[Question #519] Is test accuracy a taboo?

36 months ago
I recently had a negative herpeselect test for hsv2 after five months of celibacy.  Naturally I looked up the reliability of the test, and it compares well to the Western Blot gold standard.  The problem is that is impossible to find consistent info on the HSV Western Blot accuracy.  Univ Wash posts 99% specificity but nothing on sensitivity.   The only thing I could find was an older quote from Dr. Hook:
www.medhelp.org/posts/STDs/Dr-Hook--HSV-2-question/show/527261
And that was at an abysmal 93%... surely the 'gold standard' couldnt be missing that many folks??  Also, Dr. Handsfield mentioned the test reliability a couple times but never really gave his position on it.  Why is this data so elusive?  Is there not an official number? 

Terri Warren, RN, Nurse Practitioner
Terri Warren, RN, Nurse Practitioner
36 months ago
The issue with any gold standard test is that there is nothing better to compare it to.  So for example, we know, that compared to PCR swab, the culture swab test misses up to 75% of infection - the PCR is far superior but there is nothing better than PCR so we don't know the sensitivity.  The same idea applies to the western blot.  It has been my experience that in 33 years of practice, the western blot has missed 9 people who are swab test positive for HSV 2.  But admittedly, we don't do western blots on every person who is swab test positive because there is no need to confirm a symptomatic person with a positive swab test.
I am not reading Dr. Hook's statement as you are - are you reading that the western blot is only 93% sensitive?  I think he was referring to the screening test and in our most recent study, we found that for people who had both and ELISA and a western blot done, there was 98% agreement between the two tests for HSV 2 - the western blot picked up 2% more than the ELISA.  We are doing another study now comparing western blot and ELISA with more subjects and as those data come out, I will post something here about the results - they could vary.  I can only report to you what we have so far.

Terri
---
36 months ago
I guess I am still fuzzy on the WB then...  Has there never been a study that observed/counted the amount of PCR/culture proven patients who went on to seroconvert by WB and those who failed to?  With other tests, like HIV blood tests for example, the sensitivity is pretty readily available, yet it is mysterious for herpes...  
It is hard to gauge if your 9 patients is a big number or not, right?  Because we dont know the amount of patients that were swab tested and successfully seroconverted to compare it to?  I appreciate all the info this and other sites have, but you can see how this lack of info would be a source of anxiety for folks, right?
I was really hoping for a solid number or at least a conservative ballpark consensus among experts.  If Dr. Hook's comments are true, and the 93% is accurate for screening tests, and your study says the WB only picks up 2% more than that, that still leaves pretty many folks with false negatives.
Terri Warren, RN, Nurse Practitioner
Terri Warren, RN, Nurse Practitioner
36 months ago
In our study, 98% of those who test positive by the screening test were also positive by western blot - saying the screening test didn't miss many people compared to western blot.  Sorry if that was put in a way that was confusing.  That addresses the SENSITIVITY of the IgG screening test, compared to the gold standard.  The area of difficulty with the HSV 2 screening test is not that it misses people is that it says some people are positive who are not.  Of those who swab test positive, the western blot confirms about 99.5% of those (l. corey).  Nine patients in 33 years is an extremely tiny number.  I see literally hundreds of herpes patients in a year.  And other people in my clinic see herpes patients also.

Terri
---
36 months ago
Thank you, Terri, those are much less scary numbers than 93%...  I wonder why Dr. Hook was so low...  Who/what is (l. corey)?  Is that from a study?  If 99.5% is the real sensitivity for the WB that is pretty great really, if that is the official agreed upon number?  

One last question, since this is my last reply... I read that genital herpes is most infectious in the first year.  Does the shedding tend to drop off a lot after that first year or is it gradual?  How much higher is transmission risk in for example the second year? 
Terri Warren, RN, Nurse Practitioner
Terri Warren, RN, Nurse Practitioner
35 months ago
Dr. Lawrence Corey at University of Washington is the leading herpes researcher in the world, and it is the number that experts use, based on years of research with this test.  The 93%, as I mentioned, is likely the screening test rather than the western blot. 
Shedding is most in the first six months of infection, then there is a decrement again at 2 years, then likely again around 10 years.

Terri
---