EditversePlain language summaries of publications: format, review and where journals host them
What GPP 2022 and the Open Pharma recommendations say about plain language summaries, how to test reading level, where summaries are hosted, and a worked abstract-to-summary example.
Key points
- A plain language summary of a publication restates a peer-reviewed article for readers outside the specialty: patients, carers, clinicians in other fields, policymakers. It is different from the lay summary of trial results that EU law requires sponsors to submit.
- GPP 2022 treats the summary as part of the publication it accompanies, encourages one for all studies, and recommends one for each publication of clinical data that follows the CONSORT format.
- The Open Pharma recommendations set a minimum standard: in the style of an abstract, free of technical jargon, unbiased, non-promotional, peer reviewed, easily accessed, and meeting the technical requirements for indexing in directories such as PubMed.
- PubMed displays a plain language summary below the abstract when the publisher supplies it.
- Reading level is tested twice: with a readability formula as a screen, and with a reader who does not know the subject.
Three documents that share a name
"Plain language summary" and "lay summary" are used for at least three different documents, with different rules.
| Document | What it summarises | Governing source | Where it appears |
|---|---|---|---|
| Plain language summary within an article | One peer-reviewed publication; a short text placed with the abstract, sometimes with a graphical version | Journal instructions; GPP 2022; Open Pharma recommendations | Article page; PubMed where supplied by the publisher |
| Plain language summary of publication as a stand-alone article | One or more peer-reviewed articles, written under plain language principles and published in a peer-reviewed journal with its own DOI | Journal instructions; publisher guidance | Journal, as a separate citable item |
| Lay summary of clinical trial results | The results of a trial, whatever the outcome, for the EU database | Regulation (EU) No 536/2014, Article 37 and Annex V; EU guidance on summaries for laypersons | EU clinical trials database |
The distinction between the first two rows follows the definitions used by the cross-publisher working group that Open Pharma convened in June 2024, which treats the stand-alone summary as a distinct article type with a unique DOI.5 The third row is a regulatory document: the sponsor submits it within one year of the end of the trial, and Annex V of the Regulation lists ten elements it must contain.8 The rest of this article is about the first two.
What GPP 2022 says
GPP 2022 brought plain language summaries into the publication guideline for company-sponsored research. Its scope statement names them among the outputs it covers, as enhanced content.1 The guideline's supplement and ISMPP's published answers add the detail.12
- Summaries are encouraged for all studies. The supplement recommends one for each publication of clinical data that follows the CONSORT format, and that recommendation is meant to carry more force than the general statement that ideally each journal article would have one.
- A single stand-alone summary can explain several publications or studies, and this does not replace per-trial summaries where those are needed.
- A summary can take many forms, from a short text abstract to video, infographic or other graphical formats. A brief plain language abstract is sufficient for many publications and need not add much cost.
- Where a journal accepts a summary only after acceptance of the article, the data in it should match the main publication, and at least the lead and corresponding authors should be involved in preparing it, with other authors invited to review.
- Patients and patient advocates may be compensated at a reasonable rate for reviewing plain language summaries.
The practical result for sponsors is that summaries belong in the publication plan and budget from the start, as GPP 2022's planning principles say.
The Open Pharma recommendations
Open Pharma is a collaboration of pharmaceutical companies, other research funders, publishers, patients, academics, regulators, editors and societies. Its recommendations for plain language summaries of peer-reviewed medical journal publications were developed by its accessibility workstream, refined at an expert roundtable and put to public consultation, and published in Current Medical Research and Opinion in 2021.3 They are presented as a minimum standard that complements more detailed guidance. A summary should be:
| Recommendation | Working test |
|---|---|
| In the style of an abstract | Short continuous text that covers why the study was done, what was done, what was found and what it means |
| Free of technical jargon | A reader outside the field can restate the main finding after one reading; every unavoidable term is explained once |
| Unbiased | Same primary outcome, same direction and size of effect, harms and limitations included as in the article |
| Non-promotional | No brand names beyond what the article uses, no claims beyond the article's conclusions, no call to action |
| Peer reviewed | Submitted with the manuscript so that it is reviewed alongside it |
| Easily accessed | Free to read, on the article page, not behind the paywall |
| Indexable | Meets the technical requirements for directories such as PubMed: plain text, within the journal's word limit, supplied by the publisher in the article metadata |
The left column is taken from the published recommendations. The right column is a working interpretation for writers, not Open Pharma's text. A later ISMPP working group paper sets out the questions sponsors still need to settle when adopting summaries across a portfolio.4
Where journals host them
The US National Library of Medicine announced in February 2019 that PubMed displays plain language summaries when they are supplied by the publisher, below the abstract, and that they appear in the citation data in the other-abstract field.6 Whether a given article's summary reaches PubMed therefore depends on the journal: it has to accept a text summary and send it with the article record.
Journal practice varies, and the author instructions decide the format. One large publisher's guidance for authors gives a sense of what a PubMed-ready summary looks like: 250 words or less, in English, in plain text as continuous prose, placed after the abstract under its own heading, distinct from the abstract, factual and not speculative, in the active voice, using person-centred language, with any technical term explained on first use. The same guidance notes that a graphical summary can be published as well but cannot be indexed on PubMed, and suggests asking someone with no knowledge of the subject to read the draft.7
Where the target journal does not accept summaries, the options are a stand-alone plain language summary of publication in a journal that publishes that article type, with permission from the original publisher where copyright requires it, or hosting on a trial results or sponsor website with a link to the article. In each case the summary cites the original article and its DOI.
Step sequence
Check the journal's rules
Accepted or not, word limit, text or graphical, where it is placed, whether it is peer reviewed, and whether there is a charge.
Fix the content from the final manuscript
List the population, intervention, comparison, primary outcome, main result with numbers, key harms, main limitation and the authors' conclusion. Nothing enters the summary that is not in the article.
Draft to a four-question frame
Why was the study done? What did the researchers do? What did they find? What do the results mean?
Convert the numbers
Use natural frequencies with a common denominator, give absolute numbers alongside any relative effect, and keep the time frame with every figure.
Screen with a readability formula
Record the score and revise long sentences and multi-syllable terms.
Test with a reader
Ask someone outside the field to read it and say back the main finding and one limitation. Revise until they can.
Check against the article
A second person verifies every number and confirms that direction, size and certainty match the abstract.
Author review and approval
All authors see it; at minimum the lead and corresponding authors work on it. A patient reviewer is included where the sponsor has that process.
Submit with the manuscript
So that it is peer reviewed with the article, and update it if the article changes in revision.
Reading-level testing
Neither GPP 2022 nor the Open Pharma recommendations, as read for this article, set a numerical reading grade, so any target is a house or journal rule and should be described as such. Testing has two parts with different purposes.
Formula screen. Readability formulas such as Flesch Reading Ease and Flesch-Kincaid Grade Level estimate difficulty from sentence length and syllables per word. They are quick and repeatable, and they flag long sentences. They cannot tell whether a short word is understood, whether a number is interpretable, or whether the summary is accurate. Record the tool, the version and the score so the result can be reproduced.
Reader test. A person from the intended audience reads the draft and answers three questions without help: what was studied, what was found, and what is still uncertain. Their answers are compared with the article. This is the test that catches undefined terms, ambiguous percentages and conclusions that read more strongly than the paper.
Worked example: from abstract to summary
The example is illustrative. The trial, drug name and results are invented.
Illustrative scientific abstract (extract)
In this double-blind trial, 1,200 adults with episodic migraine were randomised 1:1 to oral examplitan 50 mg daily or placebo for 12 weeks. The primary endpoint was change from baseline in mean monthly migraine days (MMD). Mean MMD decreased by 4.1 with examplitan and 2.5 with placebo (least-squares mean difference −1.6; 95% CI −2.1 to −1.1; P<0.001). A reduction of at least 50% in MMD occurred in 46% versus 29% of participants. Adverse events were reported in 38% and 33%; the most frequent was nausea (9% vs 4%). Discontinuation due to adverse events occurred in 3% and 2%.
Illustrative plain language summary
Why was this study done? Migraine causes severe headaches that can stop people from working. Researchers wanted to know whether a new medicine, examplitan, could reduce the number of days people have migraine each month.
What did the researchers do? They asked 1,200 adults who had migraine to take a tablet every day for 12 weeks. Half took examplitan. Half took a placebo, a tablet that looks the same but contains no medicine. Neither the participants nor their doctors knew who was taking which tablet.
What did they find? Before the study, people counted their migraine days each month. After 12 weeks, people taking examplitan had about 4 fewer migraine days a month. People taking placebo had about 2 and a half fewer. So examplitan reduced migraine by about 1 and a half more days a month than placebo. Migraine days were cut by half or more in 46 out of every 100 people taking examplitan and in 29 out of every 100 taking placebo. Feeling sick was the most common side effect. It affected 9 out of every 100 people taking examplitan and 4 out of every 100 taking placebo. About 3 in 100 people stopped examplitan because of side effects, and 2 in 100 stopped placebo.
What do the results mean? In this study, examplitan reduced migraine days more than placebo over 12 weeks. The study did not look at how well it works, or how safe it is, over a longer time.
The summary is 245 words, inside the 250-word limit quoted above. Compare it with the abstract line by line: the population, duration and blinding are kept; the primary result is given as days per month and then as the difference; the responder result uses a common denominator of 100; harms are reported with the same frequencies; the confidence interval and P value are replaced by plain statements of what was found, without overstating certainty; and the closing sentence states a limit of the trial. No number appears that is not in the abstract.
Review record
| Check | Evidence kept | Done by |
|---|---|---|
| Accuracy against article | Number-by-number check sheet | Second checker |
| Balance | Harms and limitation present; conclusion matches the article's | Medical reviewer |
| Readability screen | Tool, version, score | Writer |
| Reader test | Reader's answers and changes made | Writer with lay or patient reviewer |
| Author approval | Written approval of the final text | Authors |
Editverse Medical writes plain language summaries for journals and for EU trial results through its medical communications service, with a readability check included, and prepares them alongside sponsored manuscripts under industry publications. The client supplies the accepted or final manuscript and the journal's instructions. The corporate medical writing page describes the review steps.
Is a plain language summary the same as a lay summary for the EU Clinical Trials Regulation?
No. The EU lay summary is a regulatory document about a trial's results, submitted by the sponsor to the EU database with content set by Annex V of Regulation (EU) No 536/2014. A plain language summary of a publication accompanies a journal article and follows the journal's instructions.
How long should a plain language summary be?
The journal decides. One major publisher sets 250 words or less for a text summary intended for PubMed indexing. Stand-alone summaries published as articles are longer and follow that journal's article-type rules.
Do plain language summaries appear on PubMed?
They can. PubMed displays a plain language summary below the abstract when the publisher supplies it with the article record. One publisher's author guidance notes that graphical summaries cannot be indexed there.
References
- DeTora LM, Toroser D, Sykes A, Vanderlinden C, Plunkett FJ, Lane T, et al. Good Publication Practice (GPP) Guidelines for Company-Sponsored Biomedical Research: 2022 Update. Ann Intern Med. 2022;175(9):1298-1304. PMID 36037471 · DOI
- International Society for Medical Publication Professionals. GPP 2022 frequently asked questions. ismpp.org
- Rosenberg A, Baróniková S, Feighery L, Gattrell W, Olsen RE, Watson A, et al. Open Pharma recommendations for plain language summaries of peer-reviewed medical journal publications. Curr Med Res Opin. 2021;37(11):2015-2016. PMID 34511020 · DOI
- Lobban D, Gardner J, Matheis R; ISMPP PLS Perspectives Working Group. Plain language summaries of publications of company-sponsored medical research: what key questions do we need to address? Curr Med Res Opin. 2022;38(2):189-200. PMID 34736362 · DOI
- Open Pharma. Plain language summaries of publications: guidance from the cross-publisher working group. openpharma.blog
- Collins M. PubMed Updates February 2019. NLM Technical Bulletin. 2019 Jan-Feb. Posted 6 February 2019. nlm.nih.gov
- Taylor & Francis. Plain Language Summaries: Author Guidelines for Development, Submission and Publication. authorservices.taylorandfrancis.com
- 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. eur-lex.europa.eu