A novel clinical risk prediction model for myocardial infarction, coronary revascularization, and cardiac death according to clinical, exercise, and gated SPECT variables (VH-RS).

AIMS

To create a risk score for cardiac events (CE) according to clinical, exercise, and gated SPECT variables.

METHODS AND RESULTS

We analysed 5707 consecutive patients; 3181 patients (age 64.2 ± 11 years, male 59.6%) with suspected coronary artery disease (CAD) [without previous myocardial infarction (MI) or coronary revascularization (CR)] and 2526 patients (age 63.3 ± 11 years, male 81.7%) with established CAD (with previous MI or CR). To create the Vall d'Hebron Risk Score (VH-RS), first we analyse the predictors of CE (non-fatal MI, CR, and/or cardiac death), then the probability of CE for every patient according to the predictive variables. According to risk we stratified patients into four risk levels: very low risk (VLR), low risk (LR), moderate risk (MR), and high risk (HRi) using Multiple Cox Regression analysis models. Finally, we validate the VH-RS in another prospective cohort of 734 patients. In patients with suspected CAD; age (P < 0.001); gender (P = 0.001); hyperlipidaemia (P < 0.001); nitrates (P = 0.04); ejection fraction (EF) (P = 0.001); summed stress score (P < 0.001); METs (P < 0.001); exercise angina (P = 0.006); and mm of ST segment depression (P = 0.004) were the independent predictors of CE (C-statistic: 0.8; P < 0.001). In patients with established CAD, EF (P < 0.001); summed difference score (P = 0.001); age (P < 0.001); smoker (P = 0.002); nitrates (P = 0.003); exercise angina (P = 0.001); METs (P < 0.001); and mm of ST segment depression (P = 0.011) were the independent predictors of CE (C-statistic: 0.7; P < 0.001). The risk score obtained from these variables allows the stratification of patients into four risk levels: VLR, LR, MR, and HRi.

CONCLUSIONS

The cardiac risk stratification by mean of clinical, exercise, and gated SPECT variables is an objective aid to assessing an individual's cardiac risk.