Prognostic value of liver scoring systems in predicting early outcomes after pericardiectomy: a single-center retrospective study

Pericardiectomy for constrictive pericarditis remains a high-risk operation, and early outcomes are inadequately predicted by conventional cardiac risk models despite the high prevalence of chronic hepatic congestion in this population. We performed a retrospective observational analysis of 104 consecutive patients undergoing isolated pericardiectomy between 2010 and 2024 to determine whether established hepatic scoring systems [Child-Pugh, Model for End-Stage Liver Disease (MELD) and Model for End-Stage Liver Disease-Sodium (MELD-Na)] predict early mortality and major postoperative morbidity. Preoperative clinical and biochemical variables were used to derive the hepatic scores. The primary endpoint was 30-day mortality; the secondary endpoint was new postoperative renal-replacement therapy. Thirty-day mortality was 4.8% (5/104). Patients were classified as Child-Pugh A in 66 (63.5%), B in 35 (33.7%), and C in 3 (2.9%), with corresponding mortalities of 1.5%, 2.9%, and 100% (p<0.01 for trend). Discrimination for 30-day mortality was strong for Child-Pugh [area under the receiver operating characteristic curve (AUC) 0.89, 95% confidence interval 0.73-1.00], MELD (0.85, 0.60-0.98), and MELD-Na (0.83, 0.55-0.97), whereas EuroSCORE II showed no meaningful predictive ability (AUC≈0.51). Hepatic dysfunction quantified by Child-Pugh and MELD-based scores is a strong predictor of early mortality after pericardiectomy and outperforms conventional cardiac risk models in this cohort. Incorporation of structured hepatic assessment into preoperative risk stratification may improve perioperative planning and patient counseling; however, the small number of Child-Pugh C patients and the retrospective single-center design necessitate validation in larger prospective cohorts.