Prehospital prediction of mechanical circulatory support during hospital stay in cardiogenic shock: The FLAG score.
Heart & lung · 2026-01-01 · Registry study
Abstract
BACKGROUND: Mechanical circulatory support (MCS) is a treatment option in cardiogenic shock (CS) for patients refractory to pharmacological therapy. In a prehospital setting, the ability to predict the need for MCS during hospitalization may influence the choice of the designated hospital. OBJECTIVES: Development of a prehospital risk score to predict MCS treatment during hospital stay in CS patients. METHODS: We retrospectively analysed 431 prehospitally treated patients directly admitted to two hospitals of LMU university from 2014 to 2025. Using a logistic regression model that is further refined by a sparse beam algorithm and star ray search, a risk score to predict MCS treatment during hospital stay was developed. 80 % of patients were used for the training and 20 % for the testing cohort. MCS was defined as venoarterial extracorporeal membrane oxygenation (VA-ECMO), Impella CP and Impella 2.5 (Abiomed, Danvers, USA) or left ventricular assist device (LVAD) implantation. RESULTS: MCS implantation was performed in 34.3 % of patients. The following variables of the prehospital period were included in the model by the algorithm: fibrinolysis, prolonged CPR ≥ 35 min, age ≥ 65 years and female gender (FLAG). Calculated risk for MCS ranged from 16.9 % (-3 points) to 83.1 % (+5 points). An AUROC of 0.707 was achieved in the internal validation cohort. CONCLUSIONS: The FLAG risk score is a simple, registry-based model to predict MCS treatment during hospital stay in a prehospital setting. These results should be validated and refined in external cohorts. TRIAL REGISTRATION: WHO International Clinical Trials Registry Platform Number DRKS00015860.