Retractions, Corrections and Duplicate Reports
How retractions, expressions of concern, corrections, preprints and duplicate reports work, and how to check a study before relying on it.
Content last updated . This page has not been medically reviewed.
The short answer
Published research is not frozen. Journals can retract an article, flag a concern, correct an error, or link a later version. A retracted paper stays visible but should be labeled as retracted, and its findings should not be treated as reliable evidence. Separately, the same study can appear in more than one report, which can make one study look like several.What each term means
| Term | Plain meaning | What to do with it |
|---|---|---|
| Retraction | A journal or author withdraws an article in whole or in part. The article is given the publication type "Retracted Publication." [1] | Do not use its findings as evidence; read the notice for the reason |
| Expression of concern | An editor's notice, typically written under that phrase in its title, raising a question about an article's integrity [2] | Treat the article's findings as uncertain until an outcome is published |
| Correction (erratum) | A notice fixing text or information in an earlier article. NLM counts corrections, corrigenda, addenda and partial retractions all as errata [2] | Read the correction; it may or may not change the conclusions |
| Corrected and republished | The article is republished to fix errors, and linked to the original [2] | Use the corrected version |
| Preprint | A scientific manuscript made available before peer review [1] | Treat as preliminary |
| Duplicate publication | Substantially the same material published again without acknowledging the earlier publication [1] | Count the study once |
Retractions
A retraction can happen for different reasons, and the label alone does not say which. NLM does not differentiate between articles retracted because of honest error and those retracted because of scientific misconduct or plagiarism. [2] A 2012 analysis of 2,047 retracted biomedical articles found that most retractions were attributable to misconduct rather than error, although the authors later published an erratum, so check the original. [3]Retracted articles should remain in the public domain and be clearly labeled as retracted. [4] In PubMed, the retracted article carries the publication type Retracted Publication, and its retraction notice is linked to it. [2] The validity of previous work by the author of a fraudulent paper cannot be assumed, so editors may ask for assurance or publish an expression of concern about other papers. [4]
Corrections
Not every correction is serious. Errata can correct or add text or information anywhere in an earlier article. [2] An erratum can change a typo or change a result; the notice is how you tell. The correction is linked to the article in PubMed and should be read alongside it. [2]Preprints
A preprint is made available before peer review. [1] The International Committee of Medical Journal Editors notes that there are benefits and harms in sharing findings before peer review. [5] A preprint may later be published, changed or withdrawn, so treat it as preliminary.Duplicate and multiple reports
Two different situations get mixed up. One is duplicate publication: a paper that overlaps substantially with one already published, without clear reference to it. [5] Duplicate publication of original research is a problem because it can cause inadvertent double-counting of data, or inappropriate weighting of one study's results, which distorts the available evidence. [5]The other is normal: a single study can be reported in several sources, such as a protocol paper, a results paper and a follow-up. These should be identified as one study. [6] Both situations lead to the same practical rule: count studies, not documents. NLM does not routinely examine articles for originality, so catching duplicates is not automatic. [2]
Common misunderstandings
- "Retracted means it was fraud." Not necessarily; the reason is in the notice.
- "Retracted papers disappear." They stay available and are meant to be labeled.
- "Two papers, two studies." They may be one study.
- "A preprint is a published study." It has not been through peer review.
What this means in the Research Library
Library publication records are imported from public databases. No record has been reviewed by a clinician, and imported records are not verified. Do not assume that a record shows its latest status. Before relying on any publication, open its PubMed entry, look for a retraction notice, expression of concern or correction, and check whether it is one of several reports of the same study.Related: evidence types and study designs, how to read Research Library records, and evidence labels and verification status.
Frequently asked questions
What is a retracted study?
It is an article that a journal or its authors have withdrawn, in whole or in part. [1] It should stay visible and be clearly labeled as retracted. [4]
Why are studies retracted?
For different reasons, including honest error and misconduct. NLM does not distinguish between them in its labels. [2] Read the retraction notice for the stated reason.
What is an expression of concern?
It is an editor's notice about the integrity of a published article, typically using that phrase in its title. [2] It can be followed by a retraction depending on what an investigation finds. [4]
Is a correction the same as a retraction?
No. A correction fixes text or information within an article, while a retraction withdraws it. [2] Partial retractions, such as of a single table, are treated as errata. [2]
Is a preprint reliable?
A preprint is a manuscript made available prior to peer review. [1] It can be useful but is preliminary.
Why do duplicate reports matter?
They can lead to double-counting of data or inappropriate weighting of one study's results. [5]
References
- National Library of Medicine. Publication Types (MeSH). Read 2026-10-22.
- National Library of Medicine. Errata, Retractions, and Other Linked Citations in PubMed. Read 2026-10-22.
- Fang FC, Steen RG, Casadevall A. Misconduct accounts for the majority of retracted scientific publications. Proc Natl Acad Sci U S A. 2012;109:17028-33. Read 2026-10-22.
- ICMJE. Scientific Misconduct, Expressions of Concern, and Retraction. Read 2026-10-22.
- ICMJE. Overlapping Publications. Read 2026-10-22.
- Li T, Higgins JPT, Deeks JJ. Cochrane Handbook for Systematic Reviews of Interventions, Chapter 5: Collecting data. Read 2026-10-22.