Can brain natriuretic peptide predict outcome after cardiac arrest? An observational study.

Abstract

No accurate, independent biomarker has been identified that could reliably predict neurological outcome early after cardiac arrest. We speculated that brain natriuretic peptide (BNP) measured at hospital admission may predict patient outcome. BNP-levels were measured in 155 comatose cardiac arrest survivors (108 male, 58 years [IQR 49-68]) (median time to ROSC 11min; IQR 20-30) during a 6-year study period. Cardiovascular co-morbidities and resuscitation history were assessed according to the Utstein-style and patients were followed for 6-month neurological outcome measured by cerebral performance category (CPC) and survival. Seventy patients (45%) suffered from unfavourable neurological outcome and 79 deaths (51%) occurred during the first 6 months. BNP was significantly associated with an adverse neurological outcome and mortality, independent of the prearrest health condition and cardiac arrest characteristics (median 60 pg/ml; IQR 10-230). Adjusted odds ratios for poor neurological outcome at 6 months were 1.14 (95% CI 0.51-2.53), 1.76 (95% CI 0.80-3.88) and 2.25 (95% CI 1.05-4.81), for increasing quartiles of BNP as compared to the lowest quartile. Adjusted odds ratios for mortality until 6 months were 1.09 (95% CI 0.35-3.40), 2.81 (0.80-9.90) and 4.7 (1.27-17.35) compared to the lowest quartile, respectively. Brain natriuretic peptide levels on admission predict neurological outcome at 6 months and survival after cardiac arrest.