Great debate: the new risk factor–weighted clinical likelihood model is useful to estimate the initial pre-test probability of obstructive coronary artery disease in individuals with suspected chronic coronary syndromes

For individuals with suspected chronic coronary syndrome, the 2024 ESC guidelines recommend use of a structured estimate of the probability of obstructive coronary artery disease (CAD). This ‘risk factor–weighted clinical likelihood’ (RF-CL) model is recommended as the initial step after history taking and combines age, sex, symptom characteristics, and five clinical risk factors, with coronary calcification data, if available. The resulting numerical estimate indicates an initial ‘pre-test probability’ of obstructive CAD, that has been calibrated to provide improved accuracy compared with previous models. It can help triage patients for appropriate testing and identify individuals with a very low likelihood of obstructive CAD, for whom deferral of further diagnostic tests should be considered. Designed to assess the likelihood of obstructive CAD, the RF-CL model is not designed to predict ischaemia, which may occur in the absence of obstructive coronary disease and could account for patient symptoms. The score is easy and quick to use, with extensive external validation in contemporary populations including European, North American, and Asian cohorts. However, some have questioned the practical application of the RF-CL tool, citing challenges with the specificity and clarity of patient symptoms, definition and weighting of risk factors, as well as the other ‘enrichment factors’ that can enhance the likelihood. The RF-CL model is quantitative up to 45% and then becomes semi-quantitative/qualitative. For patients considered very high likelihood, with an estimated score > 85%, invasive coronary angiography is recommended, although how this score may be reached is not entirely clear. The RF-CL model undoubtedly improves the prediction of obstructive CAD and can ‘de-risk’ a significant number of symptomatic patients safely, reducing unnecessary testing. In the development and application of such a probability estimate, there is a need to strike a good balance between simplicity and usefulness, vs increased sensitivity at the expense of greater complexity. Here, the two sides of this Great Debate are presented, to help the reader better evaluate the practical usefulness of the new RF-CL assessment in predicting the probability of obstructive CAD.