In-Hospital Mortality After Aortic Valve Replacement Surgical Versus Transcatheter: A Nationwide Study with Machine Learning-Based Risk Prediction.

INTRODUCTION

Transcatheter aortic valve replacement (TAVR) and surgical aortic valve replacement (SAVR) are two preferred methods for treating aortic stenosis. This study compared the in-hospital mortality rates associated with SAVR and TAVR.

METHODS

In accordance with the STROBE guidelines, we conducted an analytical cross-sectional study utilizing data from the National Readmission Database (NRD) spanning from 2016 to 2020. Patients who underwent TAVR or SAVR were identified using specific ICD-10 procedure codes. We employed adjusted multivariate logistic regression analysis to determine the odds ratio (OR) of death associated with SAVR compared to TAVR. Patients were followed for a year and OR with 95% CI were calculated. AI based XGBoost model were used to find the predictors of mortality.

RESULTS

Our study included a total of 361,875 patients. Among these, 182,443 underwent TAVR, while 179,432 underwent SAVR. In the TAVR group, 2,563 patients (1.4%) died during hospitalization, whereas 6,671 patients (3.7%) died in the SAVR group. The multivariate logistic regression analysis indicated that SAVR is associated with higher in-hospital mortality (OR = 2.66, 95% CI 2.51-2.81, p < 0.001) compared to TAVR. Follow up data at 30 days, three months, six months, did not find any difference in mortality between TAVR and SAVR. Notably, liver failure, HLD, renal failure, SAVR, female, CHF, and atrial fibrillation are amongst the top predictor of mortality in patients undergoing (TAVR or SAVR). While in cardiac rehospitalization, risk factors like female gender, CHF, Afib, HLD, DM have higher predictability than procedure.

CONCLUSION

Our study demonstrates that SAVR is associated with a higher likelihood of in-hospital mortality compared to TAVR, with organ dysfunction (liver failure, kidney dysfunction) as a predictor of mortality.