February 11, 2009
There's nothing like hearing the results of studies directly from those who actually conducted the research. In this interview, you'll meet one of these impressive HIV researchers and read her explanation of a study she presented at CROI 2009. After her explanation, she answers several questions about her research.
I'm Shelley Facente, and I am the coordinator for evaluation and quality assurance for the HIV Counseling, Testing and Linkage programs in the San Francisco Department of Public Health. This study looked at results for about 21,000 patient tests that were done with the OraQuick Advance Rapid HIV-1/2 Antibody Test between January 2005 and December 2007, who had either an initial HIV negative result or a preliminary positive result on the test that was proved to be a false positive after confirmation with an IFA [immunofluorescent assay] or a Western Blot.1
Shelley Facente, M.P.H.
We looked at the results of this test based on the lot number, expiration date and other information that was systematically recorded by all the technicians who had completed the tests. We computed the time to expiration for the tests; the shelf life of the OraQuick Advance is six months. We looked at the day the test was run and how much time was there until the test actually expired.
The specificity of the test, according to the manufacturer, is 99.8%; two out of 1,000 people would be expected to get a false-positive result. With a 95% confidence interval, it's as low as 99.6%, so that would be as many as 4 out of 1,000 could be expected. We looked to see whether that actually bore out in our patient population, and whether it changed over time as it went closer to the expiration.
In San Francisco, we've had a number of times when there was a cluster of false-positive results -- where the test was functioning as expected, and then all of a sudden we had a number of false positives. This particularly happened in late 2005, and we had the manufacturer come in, CDC [U.S. Centers for Disease Control and Prevention], the state's Office of AIDS, and everyone tried to figure out: What could it be? We looked at many, many different factors: the test technicians, the storage temperature, all sorts of things. We were never able to find a pattern that pointed to what this could be.
This study is the first time we actually saw a pattern. When we computed the specificity for the kits with time to expiration, we broke them into six or more months to expiration, five months, four months, three months, two months, and then within the last month before expiration. What we found was that, of the 1,108 test kits that were used within a month of expiration, 13 of those -- which is 1.17% -- were false positive, which is 98.83% specificity. Which is far lower than the manufacturer suggests should happen.
As we looked over time, we found that the closer you get to expiration, the further that specificity drops. This was statistically significant within a month, even as we adjusted for test technician and test site. Within a month before expiration, it still remains statistically significant. And it was elevated in two months and three months before expiration, but the elevated rate at three months was not statistically significant once we adjusted for test technician.
What are the recommendations, then?
Our recommendation would be that additional analyses should be conducted with other data sets, because this is one specific analysis. If results are replicated in other places with similar data, then policy and procedural changes should be considered to discontinue oral fluid testing so close to expiration.
What about the manufacturer just changing expiration dates?
It's not really feasible with this version of the test to shorten the expiration date, only because it's already so short that by the time it gets out to places, a lot of test kits would go bad.
However, the manufacturer just announced about two weeks ago, I believe, that they received FDA [U.S. Food and Drug Administration] approval for an enhanced version of this product that actually has a 12-month shelf life, some additional quality control and some changes to the manufacturing process that hopefully will make this effect go away, and will increase consistency and stability of the product over the long term. Hopefully, that product will be available starting next month. We're excited to get it and start using it, and see if this effect goes away.
This is the crux of it: It was all about expiration date?
Well, I can't say that for sure. [Laughs.] All I know is, of all the things we have tried to look at to determine a pattern in the two years that we've been trying to figure out why we see these clusters of false positives, this is the only time we've seen a pattern that had any statistical significance. I certainly can't say that it's all related to this; like I said, additional analyses would be needed to see if this is true in other locations as well.
Thank you very much.