The trusted source for
healthcare information and
Synopsis: The B-type natriuretic peptide test appears sensitive and specific for diagnosing heart failure in the urgent care setting.
Source: Dao Q, et al. J Am Coll Cardiol. 2001;37:379-385.
B-type natriuretic peptide (BNP) is secreted from the ventricles in response to volume or pressure overload and is elevated in heart failure (HF). With the development of a new rapid assay, Dao and associates sought to determine the value of BNP in the urgent care setting. A sample of 250 of 438 patients presenting to a VA urgent care center with dyspnea were studied. Patients whose dyspnea was clearly not due to HF ie chest trauma and patients with acute coronary syndromes were excluded. The patients were predominately male (94%) and most (70%) had dyspnea on exertion, but 50% had dyspnea at rest. Although the urgent care center physicians rated the probability of HF as high, medium, or low, the gold standard was 2 cardiologists who retrospectively reviewed all the available data including the hospital course and tests done after transfer from urgent care. During the initial evaluation in urgent care, blood for BNP, measured by the Biosite Diagnostics test, was obtained, but the urgent care doctors were not provided the results. The BNP results were evaluated against the cardiologists diagnosis. They diagnosed acute HF in 39%, LV dysfunction with no HF in 6%, and noncardiac dyspnea in 55%. BNP levels in the HF patients were 1076 vs. 38 pg/mL in the patients with noncardiac dyspnea (P < .001). Those with LV dysfunction but no HF had levels of 141 ¹ 31 pg/mL. At BNP levels greater than 80, HF was predicted with an accuracy of 95% and a negative predictive value of 98% for levels less than 80. Also, BNP levels were higher as the assessed severity of HF increased. In addition, BNP was highly accurate in separately pulmonary dyspnea vs. HF and noncardiac pedal edema from HF. Finally, BNP greater than 80 was the most accurate predictor of HF compared to all pertinent clinical variables available initially including ECG and chest x-ray; consideration on BNP levels would have prevented 15 unnecessary admissions, and 29 of 30 misdiagnosed cases would have been corrected. Dao et al concluded that the BNP test appears sensitive and specific for diagnosing HF in the urgent care setting.
Comment by Michael H. Crawford, MD
As I was perusing the annotated table of contents of Journal of the American College of Cardiology, I saw a full page ad for The Triage BNP Test by Biosite Diagnostics. It touted a 15-minute result with 98% accuracy for the diagnosis of HF vs. "all significant variables in patients with or without disease history." My first reaction was this is ludicrous. I know how to diagnose HF and don’t need some blood test. How accurate will it be in the real world? Will this be another "troponin" that results in endless, mindless consultations to the emergency department? Now in addition to seeing everyone with chest pain, we would have to see everyone with dyspnea. At this rate, ER doctors would soon just draw blood, await a battery of tests, then start calling consultants. I was appalled. Then I noticed the strategic placement of this ad next to the brief description of 2 studies in Journal of the American College of Cardiology using this test that suggested it was useful. As I started to calm down, I remembered those patients with COPD and CAD in whom COPD exacerbation vs. HF is always a difficult call on clinical grounds. And what about the patient who languishes intubated under the ineffectual care of intensivists in the ICU, in whom they finally order an echo, which shows an ejection fraction of 20%? Perhaps there is a use for this test. This study makes a good point that dyspnea is a nonspecific symptom, and HF is difficult to diagnose clinically in patients with pulmonary disease or in elderly deconditioned patients. The ECG and chest x-ray can be helpful but are not sufficiently diagnostic in many cases. Echocardiography is extremely useful but is logistically more difficult than a blood test and probably more costly. Thus, the rapid BNP test could be a useful triage test as this study showed. The 98% negative predictive value and sensitivity is just what the ER physicians like because it reduces missed diagnoses to a litiginously acceptable level. Fortunately, in this study the specificity and positive predictive value of BNP were also greater than 90%, which keeps the false positives that drive us consultants crazy to a minimum.
One of the major limitations of this study was that it was a convenient sample of mainly elderly men. Also, patients with obvious confounders such as acute coronary syndromes and obvious causes of dyspnea (ie pneumothorax) were excluded. Thus, the real world performance of this test is still an issue. Would it be elevated in other causes of ventricular enlargement or trauma such as hypertensive urgencies, pregnancy, and cardiac contusion? Clearly, more information is needed. However, this study suggests that this may be a useful test in selected patients.