Corrections

When we get something wrong we say so here, with the date and what changed. We do not silently edit pages.

What counts as a correction

The log

1. 2026-08-07 · Homepage, /tests/lactulose-vs-glucose/, and six other pages

   What was wrong: We published "the two common breath tests agree with each other on 8.7% of cases" as a headline statistic, including in the first paragraph of our homepage. That is not what the study found. The 8.7% is the share of the overgrowth DIAGNOSES made within a subgroup of 47 patients who took both tests, and the concordant cases were specifically those positive for methanogen overgrowth on both. It is not an agreement rate across cases, and we presented it as one.

   What we did: Found while resolving the citation itself, which we had also recorded under the wrong title: we were carrying a title that described the finding rather than the paper, which is how a citation nobody can look up survives review. The paper is Scalese et al. 2025 in the Journal of Clinical Medicine, doi 10.3390/jcm14248920. Reading the open-access full text produced the sentence we should have had from the start, including the authors’ own caveat that the subgroup was too small for a concordance index to be calculated at all. Every instance on the site now states the denominator, and the citation record carries the verbatim sentence. This is the same error we corrected in our own planning documents on 5 August: a real number, reported with the wrong denominator, which is the failure this site was built to be careful about.

2. 2026-08-05 · Site-wide, before launch

   What was wrong: Our own planning documents stated that 12,411 comments discuss SIBO and histamine intolerance together, and 29,013 discuss histamine intolerance and MCAS together. Both figures were wrong, and both were about to be published as headline statistics.

   What we did: The counts had been measured with the bare term "histamine", which matches every mention of antihistamines, of mast cells releasing histamine, and of histamine in food. It does not identify people discussing histamine intolerance as a condition. Re-measured with an explicit condition term list, the correct figures are 5,086 and 8,253: the original numbers overstated them by 2.4 and 3.5 times. Neither figure appears on the site today. We no longer publish these counts on the pages at all, because how we pick a topic is our problem rather than a reader’s, and every count we hold is now stored with the exact query that produced it.

Telling us we are wrong

If something here is incorrect, we want to know, and we would rather hear it bluntly than not at all. Point at the page and the specific claim. If you are right we will fix it and log it here with the date.