetes mellitus were independent risk factors associated with CAD severity in older menopausal women (P<0.05); the area under the ROC curve of the nomogram constructed based on the independent risk factors was 0.846 (95% CI: 0.756-0.937). The area under the ROC curve after internal validation of the nomogram by the Bootstrap method after resampling 1000 times was 0.840 (95% CI: 0.741-0.923). The calibration curve suggested that the nomogram had an excellent predictive agreement, and the DCA curve indicated that the net benefit of applying the nomogram was significantly higher than that of the "no intervention" and "all intervention" methods when the risk probability of patients with high-risk CAD severity was 0.30-0.81.

Conclusion: A personalized risk assessment model was constructed based on the risk factors of severe CAD in older menopausal women with CHD, which had good prediction efficiency based on discrimination, calibration, and clinical applicability evaluation indicators. This model could assist cardiology medical staff in screening older menopausal women with CHD who are at a high risk of severe CAD to implement targeted interventions.