Abstract & Commentary
Synopsis: At this time, BNP seems to be a useful initial test that should not be repeated during short hospital stays.
Source: Wu AHB, et al. Am J Cardiol. 2004;93: 1562-1563.
Since b-type natriuretic peptide (BNP) is elevated in patients with congestive heart failure, serial testing may be useful for determining the response to therapy. In order to establish the usefulness of serial testing, knowledge of the assay variability is required. Thus, Wu and colleagues performed a retrospective study of laboratory records at 2 large metropolitan hospitals to determine the optimal frequency of BNP measurements for this purpose. The study population consisted of 2748 samples from 1926 patients admitted for heart failure, approximately one-third of whom had a greater than or equal to 2 BNP measures done in the first week after hospitalization. Only patients with a greater than or equal to 2 BNP measures at 22 days were included. Their previous studies, in normal subjects and stable heart failure subjects, suggested that a difference of 100% in BNP values was the biologic variation cut point.
The likelihood of finding a significant difference in BNP values was lowest when blood was retested the next day (22%) or in 2 days (39%). The likelihood of finding a significant difference was highest when samples were 6 or 7 days apart (50%). Thereafter, the percentage, with a significant difference, fell off again. Overall, two-thirds of significant differences were decreases in BNP, suggesting that most patients improved. Wu et al concluded that daily, or every other day, monitoring of BNP, in patients admitted with heart failure, does not appear warranted.
Comment by Michael H. Crawford, MD
This observational study, based upon laboratory records, is sobering, considering the widespread use of serial BNP testing today. However, it is in line with other recent studies that have not shown value for the indiscriminate serial use of BNP testing in hospitalized patients. The results are also in line with our thinking about heart failure treatment. Some patients will get immediate benefit from diuretics, which may be reflected in the one-fourth to one-third of patients whose BNP values do change significantly over the first 2 days, but many hospitalized patients (50% in this study) may take 6 or 7 days to improve enough to overcome the effects of cardiac structural changes on BNP level, such as left atrial dilation. Also sobering is the magnitude of change required to be biologically significant, 100%. This value is based upon their previous work showing that a significant serial BNP change in normal subjects was 129% and for stable heart failure subjects was 77%. They used the average difference of 100% for this study. Had they used the lower value, perhaps BNP would have looked somewhat better, but I doubt it would have materially changed the results. At this time, BNP seems to be a useful initial test that should not be repeated during short hospital stays.
Dr. Crawford is Professor of Medicine, Associate Chief of Cardiology for Clinical Programs University of California, San Francisco.