Discriminatory power of scoring systems for outcome prediction in patients with extracorporeal membrane oxygenation following cardiovascular surgery†

OBJECTIVES Although extracorporeal membrane oxygenation (ECMO) represents a rapidly evolving treatment option in patients with refractory heart or lung failure, survival remains poor and appropriate risk stratification challenging because established risk prediction models have not been validated for this specific population. METHODS This observational single-centre registry included a total of 240 patients treated with venoarterial ECMO therapy following cardiovascular surgery and analysed the discriminatory power of the European System of Cardiac Operative Risk Evaluation (EuroSCORE) additive, the EuroSCORE II, the Sequential Organ Failure Assessment (SOFA) score, the Simplified Acute Physiology Score (SAPS) II, the SAPS III, the Acute Physiology and Chronic Health Evaluation (APACHE) II, the Risk of renal failure, Injury to the kidney, Failure of kidney function, Loss of kidney function and End-stage renal failure (RIFLE) classification, the survival after venoarterial ECMO (SAVE) score, the prEdictioN of Cardiogenic shock OUtcome foR AMI patients salvaGed by VA-ECMO (ENCOURAGE) score and the Society of Thoracic Surgeons (STS) risk model for outcome prediction. RESULTS During a median follow-up time of 37 months (interquartile range 19–67), 65% of the patients died. Only the SAVE score and the SAPS II were significantly associated with the 30-day mortality rate with a hazard ratio (HR) of 1.06 [95% confidence interval (CI) 1.02–1.11; P = 0.002] for the SAVE score and an HR of 1.02 (95% CI 1.01–1.03; P = 0.004) for the SAPS II with a modest discriminatory power displayed by a C-index of 0.61 and 0.57, respectively. Seven out of 10 scoring systems revealed significant association with long-term mortality, with the SAVE score and the SAPS II remaining the strongest predictors of long-term mortality with an HR of 1.06 (95% CI 1.03–1.09; P < 0.001, C-index 0.61) for the SAVE score and an HR of 1.02 (95% CI 1.01–1.03; P < 0.001, C-index 0.58) for the SAPS II. CONCLUSIONS Risk assessment based on established risk models in patients with ECMO remains difficult. Only the SAPS II and the SAVE score were exclusively found to be suitable for short- and long-term outcome prediction in this specific vulnerable patient population.