Free scoping read. A written assessment of your manuscript in two business days, at no cost.

Start one
Contents

Insights · Study guide · Types of scientific communication

Types of scientific communication in research methodology

There is an authoritative taxonomy of these forms, and most confident claims about them — about abstracts, preprints and open review — are looser than their sources.

Pubrica AcademyWritten by a research writing specialist Reviewed bya research methodology reviewer Published15 October 2026 Last checked15 October 2026 Reading time15 minutes Sources14 primary

Scientific communication is not one thing. It is a set of genres with different audiences, different standards and different standing, and the National Library of Medicine maintains a formal taxonomy of them. This page sets out what each form is, from the authorities that define it.

The forms, in one place

Table 1. Forms of scientific communication

FormWhere it appearsAudienceWhat it is
Original research articlePeer-reviewed journalResearchers in the fieldNew data, reported in full. The primary form.
ReviewPeer-reviewed journalResearchers, cliniciansSynthesises published work. A systematic review is a research method in its own right.
Case reportPeer-reviewed journalCliniciansOne patient or a small series.
PreprintPreprint serverAnyone, immediatelyA complete draft, not peer reviewed. Carries a disclosure duty on later submission.
Conference abstract or posterScientific meetingPeers at the meetingA preliminary report. Most never reach full publication.
Editorial, letter, commentPeer-reviewed journalThe journal’s readershipOpinion and correspondence, formally indexed and linked to what they discuss.
Peer review reportTo editors, sometimes publishedEditors, sometimes readersConfidential by default. Some journals publish the record.
Plain language summaryAlongside the review or on a registryPatients, carers, the publicRequired by Cochrane for every review, and by EU law for clinical trial results.
Press releaseInstitutional press officeJournalists, then the publicWhere most exaggeration in health news originates.

The formal taxonomy

NLM's publication types are the authoritative list, and they describe what an item is rather than what it is about. “Publication Type (PT) values are added to the MEDLINE citation during indexing.”2 “to indicate the article's type of information or the manner in which it is conveyed.”2 NLM groups them as “Specific genre of publication”1 and “Specific parts of publications”1

Table 2. Publication types, with NLM’s own scope notes

TypeNLM scope noteNotes
Journal Article“The predominant publication type for articles and other items indexed for NLM databases.”A container type covering original research, reviews and other reports — not a synonym for an original research article.
Review“An article or book published after examination of published material on a subject.”Synthesises existing work rather than reporting new data.
Randomized Controlled Trial“A work that reports on a clinical trial that involves at least one test treatment and one control treatment,”A publication type, not a subject heading. Searching it finds trials; the as-Topic heading finds articles about trials.
Case Reports“Clinical presentations that may be followed by evaluative studies that eventually lead to a diagnosis.”One patient or a small series.
Editorial“Work consisting of a statement of the opinions, beliefs, and policy of the editor or publisher of a journal,”Signed opinion, not peer-reviewed research.
Letter“Work consisting of written or printed communication between individuals”Correspondence, often responding to a published article.
Comment“Work consisting of a critical or explanatory note written to discuss, support, or dispute an article”Formally linked to the article it discusses.
Preprint“Scientific manuscript made available prior to PEER REVIEW.”NLM added this as a publication type; PubMed carries a filter to exclude them.

Four of the scope notes above are longer than our tooling can capture in one piece, so they are given as the opening of NLM's wording rather than in full. Read the complete notes on NLM's own page, linked in the references. Our guide to MeSH covers how publication types work in a search, including why Randomized Controlled Trial[pt] and Randomized Controlled Trials as Topic[mh] are different searches.

Research article, review, case report

A category the taxonomy does not define

A distinction worth drawing, because the taxonomy does not draw it for you. NLM has no publication type called “Original Research Article”. Journal Article is a container, annotated as covering “original full text research, review or other reports published in a journal”. So “original research article” is a convention of editorial practice rather than a formally defined category in the indexing vocabulary.

The annotation on the descriptor record: “This heading is used as a Publication Type. for original full text research, review or other reports published in a journal”3 A review, by contrast, is “An article or book published after examination of published material on a subject.”1 — so the distinguishing feature is whether the work reports new data or appraises existing work. Case reports are “Clinical presentations that may be followed by evaluative studies that eventually lead to a diagnosis.”1

For how each of these is summarised, and the length and structure conventions that go with them, see our guide to abstract formats and the guide to writing an original article.

Preprints

“Preprints are complete and public drafts of scientific articles that have not yet been peer reviewed.”4 As a MeSH publication type the definition is tighter still: “Scientific manuscript made available prior to PEER REVIEW.”1

The server's own framing of what that means: “Manuscripts submitted to bioRxiv are not certified by scientific peer review”5 — so a preprint is a complete piece of scientific communication with one specific thing missing, and it is the thing that distinguishes the literature from everything else.

NLM archives a subset: its pilot “will include preprints supported by an NIH award, contract, or intramural program.”4 So a preprint's presence or absence in PubMed Central is a statement about its funding, not its quality.

Peer review is itself a form of communication

It is easy to leave peer review out of a list of communication forms, because most of it is never published. But it is “Peer review is the critical assessment of manuscripts submitted to journals by experts”6, and the ICMJE describes it as one where “peer review is an important extension of the scientific process.”6 COPE's framing: “Peer reviewers have an important role in ensuring the integrity of the scholarly record”7

Its defining property is confidentiality. “Reviewers therefore should keep manuscripts and the information they contain strictly confidential.”6 and “Reviewers must not retain the manuscript for their personal use”6 COPE puts the same duty on the reviewer — “Reviewers should respect the confidentiality of the manuscripts they evaluate.”7 — and names the abuse it exists to prevent: “you should not agree to review a manuscript just to gain sight of it with no intention of submitting a review”7

Open peer review. One journal's version: “For research papers The BMJ has fully open peer review.”8 Research papers “usually have their prepublication history posted alongside them on bmj.com.”8 “This prepublication history comprises all previous versions of the manuscript,”8 “the report from the manuscript committee meeting, the reviewers' comments, and”8 “the authors' responses to all the comments from reviewers and editors.”8

One term, several practices

One caution on terminology. “Open peer review” is used for several different things, and the example above is a specific one: publishing the prepublication record — previous versions, the committee report, the reviewers' comments and the authors' responses. That is not the same as signed review, where reviewers' names are disclosed. The policy quoted here says nothing about reviewer identity, so if a journal advertises open peer review, read which of these it means.

What counts as prior publication

This is where the forms interact, and where two confident claims are both wrong. The ICMJE's starting definition: “Duplicate publication is publication of a paper that overlaps substantially with one already published”9, and “Prior publication may include release of information in the public domain.”9

Correcting a claim about conference abstracts

A correction to something very widely written, including in an earlier version of our own guidance. No authority states that a conference abstract “is not a publication”. What the ICMJE actually does is treat “an abstract or poster displayed at a scientific meeting” as a preliminary report, and provide that a journal may consider “a paper that has been presented at a scientific meeting but was not published in full”. That is a narrower and more useful claim: presenting does not normally bar the full paper, and the abstract is a stage rather than a nullity.

The provisions themselves: the ICMJE contemplates “a complete report that follows publication of a preliminary report, such as a letter to the editor,”9 “an abstract or poster displayed at a scientific meeting.”9, and allows a journal to consider “a paper that has been presented at a scientific meeting but was not published in full”9

And on preprints

Likewise on preprints, where both of the confident positions are wrong. The ICMJE neither forbids preprinting nor declares it harmless. It states that “posting of work as a preprint may influence a journal's interest in or priority for peer review”, and places a duty on the author: “it is the author's responsibility to inform a journal if the work has been previously posted on a preprint server”, with a link to it. Practice varies by journal — bioRxiv's own guidance is that “most basic research journals will consider manuscripts that were posted”, and that authors should contact the journal if in doubt.

Note the final consideration the ICMJE raises, which applies to all of these: dissemination outside scientific presentations “may diminish the priority journal editors assign to their work.”9 That is a judgement about editorial priority rather than a rule, and it is the real reason to think about sequence before you start publicising work.

Plain language and lay summaries

Communication aimed at people who are not researchers is not optional goodwill; in two settings we verified it is required.

Cochrane. “Every Cochrane review includes a plain language summary.”10 They are “Plain language summaries are summaries of health evidence written in a way that everyone can understand.”10 and “They explain research findings in clear, everyday language without specialist terminology or jargon.”10, intended “to help patients, caregivers and members of the public who are not health experts”10

EU clinical trials. Under Regulation (EU) No 536/2014, the summary of results “It shall be accompanied by a summary written in a manner that is understandable to laypersons.”11 “The content of that summary is set out in Annex V.”11, headed “CONTENT OF THE SUMMARY OF THE RESULTS OF THE CLINICAL TRIAL FOR LAYPERSONS”11 This is a legal obligation attached to the trial, not a publication convention.

Guidance on how to write one exists: the Good Lay Summary Practice recommendations are a “framework guidance document detailing the planning, development, translation, and dissemination of LSs”12 published in 2021, “as required by the 2014 EU Clinical Trials Regulation (CTR No 536/2014)”1211.

Something that does not exist

One thing that does not exist, and is worth knowing before you go looking: there is no reporting guideline for plain language or lay summaries. EQUATOR's own index of reporting guidelines by study type covers trials, observational studies, systematic reviews, protocols, diagnostic and prognostic studies, case reports, practice guidelines, qualitative research, animal studies, quality improvement and economic evaluations — and nothing for lay communication. The nearest instrument is the Good Lay Summary Practice guidance, which is regulatory guidance rather than a reporting guideline.

A reporting guideline, for reference, is “A checklist, flow diagram, or structured text to guide authors in reporting a specific type of research.”14 — and the index of them covers every study design and no form of lay communication14.

Two sourcing notes. The regulation was read on the UK-assimilated copy, because EUR-Lex full text is closed to automated retrieval; check it against EUR-Lex before citing it as EU law. And the Good Lay Summary Practice guidance was verified through a third party rather than the European Commission's own document, so nothing is quoted from the guidance text itself.

Press releases, and where exaggeration starts

The distortion is upstream of the journalists

The finding worth carrying from that study is about where exaggeration starts. Of 462 press releases from 20 leading UK universities, 40% contained exaggerated advice, 33% exaggerated causal claims, and 36% exaggerated inference to humans from animal research. Where the press release was exaggerated, the great majority of the resulting news stories were too; where it was not, only around a tenth to a fifth were. So the distortion is substantially upstream of the journalists — which puts it within reach of the researchers and press offices who write the release.

The figures, from a study of “Press releases (n=462) on biomedical and health related science issued by 20 leading UK universities”13: “40% (95% confidence interval 33% to 46%) of the press releases contained exaggerated advice”13, “33% (26% to 40%) contained exaggerated causal claims”13, and “36% (28% to 46%) contained exaggerated inference to humans from animal research”13 Where the release was exaggerated, “58% (95% confidence interval 48% to 68%), 81% (70% to 93%), and 86% (77% to 95%) of news stories”13 were exaggerated in the corresponding way — against roughly a tenth to a fifth where the release was not.

The practical implication for a researcher is specific and actionable: the press release is a form of scientific communication you can still influence, and it is the point at which most of the distortion enters. Read it before it goes out, and check the three things the study measured — the advice, the causal language, and any inference from animals to humans.

Working practices

How to apply this to your own work

  1. Pick the genre deliberately. An original research article, a review and a case report have different standards and different audiences, and a paper written in the wrong genre is rejected on scope.
  2. Tell the journal if your work is on a preprint server, and give the link. The ICMJE makes that the author’s responsibility, and it affects editorial priority.
  3. Do not assume a conference abstract blocks the full paper. It is a preliminary report, and journals may still consider work presented but not published in full.
  4. Treat a peer review you are given as confidential. Do not keep the manuscript, and do not accept a review in order to read the work.
  5. Check what a journal means by open peer review before assuming your name will or will not be attached.
  6. Write the plain language summary as a deliverable in its own right, not a compressed abstract. For EU trials it is a legal requirement with a prescribed content list.
  7. Read the press release before it is issued, and check the advice, the causal claims and any animal-to-human inference against what the paper actually supports.
  8. Use NLM’s publication types when you need a defensible description of a form, rather than a convention you have read elsewhere.

Take-home message

Scientific communication has a formal taxonomy — NLM publication types, each with a scope note — and most of the confident claims about it are looser than their sources. There is no NLM type called “original research article”; Journal Article is a container. A conference abstract is a preliminary report, not a nullity, and presenting does not normally bar the full paper. Preprints carry a disclosure duty rather than a prohibition. Open peer review usually means a published record, not named reviewers. Lay summaries are mandatory for Cochrane reviews and for EU trials, and have no reporting guideline. And most exaggeration in health news is already in the press release.

The National Library of Medicine maintains the authoritative taxonomy as publication types, each with its own scope note: Journal Article, Review, Randomized Controlled Trial, Case Reports, Editorial, Letter, Comment and Preprint, among others. Alongside those formal genres sit conference abstracts and posters, peer review reports, plain language summaries and press releases.

An indexing value NLM assigns to a record. “Publication Type (PT) values are added to the MEDLINE citation during indexing”, NLM states, “to indicate the article’s type of information or the manner in which it is conveyed”. It describes what an item is, as distinct from a subject heading, which describes what it is about.

Not under that name. NLM’s container type is Journal Article, annotated as covering “original full text research, review or other reports published in a journal”. So “original research article” is a convention of editorial practice rather than a formally defined category in the indexing vocabulary.

NLM’s definition: “preprints are complete and public drafts of scientific articles that have not yet been peer reviewed.” As a MeSH publication type it is “scientific manuscript made available prior to PEER REVIEW.” bioRxiv puts the consequence plainly: manuscripts posted there “are not certified by scientific peer review”.

Usually not, but you must disclose it. The ICMJE states that “posting of work as a preprint may influence a journal’s interest in or priority for peer review”, and that “it is the author’s responsibility to inform a journal if the work has been previously posted on a preprint server”, with a link. bioRxiv notes that “most basic research journals will consider” posted manuscripts, and advises contacting the journal if in doubt.

The honest answer is narrower than the usual one. No authority states that an abstract is not a publication. The ICMJE treats “an abstract or poster displayed at a scientific meeting” as a preliminary report, and provides that a journal may consider “a paper that has been presented at a scientific meeting but was not published in full”. So presenting does not normally bar the full paper.

The ICMJE defines it as “the critical assessment of manuscripts submitted to journals by experts”, and describes it as “an important extension of the scientific process”. COPE adds that “peer reviewers have an important role in ensuring the integrity of the scholarly record”. It is a form of scientific communication in its own right, usually unpublished.

Not necessarily, and the term covers several different practices. The BMJ, for instance, publishes the prepublication history — previous versions, the manuscript committee report, the reviewers’ comments and the authors’ responses. That policy says nothing about reviewer identity. If a journal advertises open peer review, check which version it means.

Yes, in two places we verified. Cochrane states that “every Cochrane review includes a plain language summary”. And under the EU Clinical Trials Regulation, the summary of results “shall be accompanied by a summary written in a manner that is understandable to laypersons”, whose content is set out in Annex V. So for EU trials it is a legal requirement.

Substantially from the press release rather than the journalist. In a study of 462 press releases from 20 leading UK universities, 40% contained exaggerated advice, 33% exaggerated causal claims and 36% exaggerated inference to humans from animal research — and where the release was exaggerated, the large majority of the resulting news stories were too.

References and sources

  1. U.S. National Library of Medicine. Publication Characteristics (Publication Types) with Scope Notes. read 15 October 2026. View on U.S. National Library of MedicineA genuine authoritative taxonomy of the forms scientific communication takes, with NLM's own scope note for each. The best single source for this page.
  2. U.S. National Library of Medicine. Publication Types. NLM Technical Bulletin, March–April 2003. read 15 October 2026. View on NLM Technical BulletinCited for what a publication type is and when it is applied. Note its count of publication types is a 2003 figure and is not given here as current.
  3. U.S. National Library of Medicine. MeSH descriptor Journal Article, unique ID D016428, tree number V02.600. MeSH Browser. read 15 October 2026. View on MeSH BrowserThe descriptor record itself. Its annotation is what shows that Journal Article is a container type rather than a synonym for original research.
  4. U.S. National Library of Medicine. Next Phase of the NIH Preprint Pilot Launching Soon. NLM Technical Bulletin, January–February 2023;(450):e2. read 15 October 2026. View on NLM Technical BulletinNLM's own definition of a preprint and the scope of its archiving pilot. Cited because PubMed Central's policy pages themselves were not reachable from here.
  5. bioRxiv. Frequently asked questions. Cold Spring Harbor Laboratory. read 15 October 2026. View on bioRxivA preprint server's own account of what posting does and does not mean for later journal submission.
  6. International Committee of Medical Journal Editors. Recommendations for the Conduct, Reporting, Editing, and Publication of Scholarly Work in Medical Journals, sections II.C.2.c and II.C.3. read 15 October 2026. View on ICMJEThe authority on what peer review is for and what a reviewer must do. Cited from the consolidated PDF, which answers several questions in one document.
  7. Committee on Publication Ethics. Ethical guidelines for peer reviewers. Version 2, September 2017. read 15 October 2026. View on COPEThe reviewer's side of the obligation, including the point about not accepting a review in order to read the manuscript.
  8. The BMJ. Peer review policy, as displayed on article peer-review pages. read 15 October 2026. View on The BMJCited for what one journal's open peer review consists of. Note it describes publishing the prepublication record and says nothing about reviewer identity, so no claim about signed reviews rests on it.
  9. International Committee of Medical Journal Editors. Overlapping Publications, sections III.D.2 and III.D.3. read 15 October 2026. View on ICMJEThe authority on prior publication. The source for two corrections: a conference abstract is a preliminary report rather than a publication barring the full paper, and preprints carry a disclosure duty rather than a prohibition.
  10. Cochrane. Evidence (plain language summaries). read 15 October 2026. View on CochraneThe source for plain language summaries being a required component of every Cochrane review. Cochrane's detailed PLS guidance pages have moved, so this is cited rather than the older training URLs.
  11. Regulation (EU) No 536/2014 of the European Parliament and of the Council of 16 April 2014 on clinical trials on medicinal products for human use, and repealing Directive 2001/20/EC. Article 37(4) and Annex V. read 15 October 2026. View on legislation.gov.ukLay summaries of trial results are a legal requirement, not a courtesy. Read on the UK-assimilated copy of the regulation because EUR-Lex full text is closed to automated retrieval; check against EUR-Lex before citing it as EU law.
  12. Good Lay Summary Practice (GLSP) recommendations, published 4 October 2021 in EudraLex Volume 10; developed by the Roadmap Initiative and adopted by the Expert Group on Clinical Trials. (Verified via ECRIN.) read 15 October 2026. View on ECRINGuidance rather than a reporting guideline. Verified through ECRIN rather than the European Commission's own document, which we could not open, so nothing is quoted from the guidance text itself.
  13. Sumner P, Vivian-Griffiths S, Boivin J, Williams A, Venetis CA, Davies A, Ogden J, Whelan L, Hughes B, Dalton B, Boy F, Chambers CD. The association between exaggeration in health related science news and academic press releases: retrospective observational study. BMJ. 2014;349:g7015. doi:10.1136/bmj.g7015 read 15 October 2026. View on Cardiff University repositoryWhere exaggeration in health science news originates. Read as the repository record's abstract; the publisher's article page did not surface in search.
  14. EQUATOR Network. What is a reporting guideline? read 15 October 2026. View on EQUATOR NetworkCited for the definition of a reporting guideline, and for the verified absence of one covering plain language or lay summaries.

Pubrica AcademyWritten and maintained by the subject matter experts of Pubrica who do this work. Articles are revised when the standard underneath them moves, and each one carries the date it was last checked.

Related: writing an original article · abstract formats · the proof stage · what editors reject papers for.

How Pubrica helps

Our medical writers work across these forms: original research articles and reviews prepared to a target journal’s instructions, conference abstracts and posters, plain language summaries written to the content requirements that apply to them, and press releases checked against what the paper actually supports.

Get a free quote

Regulations

Check your project against this rule

  1. 1Your work
  2. 2Your details
  3. 3Send
What do you need?

Material you send is held under signed confidentiality, with access limited to the assigned team and encrypted transfer. It is not used to train anything and is not passed to third-party tools. Ask for the agreement before you send anything you would rather cover first.