Abstract
BACKGROUND: Because of the increasing importance of quality assurance and individualised patient treatment, EuroSCOREs were analysed for reliability in predicting 30-day mortality in patients with postinfarction ventricular septal rupture (piVSR). To address the specific conditions of patients with piVSR, the Hannover piVSR Score (HpiVSR) was developed. METHODS: Between 2001 and 2019, 45 patients with piVSR underwent surgery. Data were collected as necessary for EuroSCORE calculation. Clinically relevant variables were validated for the HpiVSR Score using a nonparsimonious binary logistic regression model. All models were tested for their significant predictive power for 30-day mortality. Their validity was assessed using Hosmer-Lemeshow test and Nagelkerke R (2). Receiver operating characteristic curve and area under the curve were used to illustrate and quantify score accuracy. RESULTS: The specificity was 77.8% for all EuroSCOREs and 92.6% for the HpiVSR Score, and the sensitivity was in the random range for the EuroSCOREs and 83.3% for the HpiVSR Score. Accordingly, the areas under the curve were 0.676 (95% CI 0.507-0.845) for EuroSCORE II; 0.729 (95% CI 0.581-0.878) and 0.739 (95% CI 0.591-0.886) for the additive and logistic EuroSCORE, respectively; and 0.949 (95% CI 0.891-1.006) for the HpiVSR Score. CONCLUSION: The HpiVSR Score enables a more reliable and accurate prediction of 30-day mortality than the EuroSCOREs using 7 significant, objective, reliable, and preoperatively determinable variables. Because of the small sample size of the present study and the fact that only internal validation has been performed so far, the weighting of the factors of the HpiVSR Score can be adjusted after studies with larger patient samples.