Relationship between B-type natriuretic peptide and adverse outcome in patients with clinical evidence of sepsis presenting to the emergency department

Acad Emerg Med. 2011 Feb;18(2):219-22. doi: 10.1111/j.1553-2712.2010.00968.x. Epub 2011 Jan 28.

Abstract

Objectives: Myocardial dysfunction is an important aspect of sepsis pathophysiology. B-type natriuretic peptide (BNP) is a neurohormone released from the ventricles in response to myocardial stretch and volume overload. The authors hypothesized that an elevated BNP in patients presenting to the emergency department (ED) with suspected sepsis are at increased risk for development of adverse events.

Methods: This was a prospective, observational, multicenter cohort study in 10 EDs. Patients were eligible if they were older than 18 years, had two or more systemic inflammatory response syndrome (SIRS) criteria, and had suspected infection or a serum lactate level > 2.5 mmol/L. Patients were excluded if they were pregnant, had do-not-attempt-resuscitation status, sustained a cardiac arrest prior to hospital arrival, had known chronic renal insufficiency, or were on dialysis. BNP levels were obtained at arrival. The primary outcome was a composite of severe sepsis, septic shock within 72 hours, or in-hospital mortality.

Results: There were 825 patients enrolled (mean ± standard deviation [SD] age = 53.5 ± 19.6 years; 51% were female and 37% were African American). The area under the curve (AUC) for BNP to predict the triple composite outcome was 0.69, and the optimal cut-point of BNP was 49 pg/mL. Patients with a BNP > 49 pg/mL had a greater mortality rate (11.6% vs. 2.1%; p = 0.0001), a greater risk of development of severe sepsis (67.7% vs. 36.8%; p = 0.0001) and septic shock (51.7% vs. 26.4%; p = 0.0001), and a higher rate of the triple composite outcome (69% vs. 37%; unadjusted odds ratio [OR] = 1.9, 95% confidence interval [CI] = 1.6 to 2.1; p < 0.001). The sensitivity was 63% (95% CI = 58% to 67%), specificity was 69% (95% CI = 65% to 73%), negative predictive value (NPV) was 63% (95% CI = 58% to 67%), and positive predictive value (PPV) was 69% (95% CI = 65% to 74%). In multivariate modeling, after adjusting for age, sex, heart rate, white blood cell count, and creatinine, an elevated BNP was associated with increased odds of having the composite outcome. The outcome was similar in the subset of patients who did not have severe sepsis or septic shock upon arrival.

Conclusions: In patients who present to the ED with SIRS criteria and suspected infection, an elevated BNP is associated with a worse prognosis but has limited diagnostic utility.

Publication types

  • Multicenter Study

MeSH terms

  • Adult
  • Aged
  • Emergency Service, Hospital
  • Female
  • Hospital Mortality
  • Humans
  • Logistic Models
  • Male
  • Middle Aged
  • Natriuretic Peptide, Brain / blood*
  • Prospective Studies
  • ROC Curve
  • Sepsis / blood*
  • Sepsis / mortality*
  • Systemic Inflammatory Response Syndrome / blood*
  • Systemic Inflammatory Response Syndrome / mortality*
  • Treatment Failure
  • United States / epidemiology

Substances

  • Natriuretic Peptide, Brain