Submitted to: Meeting Abstract
Publication Type: Abstract only
Publication Acceptance Date: 10/13/2011
Publication Date: 11/17/2011
Citation: Ridpath, J.F. 2011. Troubleshooting the diagnostic conundrums [abstract]. In: Proceedings of the U.S. Bovine Viral Diarrhea Virus Symposium, November 17-18, 2011, San Diego, California. p. 37. Interpretive Summary:
Technical Abstract: Because so many different clinical presentations can be associated with infection with BVDV, diagnosis based on history, clinical presentation, and postmortem of affected animals is presumptive at best. Accurate and definitive detection of infection with BVDV depends on laboratory diagnosis. While there are numerous tests available for detecting BVDV and there are also numerous opinions on advantages and shortcomings of each. The basis of diagnostic tests includes the detection of replicating virus, the detection of genomic material of the BVDV or the detection of proteins associated with BVDV. Selection of the best test for a given situation requires an analysis that weighs sensitivity, specificity, risk, cost, speed, and ease of sampling. While testing can be used to determine the etiology of clinical disease on a case-by-case basis, the most effective use of BVDV diagnostics is as a component of a surveillance program. In order to design effective surveillance programs we need to identify factors present that might result in test or testing protocol failure, we need to understand the limits of tests (positive predictive value and negative predictive value) and we need to understand how to best customize testing protocols for different production settings. While it may seem counter intuitive, understanding why tests and testing protocols fail is vital to designing tests and testing protocols that work. Currently, there is controversy as to the rate of detection of acutely infected animals versus persistently infected animals and test failure rate introduced by pooling of samples. The positive predictive value (PPV) and negative predictive value (NPV) need to be calculated for any test or testing protocol. These values must be determined under the conditions that will be in effect when the test is employed. Unlike sensitivity and specificity calculations, the PPV and NPV depend not just on the intrinsic accuracy of the test but are also directly proportional to the prevalence of the disease or pathogen to be detected. The PPV and NPV calculations are particularly critical when diagnosing a low prevalence event such as BVDV persistent infection.