Biotoxin Illness Test Results

Updated on October 9, 2013


I’ve compiled the Biotoxin Illness Test Version 1 results from August 26, 2014 to October 6, 2014. A total of 153 people took the test. Of the non-moldy people that took that test, 71% (5/7) were diagnosed as mold free, with the other 29% being diagnosed as either slightly or moderately at risk of having issues with biotoxins. Out of the confirmed Veteran and Patient groups that signed up to be “Helpers” (people that took the test based upon how they felt during the block of time when they were most symptomatic), 92% (35/38) were correctly diagnosed with Biotoxin Illness (having 8 or more clusters) while only 3% of the remaining people were completely miss-diagnosed as not having any issue with biotoxins (less than 3 clusters). In general, the test looks like its working.

Classification Mold Free Newbie Student Veteran Veteran Helper Patient Patient Helper Injured
No. Tests 7 41 29 12 7 41 31 24
Ave. Clusters 2.9 10.2 11.4 11.7 12 11.2 11.1 12
Ave. Symps. 4.4 26.3 32.4 32.1 33.3 33.2 32.9 35.1
3 ≤ Clusters 5 2 0 0 0 1 1 0
4-5 Clusters 1 2 0 0 0 1 1 0
6-7 Clusters 1 1 0 1 0 1 1 1
8 ≥ Clusters 0 36 29 11 7 36 28 23


  1. Mold Free    – mold is not an issue.
  2. Newbie        – uncertain that mold may be contributing to impaired health.
  3. Student        – studied somewhat and believe mold is affecting health.
  4. Veteran        – studied a lot and improved health.
  5. Veteran Helper – studied a lot and improved health (Test Helper).
  6. Patient         – have or had many symptoms and Dr. Shoemaker labs.
  7. Patient Helper – have or had many symptoms and labs. (Test Helper)
  8. Injured         – after being exposed to indoor mold, health has declined.


Dr. Shoemaker’s symptom diagnostic algorithm uses 35 specific symptoms that are divided into 13 groups/clusters with anywhere from 1 to 5 symptoms in each cluster. These clusters were selected by a statistician in order to maximize their predictive capabilities. When a person has a one or more symptoms in a given cluster, that cluster is counted. Based upon the total number of clusters counted, the likelihood that the person is suffering from Biotoxin Illness (chronic inflammation due to an impaired immune system) is determined.

  • 3 or Less Clusters – Biotoxin Illness is not an issue
  • 4-5 Clusters – Biotoxin Illness might be an issue
  • 6-7 Clusters – Biotoxin Illness is probably an issue
  • 8 or More Clusters – Biotoxin Illness is almost certainly an issue


Below are graphs for each of the 8 classifications of people that took the test. The Veteran and Patient groups have two graphs each. One graph includes all the data for the group. The other only includes data from individuals that indicated they were “Test Helpers” – taking the test based upon how they felt prior to treatment. The 15 symptoms that are underlined were not included in Dr. Shoemaker’s algorithm but were included in this test. I added these mold symptoms out of curiosity and to see how they would compare with the other 35 symptoms used to grade the tests. (I’ve removed the extra 15 symptoms on version 2 of the test.)

1 Mold Free Aug-Oct 2014 2 Newbie Aug-Oct 2014
1 Mold Free Aug-Oct 2014 2-Newbie-Aug-Oct-2014
3 Student Aug-Oct 2014 4 Veteran Aug-Oct 2014
3-Student-Aug-Oct-2014 4-Veteran-Aug-Oct-2014
5 Veteran Helper Aug-Oct 2014 6 Patient Aug-Oct 2014
5-Veteran-Helper-Aug-Oct-2014 6-Patient-Aug-Oct-2014
7 Patient Helper Aug-Oct 2014 8 InjuredAug-Oct 2014
7-Patient-Helper-Aug-Oct-2014 8-InjuredAug-Oct-2014


For the purposes of evaluating the test, only data from the Mold Free, Veteran and Patient groups were evaluated. More specifically, only Veteran and Patient people that also selected to be a “Test Helper” were used to determine test accuracy. Non-Helpers in these two groups may have been taking the test based upon their symptoms after treatment and as such, their data is less reliable. Likewise, folks in the other 3 groups (Newbie, Student, and Injured) were not included in part because it’s not clear what’s behind their symptoms – mold, Lyme, parasites, and so on.


  1. There is a glaring difference in the number of symptoms between moldy and mold-free test participants – 33 to 4. This is consistent with Dr. Shoemaker’s work albeit higher (22 to 3). Note: Remember that for the purposes of evaluation, only Veteran and Patient Helper groups were considered.
  2. According to Dr. Shoemaker’s data, mold-free people have about 3 symptoms on average. Although the number of mold-free people that took the test was small, the average number of symptoms was 4.4. This seems a little high but reasonable.
  3. The Mold Free group should have had a higher number diagnosed as non-moldy – only 71% were correctly diagnosed. However, without testing, we can’t be certain these people truly are non-moldy. They may be suffering from Biotoxin Illness and not know it. Furthermore, the test group was quite small so a single outliner can throw the percentage off by quite a bit.
  4. Based upon Dr. Shoemaker’s work, moldy people have about 22 symptoms on average. In comparison, the average number of symptoms for folks with Biotoxin Illness (Veteran and Patient groups) that took the test was 33. This seems too high. I can think of 4 possible reasons.
    1. Reporting Inaccuracy – didn’t read the “more…” information, was too long ago to remember, rushed the test, etc.
    2. Test Methodology – When I was examined by Dr. Shoemaker, he did not go through a checklist. Instead, I mailed in my history in advance to the first office visit. Later, I recounted my story in his office. He asked plenty of questions but only a few were about symptoms. The test goes through all 35 relevant symptoms. I suspect scores are higher in part because of this difference in format.
    3. Timeframe – It may be that at the time folks were evaluated by Dr. Shoemaker, not all symptoms had presented. The test allows folks to recount all there symptoms over a possibly longer period of time. Additionally, most folks are no doubt feeling better and consequently their brains are now able to recall more of their symptoms.
    4. Internet Type – I suspect that the Veteran and Patient folks that are also still actively involved in learning about mold and consequently found their way to the test tend to be more beat up by mold. Its not hard to imagine that folks with less problematic genes and fewer symptoms simply got better and moved on with their lives.
  5. It was interesting to see that 3 mold symptoms that were not included in Dr. Shoemaker’s algorithm, nonetheless scored very high with moldy people and very low for mold-free folks – noise sensitivity, poor sleep, and wired-and-tired. This suggests the test algorithm could perhaps be tweaked further.
  6. Not surprisingly, people with more experience became more confident that they were moldy and were willing to become “Test Helpers” – there were fewer Newbie/Student/Injured helpers compared to Veteran/Patient helpers.

Final Words

The test seems to be working albeit with a slight bias toward over-diagnosing Biotoxin Illness – although this bias may simply be due to the differences noted. In an effort to address this bias, I’ve added notes for people to take their time and make sure to read the “more…” information. Additionally, I’m considering adding comments that try to describe not only what a symptom is all about but also, what it’s not. For example, “Chronic Fatigue – this does not include feeling tired after a days work” and so on. I’ve launched version 2 of the test and have improved the symptom descriptions based upon input from moldy family members and the single person that contributed – thanks Gail G. 🙂 If anyone has additional insights or suggestions, please post them.

Microsoft Excel Test Data


Current Version 2 Data in Raw CSV Format