External Validation of the CAST and rCAST Score in Patients With Out‐of‐Hospital Cardiac Arrest Who Underwent Extracorporeal Cardiopulmonary Resuscitation: A Secondary Analysis of the SAVE‐J II Study

Abstract

Background Risk stratification is important in patients with post–cardiac arrest syndrome. The Post‐Cardiac Arrest Syndrome for Therapeutic Hypothermia (CAST) and revised CAST (rCAST) scores have been well validated for predicting neurological outcomes, particularly for conventionally resuscitated patients with post–cardiac arrest syndrome. However, no studies have evaluated patients undergoing extracorporeal cardiopulmonary resuscitation. Methods and Results Adult patients with out‐of‐hospital cardiac arrest who underwent extracorporeal cardiopulmonary resuscitation were analyzed in this retrospective observational multicenter cohort study. We validated the accuracy of the CAST/rCAST scores for predicting neurological outcomes at 30 days. Moreover, we compared the predictive performance of these scores with the TiPS65 risk score derived from patients with out‐of‐hospital cardiac arrest who were resuscitated using extracorporeal cardiopulmonary resuscitation. A total of 1135 patients were analyzed. The proportion of patients with favorable neurological outcomes was 16.6%. In the external validation, the area under the receiver operating characteristic curve of the CAST score was significantly higher than that of the rCAST score (area under the receiver operating characteristic curve 0.677 versus 0.603; P <0.001), but there was no significant difference with that of the TiPS65 score (versus 0.633; P =0.154). Both CAST/rCAST risk scores showed good calibration (Hosmer–Lemeshow test: P =0.726 and 0.674), and the CAST score showed significantly better predictability in net reclassification compared with the rCAST ( P <0.001) and TiPS65 scores ( P =0.001). Conclusions The prognostic accuracy of the CAST score was significantly better than that of other risk scores in net reclassification. The CAST score may help to predict neurological outcomes in patients with out‐of‐hospital cardiac arrest who undergo extracorporeal cardiopulmonary resuscitation. However, the predictive value of the CAST score was not sufficiently high for clinical application. Registration URL: https://center6.umin.ac.jp/cgi‐open‐bin/ctr_e/ctr_view.cgi?recptno=R000041577 ; Unique identifier: UMIN000036490.