Comparison of return of spontaneous circulation prediction scores in patients with cardiac arrest during ambulance transport

Background Cardiac arrest during ambulance transport is a complex situation that has features of both in-hospital cardiac arrest (IHCA) and out-of-hospital cardiac arrest (OHCA) but lacks a clear classification. This study aimed to evaluate the diagnostic performance of prehospital and in-hospital ROSC prediction tools in patients experiencing cardiac arrest during ambulance transport. Methods A retrospective study was conducted with patients experiencing cardiac arrest during transport. Demographic, clinical, and treatment data were collected, including pre-arrest consciousness, arrest rhythm, and cardiopulmonary resuscitation duration. Four ROSC prediction scores (Prehospital-ROSC, the ROSC after cardiac arrest, Utstein-Based ROSC, and The Cardiac Arrest Survival Post-Resuscitation In-Hospital scores) were used to assess the patients. The AUROCs of the scores were compared to evaluate their diagnostic accuracy. Results Patients were categorized into two groups based on ROSC: No-ROSC (n = 248, 75.2%) and ROSC (n = 82, 24.8%). The ROSC group had significantly more shockable rhythms (81.7% vs. 22.2%, p < 0.001) and a higher proportion of cardiac etiology (p = 0.015) compared to the No-ROSC group. The time for the ambulance to reach the patient did not significantly differ between groups (p = 0.140), but the time spent in the ambulance before arrest was significantly shorter in the ROSC group (p = 0.026). The prehospital-ROSC score had the highest diagnostic performance (AUROC 0.791), followed by The ROSC after cardiac arrest score (0.754) and The Utstein-Based ROSC score (0.716). The Cardiac Arrest Survival Post-Resuscitation In-Hospital score had the lowest performance (0.658). Prehospital-ROSC score outperformed the Utstein-Based ROSC score (p = 0.005), and the Cardiac Arrest Survival Post-Resuscitation In-Hospital score was significantly lower than both ROSC after cardiac arrest score (p = 0.031) and the prehospital-ROSC score (p < 0.001). Conclusion Prehospital-ROSC score was the most accurate predictor of ROSC in patients experiencing cardiac arrest during ambulance transport, while the Cardiac Arrest Survival Post-Resuscitation In-Hospital score demonstrated lower predictive accuracy.