Real-Time Predictors of Return of Spontaneous Circulation in an Emergency Setting: A Five-Year Retrospective Study

Abstract

Background: Cardiopulmonary resuscitation (CPR) is a highly effort-intensive intervention and, in cases of cardiac arrest, the ability to predict a return of spontaneous circulation (ROSC) is of great importance for the efficient use of resources. This real-time assessment approach offers a practical advantage by increasing the applicability of prognostic models during acute resuscitation in an emergency department. Method: In this study, the data of patients who underwent CPR in the emergency department of a tertiary care hospital between 1 June 2019 and 1 June 2024 and underwent cardiopulmonary resuscitation were retrospectively analyzed. The patients’ demographics, comorbidities, CPR characteristics, and laboratory findings were evaluated using logistic regression and ROC curve analysis to identify the predictors of ROSC. Result: Our study revealed that cases with shockable rhythms and a shorter CPR duration were more likely to achieve ROSC. Elevated levels of albumin, creatine kinase, glucose, hemoglobin, and white blood cells were significantly associated with ROSC, while higher levels of creatinine, base excess, and eosinophils were more common in non-survivors. Atrial fibrillation and neurodegenerative disease were associated with lower ROSC rates. Conclusions: Although the criteria for the termination of cardiac arrest resuscitation are not definitive, certain patient characteristics and laboratory findings may guide the prediction of ROSC or the identification of cases requiring prolonged CPR. The integration of these real-time predictors into clinical algorithms may support decision making in crowded emergency departments.