The combination of Euroscore II with a new variable "Nongraftable coronary artery lesions" improves the discriminating power of the test in surgical patients with postinfarction ventricular septal defect and ventricular aneurysm.

Background: We aimed to establish whether Euroscore II can be used for the prediction of hospital mortality in surgical patients with postinfarction intraventricular septal defect (PIVSD) and ventricular aneurysm (VA), and coexisting coronary artery lesions (CALs), and identify perioperative mortality risk factors to improve the discriminating power of Euroscore II.

Methods: This was a retrospective observational study. The inclusion criterion was PIVSD. Exclusion criteria were previous CABG, conservative treatment, percutaneous transcatheter closure of PIVSDs, and PIVSDs with normal coronary arteries on coronary angiography.

Results: Among 53 patients with PIVSDs and VAs who met eligibility criteria, 12 (22.6%) patients died in the hospital. Logistic regression demonstrated that Euroscore II was associated with in-hospital mortality (odds ratio [OR] = 1.13; 95% confidence interval [CI]: 1.03-1.23; p = .006), well-calibrated (Hosmer-Lemeshow χ (8) = 9.75; p = .283), and had fair discriminating power, area under receiver operating characteristic curve (AUC) = 77% (95% CI: 58%-96%). A newly identified variable "Nongraftable CALs" was associated with in-hospital mortality (OR = 6.65; 95% CI: 1.24-35.53; p = .027), and had a fair discriminating power, AUC = 70% (95% CI: 54%-85%). When Euroscore II and Nongraftable CALs were combined, the discriminating power of the test increased to 83% (95% CI: 71%-95%), p = .036.

Conclusion: Euroscore II has adequate discriminating power and good calibration in predicting in-hospital mortality of surgical patients with PIVSDs and VAs. The combination of Euroscore II with a new variable "Nongraftable CALs" significantly improves the performance of the model.