Corrections

The About page said this site is supported by display advertising. No advertising runs on it, and none has: no ad network is configured, no ad script loads, and the HTML contains no ad markup on any page.

Why: Found while checking whether the cookie banner described what the site actually does. It did not either: it said declining meant ads would not be personalised, when declining stops the ad script and Google Analytics from loading at all. Both now say what happens, and the banner names Plausible, which is cookieless and runs either way — the privacy policy had always disclosed that correctly, and only the banner implied every analytics tool sat behind the choice. Who funds a health site is among the first things a careful reader checks, which is why this is logged rather than quietly edited.

This page stated that no article on the site claims medical review. Every article was in fact emitting a lastReviewed date in its structured data, which asserts exactly that claim to search engines and anything else reading the page automatically.

Why: Found while auditing how the site describes itself. schema.org defines lastReviewed as the date the content was last reviewed for accuracy or completeness; it was being set from the frontmatter's updated date, so fixing a link or adding an FAQ block silently asserted a medical accuracy review nobody performed. The visible byline had always been gated correctly — this site does not name reviewers it does not have — and the structured data had simply never been held to the same rule. Two tool pages carried a hardcoded lastReviewed for the same reason and it has been removed from both.

A higher tinnitus rate in men was listed as one of the findings that shows up reliably across nearly every dataset. The largest meta-analysis on the question found no significant difference by sex, so it is not a consistent cross-study pattern.

Why: Found while adding the 2022 JAMA Neurology meta-analysis of global tinnitus prevalence, which this site had gestured at without citing or quoting. Its 83 pooled studies report 14.1% among men and 13.1% among women, explicitly not a significant difference. The male excess is real in particular cohorts; presenting it as a pattern that holds across datasets was the error, and the same overstatement was corrected in the what-is-tinnitus article at the same time.

The article stated a causal safety risk — vertebral artery injury from cervical manipulation — that the strongest available review does not support. That review found an association of very low quality and concluded there was no evidence for causation.

Why: Found while replacing a source. The article listed one of this site's own pages in its Sources block, which is circular; sourcing the claims properly meant reading the actual literature on cervical manipulation and artery dissection, and the best available review says something materially weaker and more interesting than the standard phrasing the article had used. The revised text states neither that it is dangerous nor that it is safe, because very low quality evidence supports neither.

An 80% habituation figure was reported as a share of all patients within 12 months. In the underlying study it is 80% of the 68% whose annoyance improved, over a mean follow-up of 33 months — a materially smaller and slower result.

Why: Found while re-sourcing the article. The claim had rested on two clinic marketing pages; reading the actual study behind the figure showed the number had been flattened into a simpler, more encouraging form than the data supports. The article now states the two milestones separately, because they are separate.

The article stated that VA had eliminated tinnitus as a standalone disability rating for new claims from April 2025. It had not. Diagnostic Code 6260 is still in the regulation.

Why: Found during a source audit. The claim rested only on law-firm and claims-service marketing pages. Checking 38 CFR 4.87 as published by GovInfo showed the standalone rating still in force, and the article now cites the regulation itself.

A compensation figure was attributed to VA's annual benefits reporting but could not be traced to it, and it disagreed with the figure used in another article on this site.

Why: Two articles on this site gave different totals. The lower figure is the one traceable to a VA page with a year attached; the higher one appeared only in secondary summaries without a year. The article now says so rather than picking the larger number.