# Temporal Cardiorenal Dynamics and Mortality Prediction After TAVR: The Prognostic Value of the 48–72 h BUN/EF Ratio

## TL;DR

This study shows that a simple ratio of blood urea nitrogen to heart function measured 48–72 hours after a heart valve procedure can predict early death risk, while the baseline ratio helps identify long-term risks.

## Contribution

The 48–72 h BUN/EF ratio is introduced as a novel dynamic biomarker for early mortality prediction after TAVR.

## Key findings

- The 48–72 h BUN/EF ratio outperformed baseline in predicting in-hospital mortality (AUC 0.826 vs. 0.743).
- Baseline BUN/EF improved risk model performance when added to EuroSCORE II (AUC 0.712 vs. 0.668).
- Temporal BUN/EF showed a non-linear association with mortality, while baseline showed a linear relationship.

## Abstract

Background: Renal and cardiac dysfunction are major determinants of adverse outcomes following transcatheter aortic valve replacement (TAVR). The ratio of blood urea nitrogen to left ventricular ejection fraction (BUN/EF) integrates renal and cardiac status into a single physiological index. This study aimed to evaluate the prognostic value of both baseline and temporal (48–72 h) BUN/EF ratios for predicting mortality after TAVR. Methods: A total of 429 patients (mean age 76 ± 8 years; 51% female) who underwent TAVR for severe aortic stenosis between 2017 and 2025 were retrospectively analyzed. The primary endpoint was long-term all-cause mortality; in-hospital mortality was secondary. Receiver operating characteristic (ROC) curves, Cox regression, and reclassification metrics (NRI, IDI) assessed prognostic performance. Restricted cubic spline (RCS) analysis explored non-linear associations. Results: During a median follow-up of 733 days, overall and in-hospital mortality rates were 37.8% and 7.9%, respectively. Both baseline and 48–72 h BUN/EF ratios were independently associated with mortality (HR = 3.46 and 3.79 per 1 SD increase; both p < 0.001). The temporal ratio showed superior discrimination for in-hospital mortality (AUC = 0.826 vs. 0.743, p = 0.007). Adding baseline BUN/EF to EuroSCORE II significantly improved model performance (AUC 0.712 vs. 0.668, p = 0.031; NRI = 0.33; IDI = 0.067). RCS analysis revealed a linear relationship for baseline and a steep, non-linear association for temporal ratios with mortality risk. Conclusions: The 48–72 h BUN/EF ratio is a robust dynamic biomarker that predicts early mortality after TAVR, while baseline BUN/EF identifies patients at long-term risk. Integrating this simple bedside index into risk algorithms may refine postoperative monitoring and improve outcome prediction in TAVR populations.