A new study compared four point-of-care tests for FeLV and FIV from IDEXX, Witness, Anigen and VetScan. These tests are often the only test used for retroviruses in cats, given the lack of good “gold standard” confirmatory tests.
Shelters may euthanize cats that test positive, or, for those that adopt out positive cats, a positive result can present significant barriers to adoption. So it’s really important to be confident that a positive test result really is correct.
Statistically, IDEXX SNAP was the best test for FeLV, with no significant differences for FIV between the four tests. The IDEXX test had sensitivity and specificity of 100% for FeLV, and 97.9%/99.0% for FIV. For FeLV, Anigen was next best, then Witness, then VetScan trailing a fair way behind.
A limitation of the study noted by the authors was the above-mentioned lack of a gold standard – but it was a well-designed study with a good chance that their positive samples really were positive and negatives really negative. The researchers used plasma and serum for testing, while anticoagulated whole blood is typically used in shelters, but this should not have substantially affected the results.
So far so good. But here’s the issue: The lower the prevalence of disease in the test population, the lower the predictive value* of a test. For FeLV, the positive predictive value of the IDEXX SNAP was very high, but for the other 3 tests it was quite poor in a low prevalence population . For FIV, the positive predictive value was lower overall, but best for the Witness test.
So how can we avoid believing results that aren’t correct, in a low prevalence population? First of all, use the best test available. Then, remember that the higher the prevalence in the test population, the better the predictive value – so consider selecting which cats to test instead of testing all of them. Look at the cat and everything you know about it. For a beat-up older tomcat covered in scars, a positive result is likely to be real. For an indoor spayed female with no known risk factors, it would be reasonable confirm the diagnosis by running a different test or a different type of test.
- Levy JK et al. Performance of 4 Point-of-Care Screening Tests for Feline Leukemia Virus and Feline Immunodeficiency Virus. Journal of Veterinary Internal Medicine 31(2):521-526; 2017. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5354053/
* Positive predictive value: Given a test is positive, the likelihood disease is truly present.
Negative predictive value: Given a test is negative, the likelihood disease is truly not present