Prediction of In‐Hospital Mortality for Ischemic Cardiogenic Shock Requiring Venoarterial Extracorporeal Membrane Oxygenation

Abstract

Background Clinical outcome of ischemic cardiogenic shock (CS) requiring extracorporeal membrane oxygenation is highly variable, necessitating appropriate assessment of prognosis. However, a systemic predictive model estimating the mortality of refractory ischemic CS is lacking. The PRECISE (Prediction of In‐Hospital Mortality for Patients With Refractory Ischemic Cardiogenic Shock Requiring Veno‐Arterial Extracorporeal Membrane Oxygenation Support) score was developed to predict the prognosis of refractory ischemic CS due to acute myocardial infarction. Methods and Results Data were obtained from the multicenter CS registry RESCUE (Retrospective and Prospective Observational Study to Investigate Clinical Outcomes and Efficacy of Left Ventricular Assist Device for Korean Patients With Cardiogenic Shock) that consists of 322 patients with acute myocardial infarction complicated by refractory ischemic CS requiring extracorporeal membrane oxygenation support. Fifteen parameters were selected to assess in‐hospital mortality. The developed model was validated internally and externally using an independent external cohort (n=138). Among 322 patients, 138 (42.9%) survived postdischarge. Fifteen predictors were included for model development: age, diastolic blood pressure, hypertension, chronic kidney disease, peak lactic acid, serum creatinine, lowest left ventricular ejection fraction, vasoactive inotropic score, shock to extracorporeal membrane oxygenation insertion time, extracorporeal cardiopulmonary resuscitation, use of intra‐aortic balloon pump, continuous renal replacement therapy, mechanical ventilator, successful coronary revascularization, and staged percutaneous coronary intervention. The PRECISE score yielded a high area under the receiver‐operating characteristic curve (0.894 [95% CI, 0.860–0.927]). External validation and calibration resulted in competent sensitivity (area under the receiver‐operating characteristic curve, 0.895 [95% CI, 0.853–0.930]). Conclusions The PRECISE score demonstrated high predictive performance and directly translates into the expected in‐hospital mortality rate. The PRECISE score may be used to support clinical decision‐making in ischemic CS ( www.theprecisescore.com ). Registration URL: https://www.clinicaltrials.gov ; Unique identifier: NCT02985008.