Stratificazione prognostica nei pazienti post-arresto cardiaco attraverso un approccio multiparametrico: analisi retrospettiva di un singolo centro.

Background: Prognostic assessment after cardiac arrest remains a major clinical challenge due to the heterogeneous post–resuscitation syndrome. Early identification of patients at high risk of in-hospital mortality or poor neurological outcome is essential to guide management strategies and optimize intensive care resources. Objectives: This study aimed to identify independent predictors of in-hospital mortality and unfavorable neurological outcome in post–cardiac arrest patients admitted to the Cardiac Intensive Care Unit (CICU) and to evaluate the incremental prognostic value of integrating neurological markers into a multiparametric predictive model. Methods: We conducted a retrospective, single-center observational study including 260 consecutive patients admitted to the CICU of the University Hospital of Padua between January 2018 and August 2025, after in-hospital or out-of-hospital cardiac arrest with return of spontaneous circulation (ROSC). Four multivariate logistic regression models were developed to identify predictors of (1) in-hospital mortality and (2) unfavorable neurological outcome, defined as a Cerebral Performance Category (CPC) ≥ 3. Model performance and discrimination were assessed using receiver operating characteristic (ROC) curve analysis. Results: Serum lactate on admission, age, left ventricular ejection fraction and the number of epinephrine doses administered during resuscitation were identified as independent predictors of in-hospital mortality (AUC 0.834). The addition of neurological markers, such as peak neuron-specific enolase (NSE) and the presence of myoclonus or epilepsy significantly improved the model’s discriminative ability (AUC 0.882). For neurological outcome, serum lactate and epinephrine administration were independent predictors of poor CPC (AUC 0.798), while incorporation of NSE and neurological features further enhanced predictive accuracy (AUC 0.928). Conclusions: Elevated serum lactate at admission, older age, lower LVEF and higher epinephrine dose requirements independently predicted in-hospital mortality. The integration of early neurological markers, such as peak NSE and myoclonus/epilepsy, further improved prognostic accuracy for both mortality and neurological outcome. These findings support a multimodal prognostic approach combining systemic and neuro-specific variables to refine early risk stratification in post-ROSC patients.