The predictive model of intra-arrest blood markers for the sustained return of spontaneous circulation in patients with out-of-hospital cardiac arrest

Objective This study aimed to investigate the predictive value of intra-arrest blood markers, including blood routine, blood biochemistry, and coagulation profile, during cardiopulmonary resuscitation (CPR) for the return of spontaneous circulation (ROSC) in patients with out-of-hospital cardiac arrest (OHCA). Methods A retrospective cohort of patients who experienced OHCA and admitted to the emergency department of a tertiary hospital in Zhejiang Province, China, between January 2021 and December 2025 was assembled. Their clinical and demographic data were collected, and missing values were addressed using multiple imputation. Physiologic laboratory markers obtained during resuscitation were compared between the sustained ROSC and non-ROSC groups. Candidate predictors were selected based on clinical relevance, previous literature, availability at the intended prediction time point, and univariable screening, with age and gender retained as prespecified demographic covariates. Model discrimination for the prediction of sustained ROSC was evaluated using a receiver operating characteristic analysis and the area under the curve (AUC). Internal validation was performed using bootstrap resampling to assess model stability and quantify potential overfitting. Results A total of 283 patients were included, and among them, 94 (33.22%) achieved sustained ROSC. Missing data were handled using multiple imputation ( m = 20). In univariable analyses, the sustained ROSC group had higher PCO 2 , platelet count, white blood cell count, and absolute lymphocyte count; higher admission glucose, fibrinogen, and serum calcium levels; shorter chest compression duration and prothrombin time; and lower epinephrine dose, serum potassium level, and international normalized ratio than the non-sustained ROSC group (all P < 0.05). A multivariable logistic regression model incorporating age; gender; white blood cell count; platelet count; blood PCO 2 measured during CPR; capillary blood glucose level on admission; and fibrinogen, serum potassium, and serum calcium levels yielded a mean apparent AUC of 0.832 across the imputed datasets. After internal validation with 500 bootstrap resamples, the optimism-corrected AUC was 0.807 and the bootstrap-derived calibration slope was 0.853. The final model is given as follows: logit[P(ROSC=1)] = −2.65−0.002 × age−0.790 × gender + 0.076 × white blood cell count + 0.001 × platelet count + 0.010 × PCO 2 −0.011 × admission glucose + 0.145 × fibrinogen−0.247 × serum potassium+0.633 × serum calcium. Conclusion In patients who experienced OHCA, higher white blood cell count and lower serum potassium levels were associated with a higher incidence of sustained ROSC. Female patients were more likely to achieve sustained ROSC than male patients. Although the model showed certain predictive value for sustained ROSC, further multicenter, large-scale studies are needed in the future.