A systematic screening and heart team approach contributes to unravel novel risk factors in revascularisation candidates of complex coronary artery disease: a machine learning approach

Abstract

In randomized control trials, established baseline risk factors for mortality of patients with 3VD with/without LM disease candidate for mechanical revascularization may diverge from risks encountered in daily clinical practice. By creating a specific linear model based on Cox regression after LASSO selection of new risk factors specific to the registry, we considerably improved the accuracy of the probabilistic formula in internal validation. In the present registry that included a population of all comers amenable to either surgical or percutaneous revascularization, systematic screening by Heart Team and machine learning approach enabled the detection of specific risks not necessarily collected in randomized control trials with numerous restrictive inclusion and exclusion criteria.