This page: How reliable are the tests for FIV? - If your cat has tested positive for FIV, find out if that is safe, and how you can check if you are not sure
More about FIV - Testing
Testing for FIV
There are a number of tests available for establishing FIV status. Those most often used are indirect tests, ie they don't test for the virus itself, but for the antibodies to the virus that are produced by the cat's immune system once it becomes infected.
It can take several weeks for the immune system to produce enough antibodies to be measured, so there is a delay of several weeks between a cat being infected and being able to be tested accurately.
---------------------------------General screening test:
Elisa (Enzyme-Linked ImmunoSorbent Assay)
The elisa test is the most common, in that it is the "quick" test that vets can carry out in-house, and therefore get results within minutes.
The elisa is sometimes called a "snap", "cite" or "combo" test.
It tests for both FIV and FeLV at the same time, although different in what is tested - (antibodies for FIV and antigens for FeLV). The test is a good "general indicator", but is not always to be relied upon as it is a sensitive test which is therefore liable to give a false positive, so should be confirmed with one of the other (slower) tests.
You will see various claims to the accuracy of the Elisa test; in fact it is not a simple thing to give a percentage of accuracy - (see explanation at foot of page)
Perhaps the greatest danger for false positives from the ELISA test are due to operator error. The test only takes a few minutes to carry out, but the timing is important; if the test is left too long it can change the result! A true negative result, can turn to a false positive if the test is left too long before checking - a possibly fatal mistake! There are also potential problems with storage of the test materials and temperature in use and storage; so the reliability of the result has as much to do with the way it is used, as the test itself.
This is why it is essential that a positive test is confirmed with one of the other test systems.
IFA (ImmunoFluorescent Assay)
The IFA test is performed in a specialist laboratory, so your vet will need to send a blood sample away for the test, and this can take around a week to give results.
Western Blot (Protein Immunoblot)
The Western Blot is another test that needs your vet to send a sample of blood away
Each of these tests uses slightly different techniques, the details of which are not important to the cat owner, other than to know that a positive result from an IFA or Western Blot test is considered very reliable, assuming that potential infection took place more than 3 weeks previously, to allow time for the antibodies to develop.
Testing direct - for the virus
There are other tests that do actually test for the virus, but these are less frequently used, for different reasons.
PCR (Polymerase Chain Reaction)
The PCR test actually tests for the DNA of the virus itself, so can be used for kittens as well as adults; it can also be used at an earlier stage after infection, as it does not need to wait for the antibodies to form (which can take several weeks). There are not many labs that do the PCR test, so, depending on where you are, it may be difficult to arrange. The main problem with PCR is that it is known to give a high rate of false negatives; reasons are many and complex, but it does mean that it cannot be relied upon to fully eliminate the chance of infection. A positive result from a PCR is however, considered reliable.
One reason for false negatives is that PCR may not detect certain strains of FIV, so if it is used to check whether kittens actually have the virus, it is advisable, if possible, also to test the mother, to ensure the test can identify that particular strain; this will ascertain if a negative result is more or less likely to be reliable.
Virus isolation is the most reliable test of all, but it does involve prolonged laboratory work over several weeks, and consequently is realistically only used in research situations.
Accuracy of Elisa testing for FIV
The Elisa (in-house) test, even when carried out correctly, is not 100% accurate, but it is not possible to give a simple percentage as to its actual accuracy in use - for example:
If the test itself is only slightly inaccurate, say 99% accurate; then if 100 cats are tested, there will statistically be one false positive.
Now, what this means in practice, is dependent on the prevalence of the virus in the local cat population.
Estimates vary from 1% to 14% of healthy cats in an area that might be FIV+, and up to 44% of ill cats might be FIV+
Assuming for the purpose of this example that these estimates are accurate, then in the least infected populations where there is just 1% prevalence, there will statistically be one true positive in 100 cats tested; but then there will also be the one false positive (as described above); so 2 cats will test positive in the 100 tested cats, but only one will be true - that's an accuracy of only 50%
However, if on the other hand, there are 14% infected cats, then the same test of 100 cats will show 14 true positives and one false positive - - 14 out of 15 will be accurate - that's an accuracy rate of 93%.
And if we take the quoted 44% of ill cats being positive, the same test of 100 would give 44 true positives and one false positive - 44 out of 45 is nearly 98% accurate.
So you can see that any accuracy figure is highly variable depending on circumstances.
Remember also, that the way the test is carried out (timing etc), can affect the result - so a busy surgery might potentially not give the degree of attention required, which could then lead to further false positives.
What does this mean?
Not a lot more than common sense tells us - that the test is more likely to be accurate in high FIV-risk cats, and less so in low FIV-risk cats.
This is why you should always have a confirmatory lab test done on a positive result.
<back to top>