Geriatric nutritional risk index improves risk scoring for mortality after TAVR compared to established scores

Risk scoring prior TAVR is based on the EuroSCORE II (European System for Cardiac Operative Risk Evaluation) and the STS-score (Society of Thoracic Surgeons) which are complex, and partly prone to investigator bias. The geriatric nutritional risk index (GNRI) can be calculated by five parameters via publicly available formula (age, height, weight, sex, serum albumin), whereas EuroSCORE II and the STS-score require 21 and 69 variables, respectively. The study compares the efficiency of GNRI in predicting 30-day mortality compared to the EuroSCORE II and the STS-score. Furthermore, GNRI risk classes were analysed in the long-term.

3.470 consecutive patients who underwent TAVR between 2010 and 2023 at our institution were analysed. GNRI calculation produces a linear parameter that can be divided in four risk groups.

ROC (receiver operating characteristic) curve analysis demonstrated no difference in predicting 30-day mortality between GNRI vs. EuroSCORE II (AUC = 0.72 vs. 0.69, p = 0.3) and GNRI vs. STS-score (AUC = 0.72 vs. 0.72, p = 1.0). The Hosmer-Lemeshow test indicated good calibration for the GNRI model (p = 0.3793). After adjustment for preoperative demographic characteristics, Cox regression analysis for overall survival after TAVR reveals for the major risk group [21 patients; HR = 4.624; CI95%(2.881-7.422); p < 0.0001], the moderate risk group [198 patients (5.7%), HR = 2.201; CI95%[1.821-2.660]; p < 0.0001], and the low risk group [452 patients (13.0%); HR = 1.831; CI95%[1.597-2.1]; p < 0.0001], respectively.

The GNRI is an objective publicly available score that simplifies risk assessment prior TAVR without any loss of precision compared to the EuroSCORE II and the STS-score.