Interpreting HIV and testing statistics (HIV STATISTICS)
Oct 10, 2008
Greetings. I recently tested HIV+ on a rapid blood test, and am awaiting Western Blot results. Needless to say, I'm extremely worried, almost resigned. I've been reading about testing statistics and through questions on this forum, and have a few questions.
Firstly, I notice that a lot of people have questions about indeterminate results, or results on separate tests that bring back different results. I notice that in some answers, unless tests are conclusively negative or positive, that diagnosis can include risk assessment.
Well, I can pinpoint my exposure, as I had only one sexual encounter within the window period. The man I was with was rubbing and probing around my anus without a condom, inserted about a quarter of the way in, and then I stopped him and said he needed to use a condom, and we didn't even have oral sex. So, this might be seen as risky as unprotected anal receptive sex, as I was exposed to his precum, but it did not last long, he did not insert all the way, and did not ejaculate in me. This is the closest I've had to unsafe sex in years, and at the time, considered it foreplay and didn't worry. Two weeks later, I had classic sero-conversion sickness for 5 days, and five weeks after the exposure, tested preliminary positive.
Now, I am almost sadly sure that this means I am HIV+, just because the timing of the positive result is too perfect. However, the day I got sick was the same day that almost everyone I knew felt feverish and achy due to a sudden drop in temperature (perhaps a "bug" that was going around).
As I've read, ELISA tests are 99.9% sensitive and 99.8% specific, which means there's only a 3 in 1000 chance that I got a false positive. But then I read about prevalence at the testing site (i.e. I went to a supposedly low prevalence site, so let's say there will be 20 true positives and 3 false positives out of 1000 tests).
And of course, all I've been inundated with in education, HIV campaigns, and even on this site, is that in the risk assessment of my particular case, since I have not had unprotected sex otherwise . . . that I might be considered in a lower risk group. I've read over and over that transmission rates of receptive anal unprotected sex are about 1% (i.e. even if you're having risky sex it might take multiple exposures).
So I'm confounded by the fact that I may have gotten HIV by a single half-assed (literally) exposure, despite making such a "successful" effort to stay away from the drug scene, be very healthy, very direct and open in my communication with sexual partners, and vigilant about safe sex.
Obviously, I want to have hope, despite the statistics. I've read about quite a few false positive stories here, and I secretly *want* to hear the rapid blood test might be unreliable, that I might even have mono or just had a regular flu. I'm not finding my help on the web as I've found mostly outdated articles or answers that are too general. So, if you would be so kind, I have some specific questions.
- Would you say that the .3% false positive ELISA rate is accurate? If so, does this statistic mean that the .3% *includes* human error and all the possible complications like Epstein Barr and pregnancy or a recent immune system event that can cause a false positive, or does it mean that on top of that, the test itself, if performed perfectly, has a .3% fail rate?
- When it is said that the estimated rate of transmission via unprotected receptive anal sex is 1%, does this mean that if you are actually exposed to HIV, you will get it 1 out of 100 times; or does it mean that *real* exposure actually doesn't occur 99 out these 100 times? (In addition, can an actual positive status be result of "cumulative" exposure?) If it's the former, as I think we are lead to believe when reading statistics, then wouldn't my luck just be truly awful and almost defy the statistics?
- I know that I just must wait for my Western Blot confirmation results. But in that you've given your assessment based on risk factors, and with your knowledge of testing -- would you say I have any reason to hope that my preliminary test is a false positive?
Response from Dr. Frascino
I'll respond to your three specific questions first:
1. The sensitivity and specificity of a test assume the test is performed properly and free from confounding variables.
2. The estimated statistical risk assumes unprotected sex with a partner confirmed to be HIV positive. Please note, however, these general population estimated-risk statistics are useful in assessing relative risk and cannot be applied to a single specific encounter. (See below.)
3. Yes, there is a distinct possibility your initial rapid test was falsely reactive. Remember that a reactive (positive) rapid test is only considered to be "preliminarily" positive. If it is followed by a negative Western Blot, the two tests together are considered a "negative HIV test."
Your HIV-transmission risk was low, but not completely nonexistent. Write back with your Western Blot results. We are all rooting for you! However, no matter what the outcome, I'm here if you need me, OK?
Are you following what goes on in Mexico? (HIV STATISTICS) Aug 11, 2008
In AIDS conference in Mexico they said that the rates of transmission that the CDC quotes could be way off and actually be much higher than stated. They said some instances of Vaginal sex could be as low as 1 in 10 exposures and anal sex 1 in 3 exposures. I have always followed your advise and use a condom for both these acts( wife poz)but I am wondering if this were to be true could getting oral maybe turn into the 1 in a 1000. We never use condom when I get oral from her and really would just love to here your expert opinion on this.I know you said getting oral is a extremely low risk(theoretical under extenuating circumstances) but want to make sure this would not make you change your mind or at least think twice.(meaning do you think it could also be riskier than everyone has thought)THANKS, Here is the article
Response from Dr. Frascino
Not only am I following what's going on in Mexico City, I'm actually in Mexico City participating in the conference! There have been over 5,000 presentations, including the one you reference. The take-home message of this study of HIV infectivity is that population based HIV statistical estimates can not be applied as actual risk to one specific sexual action. I've made this point many times in the forum (see below). The presentation from the group at North Carolina focused on the effect of co-factors that can affect the risk of heterosexual HIV transmission (circumcision, genital ulcer disease, anal rather than vaginal sex and other biological cofactors). I'll reprint the article you reference below.
Condom broke exposing my turkey now i'm worried sick (HIV STATISTICS) Nov 26, 2007
Hello Doc. Hopefully you can help me with my fears. Eight days ago I had protected anal sex with another gay male of unknown HIV status. I was the top. After completion, which lasted less than five minutes, I pulled out and noticed the condom i was wearing had broke. I immediately went into a panic. I checked my penis and there was no sign of blood or anything else. I do not know at what point the condom broke. As the top, how much should I be worried? What are the chances that I might have contracted HIV, if indeed this person is positive and that's an assumption. Please help. I'm stressing out here and scared to death and depressed. What are your thoughts? Should I be worried? Thanks.
Response from Dr. Frascino
So it appears the "turkey" got exposed while you were doing the "stuffing!"
The estimated per-act risk for acquisition of HIV from unprotected (or broken-condom) insertive anal sex with a partner confirmed to be positive is 6.5 per 10,000 exposures. However, please note these are merely estimated risk statistics and cannot be applied as an actual statistical risk for your particular situation. (See below.)
Should you be worried? No, not excessively. But you should get a rapid HIV test at the three-month mark, OK?
HIV STATISTICS Sep 13, 2007
Ive written to you many times over the past 3 years and youve answered several of my questions. Thank you! What I really need to know now is how accurate are your statistics about oral and anal sex. Is it really 1 per 10,000 for oral and 50 per 10,000 for anal? Im trying to explain to my negative partner exactly what our specific risk is. Neither one of us are math whiz-kids but this seems reasonably straight forward. He could expect to become infected once for every 10,000 blowjobs. Right?
Thanks Dr. Bob
Response from Dr. Frascino
"He could expect to become infected once for every 10,000 blowjobs. Right?"??? Well actually no, that would be a wrong conclusion to draw from those statistics!!! I've covered this topic numerous times in the past, but I know HIV statistics can be a confusing topic. So even though this questions has now become a QTND (question that never dies) with an ATNC (Answer that never changes), I'll try to explain the limitations of these statistics once again.
The statistics I quoted are "estimated per-act risk statistics for acquisition of HIV by various exposure routes" published in a CDC document. These statistics were generated by combining a variety of published reports and did not control for many different potential variables that occur in different populations and among individuals. In other words, these statistics are primarily useful in determining relative risk, but not specific risk or actual risk for any individual. The reason for this is that any specific sexual coupling has a wide variety of variables to take into consideration when attempting to quantify specific HIV-transmission risk. These would include both viral factors, such as viral strain and viral load, as well as host factors, such as immune integrity, concurrent illnesses, circumcised/uncircumcised, genetic susceptibility, etc. Add to this nonspecific factors/extenuating circumstances, such as roughness of the encounter possibly causing trauma to mucous membranes, menstruation, etc., and perhaps you will begin to see the difficulty in providing transmission-risk statistics for any specific coupling. Also I should point out we cannot conduct prospective controlled epidemiological studies to try to account for theses variables, as that would be unethical. There are some published reports that address risk associated with specific sexual practices that control for some variables, but these studies usually have relatively small sample sizes and again are not applicable to everyone's specific situation. Another reference that I quote frequently is http://hivinsite.ucsf.edu/InSite?page=kb-07-02-02 (SAFER SEX METHODS). If you review the specific epidemiologic studies in this well referenced report, you'll get a better understanding of the complexity involved in these issues.
So why do I quote the statistics that I do? Good question! The main reason is that I am constantly barraged by anxious wrecks desperately trying to quantify their risk. I use the CDC statistics, because CDC is a very conservative organization and the numbers they generated are an amalgamation of many studies. They also standardized the relative risk to a common denominator ("10,000 exposures to an infected source"), which allows us to discuss relative risk. For instance, unprotected receptive anal sex is approximately 10 times more risky than unprotected insertive penile-vaginal sex, which in turn is approximately 10 times more risky than unprotected insertive oral sex.
I hope that this will help clarify the limitations of these estimated HIV-transmission risk statistics.
The bottom line is really much more concrete and easy to comprehend. If someone has placed himself or herself at risk for HIV, he or she should be HIV tested. Period. End of story.
I can just about hear all the paranoid panicky worried wells beginning to type away furiously, providing me with a blow-by-blow of their latest blow-by-blow and begging for me to quantify their specific risk. But unfortunately, unless the other person they were having sex with was me, I will not have enough specific detail to give them an accurate response. Hell, even if it were me, I still might not be able to give a completely accurate risk quantification!
Commonly Used HIV Infectivity Rate Misses Risks
Tuesday, August 5, 2008; 12:00 AM
TUESDAY, Aug. 5 (HealthDay News) -- A widely used HIV infectivity rate doesn't take into account multiple risk factors, say U.S. researchers who reviewed published data.
The heterosexual infectivity of HIV (the virus that causes AIDS) is often cited as a fixed value of one transmission per 1,000 sexual contacts. However, most studies estimating this value were conducted among stable couples with a low prevalence of risk factors, which can increase the risk of HIV transmission by several to several hundred times, according to Kimberly Powers, of the University of North Carolina, and colleagues.
They reviewed published data on HIV (up to April 2008) in order to estimate the effects of transmission co-factors that can affect the risk of heterosexual HIV transmission.
They found wide variations, with estimates ranging from zero transmissions after more than 100 penile-vaginal contacts in some serodiscordant couples (one partner HIV-positive and one partner HIV-negative), to one transmission for every 3.1 episodes of heterosexual anal intercourse, which is more than 300 times the commonly-cited fixed heterosexual infectivity rate.
Other infectivity differences, expressed as number of transmissions per 1,000 contacts, were:
13.2 for uncircumcised susceptible males vs. 5.1 for circumcised males.7.5 for susceptible people with genital ulcer disease vs. 1.5 for those without such disease.3.2 for early-stage index cases, 2.6 for late stage index cases, and 0.7 for mid-stage index cases.
"The use of a single, 'one-size-fits-all' value for the heterosexual infectivity of HIV-1 obscures important differences associated with transmission cofactors. Perhaps more importantly, the particular value of 0001 (i.e., one infection per 1000 contacts between infected and uninfected individuals) that is commonly used seems to represent a lower bound. As such, this value substantially underestimates the infectivity of HIV-1 in many heterosexual contexts...heterosexual infectivity can exceed 01 (one transmission per 10 contacts) for penile-vaginal contact or even 03 (one transmission per three contacts) for penile-anal contact. Claims in both the popular media and the peer-reviewed literature that HIV is very difficult to transmit heterosexually are dangerous in any context where the possibility of HIV exposure exists," the review authors wrote.
"Improved infectivity estimates -- especially more detailed estimates that quantify the amplifying effects of biological cofactors -- will help us to grasp the magnitude of the HIV epidemic, accurately communicate the level of risk involved in heterosexual sex, and identify the best possible intervention strategies for slowing the epidemic's spread," they concluded.
"Infectivity studies are very difficult to conduct, and that few studies exist as a result. Many of the studies producing the published estimates suffered from at least one potential bias. Therefore, while our study documents the considerable heterogeneity of the heterosexual infectivity of HIV-1 and provides some explanations for this heterogeneity, considerable uncertainty remains."?
The findings were expected to be announced at the International AIDS Conference in Mexico City. The review was published online inThe Lancet Infectious Diseasesand was expected to be in the September print issue of the journal.
The U.S. Centers for Disease Control and Prevention has more about HIV/AIDS.
SOURCE:The Lancet Infectious Diseases, news release, Aug. 5, 2008
Get Email Notifications When This Forum Updates or Subscribe With RSS
This forum is designed for educational purposes only, and experts are not rendering medical, mental health, legal or other professional advice or services. If you have or suspect you may have a medical, mental health, legal or other problem that requires advice, consult your own caregiver, attorney or other qualified professional.
Experts appearing on this page are independent and are solely responsible for editing and fact-checking their material. Neither TheBody.com nor any advertiser is the publisher or speaker of posted visitors' questions or the experts' material.