Prediction models as gatekeepers for diagnostic testing in angina patients with suspected chronic coronary syndrome

Abstract

Aims: Assessment of pre-test probability (PTP) is an important gatekeeper when selecting patients for diagnostic testing for coronary artery disease (CAD). The 2019 European Society of Cardiology (ESC) guidelines recommend upgrading PTP based on clinical risk factors but provide no estimates of how these affect PTP. We aimed to validate two published PTP models in a contemporary low-CAD-prevalence cohort and compare with the ESC 2019 PTP. Methods and results: Previously published basic and clinical prediction models and the ESC 2019 PTP were validated in 42 328 patients (54% women) ≥30 years old without previous CAD referred for cardiac computed tomography angiography in a region of Denmark from 2008 to 2017. Obstructive CAD prevalence was 8.8%. The ESC 2019 PTP and basic model included angina symptoms, sex, and age, while the clinical model added diabetes mellitus family history of CAD, and dyslipidaemia. Discrimination was good for all three models [area under the receiver operating curve (AUC) 0.76, 95% confidence interval (CI) (0.75-0.77), 0.74 (0.73-0.75), and 0.76 (0.75-0.76), respectively]. Using the clinically relevant low predicted probability ≤5% of CAD cut-off, the clinical and basic models were well calibrated, whereas the ESC 2019 PTP overestimated CAD prevalence. At a cut-off of ≤5%, the clinical model ruled out 36.2% more patients than the ESC 2019 PTP, n = 23 592 (55.7%) vs. n = 8 245 (19.5%), while missing 824 (22.2%) vs. 132 (3.6%) cases of obstructive CAD. Conclusion: A prediction model for CAD including cardiovascular risk factors was successfully validated. Implementation of this model would reduce the need for diagnostic testing and serve as gatekeeper if accepting a watchful waiting strategy for one-fifth of the patients.