In the March 28 issue of the ARCHIVES, Battalgia et al1 raised 3 issues related to our recent review of the diagnostic accuracy of natriuretic peptides.2 First, they were aware of 4 studies of the rapid enzyme-linked immunosorbent assay (ELISA) form of the test that were not included in our review. It is not clear from their letter if these studies were published before or after the search date for our review, December 2002. We excluded 5 studies from our review that had used the rapid ELISA form of the test because the dates of recruitment and the recruitment sites overlapped with those of the studies by Maisel et al, 2002,3 and Krishnaswamy et al, 2001.4 We have also recently updated the review (search date, March 2004) and found a further 2 studies that used the rapid ELISA form of the test. We agree with the correspondents that the type of assay may be one reason for the significant heterogeneity seen in the test characteristics between the studies. One difficulty in determining this, however, is that there are no studies, to our knowledge, comparing these 2 forms of the test within 1 population. The results of any comparison made between studies would need to be interpreted with considerable caution. Of interest, we are currently conducting a study of the variability of 2 versions of the test. The 2 forms of the test are showing considerable and inconsistent variability.
Doust JA, Glasziou PP, Dobson AJ. Diagnostic Accuracy of Natriuretic Peptides for Heart Failure—Reply. Arch Intern Med. 2005;165(10):1199. doi:10.1001/archinte.165.10.1199-a