Nomogram for early major adverse event in infants after cardiac surgery: a retrospective study

Abstract

Background Early major adverse event has a high mortality rate after cardiac surgery. In this study, our goal was to investigate the risk factors associated with early MAE in infants after cardiac surgery, develop a prediction model, and assess its accuracy in predicting outcomes. Methods A model was constructed incorporating 766 patients at our Hospital from January 2020 to December 2021. Participants were randomly divided into modelling and validation group using a 7:3 ratio. We utilized the least absolute shrinkage and selection operator regression analysis to screen the variables, and then conducted a multiple logistic regression analysis to create a prediction nomogram. Results The risk factors of MAE were weight, aortic clamp time, postoperative 8th hour lactate, off-CPB blood glucose and postoperative 4 hours urine output. The Hosmer−Lemeshowtest demonstrated that the model was a good fit (χ²=6.105, p=0.636). The clinical decision curve analysis showed significantly better net benefit in the predictive model, as well as that in the validation cohort. Conclusion The prediction model based on perioperative factors was developed to screen the occurrence of early MAE in infants after cardiac surgery. It provided physicians with an effective tool for the early prediction, and took timely preventive measures.