Patient Characteristics and Surgical Variables associated with Intraoperative Reduced Right Ventricular Function

Abstract Objective Perioperative right ventricular (RV) function is a significant predictor of patient outcomes after cardiac surgery. This prospective study aimed to identify perioperative factors associated with reduced intraoperative RV function. Methods RV function was assessed at the beginning and end of surgery by standardized transesophageal echocardiographic measurements including tricuspid anular plane systolic excursion (TAPSE), peak systolic longitudinal RV strain, and fractional area change (FAC) in 109 adult patients undergoing cardiac surgery at Cleveland Clinic. Associations between RV function and 33 patient characteristics and perioperative factors were analyzed by random forest machine learning. The relative importance of each variable in predicting RV function at end of surgery was determined. Results Longer aortic clamp duration and lower baseline RV function were highly important variables for predicting worse RV function measured by TAPSE, RV strain, and FAC at end of surgery. For example, RV function after longer aortic clamp times of 100-120 min was worse (median[Q1,Q3] TAPSE 0.97[0.89, 1.09]cm) compared with RV function after shorter aortic clamp times of 50-70 minutes (TAPSE 1.49[1.3, 1.70]cm; p=0.001). RV strain at end of surgery was reduced in patients with worse baseline RV function compared with those with higher baseline RV function (end of surgery RV strain in lowest quartile -13.7[-16.6, -12.4]% vs. highest quartile -17.7[-18.6, -15.3]% of baseline RV function; p=0.043). Conclusion Intraoperative decline in RV function is associated with longer aortic clamp time and worse baseline RV function. Efforts to optimize these factors, including better myocardial protection strategies, may improve perioperative RV function.