A comparative study of the risk stratification models for pediatric cardiac surgery

Objective The objective of the study was to compare Risk Assessment for Congenital Heart Surgery (RACHS-1), Aristotle Basic Complexity (ABC) and Society of Thoracic Surgeons – European Association for Cardiothoracic Surgery (STS-EACTS) complexity scoring models for predicting outcome after surgery for congenital heart disease. Methods This retrospective study included children <18 years. Procedures were categorized based on RACHS-1, ABC and STS-EACTS system. Outcome indicators were prolonged length of ICU stay (upper 25th percentile) and hospital mortality. The stratification models were tested for calibration using Hosmer‑Lemeshow modification of chi-square test and for discrimination using Receiver Operating Characteristic (ROC) curve. Area under the curve (AUC) of individual ROC curves was compared using z-statistics. Results The study included 920 patients. All 3 models showed good fit for both prolonged ICU stay and mortality on calibration. STS-EACTS outclassed RACHS-1 and ABC models with AUC of 0.759 for prolonged PLOS and 0.870 for hospital mortality. AUC of ROC curve for STS-EACTS was significantly higher than RACHS-1 model for both prolonged PLOS (p – 0.046) and hospital mortality (p – 0.015). No significant difference was observed between the AUC of ROC curves of other models. Conclusion Risk stratification for pediatric heart surgery is a useful tool to predict the outcome. STS-EACTS risk stratification model has the best discriminative power.