Authors: Xiaxia Jin, Qiaofei Zheng, Ying Cheng, Lingling Hu, Wenhui Yang, Jun Li, Tao Li
Journal: American Journal of Emergency Medicine, October 2024
Conclusion: BNP levels are a significant predictor of 30-day mortality in patients with cardiac arrest who regain ROSC. Incorporating BNP levels into predictive models enhances the ability to predict 30-day mortality, suggesting that BNP could be a useful biomarker in post-ROSC care.
Methods: This retrospective study analyzed 260 cardiac arrest patients of presumed cardiac etiology who achieved ROSC, from November 2013 to June 2022 at two tertiary hospitals. Cox regression models and nomogram models were used to evaluate the association between BNP levels and 30-day mortality. The study also employed net reclassification improvement (NRI) and integrated discrimination improvement (IDI) to assess the value of BNP in predicting 30-day mortality compared to a baseline model.
- Model 1 included the variables from Model 2 and BNP levels.
- Model 2 included male sex, age, non-shockable rhythm, epinephrine use, and time to ROSC >30 minutes.
Results:
- BNP level was a significant predictor of 30-day mortality (HR = 1.441; 95% CI: 1.198-1.734).
- The area under the curve (AUC) for BNP alone was 0.813, and for Model 2 it was 0.834.
- Model 1, which included BNP, showed a superior predictive value with an AUC of 0.887 (95% CI: 0.836-0.939).
- Compared to Model 2, Model 1 improved mortality prediction, as demonstrated by a higher NRI (0.451; 95% CI: 0.267-0.577) and IDI (0.109; 95% CI: 0.035-0.191).