How We Judge Tinnitus Claims: Evidence Verdicts
Every claim page ends with one of five verdicts. They describe the evidence, not the people making a claim.
Supported
Good-quality evidence, ideally repeated by independent research groups, supports the claim as it is usually stated.
What it does not mean. It does not mean the approach works for everyone, is risk-free, or that the evidence cannot change.
Partly Supported
A real part of the claim holds up, but the claim as it is usually made goes further than the evidence does.
What it does not mean. It does not mean the claim is “half right” in a way that makes the rest acceptable. We say which part holds up and which does not.
Misleading
The claim contains a real kernel, but the usual way it is presented leads people to a wrong conclusion, for example by treating temporary relief as a cure, a result in mice as a human treatment, or a questionnaire improvement as a quieter sound.
What it does not mean. It does not mean the person making the claim intends to mislead. We judge the claim, not the speaker.
Unsupported
We found no good evidence for the claim. This covers two quite different situations: the claim was tested and did not show a benefit, or it has not been tested well enough to know. Each claim page says which applies.
What it does not mean. Unsupported does not necessarily mean disproven. It means the evidence needed to believe the claim is missing. A claim can be unsupported today and supported after better research.
Insufficient Evidence
There is too little, too small or too inconsistent evidence to say yes or no. Early or conflicting findings are in this group.
What it does not mean. It does not mean the approach fails, or that it works. It means we cannot tell yet, and more research could change the answer.
Six outcomes that claims blur together
A claim can be true about one of these and misleading about another. Each claim page shows which applies.
| Outcome | What it means |
|---|---|
| Loudness | How loud or intrusive the tinnitus sound itself is, usually rated by the person or matched to an external sound. A treatment that lowers loudness changes the sound. |
| Distress | How much tinnitus upsets, worries or frustrates a person. It is measured by questionnaires (for example the Tinnitus Handicap Inventory) and can improve while the sound stays the same. |
| Functional impact | How much tinnitus interferes with sleep, concentration, work, hearing in noise and relationships (for example the Tinnitus Functional Index). |
| Temporary relief | Feeling better for a short time, such as while a masking sound plays or for hours afterwards. It is real, but it is not the same as lasting change. |
| Sustained improvement | Benefit that is still present weeks or months after a treatment, measured against a comparison group. |
| Cure | Tinnitus that is gone and stays gone because of the treatment, shown in controlled studies and repeated independently. No treatment has met this standard for tinnitus. |
How we assign a verdict
- We judge each claim as it is commonly stated, not the topic in general. A topic can contain true statements and misleading ones, so each page also gives a verdict for each part of the claim.
- We rely on primary sources: the trials and reviews themselves, guideline texts, regulator records and trial registries. Every page lists them, with links.
- For most published studies, we reviewed the abstract and information indexed by PubMed rather than the full article, so we do not claim to have read every study in full. When a page relies on a long guideline or report, it says which part we read; where we read only a summary or an abstract, we say so and we do not describe the full document.
- We keep loudness, distress, functional impact, temporary relief, sustained improvement and cure apart, because a claim often wins its credibility by blurring them.
- We date every page. “Evidence through” is the date up to which we searched; “last reviewed” is when we last checked the page against its sources. A “first published” date appears only once a page is public. Verdicts can change when new evidence appears, and corrections are logged on our methodology page.
- We name a product, company or person only when a regulator, a published study or a documented fact-check supports the specific statement, and we do not accuse anyone of fraud. We distinguish “tested and shown not to help” from “never tested”.
Back to Tinnitus Claims: Fact vs. Evidence · Our evidence methodology · Corrections