HSV-2 Transmission Risk and Condoms

[JohnB5000]

I have seen experts say condoms reduce risk 50%, or 30%, or 90% - all based on different studies.

But there is a new (2013) meat-study using case-crossover analysis that shows an 80% risk reduction.  This study is done by several of the same researchers who did the previous meta-study, but is statistically more robust.

Is this study (80%) the best estimate we have of risk reduction from condom use?

The study "Case-crossover analysis of condom use and HSV-2 acquisition" is linked here: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3334342/

Thank you

[Terri Warren, RN, Adult Nurse Practitioner]

I don't know about you, but the statistics in that paper spin around in my head.  There is such dense data in that paper!  I wrote to Anna Wald about this paper, one of the authors and the country's leading expert on genital herpes.  I asked her what she would now describe as the reduction of transmission risk with the use of condoms and she said about 50%.  This is what we have always told people, and I don't see her advice as different, even with the publication of this paper.

I'm wondering if you might be a person who has called my clinic a few times and spoken with Sheryl about these statistics.  John B, 50% is the best guess that we can give you at the point, even including this paper.  

If you ask me more, I am afraid I must give you the same answer.  

Terri

[Terri Warren, RN, Adult Nurse Practitioner]

John, I believe that you likely have every article written about herpes and condom use already, correct?  When we try to advise people on the risk of acquiring herpes, given XYZ, I think both Anna and I and Drs. Hook and Handsfield all try to look at averages of the studies done on this topic.  The same is true of the topic of the likelihood of transmission of herpes to a partner.  Those numbers are even more all over the place, so we try to come up with averages.

I don't know if you have herpes and you want to reassure a partner or what your situation is exactly.  But if you do have herpes, I'm certain you will talk about this research with a partner and share your thoughts about this 2013 study.  

So rather than saying that we kind of want the easy way out so we give a number like 50%, I hope you will realize that this one study show one set of numbers but other show different numbers, and though this study is strong, we can't say that the others have no value because their results are different.

John B, when is comes to this topic, I will defer to the author of this study, Anna Wald.  Upon my specifically asking Anna does this study result in a different number that she believes more accurately reflects the reduction in transmission of herpes, and she continues to say 50%, then that's the number that I will share with you and with others.  

Given that, you are certainly free to sort out this information in the way that you feel is most reflective of all the research studies on this topic.  Thank you so much for your input on this topic.  It is clear that you've given this a lot of thought and research.

Terri

[JohnB5000]

Terri, do you have any data or research you could mention that would support the 50% risk reduction estimate for condoms? I would love to be able to understand the data behind the number (or to read the study it comes from).

Thank again.

[JohnB5000]

Terri - Thank you for the information and for reviewing the paper - It is hard sometimes to make sense of the data, but I appreciate everyone's assistance very much.

[Lucky49er]

Thank you so much Terri, for your reply. I agree the data was dense in the paper, but the statistical conclusions are as compelling and robust as any other expert study I have seen so far (in some ways much more so). What first caught my attention was the conclusion: "The confidence interval surrounding the OR for the risk of acquisition with protected acts included 1.0; as an OR of 1.0 corresponds to 100% efficacy, these findings are consistent with but do not demonstrate complete protection."

Now, I understand that "50%" is a useful recommendation to the public to convey "some protection but not total protection."

But for the single people in this study, the risk reduction from condoms was 86% per act, and applying a 95% confidence interval indicated that "these findings are consistent with but do not demonstrate complete protection" - ie the upper bound was over 100%, which is by definition complete protection.

We know that the single people in this study (a meta study of all the data available) were protected by condoms by much more than just 50% (quite obviously) - indeed, the risk reduction for singles was 86% for protected vs unprotected acts (all of the numbers are in table 3a of the study, for reference).

We could still say "50%" because we don't know for sure - except that we do know for sure that it isn't "50% exactly for everyone", it is a very simplistic summation, and it is clearly not supported if we are talking about the data for single people in this study. And 50% is a lot different than "30%" and a lot different for "86%".

I think people in the medical community in general should be more forthcoming about where the numbers they cite come from and the confidence intervals behind them (saying 50%, because maybe it's nothing and maybe its 100%, is the same as saying "I don't know" or "its a coin toss" - I think we (you, me, the public) have access to data that is a lot better than that).

Maybe we could say "the 2013 study showed single participants reduced risk per act by 86% by using condoms, and the 95% confidence interval included 100%+ efficacy on the higher-end, and as low as 40% on the lower-end" . . . now that means something to people. People can understand that condoms are a really good idea, not "a coin toss" or "50%" protective". We know that "50%" is within the range of the 95% confidence interval here, but it is not in the middle of the range, it is near the very lowest edge of the confidence interval - in fact "100% efficacy" is statistically much more likely than "50% efficacy", based on the data analysis for single people in this study.

There are caveats, but anyone who takes the time to read this study can find a more accurate answer than just "50%" . . . I have also seen you personally reference "30%" in the not-to-distant past (I assume this was based on the older meta study - which preceded this one - it was probably the best research available when it came out, so clearly no fault of yours to rely on it) . . . but if you have updated to "50%" now as a result of this conversation and this new study . . .Then I think it's a great step in the right direction, and I commend you for your willingness to review the more recent research that you may not have had a chance to read before, and take it into consideration. If condoms are more effective than people thought, then the public needs to know - because one of the best things that can happen is more people being encouraged about using condoms more of the time.

I acknowledge that there is a difference between guideline recommendations to the public and "detailed implications of the latest research" - but guidelines do change over time and they need to be updated sometimes. This 2013 study is much more statistically relevant than a simple "50%" that doesn't have any back-up numbers or a confidence interval behind it. It's easy to say "50%" and "keep it simple", but without any data to back it up, I don't find it convincing.

The CDC still doesn't recommend standard testing for HSV2, but hopefully someday they will read the research that is out there and realize that standard testing would educate people, do much more good than harm, and help prevent the spread of this disease (and other more serious ones, where HSV2 is a risk factor) - but so far, they have been very slow to move to protect public health on this issue. I wish everyone would try to be more data-oriented and open minded - knowledge is power, after all (for the public and for everyone else too).

Thanks again for your thoughts.

[JohnB5000]

* should say "meta-study" not "meat-study", haha