Chapter 21 · Evidence, ethics and publishing

How do you write it up?

Thesis and paper format, the abstract, Vancouver and APA referencing.

16 TNMGRMU questions answered hereset 18 times, 2009–2025

The problem first

You are a final-year MD resident about to submit your thesis. While revising the review of literature you added three papers, and every citation number after them increased by three. The reference list at the back was not renumbered.

Your guide opens a page of the discussion. Reference 14 in the text is a randomised trial, but number 14 in the list is a pharmacology textbook. Your guide asks one question.

"Which source supports this sentence?"

Start here

What goes where in a thesis or a paper?

The text of a research article is usually divided into Introduction, Methods, Results and Discussion (IMRAD). The ICMJE describes this order as a reflection of the process of scientific discovery, not an arbitrary format.

Title and abstractIntroductionwhat is already knownwhat is not knownyour questionMethodswhat was doneResultswhat was foundDiscussionyour main findingcompared with other studieslimitations and meaning for practiceReferencesWhy was it needed?What was done?What was found?What does it mean?wide:the topicnarrow:your questionyour studyonlywide again:practice
The text is wide at the top, narrow in the middle and wide at the bottom. The introduction moves from the general topic to the gap in knowledge and then to your question. Methods and results describe only your study. The discussion starts with your main finding, compares it with other studies, states the limitations and ends with what it means for practice.

A thesis follows the same order with more chapters. A common arrangement is: Introduction, Aims and objectives, Review of literature, Materials and methods, Results, Discussion, Summary and conclusion, Bibliography, and Annexures (ethics approval, consent form, proforma, master chart). Your university's thesis regulations decide the exact order and the page limits. The plan for the study itself is covered in the research protocol chapter.

Interactive

Sort each sentence into Introduction, Methods, Results or Discussion

Try this: choose the section for each sentence, then read the reason. Watch for the two sentences about the difference between the drugs: one belongs in the results and one in the discussion.

Teaching example: sentences from a made-up trial report, in mixed order. Drug A and drug B are not real drugs.

  1. The mean pain score at 24 hours was 3.1 with drug A and 4.0 with drug B.

  2. No trial has compared the two drugs after this operation.

  3. The trial was done in one hospital, so the result may not apply to other settings.

  4. Adults were randomly allocated to drug A or drug B.

  5. Drug A and drug B are both used for pain after laparoscopic cholecystectomy.

  6. Nausea occurred in 9 patients given drug A and in 11 given drug B.

  7. Surgeons may consider drug A for patients having this operation.

  8. This trial asked whether drug A gives lower pain scores than drug B at 24 hours.

  9. A nurse who did not know the allocation recorded the pain score on a 0 to 10 scale.

  10. Drug A gave lower pain scores than drug B, a smaller difference than earlier studies had reported.

Comparison

Thesis or journal paper: what changes?

The content is the same study. The reader, the length and the rules are different.

FeatureThesisJournal paper
Written forExaminers, for a degreeEditors, peer reviewers and readers of the journal
Rules set byThe university thesis regulationsThe journal instructions to authors and the ICMJE Recommendations
LengthLong, often with page limitsShort, with a word limit
Review of literatureA separate chapterA few paragraphs in the introduction and discussion
Extra materialAnnexures: consent form, proforma, master chartOnline supplementary files
Reporting checklistUseful for checking that nothing is left outCONSORT 2025 recommends submitting the completed checklist

Scientific writing

What makes scientific writing different?

A reader must be able to understand what you did, check it and repeat it. Scientific writing is accurate, clear, brief and objective.

It was observed that there was a reduction in the pain scores of thepatients who were in the group that received drug A when a comparisonwas made with the group that received drug B.36 wordsPatients given drug A had lower pain scores than patients given drug B.13 wordsthe same finding in about a third of the words
Teaching example. Both sentences report the same finding. The first uses 36 words, with a vague opening ("it was observed that") and a verb turned into a noun ("a comparison was made"). The second uses 13 words. In a real paper, add the numbers: the mean scores, the difference and its confidence interval.
  • Accurate: every number in the text matches the tables, and every citation says what you claim it says.
  • Clear: one idea per sentence, and technical terms defined at first use.
  • Brief: no words that add no information.
  • Precise: exact numbers with units and confidence intervals, not "many patients improved".
  • Objective: report what the data show, including results that did not support the hypothesis.
  • Logical: sections in IMRAD order, and paragraphs that follow from one another.
  • Consistent: the same term for the same thing throughout (do not switch between "cases" and "subjects").
  • Tense: past tense for what you did and found, present tense for established knowledge.
  • Honest: sources acknowledged, no copying and no invented data. See the publishing and IPR chapter.

The abstract

What must an abstract contain?

The ICMJE notes that the abstract is the only substantive part of an article indexed in many electronic databases, and the only part many readers read. It must reflect the article accurately.

Unstructuredthe result is somewhere in the paragraphStructuredBackgroundMethodsResultsConclusionsthe result is under Results
The same content in two forms. In a structured abstract the reader finds the result under its heading. Structured abstracts were proposed for clinical articles in 1987 and revised in 1990, and there is considerable evidence that they communicate more effectively than traditional abstracts (Altman and Simera, 2016).

The essential features (ICMJE Recommendations)

  1. The context or background of the study.
  2. The purpose of the study.
  3. Basic procedures: selection of participants, settings, measurements and analytical methods.
  4. Main findings, with specific effect sizes and their statistical and clinical significance where possible.
  5. Principal conclusions.
  6. Emphasis on new and important aspects, with important limitations noted and no overinterpretation.
  7. For a trial, the registration number at the end of the abstract.

Writing the abstract

  1. Write it last, after the results and discussion are final.
  2. Use the headings and word limit of the journal or university.
  3. Give the main result as numbers, for example a difference with its 95% confidence interval, not only "significant".
  4. Check that every number matches the results section. The ICMJE warns that information in abstracts often differs from the text.
  5. Do not include material that is not in the paper, and do not cite references.
  6. Make the conclusion follow from the result you reported.
  7. Add keywords, for example MeSH terms, so that the paper can be found in searches.

Comparison

Structured or unstructured abstract?

FeatureStructuredUnstructured
FormShort paragraphs under headingsOne continuous paragraph
Typical headingsBackground, Methods, Results, Conclusions (some journals use Objective, Design, Setting, Participants, Interventions, Main outcome measures)None
Required forOriginal research, systematic reviews and meta-analyses (ICMJE)Articles for which the journal does not ask for headings
AdvantageEasier to scan, and missing items are easy to seeFlexible for articles that do not follow IMRAD
DisadvantageHeadings use part of the word limitKey items, such as the design or the numbers, are easily left out

The research report

What is a research report?

A research report is a written account of a completed study: the question, what was done, what was found and what it means. Its form depends on the reader.

TypeReaderForm
Thesis or dissertationExaminersFull chapters with review of literature and annexures
Journal articleResearchers and cliniciansIMRAD, word limit, reporting guideline
Technical reportFunder, sponsor or ethics committeeInterim or final report with full methods and data
Popular reportPatients, public, communityPlain language, main finding and what it means for them
Policy briefHealth officials and administratorsOne or two pages with the finding and recommended action

Every type has the same three parts: preliminary pages (title, abstract or summary, contents), the main text (introduction, methods, results, discussion, conclusions and recommendations) and end matter (references, appendices).

Dissemination

How do findings reach the people who can use them?

Dissemination is the planned communication of findings to the people who can act on them. Utilisation is the use of those findings to change practice, policy or further research.

Your findingthesis, datasetResearchersjournal paper, conferenceCliniciansCME talk, department meetingPatients and publicplain-language summaryPolicy makersone-page brief with actionsUtilisationpractice or policychangesdisseminationuse
Each audience needs a different form. It is often stated that research evidence takes an average of 17 years to reach clinical practice. A 2011 review of 23 papers on such time lags found that the studies measured different things at different stages, so the figure is of limited use (Morris, Wooding and Grant, 2011). A finding that is not communicated cannot be used.

Knowledge dissemination, step by step

  1. Plan it at the protocol stage: who needs the result, and what could they change?
  2. Involve the users early: clinicians, programme officers, patient groups.
  3. Write a separate message for each audience, in the form that audience reads.
  4. Choose the channels: journal, conference, CME, departmental meeting, community meeting, summary for participants.
  5. Report negative results as well as positive ones, so that reviews and guidelines are not biased.
  6. Check later whether the finding was used.

Methods of illustration

Which illustration shows which kind of result?

Tables give exact values. Graphs show patterns. Diagrams show a process. The detailed rules are in the data presentation chapter.

You want to showUse
Exact values for several variables, such as baseline characteristicsTable
Counts or percentages in categoriesBar chart
Distribution of a continuous measurementHistogram or box plot
Relation between two measurementsScatter plot
Change over timeLine graph
Flow of participants through a studyFlow diagram (CONSORT, PRISMA)
Clinical findings, operative steps, histologyPhotograph or line drawing, with consent and identity removed

Number tables and figures in the order they are first cited, give each a title that makes sense without the text, and do not repeat the same data in a table and a graph. See the data presentation chapter.

Two types of referencing

How does a citation lead the reader to its source?

Every referencing style has two parts: a marker in the text and a full entry in a list. Vancouver and APA differ in what the marker is and in how the list is ordered.

Vancouvernumbers, in order of first citation(1)(2)(1)same numberReferences1. Sollaci LB, Pereira MG … 20042. Haynes RB, Mulrow CD … 1990APAauthor and year, list in alphabetical order(Sollaci & Pereira, 2004)(Haynes et al., 1990)ReferencesHaynes, R. B., Mulrow, C. D., … (1990)Sollaci, L. B., & Pereira, M. G. (2004)list follows the order of citationlist follows the alphabet
Grey bars are lines of text, and the list entries are shortened. Vancouver numbers each source when it is first cited, and a source cited again keeps its first number. APA puts the author and year in the text and lists the sources in alphabetical order, so Haynes comes before Sollaci although Sollaci was cited first.

Interactive

Add a citation and watch the Vancouver numbers change while APA stays the same

Try this: add the new citation in sentence 2 and watch the Vancouver numbers after it change. Then move it to the end.

The sources are real papers cited in this chapter. Titles are left out of the lists to save space.

New citation (Altman and Simera, 2016)
0Vancouver numbers changed · 0 APA markers changed

Sollaci and Pereira are cited twice. In Vancouver style the second citation keeps its first number, 1.

Vancouver

By the 1980s, IMRAD was the only pattern for original papers in four major general medical journals (1). A revised proposal set out the headings of a structured abstract in 1990 (2). The often-quoted average of 17 years from research to practice rests on studies that measured different things (3). The four journals studied were BMJ, JAMA, The Lancet and NEJM (1).

  1. Sollaci LB, Pereira MG. J Med Libr Assoc. 2004;92:364-7.
  2. Haynes RB, Mulrow CD, Huth EJ, Altman DG, Gardner MJ. Ann Intern Med. 1990;113:69-76.
  3. Morris ZS, Wooding S, Grant J. J R Soc Med. 2011;104:510-20.

APA

By the 1980s, IMRAD was the only pattern for original papers in four major general medical journals (Sollaci & Pereira, 2004). A revised proposal set out the headings of a structured abstract in 1990 (Haynes et al., 1990). The often-quoted average of 17 years from research to practice rests on studies that measured different things (Morris et al., 2011). The four journals studied were BMJ, JAMA, The Lancet and NEJM (Sollaci & Pereira, 2004).

  • Haynes, R. B., Mulrow, C. D., Huth, E. J., Altman, D. G., & Gardner, M. J. (1990).
  • Morris, Z. S., Wooding, S., & Grant, J. (2011).
  • Sollaci, L. B., & Pereira, M. G. (2004).

Vancouver style

How do you write a reference in Vancouver style?

Vancouver style is the numbered style set out by the International Committee of Medical Journal Editors (ICMJE), with reference formats from the US National Library of Medicine (NLM). The detailed rules are in NLM's Citing Medicine, 2nd edition.

VancouverHaynes RB, Mulrow CD, Huth EJ, Altman DG, Gardner MJ.authors: surname and initialsMore informative abstracts revisited.title, sentence caseAnn Intern Med.journal, MEDLINE abbreviation1990;113(1):69-76.year;volume(issue):pagesAPAHaynes, R. B., Mulrow, C. D., Huth, E. J., Altman, D. G.,authors: surname, then initials& Gardner, M. J.& before the last author(1990).year in bracketsMore informative abstracts revisited.title, sentence caseAnnals of Internal Medicine, 113(1), 69–76.journal in full, volume in italicshttps://doi.org/10.7326/0003-4819-113-1-69DOI as a link
The same article in both styles. Vancouver abbreviates the journal and joins year, volume, issue and pages with punctuation. APA spells out the journal, puts the year after the authors and ends with the DOI.

The rules

  1. Number references consecutively in the order in which they are first mentioned in the text.
  2. In the text, identify references by Arabic numerals in parentheses, for example (1). A source cited again keeps its first number.
  3. References cited only in a table or figure legend take numbers from the point where that table or figure is first cited in the text.
  4. Authors: surname followed by initials without full stops, separated by commas. List the first six authors, then "et al."
  5. Article title in sentence case: capital letters only for the first word and proper nouns.
  6. Journal title abbreviated as in MEDLINE (the NLM Catalog), for example Ann Intern Med.
  7. Year;volume(issue):pages, with repeated digits of the last page left out (69-76, 284-7).
  8. Books: author, title, edition, place of publication, publisher and year.
  9. Material on the internet: add [Internet], the date you cited it and "Available from:" with the address.

Worked examples

SourceVancouver reference
Journal article, up to six authorsHaynes RB, Mulrow CD, Huth EJ, Altman DG, Gardner MJ. More informative abstracts revisited. Ann Intern Med. 1990;113(1):69-76.
Journal article, more than six authorsChan AW, Tetzlaff JM, Altman DG, Laupacis A, Gøtzsche PC, Krleža-Jerić K, et al. SPIRIT 2013 statement: defining standard protocol items for clinical trials. Ann Intern Med. 2013;158(3):200-7.
BookLwanga SK, Lemeshow S. Sample size determination in health studies: a practical manual. Geneva: World Health Organization; 1991.
Book or guide on a websitePatrias K. Citing medicine: the NLM style guide for authors, editors, and publishers [Internet]. 2nd ed. Wendling DL, technical editor. Bethesda (MD): National Library of Medicine (US); 2007 - [updated 2015 Oct 2; cited 2026 Sep 14]. Available from: http://www.nlm.nih.gov/citingmedicine

The website example is the form NLM itself gives for citing Citing Medicine. Replace the cited date with the date you consulted the page.

Advantages

  • A number takes little space and does not interrupt the sentence.
  • The list is in the order of reading, so the reader finds each reference quickly.
  • It is the style of the ICMJE Recommendations, which a large number of medical journals follow.
  • The journal abbreviations match those used in MEDLINE and PubMed.

Disadvantages

  • A number tells the reader nothing about who wrote the source or when.
  • Adding or removing a citation changes every number after it, in the text and in the list.
  • The list is not alphabetical, so finding a particular author means reading the whole list.
  • Journals vary in details, for example whether electronic sources are cited in the text or in the numbered list (ICMJE), so check the instructions to authors.

APA style

How do you write a reference in APA style?

APA style is an author-date system from the American Psychological Association. The current edition of its Publication Manual is the 7th, and the APA Style website still gave 7th edition guidance in 2026.

The rules

  1. In the text, give the author surname and year: (Sollaci & Pereira, 2004) in brackets, or Sollaci and Pereira (2004) in a sentence.
  2. For three or more authors, give the first author and "et al." in every citation, including the first: (Haynes et al., 1990).
  3. The reference list is in alphabetical order of the first author's surname, with a hanging indent.
  4. Authors: surname, comma, initials with a space between them. Use & before the last author. List up to 20 authors. With 21 or more, give the first 19, an ellipsis, then the last author.
  5. Year in parentheses after the authors.
  6. Article title in sentence case, not italic. Journal title in full and in italics, volume in italics, issue in parentheses.
  7. Include the issue number, and give the DOI as a link (https://doi.org/...).
  8. Books: title in italics, edition in parentheses, publisher. Do not include the place of publication.
  9. Webpages: author, date, title in italics, site name (left out when it is the same as the author) and URL.

Worked examples

SourceAPA reference (7th edition)
Journal articleHaynes, R. B., Mulrow, C. D., Huth, E. J., Altman, D. G., & Gardner, M. J. (1990). More informative abstracts revisited. Annals of Internal Medicine, 113(1), 69–76. https://doi.org/10.7326/0003-4819-113-1-69
BookLwanga, S. K., & Lemeshow, S. (1991). Sample size determination in health studies: A practical manual. World Health Organization.
Online document by an organisationInternational Committee of Medical Journal Editors. (2026, January). Recommendations for the conduct, reporting, editing, and publication of scholarly work in medical journals. https://www.icmje.org/icmje-recommendations.pdf

In the text these become (Haynes et al., 1990), (Lwanga & Lemeshow, 1991) and (International Committee of Medical Journal Editors, 2026).

Comparison

Vancouver or APA: how do they differ?

Use the style your university or journal asks for. Know both for the exam.

FeatureVancouverAPA (7th edition)
TypeNumbered (citation-sequence)Author-date
Marker in textArabic numeral in parentheses: (1)(Haynes et al., 1990)
Order of listOrder of first citationAlphabetical by first author
Authors listedFirst six, then et al.Up to 20; with 21 or more, first 19 … last
InitialsHaynes RBHaynes, R. B.
Journal titleMEDLINE abbreviation, not italicFull title, italic
YearAfter the journalAfter the authors, in parentheses
DOIOptionalIncluded when the work has one
Place of publisher (books)IncludedNot included
Governing sourceICMJE Recommendations, NLM Citing MedicineAPA Publication Manual, 7th edition
Used mainly inMedical journals that follow the ICMJE RecommendationsSocial and behavioural sciences, nursing, education
Memory trick

Vancouver: a number, in the order you first cite the source. APA: a name and a year, with the list from A to Z.

Interactive

Edit a reference and watch it change in Vancouver and APA style

Try this: choose the paper with 26 authors. Vancouver lists six and then "et al.", while APA lists the first 19, an ellipsis and the last author.

Example paper

Both are real papers. Edit any field to see how each style changes.

5 authors

Type it in sentence case

Used by Vancouver

Used by APA

Leave empty if the journal has none

Only if the journal uses one instead of pages

For example 10.1136/bmj.n71

Vancouver

Haynes RB, Mulrow CD, Huth EJ, Altman DG, Gardner MJ. More informative abstracts revisited. Ann Intern Med. 1990;113(1):69-76.

5 authors: all are listed. Initials without full stops, journal abbreviated, year;volume(issue):pages with repeated digits of the last page left out.

APA (7th edition)

Haynes, R. B., Mulrow, C. D., Huth, E. J., Altman, D. G., & Gardner, M. J. (1990). More informative abstracts revisited. Annals of Internal Medicine, 113(1), 69–76. https://doi.org/10.7326/0003-4819-113-1-69

5 authors: all are listed, with & before the last. Year after the authors, journal and volume in italics, DOI as a link.

Reporting guidelines

What are STROBE, SPIRIT and MOOSE?

A reporting guideline is a checklist of what a report must contain for a given type of study. It says what to report. It does not say how to design the study, and it is not a score of study quality.

GuidelineFull nameForContent
STROBE (2007)Strengthening the Reporting of Observational Studies in EpidemiologyCohort, case-control and cross-sectional studies22 items: 18 common to all three designs, 4 specific to one
SPIRIT 2013, SPIRIT 2025Standard Protocol Items: Recommendations for Interventional TrialsProtocols of clinical trials33 items (2013). 34 minimum items and a schedule diagram (2025)
MOOSE (2000)Meta-analysis Of Observational Studies in EpidemiologyMeta-analyses of observational studiesChecklist for background, search strategy, methods, results, discussion and conclusion
CONSORT 2025Consolidated Standards of Reporting TrialsReports of randomised trials30 items and a participant flow diagram
PRISMA 2020Preferred Reporting Items for Systematic reviews and Meta-AnalysesSystematic reviews27 items, abstract checklist, flow diagrams
STARD 2015Standards for Reporting Diagnostic AccuracyDiagnostic accuracy studies30 items
CARE (2013)CAse REport guidelinesCase reports13 items

The EQUATOR Network (Enhancing the QUAlity and Transparency Of health Research) is an international initiative that promotes accurate reporting and wider use of reporting guidelines. The project began in March 2006 and the Network was officially launched in London in June 2008. Its free online library lists reporting guidelines by study type. The ICMJE Recommendations direct authors to it.

Decide

Which reporting guideline fits your study?

Answer one or two questions.

What are you writing?

Which guideline: all possible answers
  • SPIRIT 2025: 34 minimum items for a trial protocol, with a diagram of the schedule of enrolment, interventions and assessments.
  • CONSORT 2025: A 30-item checklist and a diagram of the flow of participants through the trial.
  • STROBE: 22 items. 18 apply to cohort, case-control and cross-sectional studies, and 4 are specific to one of these designs.
  • STARD 2015: 30 essential items for a report of a diagnostic accuracy study.
  • PRISMA 2020: A 27-item checklist, an abstract checklist and flow diagrams for the selection of studies.
  • MOOSE: A checklist for the background, search strategy, methods, results, discussion and conclusion of a meta-analysis of observational studies.
  • CARE: A 13-item checklist for case reports, including a timeline, the patient perspective and informed consent.
  • EQUATOR Network library: Search the free EQUATOR Network database of reporting guidelines by study type.

By specialty

Writing and reporting problems by specialty

Your thesis is a randomised trial of two oral drugs for type 2 diabetes. Your guide also wants a manuscript for a journal.

Write the paper against the CONSORT 2025 checklist and note the page on which each of its 30 items is reported.

Common myths

Common mistaken beliefs

  • Myth: The abstract is an introduction to the paper.

    It summarises the whole study: purpose, basic procedures, main findings with numbers and the principal conclusion.

  • Myth: You must write the sections in IMRAD order.

    IMRAD is the order in which the paper is read. Many authors write the methods and results first and the abstract last.

  • Myth: Reading the abstract is enough before citing a paper.

    The ICMJE warns that information in abstracts often differs from the text. Read the full paper before you cite it.

  • Myth: Vancouver and APA differ only in punctuation.

    They link text to sources differently. Vancouver uses numbers in order of first citation. APA uses author and year with an alphabetical list.

  • Myth: A STROBE checklist score shows how good an observational study is.

    STROBE is a reporting guideline. In a sample of 19 systematic reviews that cited it, 10 used it inappropriately as a tool to assess study quality (da Costa et al., 2011).

  • Myth: Once the paper is published, dissemination is complete.

    Participants, clinicians and health officials rarely read journals. Each needs the finding in a form they use.

Why this matters to you

Five consequences of poor writing and referencing

  1. Thesis returned for corrections

    Citation numbers in the text do not match the reference list, or the format does not follow the university regulations.

  2. An abstract that does not match the results

    The examiner reads the abstract first. A benefit claimed there but absent from the results table is noticed at once.

  3. Unanswered viva questions

    You cannot say what a structured abstract contains, or why Vancouver numbers follow the order of citation.

  4. Paper sent back by the journal

    Methods written without the reporting guideline leave out items that peer reviewers ask for.

  5. Findings that never reach patients

    A result that stays in the department library does not change what clinicians do.

History

History of writing and referencing standards

  1. 1929

    Bentley and colleagues, Psychological Bulletin

    Published a seven-page article, "Instructions in regard to preparation of manuscript". APA describes it as the start of what became APA Style.

  2. 1978

    Editors of general medical journals, Vancouver

    A small group of editors met in Vancouver, British Columbia, to agree on a format for manuscripts. The group became known as the Vancouver Group.

  3. 1979

    The Vancouver Group

    Published the Uniform Requirements for Manuscripts Submitted to Biomedical Journals, with reference formats developed by the US National Library of Medicine. The group grew into the ICMJE.

  4. 1987

    Ad Hoc Working Group for Critical Appraisal of the Medical Literature

    Proposed structured abstracts for clinical articles in Annals of Internal Medicine.

  5. 1990

    Haynes, Mulrow, Huth, Altman and Gardner

    Revised the proposal. Structured abstracts of original studies were to state objective, design, setting, participants, interventions, main outcome measures, results and conclusions.

  6. 1996

    Begg and colleagues

    Published the CONSORT statement for reports of randomised trials. It was updated in 2001, 2010 and 2025.

  7. 2000

    Stroup and colleagues

    Published MOOSE, developed at a workshop held in Atlanta in April 1997.

  8. 2004

    Sollaci and Pereira

    Surveyed articles in BMJ, JAMA, The Lancet and NEJM from 1935 to 1985. IMRAD was first used in the 1940s, reached 80% in the 1970s and was the only pattern in original papers in the 1980s.

  9. 2007

    von Elm and colleagues, and NLM

    STROBE was published with 22 items. NLM published the 2nd edition of Citing Medicine in the same year.

  10. 2008

    EQUATOR Network

    Officially launched at its inaugural meeting in London in June 2008, two years after the project began in March 2006.

  11. 2013

    Chan and colleagues

    Published SPIRIT 2013, a 33-item checklist for clinical trial protocols.

  12. 2019

    American Psychological Association

    Announced the 7th edition of the Publication Manual, to go on sale in October 2019. It remains the current edition.

  13. 2025

    SPIRIT and CONSORT groups

    Published SPIRIT 2025 (34 items) and CONSORT 2025 (30 items), each with a new section on open science.

Guidelines

Writing and referencing in guidelines

ICMJE Recommendations (updated January 2026): manuscript preparation

The IMRAD structure is "not an arbitrary publication format but a reflection of the process of scientific discovery."

ICMJE Recommendations: abstract

"Original research, systematic reviews, and meta-analyses require structured abstracts." The abstract should give the context, purpose, basic procedures, main findings and principal conclusions, note important limitations and not overinterpret the findings.

ICMJE Recommendations: references

References are numbered consecutively in the order in which they are first mentioned, identified by Arabic numerals in parentheses, with journal titles abbreviated as in MEDLINE. They follow NLM's Sample References and Citing Medicine, 2nd edition.

NLM Samples of Formatted References

"List the first six authors followed by et al." NLM itself lists all authors in its databases.

APA Publication Manual, 7th edition (APA Style website)

Provide surnames and initials for up to and including 20 authors. For a work with three or more authors, include only the first author plus "et al." in every in-text citation.

STROBE (2007), SPIRIT 2025, MOOSE (2000)

STROBE: 22 items for cohort, case-control and cross-sectional studies. SPIRIT 2025: 34 minimum items for a trial protocol. MOOSE: a checklist for reporting meta-analyses of observational studies in epidemiology.

EQUATOR Network

An international initiative to improve the reliability and value of published health research by promoting transparent and accurate reporting and wider use of reporting guidelines.

Steps

Writing a thesis or paper, step by step

  1. Read the university thesis regulations or the journal's instructions to authors before you write.
  2. Choose the reporting guideline for your design and keep its checklist open.
  3. Make the tables and figures first.
  4. Write the methods, then the results.
  5. Write the introduction: what is known, what is not known, and your question.
  6. Write the discussion: main finding, comparison with other studies, limitations and meaning for practice.
  7. Write the abstract and the title last, and check every number against the results.
  8. Insert references in the required style, ideally with reference manager software, and check each against its source.
  9. Complete the checklist with the page number where each item is reported.

Reading a paper

Five checks on how a paper is written

  • Do the numbers in the abstract match the results section?
  • Is the study design named, and does the report cover the items of the matching reporting guideline?
  • Could you repeat the study from the methods section?
  • Are the limitations stated in the discussion?
  • Take one important citation: does the cited paper say what the text claims it says?

Why is it called that?

Word surgery

“Abstract”

In everyday speech

Not concrete, hard to picture, as in abstract art.

In research

A short summary of the whole study that appears before the paper and is indexed on its own in databases.

“Discussion”

In everyday speech

A conversation or a debate.

In research

The section that interprets the results: what they mean, how they compare with other studies and what their limits are.

“Reference”

In everyday speech

A person who vouches for you on a job application.

In research

The full details of a source, written so that a reader can find it.

Memory trick

IMRAD answers four questions in order. Introduction: why was the study needed? Methods: what was done? Results: what was found? Discussion: what does it mean?

Abstract

Latin abstractus: "drawn away"

The main points drawn out of the full paper.

Thesis

Greek thesis: "a setting down, a placing", later "a proposition"

Used for a dissertation presented for a university degree since the 1650s.

Method

Greek methodos: "pursuit, a following after", from meta + hodos, "way"

The way the study was done, described so that another researcher could follow it.

Discussion

Latin discutere: "strike asunder", in Late Latin "to examine, investigate"

The section in which the results are examined.

Cite

Latin citare: "to summon, call"

To call on a source as support for a statement.

Reference

from Latin referre: "to carry back"

An entry that takes the reader back to the source.

Vancouver style

named after the Vancouver Group, editors who met in Vancouver, British Columbia, in 1978

The numbered style of the Uniform Requirements, now the ICMJE Recommendations.

IMRAD

Introduction · Methods · Results · And · Discussion

The standard order of the text of a research article.

Exam answers

Model answer outlines

Logical format for writing a research thesis and a research paper

Ph.D. RM 2015, 2017

Principle: the text of a research article is usually divided into Introduction, Methods, Results and Discussion (IMRAD), which follows the process of scientific discovery (ICMJE). The text narrows from the topic to the question, stays on the study, then widens to practice (hourglass shape).

  1. Preliminary pages (thesis): title page, declaration, certificates from the guide and head of department, acknowledgements, contents, lists of tables, figures and abbreviations.
  2. Title and abstract: an informative title, and a structured abstract with keywords.
  3. Introduction: what is known, the gap and the aim. A thesis adds separate chapters for aims and objectives and review of literature.
  4. Materials and methods: design, setting and period, participants and eligibility, sample size and sampling, intervention or exposure, outcomes, data collection tools, statistical analysis, ethics approval and consent.
  5. Results: participant flow, baseline characteristics, primary then secondary outcomes, in text, tables and figures, without interpretation.
  6. Discussion: main finding, comparison with other studies, possible explanations, strengths and limitations, implications.
  7. Summary and conclusion: answers each objective, with no new data.
  8. References: in the required style, Vancouver or APA.
  9. Annexures (thesis): ethics committee approval, participant information sheet, consent form, proforma, master chart.

For a journal paper, follow the journal instructions and the reporting guideline for the design (CONSORT 2025, STROBE, PRISMA 2020, STARD 2015, CARE).

Characteristics of scientific writing

Ph.D. RM 2016, 2017
  • Accuracy: numbers match the tables, citations support the statements.
  • Clarity: short sentences, one idea each, terms defined at first use.
  • Brevity: no redundant words (a 36-word sentence can often say the same in 13).
  • Precision: exact values, units, confidence intervals.
  • Objectivity: reports what the data show, including negative results, without exaggeration.
  • Logical structure: IMRAD, with each paragraph following from the last.
  • Consistency: the same term, abbreviation and unit throughout.
  • Correct tense: past for methods and results, present for established facts.
  • Reproducibility: methods detailed enough for another researcher to repeat the study.
  • Honesty: sources acknowledged, no plagiarism, fabrication or falsification.
  • Conformity: follows the university or journal format and the relevant reporting guideline.

Essential features of an abstract, and the art of writing an abstract

Ph.D. RM 2017, 2019, 2020

Definition: a brief, self-contained summary of the whole study. It is often the only part indexed in databases and the only part many readers read (ICMJE).

Types: structured, with headings (required by the ICMJE for original research, systematic reviews and meta-analyses), and unstructured, a single paragraph.

Essential features (ICMJE):

  1. Context or background.
  2. Purpose or objective.
  3. Basic procedures: participants and their selection, setting, measurements, analytical methods.
  4. Main findings with specific effect sizes and their statistical and clinical significance.
  5. Principal conclusions, with new and important aspects emphasised.
  6. Important limitations, with no overinterpretation.
  7. Trial registration number at the end, for trials.

Headings proposed by Haynes et al. (1990): objective, design, setting, participants, interventions, main outcome measures, results, conclusions.

The art of writing it:

  • Write it last, within the word limit and headings of the journal.
  • Give numbers (effect size, confidence interval), not only "significant".
  • Check each number against the text, because abstracts often differ from the text (ICMJE).
  • No references, no undefined abbreviations, no information that is not in the paper.
  • A conclusion that follows from the reported result.
  • Keywords, for example MeSH terms.

Research report

Ph.D. RM 2016

Definition: a written account of a completed study, stating the question, methods, findings and their meaning, in a form suited to its reader.

Types: thesis or dissertation, journal article, technical report (interim or final report to a funder or sponsor), popular report for the public, policy brief.

  1. Preliminary pages: title page, acknowledgements, contents, abstract or executive summary.
  2. Main text: introduction and objectives, review of literature, methods, results, discussion, conclusions and recommendations.
  3. End matter: references, appendices (tools, consent forms, additional tables).

Qualities: accurate, complete, clear, objective, reproducible, timely, and written for its reader.

Steps: outline, tables and figures, first draft (methods and results first), revision, abstract, references, check against the reporting guideline, final proofreading.

Dissemination and utilisation of research findings, and knowledge dissemination

Ph.D. RM 2014, 2016

Dissemination: the planned communication of findings to audiences who can use them. Utilisation: the use of findings to change clinical practice, programmes, policy or further research.

AudienceChannel
ResearchersJournal article, conference oral or poster presentation, data sharing
CliniciansCME, departmental meetings, clinical guidelines
Patients, participants, communityPlain-language summary, community meeting, media
Policy makers and administratorsPolicy brief, technical report
  1. Plan dissemination at the protocol stage.
  2. Identify the users and involve them early.
  3. Tailor the message and format to each audience.
  4. Publish all results, including negative ones, to avoid publication bias.
  5. Evaluate whether the findings were used.

Barriers: delay between research and practice (a 17-year average is often quoted, but a 2011 review found the estimates measure different stages), technical language, paywalls, unpublished negative studies, lack of contact with users. Facilitators: open access, summaries for each audience, systematic reviews and guidelines, champions within the health system.

Vancouver system of referencing: rules, advantages and disadvantages

Ph.D. RM 2013, 2014, 2015, 2018

Origin: editors of general medical journals met in Vancouver in 1978. Their Uniform Requirements (1979) used reference formats from the US National Library of Medicine. The group became the ICMJE, and the document is now the ICMJE Recommendations. Detailed rules: NLM's Citing Medicine, 2nd edition.

Rules:

  1. References numbered consecutively in order of first mention, with Arabic numerals in parentheses in the text. A repeated citation keeps its number.
  2. Authors: surname and initials without full stops. First six, then et al.
  3. Title in sentence case. Journal abbreviated as in MEDLINE.
  4. Year;volume(issue):pages.
  5. Books: author, title, edition, place, publisher, year. Internet sources: [Internet], cited date, Available from: URL.

Examples:

  • Haynes RB, Mulrow CD, Huth EJ, Altman DG, Gardner MJ. More informative abstracts revisited. Ann Intern Med. 1990;113(1):69-76.
  • Lwanga SK, Lemeshow S. Sample size determination in health studies: a practical manual. Geneva: World Health Organization; 1991.
  • Patrias K. Citing medicine: the NLM style guide for authors, editors, and publishers [Internet]. 2nd ed. Wendling DL, technical editor. Bethesda (MD): National Library of Medicine (US); 2007 - [updated 2015 Oct 2; cited 2026 Sep 14]. Available from: http://www.nlm.nih.gov/citingmedicine

Advantages: short markers that do not interrupt reading, list in reading order, standard for ICMJE-following medical journals, abbreviations match MEDLINE.

Disadvantages: the marker gives no author or year, adding a citation renumbers the rest, the list is not alphabetical, and journals vary in details.

APA style of referencing

Ph.D. RM 2016

Definition: the author-date style of the American Psychological Association. Current edition: Publication Manual, 7th edition.

  1. In text: (Author, year). Two authors: (Sollaci & Pereira, 2004). Three or more: (Haynes et al., 1990) from the first citation.
  2. Reference list: alphabetical by first author, hanging indent.
  3. Authors: surname, initials. & before the last. Up to 20 authors. With 21 or more, first 19, ellipsis, last author.
  4. Date: year in parentheses after the authors.
  5. Journal article: title in sentence case, journal and volume in italics, issue in parentheses, page range, DOI as https://doi.org/ link.
  6. Book: title in italics, edition, publisher. No place of publication.
  7. Webpage: author, date, title in italics, site name if different from the author, URL.

Example: Haynes, R. B., Mulrow, C. D., Huth, E. J., Altman, D. G., & Gardner, M. J. (1990). More informative abstracts revisited. Annals of Internal Medicine, 113(1), 69–76. https://doi.org/10.7326/0003-4819-113-1-69

Advantages: author and year visible while reading, no renumbering when sources are added. Disadvantages: long markers interrupt the text, especially with several sources, and more words are used.

STROBE, SPIRIT, MOOSE – explain

MD Pharmacology 2024

Reporting guidelines are checklists of the minimum information a report must contain for a given study design. They improve completeness and transparency. They are not quality scores and do not tell you how to design a study. All are listed in the EQUATOR Network library.

STROBESPIRITMOOSE
Full nameStrengthening the Reporting of Observational Studies in EpidemiologyStandard Protocol Items: Recommendations for Interventional TrialsMeta-analysis Of Observational Studies in Epidemiology
Applies toCohort, case-control, cross-sectional studiesClinical trial protocolsMeta-analyses of observational studies
Published2007 (von Elm et al.)2013 (Chan et al.), updated 20252000 (Stroup et al., JAMA)
Content22 items: 18 common, 4 design-specific, covering title, abstract, introduction, methods, results, discussion33 items (2013). 34 minimum items with a schedule diagram (2025), including a new open science section and patient and public involvementChecklist for background, search strategy, methods, results, discussion, conclusion
Developed byWorkshop in September 2004 and later revisionSystematic process. 2025 update used a Delphi survey of 317 participants and a meeting of 30 expertsWorkshop in Atlanta, April 1997, with 27 participants

Common misuse: in a sample of 19 systematic reviews citing STROBE, 10 used it inappropriately as a tool to assess study quality (da Costa et al., 2011).

Related: CONSORT 2025 (trial reports, 30 items), PRISMA 2020 (systematic reviews, 27 items), STARD 2015 (diagnostic accuracy, 30 items), CARE (case reports, 13 items).

Principles and techniques of writing the research proposal

Ph.D. RM 2025

Principles: a clear and important question, feasible with the time, patients and money available, ethically acceptable, with outcomes, sample size and analysis decided in advance, written so that another person could run the study from it.

  1. Title, investigators, site.
  2. Background and rationale: what is known, the gap, why the study is needed.
  3. Research question, hypothesis, primary and secondary objectives.
  4. Methods: design, setting, duration, participants and eligibility criteria, sampling, sample size with its assumptions, intervention or exposure, outcome definitions, data collection tools, statistical analysis plan.
  5. Ethics: ethics committee approval, participant information sheet, informed consent, confidentiality, and trial registration for a clinical trial.
  6. Work plan: timeline (Gantt chart), budget and funding.
  7. References and annexures: proforma, consent forms, questionnaires.

Techniques: for a trial, follow SPIRIT 2025 (34 items). Write in the future tense. Pilot the data collection form. Get a statistician to check the sample size and analysis. Follow the institution's format. See the research protocol chapter and the research question chapter.

Phases of a clinical trial, with particular reference to phase II trials

MD Pharmacology 2009

Source: ICH E8, General Considerations for Clinical Trials (1997). A study type can occur in more than one phase, for example human pharmacology studies later in development. Full detail is in the clinical trials chapter.

  1. Phase I (human pharmacology): starts with the first administration of the new drug to humans. Assesses initial safety and tolerability, pharmacokinetics and pharmacodynamics. May be done in healthy volunteers or certain types of patients. Drugs with significant potential toxicity, such as cytotoxic drugs, are usually studied in patients.
  2. Phase II (therapeutic exploratory): starts with studies whose primary objective is to explore therapeutic efficacy in patients.
    • Early studies may use various designs, including concurrent controls and comparison with baseline. Later studies are usually randomised and concurrently controlled.
    • Patients are selected by relatively narrow criteria, giving a relatively homogeneous population, and are closely monitored.
    • A main goal is to determine the dose or doses and the regimen for phase III. Early studies often use dose escalation designs, and later studies may confirm the dose-response relationship with parallel dose-response designs.
    • Doses are usually, but not always, lower than the highest doses used in phase I.
    • Other objectives: evaluating possible endpoints, therapeutic regimens and target populations (for example mild versus severe disease) for later study.
  3. Phase III (therapeutic confirmatory): confirms the evidence from phase II that the drug is safe and effective for the intended indication and population. Provides the basis for marketing approval.
  4. Phase IV (therapeutic use): all studies after approval related to the approved indication, other than routine surveillance, such as additional interaction, dose-response, safety, and morbidity or mortality studies.

Reporting: trial reports follow CONSORT 2025, and a CONSORT extension exists for early phase trials. Protocols follow SPIRIT 2025.

The one thing to remember

Write so that a reader can check every statement: each number against your results, and each citation against the source it names.

The sentences in the scientific writing figure are a made-up teaching example. The worked references are real publications.

Sources: International Committee of Medical Journal Editors. Recommendations for the Conduct, Reporting, Editing, and Publication of Scholarly Work in Medical Journals, updated January 2026 (www.icmje.org). ICMJE. Uniform Requirements for Manuscripts Submitted to Biomedical Journals, updated October 2008 (ICMJE archives). International Steering Committee. Uniform requirements for manuscripts submitted to biomedical journals. Ann Intern Med 1979;90:95–9 (doi:10.7326/0003-4819-90-1-95). National Library of Medicine. Samples of Formatted References for Authors of Journal Articles (www.nlm.nih.gov/bsd/uniform_requirements.html). Patrias K. Citing Medicine, 2nd ed. Bethesda (MD): NLM; 2007 (NBK7256). American Psychological Association. Publication Manual, 7th ed., and APA Style reference and citation guidance (apastyle.apa.org). Bentley M et al. Psychol Bull 1929;26:57–63 (doi:10.1037/h0071487). Sollaci LB, Pereira MG. J Med Libr Assoc 2004;92:364–7 (PMID 15243643). Ad Hoc Working Group for Critical Appraisal of the Medical Literature. Ann Intern Med 1987;106:598–604 (doi:10.7326/0003-4819-106-4-598). Haynes RB et al. Ann Intern Med 1990;113:69–76 (doi:10.7326/0003-4819-113-1-69). Altman DG, Simera I. J R Soc Med 2016;109:67–77 (doi:10.1177/0141076815625599). Morris ZS, Wooding S, Grant J. J R Soc Med 2011;104:510–20 (doi:10.1258/jrsm.2011.110180). Begg C et al. JAMA 1996;276:637–9 (doi:10.1001/jama.276.8.637). Stroup DF et al. JAMA 2000;283:2008–12 (doi:10.1001/jama.283.15.2008). von Elm E et al. Lancet 2007;370:1453–7 (doi:10.1016/S0140-6736(07)61602-X). da Costa BR et al. BMJ Open 2011;1:e000048 (doi:10.1136/bmjopen-2010-000048). Chan AW et al. Ann Intern Med 2013;158:200–7 (doi:10.7326/0003-4819-158-3-201302050-00583). Chan AW et al. SPIRIT 2025 statement. BMJ 2025;389:e081477 (doi:10.1136/bmj-2024-081477). Hopewell S et al. CONSORT 2025 statement. BMJ 2025;389:e081123 (doi:10.1136/bmj-2024-081123). Page MJ et al. PRISMA 2020 statement. BMJ 2021;372:n71 (doi:10.1136/bmj.n71). Bossuyt PM et al. STARD 2015. BMJ 2015;351:h5527 (doi:10.1136/bmj.h5527). Gagnier JJ et al. CARE guidelines. J Med Case Rep 2013;7:223 (doi:10.1186/1752-1947-7-223). EQUATOR Network, About us and History (www.equator-network.org). ICH E8 General Considerations for Clinical Trials, 1997. Lwanga SK, Lemeshow S. Sample size determination in health studies. Geneva: WHO; 1991. ISBN 92 4 154405 8.

Past papers

Every TNMGRMU question on writing and referencing

Each one is answered in this chapter. Tap a sitting to open the university's own question paper.

  1. ×2What are the features that are essential for an abstract?Ph.D. RM · Dec 2020Ph.D. RM · Jul 2017
  2. ×2Reference writing in Vancouver stylePh.D. RM · Jan 2018Ph.D. RM · Jul 2014
  3. ×1Principles and Techniques of writing the Research ProposalPh.D. RM · Jan 2025
  4. ×1STROBE, SPIRIT, MOOSE – explainMD Pharmacology · Oct 2024
  5. ×1Describe the essential features required in an abstractPh.D. RM · Jan 2019
  6. ×1Art of writing an abstractPh.D. RM · Jul 2019
  7. ×1Describe the characteristics and logical format for writing a research thesis and research papersPh.D. RM · Jan 2017
  8. ×1Characteristics of Scientific writingPh.D. RM · Jan 2016
  9. ×1Research reportPh.D. RM · Jul 2016
  10. ×1APA style of ReferencingPh.D. RM · Jul 2016
  11. ×1Knowledge disseminationPh.D. RM · Jul 2016
  12. ×1Referencing styles – VancouverPh.D. RM · Jan 2015
  13. ×1Describe in detail the logical format for writing a research workPh.D. RM · Jul 2015
  14. ×1Dissemination and Utilization of research findingsPh.D. RM · Jan 2014
  15. ×1Discuss the Vancouver System of referencing, its advantages and disadvantagesPh.D. RM · Jul 2013
  16. ×1Describe different phases of a clinical trial with particular reference to phase II trialsMD Pharmacology · Sep 2009