Air pollution and COVID-19 infection affect the pathogenesis of cardiovascular disease. The impact of these factors on the course of ACS treatment is not well defined. The purpose of this study was to evaluate the effects of air pollution, COVID-19 infection, and selected clinical factors on the occurrence of perioperative death in patients with acute coronary syndrome (ACS) by developing a neural network model. This retrospective study included 53,076 patients with ACS from the ORPKI registry (National Registry of Invasive Cardiology Procedures) including 2395 COVID-19 (+) patients and 34,547 COVID-19 (-) patients. The neural network model developed included 57 variables, had high performance in predicting perioperative patient death, and had an error risk of 0.03%. Based on the analysis of the effect of permutation on the variable, the variables with the greatest impact on the prediction of perioperative death were identified to be vascular access, critical stenosis of the left main coronary artery (LMCA) or left anterior descending coronary artery (LAD). Air pollutants and COVID-19 had weaker effects on end-point prediction. The neural network model developed has high performance in predicting the occurrence of perioperative death. Although COVID-19 and air pollutants affect the prediction of perioperative death, the key predictors remain vascular access and critical LMCA or LAD stenosis.
Keywords: COVID-19 infection; acute coronary syndrome; air pollution; critical left anterior descending coronary artery (LAD); left main coronary artery (LMCA) stenosis; neural network.