• Skip to main content

Everyday Prevention Evidence

Read prevention guidance with context, numbers, and clear limits.

  • Home
  • Start Here
  • Research
  • Editorial
  • About
  • Contact

Corrections, Expressions of Concern, and Retractions: What They Mean for a Health Study

posted on September 9, 2026

Updated October 9, 2026

By the Everyday Prevention Evidence Team

When a health study you’re reading about gets a “correction,” an “expression of concern,” or a “retraction,” these are three different things — and they don’t all mean the same level of trouble. A correction usually fixes a small, isolated error. An expression of concern flags unresolved doubt about the work while an investigation continues. A retraction means editors no longer stand behind the findings. Knowing which one you’re looking at tells you how much weight to still give the study.

The three notices, from least to most serious

Journals and research databases like PubMed track these notices and link them to the original article. Here’s what each one signals, roughly in order of severity.

Correction (also called an erratum)

A correction fixes a specific, limited error — a wrong number in a table, a mislabeled figure, an author’s name spelled incorrectly. The rest of the study’s findings and conclusions are still considered reliable. The original article stays published; the correction is added alongside it.

Expression of concern

An expression of concern is a “we’re not sure yet” notice. According to the Committee on Publication Ethics (COPE), editors use this when credible concerns have been raised about a study but the evidence isn’t clear enough yet to decide whether the findings are unreliable — often because an investigation is still underway or the authors haven’t yet supplied information the editors asked for. It can be a temporary placeholder that later turns into a correction or a retraction, or it can stand as the final word on a paper that’s never fully resolved.

Retraction

A retraction is the most serious notice. Editors issue one when they no longer trust the results and conclusions in a paper — because of major errors, data or image problems, plagiarism, undisclosed conflicts of interest, or research misconduct. Per COPE guidelines, the goal of a retraction is to correct the scientific record, not to punish the authors. The original article usually stays online, but it’s now marked as retracted and should no longer be treated as reliable evidence.

What each notice does — and doesn’t — prove

It’s easy to read too much into a notice, in either direction. Here’s a quick way to keep it straight:

  • A correction does not mean the study is untrustworthy. It means one specific detail was fixed. The core findings usually still stand.
  • An expression of concern does not mean the study is wrong. It means something is unresolved. The findings might hold up, or they might not — that’s exactly why the notice exists.
  • A retraction does not automatically mean fraud. Retractions also happen for honest mistakes, like a flawed analysis method or contaminated samples. But once a paper is retracted, its conclusions shouldn’t be cited as evidence going forward.
  • None of these notices tell you why on their own. The notice itself usually states a reason — read it, don’t assume.

How to check whether a study has one of these notices

Before you share a health claim based on a study, it’s worth a quick check:

  • Search for the study on PubMed and open its citation page. Notices are linked to the original record when one exists.
  • Look at the top of the record for words like “Retracted,” “Correction,” or “Expression of Concern” attached to the title.
  • If you found the study through a news article, a summary, or a database other than the publisher’s own site, be aware that corrections don’t always reach every copy immediately. PubMed itself notes that when it corrects a citation, other databases and vendors that use PubMed’s data may not reflect the change right away — so the version you’re reading somewhere else could be out of date even after PubMed has updated its record.
  • If something in the citation looks wrong to you and it’s a genuine indexing error, PubMed’s own guidance is to contact the journal publisher directly for content errors, or the NLM Help Desk for indexing errors — that’s not something an individual reader can fix themselves, but it’s useful to know the correction path exists.

A short checklist before you share a claim

  1. Find the original study, not just a summary of it.
  2. Check the top of its record for a correction, expression of concern, or retraction notice.
  3. If there’s a notice, read what it actually says before deciding whether the finding still holds.
  4. If it’s a correction to a minor detail unrelated to the claim you care about, the finding likely still stands.
  5. If it’s an expression of concern or a retraction touching the finding you’re relying on, treat the claim as unresolved or unsupported until it’s clarified.
  6. When in doubt, look for whether other, independent studies reach a similar conclusion — one study, corrected or not, is rarely the whole picture.

When there’s no notice at all

Most studies never get corrected, flagged, or retracted, and the absence of a notice isn’t proof of anything special — it just means no formal concern has been raised yet. It’s still worth asking basic questions: How large was the study? Was it in humans? Has it been repeated by other researchers? Publication status is one useful signal among several, not a stand-in for reading the study itself. Our page on how we research and use sources walks through the fuller picture of separating established guidance from early or uncertain findings.

A note on what this page can’t do

This guide explains how publication notices work in general. It can’t tell you whether a specific study relevant to your own health decision is reliable enough to act on — that’s a conversation for a doctor or other qualified clinician, who can weigh a claim against your individual situation. If you’re new to reading evidence this way, our Start Here page is a good place to begin.

Educational information disclaimer

This article is for general educational purposes only and does not provide medical advice, diagnosis, or treatment recommendations. It does not tell you whether to start, stop, or change any medication or treatment. Everyday Prevention Evidence is an independent editorial publication and is not affiliated with the former physician or vaccination campaign that previously used this domain. If you have a medical concern, talk to a qualified healthcare provider. If you are experiencing a medical emergency, contact your local emergency services immediately.

By Everyday Prevention Evidence Editorial Team. Last updated September 10, 2026.

Related guides

  • What a P-Value Can—and Cannot—Tell You About a Health Study
  • Why the Outcome Measured Matters in a Health Study
  • What a Confidence Interval Can Tell You About a Health Study
  • How to Read a Health Study Abstract Without Stopping There

Filed Under: study and risk literacy

Reader Interactions

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

Everyday Prevention Evidence

Read prevention guidance with context, numbers, and clear limits.

Topics

  • Start Here
  • Reading Study Numbers
  • Study Design and Evidence Strength
  • Checking Sources and Reports
  • Applying Evidence to You

Publication

  • About
  • Editorial Policy
  • Review Process
  • Corrections
  • Independence Notice
  • Contact

Policies

  • Privacy Policy
  • Terms of Use
  • Medical Disclaimer
  • Affiliate Disclosure
  • Accessibility

Everyday Prevention Evidence is an independent editorial publication at UU4VE.org. It is not affiliated with the former physician or vaccination campaign, and UU4VE is not used here as an expansion of that former organization's name.

Questions about the operator of this site: contact@uu4ve.org or the Contact page.

© 2026 Everyday Prevention Evidence