A study from Madrid has found that 92% of people taking an HIV test could obtain a valid result by testing themselves with the kind of fingerprick point-of-care blood test used at clinics, without any instruction except the brochure the test came with.
The Madrid study also tested people’s ability to recognise positive, negative and invalid test results when shown photos of typical tests. Just over 5% of negative test results were interpreted incorrectly (2.7% as false-positive) and 3.6% of positive test results were interpreted incorrectly (1.1% as false-negative).
Participants were not told their actual test results till the very end of the testing process. Nine out of 519 participants (1.7% or one in 58 participants) had HIV. Out of these, eight had correctly interpreted a picture similar to what their own test result looked like, while the ninth was uncertain whether the picture closest to his own test result was positive or invalid.
After self-testing, 84% of participants said they were more motivated to self-test for HIV again.
- Another study in the same project assessed participants' knowledge of post-exposure prophylaxis (PEP) - see this report.
How the study was conducted
The study set up three testing sites in five locations in the Madrid region – in Chueca, the city’s central gay district, next to two railway stations, and in two university campuses. The test sites were in tents; the researchers note that the trial period, October 2009 to February 2010, was “a particularly cold and rainy winter,” and suggest that tests taken in the comfort of people’s homes might produce more reliable results.
The study was carefully structured so that participants received no instruction from staff in how to use the test. First, 313 out of 519 participants were given a brochure with written instructions and illustrations on how to use the test. They then used the kit to test themselves with staff observing but not helping, as if they were doing it unassisted at home. Another 207 participants were, for purposes of comparison, given help and advice when needed.
The test used was the Determine HIV1-2 antibody/p24 antigen blood test. It consists of pricking a finger with a small lancet and squeezing two drops of blood into the two test windows that detect HIV antibodies and p24 antigen. In the test kit supplied to clinics, medics are supplied with a small glass capillary tube to pick up a precise quantity of blood, but untrained participants found this very difficult to use, and the instruction to deposit “two large drops of blood” appeared to provide adequate results.
After being asked about their experience of doing the test, participants then saw a healthcare worker who, after providing HIV counselling required by Spanish law, took a second confirmatory blood test using the same Determine test kit, only this time using the capillary tube.
The Determine test takes about 20 minutes to provide results. During this time, participants were each shown six photos of various positive, negative and invalid test results and asked to interpret them based on their reading of the brochure. A variety of sets of six photos were used in order to generate a large data set.
Participants were also asked to give feedback during this point on their experience of taking the test and of interpreting the results, whether they were motivated to self-test again and how much they might be prepared to pay if the Determine test was sold over the counter. They also completed a paper questionnaire on demographic factors and risk behaviour.
Finally and only at the very end of the process, they were given the result from the staff-administered test, with appropriate counselling and referral.
Results in detail
Recruitment was entirely voluntary: participants simply walked into a testing tent. Fifty-six per cent of participants were recruited at the single tent in Chueca. Despite this being the ‘gay village’ only 29% of all test participants defined as gay men; 34% were women. Forty-seven per cent of all participants had taken an HIV test before, and 51% of those who performed the test unaided (not a significant difference). A quarter of enquirers decided not to participate (19% of those who tested unassisted). There were no demographic differences between those who participated and those who decided not to.
Eight per cent (25) of those who took the test unassisted produced an invalid result; gay men were less than a third as likely to produce an invalid result as other participants.
About five per cent of participants interpreted at least one test result picture wrongly as positive, negative or invalid when it in fact showed a different result. About 2.7% of participants wrongly identified invalid or negative results as positive; 1.5% identified invalid results as negative and 1.1% identified a positive result as negative. People over 30, compared with people under this age, were twice as likely to interpret a test result wrongly and people not born in Spain or with no university education somewhat more likely.
Ninety-three per cent of participants who produced valid results and 80% who produced invalid results said self-administering the test was “very easy” or “quite easy” and 98% of those who correctly identified all six test result photos and 77% who made at least one mistake said interpreting photos was very or quite easy.
There were a couple of individual cases where people experienced some psychological distress: in one case the person felt dizzy after drawing blood and in the second the person became frightened of the possibility of a positive result after taking the test and while looking at the photos of results.
No-one said after self-administering a test that it had decreased their interest in using such a test again and 80% said it had increased their interest.
When asked how much they might be prepared to pay for such a test, a third said ten to 19 euros, a quarter said 20 to 29 euros and one in five said €30.00 or more. The OraQuick test is now on sale for $40.00 in the US, which at today’s exchange rate is about €31.00 or £25.00. The researchers comment that, given the recent questions about the validity of the p24 antigen test, it should probably not be included in a blood test designed for home use, which might make it cheaper.
Implications
This study is important partly because the blood test concerned has a shorter ‘window period’ than OraQuick, the HIV test using oral fluid sampling that has been approved for home-testing use in the US and is now on sale over the counter. OraQuick has a considerable window period. In the trial that led to Oraquick's approval, 7% of people who really did have HIV received false-negative results.
The window period for the Determine HIV antibody/p24 antigen blood test, used in this study, is likely to be somewhat shorter.
As the researchers say, “This is the first published study showing that a high percentage of HIV-negative people are able to perform a blood-based point-of-care test and read the results correctly.”
They urge further research into self-testing in different populations and with different kits.
De la Fuente L et al. Are participants in a street-based HIV testing program able to perform their own rapid test and interpret the results? PLoS One 7(10): e46555 online publication. October 2012.