External validation of the CRAX2MACE model in an Italian cohort of patients with suspected coronary artery disease undergoing stress myocardial perfusion imaging

Background Prevention and development of diagnostic and therapeutic techniques reduced morbidity and mortality for coronary artery disease (CAD). In this context, the cardiovascular risk assessment for major adverse cardiac events (MACE) at 2-year (CRAX2MACE) model for prediction of 2-year major adverse cardiac events was developed. We performed an external validation of this model. Methods We included 1003 patients with suspected CAD undergoing stress-rest single-photon emission computed tomography myocardial perfusion imaging at our academic center between March 2015 and April 2019. Results Considering the occurrence of MACE (death from any cause, acute myocardial infarction, or late coronary revascularization), for the CRAX2MACE model the area under the receiver operating characteristic curve was 0.612 and the Brier score was 0.061. The Hosmer-Lemeshow test estimated a non-optimal fit (chi(2) 28, P < .001). Considering only hard events (cardiac death, acute myocardial infarction), the external validation of the CRAX2MACE model revealed a Brier score of 0.053 and an area under the receiver operating characteristic curve of 0.621. Hosmer-Lemeshow test was calculated by deciles and showed a poor fit (chi(2) 31, P < .001). Conclusion CRAX2MACE model had a limited value for predicting 2-year major adverse cardiovascular events in an external validation cohort of patients with suspected CAD.

Automated summary

The external validation of the CRAX2MACE model revealed limited value in predicting 2-year major adverse cardiovascular events for suspected CAD patients, showing poor fit and accuracy especially in predicting hard events.