Why Tests and Q #6 Should Both Be Served with Salt

Note: I am going to skip blogging about question #5 (how long people have been in treatment) for now. I'll be using it mainly as a cross-reference for other questions, and might circle back around to it later as its own topic. Onward!

++++
How we actually get diagnosed is a very important topic for the lyme community. Understanding the most effective ways to diagnose lyme can help doctors make better decisions in the moment of how to approach a suspicion of lyme, and these better decisions can ultimately lead to better outcomes.

Unfortunately, I think I blew this one in the set up. (This is why I'm desperately seeking someone with social research experience to help next time--to stop me from doing goofy stuff that makes the data worthless... Are you my sociologist?)

Here's the answers as they came through on the survey:

 Here's the problems with how I set this up:

1) I made an assumption that, because the phrase "CDC positive" gets bantered around a lot, people would understand what I was asking and be able to answer this question appropriately (or could get the information from the link I provided). Of course, this was silly, because I had to go looking and wracking my own brains to try to get it, so why would I think other folks had a better handle on it? This confusion became clear when reading some of the comments, as well as the fact that a few people checked multiple boxes on the three categories of positives. It also just doesn't seem that likely that we have this many CDC positives given our sample size (unless of course the CDC numbers are going to be a lot higher this year).

2) I realized afterward that no one is entirely without a clinical part to the diagnosis, and that people would probably relate to this answer in very different ways. What I mean by that is that no one goes to the doctor feeling perfect and, just for the heck of it, asks to have a lyme test run... there's always a clinical aspect. Someone may not have said that's how they were diagnosed unless it was the only choice on my list that fit, but then again, others checked multiple things, including clinical. We tend to think of testing as a more "real" or "definitive" diagnostic tool, and so not think of the clinical part as being as "good" of an answer. This is a bias in western medicine, and one I walked right into.

3) I should have phrased the answer about bullseye rashes the same in this question as I did in a later question, because the results from the two questions were 6% points different. Maybe that's an OK error of margin, or maybe not. Either way, it was sloppy of me and puts us on less firm footing.

So I want us to view this question (more than any other in the survey) with a very large grain of salt. See, there's plenty for everyone:

OK, so all that said, the thing I find most interesting is the low number of bullseyes. I read on wikipedia today that "only about 80%" of people get the rash. "Only?!?" My survey puts this at between 14.9% (on this question) and 20.7% (on the later one). Even if my set up was wonky, I feel very confident in saying that the number is nowhere close to 80%. Does anyone know how to get wikipedia to change their listings?


What's the deal with the CDC anyway?

Given the confusion a lot of us seem to have about the whole CDC thing, I'm going to focus there for a chunk of today's blog.

The way the system works, if your test comes back fitting the CDC criteria, it is supposed to get reported to the CDC and you become part of their official count (22,750 cases in 2010.) If we knew very accurately the percentage of CDC positives in our study, then we could do some fancy schmancy math stuff and extrapolate out how many overall cases there might be in the US. It would have been fun. (And probably represents my own delusions of grandeur.) Thus my motivation for asking about it.

Here's what the CDC means by positive (snagged directly from their website):

In addition, by looking at the information provided with my son's test results (from Igenex), here's what the IgG and IgM are supposed to look like for that second (Western Blot) test. You need 2 positive bands on the IgM (out of 12 they still test for) and 5 positive bands on the IgG (out of 12.) These standards vary from country to country, too. The number of bands that qualifies as positive in the US isn't the same as it is in Germany or Scotland. (OK, glaring problem #1: how many of get a positive and then are going to turn around and spend the money, time and puncture wounds to run another test?)

Now the CDC even says that doctors shouldn't be limited by their testing standards in daily work with patients. In fact, here's what they do say about diagnosis: "Lyme disease is diagnosed based on symptoms, physical findings (e.g., rash), and the possibility of exposure to infected ticks; laboratory testing is helpful if used correctly and performed with validated methods." That's right, the CDC doesn't emphasize testing as the primary mode of diagnosis, but rather clinical observation. (Which is a good thing when it comes down to it, since so many of us test negative.)

Makes me think maybe we should stop picking on them so much. OK, well, maybe a little--there's still some goofy stuff on their website.

Did you know that Western Blot testing was once considered to be more accurate (and still could be without a change in the technology we use)? The reason is that more bands used to be looked at and count. Igenex testing is considered more accurate in part because they look at more bands... it is a simple concept--look for a wider range, and you are more likely to find what is there; this is sound science so long as they are all lyme.

There are many strains of "wild" lyme... a lot more than can easily be studied in labs. (And they are apparently very difficult to cultivate in a lab setting... the folks working our our behalf don't have an easy time of it!) And the lyme spirochyte is very good at adapting, so this problem isn't likely to get more simple. The chances of your particular strain showing up on enough bands to qualify as CDC positive are all over the map. If you get infected in an area that is dominated by the spiro bugs that the bands were originally based on (and they haven't evolved much in the wild since the tests were created) then you stand a much better chance of throwing a positive. For instance, here in Missouri, our local variant is often referred to as Master's Disease, not even lyme. I've been told (but haven't confirmed) that our local bugs don't show up most of the time on tests. So we have to take it with a grain of salt.

One of the complications with lyme is that evolution doesn't stop, and spirochytes have been at it for a very, very long time and learned a lot of tricks... you can expect new strains to be emerging in an ongoing way. And we won't really know how quickly the older research loses its relevance.

Dr. Stephen Buhner, on p. 66 of his book Healing Lyme, offers a simplified way that doctors can read a good Western Blot test. He says,"... a Western Blot assay with a minimum of two bands, one being 41kd and one other being lyme specific, is an excellent indication of infection." (I thought this was interesting, because that was the exact profile of my son's test, and the hospital called it a negative.) He also quotes from a study that tells us that, of confirmed lyme cases being studied, "...4.8% of the cases no IgG bands were present and in 26.2% no IgM bands were present."* So even this more useful "excellent" indicator should be taken within the context that some people with lyme won't present that way. And it means that if you get back a negative, look at it as one more chance for salt.

Going back to the CDC diagnostic approach, clinical diagnosis should always lead the way, with testing being seen as a back up. And when it comes to testing, there's enough grains of salt available to raise the blood pressure of the dead.

+++++++++++++++++

*Hernandez-Novoa, B. et al. Utility of a commercial immunoblot kit (BAG-Borrelia blot) in the diagnosis of preliminary stages of lyme disease, Diagn Microbiol Infect Dis 2003.

Transparency: My positive test didn't rate CDC; mainly I was a clinical diagnosis. No bullseye.