Blazon Scientific logo

Scientific manuscript writing guide

How to Write a Scientific Manuscript for Publication

Written by Thomas Solomon, PhD

A scientific manuscript does more than describe a completed research project. It explains why the work was undertaken, how it addressed a defined question, what the researchers found and what those findings mean in the context of existing evidence. Readers should be able to follow that explanation and judge how much confidence to place in its conclusions.

Writing a good manuscript therefore requires more than polished English. The research question, hypothesis where applicable, study design, methods, analyses, results and interpretation need to be consistent. The writing should make those relationships clear without hiding unexpected findings, exaggerating the contribution or overlooking important limitations.

This guide explains how to write an original scientific manuscript for publication, particularly in clinical, biomedical and health research. It covers planning the paper, writing each section, preparing tables and figures, using reporting guidance and checking the finished manuscript. A hypothetical exercise-and-hypertension trial provides a continuing example.

Table of contents
The research workflow

A clinical research study involves a series of connected decisions, from defining the question to publishing the findings. The six stages below show how the parts of a study fit together and where manuscript writing belongs.

  1. Develop the research question
  2. Plan the study
  3. Prepare and conduct the study
  4. Collect and manage data
  5. Analyse and interpret
  6. Write, submit and publish

Manuscript preparation is not isolated from the earlier stages. Protocols, analyses and results need to align with what authors report; important inconsistencies should be resolved or transparently explained rather than edited away.

1. Before you write: understand the study and its contribution

Before drafting, make sure you can explain the study in a few sentences. What uncertainty motivated the work? What question did the researchers investigate? What were the principal findings, and what do they add to the evidence already available? If these questions are difficult to answer, the first task may be to clarify the study’s objectives and analyses rather than polish the opening paragraph.

Identify the scientific contribution

A worthwhile paper need not demonstrate a completely new phenomenon. Research may test an earlier finding more rigorously, extend evidence to another population, improve measurement, prolong follow-up or resolve a specific uncertainty. Replication and confirmatory studies also make useful contributions, particularly where previous evidence is inconsistent or difficult to reproduce.

Conversely, a claim that no previous study has investigated a question requires a defensible search and assessment of the literature. Describe originality proportionately rather than inventing a gap or presenting incremental evidence as a breakthrough.

Keep the primary question visible

Consider a hypothetical 12-week randomised trial comparing supervised exercise with usual care in adults with hypertension. Its primary objective is to estimate the between-group effect on clinic systolic blood pressure at 12 weeks. Secondary outcomes might include aerobic fitness and physical activity. An interesting secondary result does not become the primary finding because it has a smaller P value or makes a more appealing headline.

Unexpected findings may be important, but authors should distinguish prespecified analyses from exploratory investigations undertaken after examining the data. This distinction will matter when writing both Results and Discussion.

Choose a practical drafting order

Most original research reports are organised as Introduction, Methods, Results and Discussion, often abbreviated to IMRaD. That is the order readers see, not necessarily the order authors should write. Some researchers organise the main tables and figures first, then draft Methods and Results before developing the Introduction and Discussion. Others already have a protocol containing much of the Methods.

There is no compulsory writing sequence. Choose one that makes the research easier to explain accurately. The abstract is often best finalised only after the main text and numerical displays have stabilised.

2. Choose the journal and reporting guideline

Select a journal whose remit, readership and article categories suit the research. Consider editorial standards, indexing, access arrangements and publication costs as well as the relevance of the findings to its readers. Journal impact factor alone is a poor proxy for whether a particular paper belongs in a journal.

Read the journal’s actual instructions

Check the target journal’s current requirements for article type, length, abstract structure, title page, figures, tables, supplementary material, reference style and mandatory declarations. There is no universal 250-word abstract rule or fixed number of permitted keywords. The journal’s instructions are the practical authority for its submission process.

The International Committee of Medical Journal Editors (ICMJE) Recommendations provide a widely used framework for biomedical publication responsibilities, but they do not remove the need to consult the individual journal’s policies.

Match the reporting guideline to the study

Reporting guidelines help authors describe research transparently. The Consolidated Standards of Reporting Trials (CONSORT) 2025 statement is relevant to randomised trials; the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement applies to major observational designs. Other frameworks include PRISMA for systematic reviews, STARD for diagnostic accuracy studies, TRIPOD for prediction models and ARRIVE for animal research.

Check for appropriate extensions and consult the EQUATOR Network reporting-guideline library. Select guidance because it matches the research, not because it is familiar. A completed checklist improves reporting completeness but does not establish that the study design, conduct or analysis was sound.

3. Plan the manuscript's scientific argument

A manuscript is a connected scientific explanation, not a collection of independently written sections. Readers should be able to follow a line from the research question and hypothesis, where relevant, through the methods and analyses to the findings, interpretation and conclusion. Problems arise when the Introduction promises one investigation, the Methods describe another and the Discussion draws conclusions about questions the study did not address.

Build a manuscript consistency map

Before drafting, link the main objective and any prespecified hypothesis to the corresponding methods, result and interpretation. The table below illustrates how this works in our hypothetical exercise trial.

ElementManuscript content
Research questionDoes supervised exercise affect clinic systolic blood pressure at 12 weeks compared with usual care?
Primary objectiveEstimate the between-group treatment effect at 12 weeks
HypothesisParticipants assigned to supervised exercise will have lower mean systolic blood pressure than those assigned to usual care
Design and outcomeIndividually randomised parallel-group trial; clinic systolic blood pressure at 12 weeks
Primary analysisPrespecified between-group comparison adjusted for baseline systolic blood pressure
Main resultAdjusted between-group estimate and confidence interval
DiscussionReturn to the original objective and hypothesis; consider magnitude, uncertainty, other evidence and limitations
ConclusionProvide a proportionate answer to the primary research question

Where a protocol, trial registration or statistical analysis plan exists, compare it with the manuscript. Important deviations need an explanation, not a retrospective rewriting of the original objectives.

Plan the principal displays

Outline the main tables and figures before writing a lengthy Results section. This can reveal unclear denominators, inconsistent outcome names or gaps between the model outputs and the claims authors hope to make. Include displays because they communicate information readers need, not merely because the journal permits them.

4. How to write the Introduction

The Introduction should explain why the investigation was needed and what it was designed to find out. A useful logical progression is established knowledge, remaining uncertainty, rationale, and finally the research objectives and relevant hypotheses. It does not need to become an exhaustive history of the topic.

Establish the relevant evidence

Begin with the scientific or clinical problem and the best evidence directly related to the question. For our exercise trial, this might involve the importance of hypertension and previous evidence concerning exercise and blood-pressure control. Prioritise relevant studies and reviews, but examine their populations, methods and limitations rather than treating their findings as interchangeable.

Identify a defensible gap

Explain precisely what remains uncertain and how the current study could address it. Perhaps previous exercise trials provided limited evidence for the population being studied, used different training regimens or lacked consistent outcome measurement. Avoid a vague assertion that further research is needed, and avoid claiming to be first without an adequate assessment of the literature.

State the objective and, when relevant, the hypothesis

The Introduction should explicitly state the study’s objective and any hypothesis that was prespecified or is otherwise appropriate to the design. The objective describes what the researchers investigated; a hypothesis states an expected effect, difference or relationship. These are related, but not identical.

Objective: To estimate the effect of a 12-week supervised exercise programme, compared with usual care, on clinic systolic blood pressure in adults with hypertension.

Hypothesis: We hypothesised that participants assigned to supervised exercise would have lower mean systolic blood pressure at 12 weeks than participants assigned to usual care.

If the hypothesis was defined before the results were known, preserve its original meaning. Do not rewrite an unsuccessful prediction after inspecting the data. Not every descriptive, exploratory, qualitative or methodological study calls for a directional hypothesis; where none exists, clearly state the aims or research questions instead.

Readers should reach the end of the Introduction understanding the main question and what would count as evidence relevant to answering it.

5. How to write the Methods

The Methods explain how the work was conducted and how the findings were obtained. Their purpose is to allow qualified readers to assess validity, identify important sources of uncertainty and understand what reproducing or independently examining the work would involve. Exact reproduction may also require protocols, data, code or supporting documentation beyond a journal article, but essential methodological information should not be hidden behind vague citations.

Describe the design, setting and participants

Identify the study design accurately, then describe recruitment or sampling, setting, relevant dates, eligibility criteria and characteristics needed to understand the target population. Randomised studies should describe allocation, concealment and blinding where applicable; observational studies require suitable detail on selection, exposure measurement and important potential confounders. The necessary content differs by design.

Explain interventions and procedures

For our trial, “supervised exercise three times weekly” would be insufficient on its own. Readers may need details about exercise mode, session duration, intensity, progression, supervision, adherence and the control condition. Laboratory work may instead require information about specimens, reagents, instrumentation, experimental controls and quality checks.

Include methodological details that influence interpretation or reproducibility, even when shortening the Methods would make the prose appear more elegant. Supplementary material can house extensive procedural detail, but it should support rather than replace an intelligible main Methods section.

Define outcomes and measurement procedures

Identify primary and secondary outcomes, measurement methods, timepoints and appropriate units. For clinic blood pressure, this could include the device, participant positioning, rest period and handling of repeated readings. Measurements should correspond to the scientific objective and the analyses reported later.

Report ethics and governance accurately

Include relevant ethics approval, consent, registration and other required governance details as appropriate to the study. Never invent a protocol number, approval identifier or consent statement to complete the paper. Check unclear information with the researchers or responsible institution.

Describe the statistical analysis

Explain how the analyses address the objectives. This may require the analysis population, experimental or observational units, model specification, adjustment variables, repeated-measures structure, handling of missing data, sensitivity analyses and software information. “Appropriate statistical tests were used” is not an adequate substitute for a description of what was actually done.

In the example trial, the primary comparison might use a prespecified linear regression with follow-up systolic blood pressure as the outcome, randomised group as the treatment indicator and baseline blood pressure as an adjustment variable. The report should be consistent with the analysis plan, including any material changes and their timing. See How to Write a Research Protocol and How to Write a Statistical Analysis Plan.

6. How to write the Results

The Results should show what the study found and how those findings address the objectives. A useful sequence is study flow and sample characteristics, the primary outcome, then relevant secondary and exploratory analyses. Other study designs may require a different order, but readers should not need to reconstruct the principal findings from scattered numbers.

Explain who or what contributed data

Report how many participants, samples or experimental units entered the study and contributed to each relevant analysis. Explain important changes in numbers, the reasons for losses where known and the appropriate denominators. In laboratory experiments, differentiate independent biological replication from repeated technical readings.

Present the primary result clearly

The principal result should address the objective and any hypothesis stated in the Introduction. Suppose our fictional exercise trial estimated an adjusted between-group difference of −4 mmHg in systolic blood pressure at 12 weeks, with a 95% confidence interval from −7 to −1 mmHg. An illustrative Results statement is:

At 12 weeks, adjusted mean systolic blood pressure was lower in the exercise group than in the usual-care group (between-group difference −4 mmHg, 95% confidence interval −7 to −1 mmHg).

All values in this worked example are invented for teaching purposes and are not findings from a real trial. A complete report would also include appropriate numbers analysed and other required statistical details. The estimated contrast and uncertainty generally tell readers more than a statement that the intervention “improved significantly”.

Report secondary and inconclusive findings fairly

Do not promote a favourable secondary analysis to the status of primary evidence because its P value is small. Label exploratory analyses, include important unexpected or non-significant findings, and distinguish absence of strong evidence from evidence of no effect. Interpretation of mechanisms and clinical implications belongs principally in the Discussion.

7. How to prepare tables and figures

Tables and figures are part of the scientific report, not decoration. A table helps readers examine precise values across groups or outcomes; a figure may better reveal distributions, individual variation or change over time. Choose the format that makes the relevant information easiest to evaluate, rather than presenting the same complete dataset in several different forms.

Write informative tables

Use specific titles, clear columns, units and footnotes. Explain abbreviations and what each statistic represents; report relevant denominators, estimates and uncertainty where appropriate. Excessive decimal places may convey false precision, and symbols marking significance are not a substitute for interpretable comparisons.

Make figures understandable on their own

Identify the plotted quantity, groups, axes, units and any uncertainty measure. For repeated assessments, distinguish observed descriptive means from model-adjusted estimates. A legend should permit a reasonable interpretation without forcing readers to search repeatedly through the Results, while avoiding unnecessary duplication.

Check every representation of a result

Trace the principal outcomes from the Methods through Results text, tables, figures, abstract and supplementary material. Confirm that names, group labels, timepoints, units, sample counts, estimates and uncertainty are consistent. An apparent discrepancy could reflect rounding, a transcription mistake or a valid difference between adjusted and unadjusted results. Establish which before changing the manuscript or implying that the data are unreliable.

8. How to write the Discussion

The Discussion should explain what the findings mean in relation to the original objectives and hypotheses, then consider them in the context of other evidence, possible explanations and limitations. It should not simply repeat the Results in longer sentences or follow interesting secondary findings so far that the original scientific question disappears.

Return explicitly to the objectives and hypotheses

Begin with the primary question from the Introduction. For our hypothetical trial, the objective concerned systolic blood pressure at 12 weeks and the directional hypothesis predicted lower pressure after supervised exercise. The Discussion should say whether the observed result was consistent with that expectation, while describing the estimated magnitude and uncertainty.

For example: The observed difference was in the direction predicted by the prespecified hypothesis, with lower systolic blood pressure at 12 weeks among participants assigned to exercise. The magnitude and precision of the estimate, alongside the study limitations, inform how confidently we can interpret this finding.

If results do not support the prediction, state that plainly. If the estimate is too imprecise to distinguish meaningful alternatives, explain the uncertainty instead of declaring the hypothesis true or false. For studies without hypotheses, return to the prespecified aims or research questions.

Compare results with the existing evidence

Explain whether findings confirm, extend or differ from previous studies. Select literature relevant to the population, intervention or exposure, outcomes and design. Where results disagree, consider differences in measurement, methods, follow-up and sampling rather than selectively citing studies that support a preferred narrative.

Consider plausible alternative explanations

Interpretations should fit the design. Observational findings may be affected by confounding, selection or reverse causation. Randomised trials may still have problems involving adherence, missing data or outcome measurement. Proposed mechanisms are hypotheses unless the study genuinely evaluated them. Distinguish what the results demonstrate from what they merely make plausible.

Assess strengths, limitations and generalisability

Describe strengths that materially increase confidence and limitations that materially constrain inference. GeneralisabilityHow well findings may apply to people, settings or circumstances beyond those directly studied. is also relevant: a short follow-up, for instance, may limit conclusions about long-term blood-pressure control even if the 12-week result is estimated with reasonable precision. Explain how limitations could affect the interpretation rather than listing every imaginable imperfection.

Distinguish statistical significance from scientific importance

A small P value does not establish that an effect is clinically meaningful, and a non-significant result is not evidence of equivalence by itself. Consider the size and precision of the estimate together with the clinical or scientific context. These principles are explored in the forthcoming companion guide on reporting statistical results; until then, see How to Choose the Right Statistical Test for analysis-selection and interpretation considerations.

Close by returning to the original aims

End the Discussion by connecting the findings back to the study objectives and, where relevant, hypotheses. Explain the contribution, remaining uncertainties and practical implications without overstating clinical application. Future research suggestions should arise from identifiable unanswered questions, not a generic final sentence.

9. How to write the conclusion

The conclusion should offer a concise, proportionate answer to the original research question. It should not introduce new evidence, imply that a hypothesis has been conclusively proved or broaden conclusions beyond the population, design and outcomes studied.

For our fictional trial, an example is: In this hypothetical 12-week randomised trial, supervised exercise resulted in lower clinic systolic blood pressure at follow-up than usual care. The findings address the prespecified primary objective but do not establish whether the effect persists beyond the study period.

For observational work, causal language may require greater caution unless supported by an appropriate causal framework and analysis. Where a formal hypothesis was tested, its relationship to the findings should already be explained in the Discussion. Don’t end every paper by declaring that more research is needed; when it is, say what uncertainty remains.

10. How to write the title, abstract and keywords

The title and abstract are often encountered independently of the full article. They should faithfully represent the question, methods and principal findings, not a more promotional version of the research.

Choose a specific title

Where suitable, identify the population, intervention or exposure, main outcome and design. For our example, Supervised Exercise and Systolic Blood Pressure in Adults With Hypertension: A Randomised Controlled Trial is more informative than an exaggerated promise of remarkable cardiovascular benefits. Follow the journal’s title conventions.

Write the abstract from the final manuscript

The abstract should accurately summarise the aims, design, participants, methods, main numerical results and conclusion. If the study had a prespecified hypothesis, its stated objective and key findings must remain consistent with that hypothesis. Use the target journal’s required headings and check every number against the final Results.

Avoid foregrounding appealing secondary results while omitting the primary estimate or relevant uncertainty. Include registration and other information where required for the study type.

Choose informative keywords

Select terms that genuinely identify the study’s condition, intervention or exposure, outcome and design. Established biomedical vocabulary, including Medical Subject Headings (MeSH), can help. Avoid unrelated fashionable phrases and unnecessary invented abbreviations.

11. Prepare references and supporting declarations

Submission-ready manuscripts include accurate references and the statements required to account for who conducted the research, how it was supported and what information readers can access. Requirements vary by journal and study design.

Verify the references

Check that influential cited papers actually support the attached claims, including their populations, outcomes and inferential limitations. A study reporting an association does not automatically support a causal statement. Consider corrections or retractions where relevant; choose studies, reviews, methodological guidance and other authoritative sources according to the question, not to reach a quota of references.

Document authorship and contributions

The ICMJE authorship criteria concern substantive contribution, drafting or critical revision, final approval and accountability. The Contributor Role Taxonomy (CRediT) offers standard labels for roles including investigation, formal analysis and writing, but assigning a CRediT role does not by itself establish authorship. Confirm contributors’ responsibilities and agreement before submission.

Complete the required declarations

Provide accurate ethics approval, consent, registration, funding, competing-interest and data-availability information where appropriate. Verify what actually occurred and whether access to data is constrained by privacy, consent or governance requirements. Never invent details to fill a template.

Handle AI-assisted work responsibly

If artificial intelligence was used during research or manuscript preparation, follow the journal’s current disclosure rules and applicable ICMJE guidance on AI use by authors. Authors remain responsible for originality, numerical accuracy, citations, interpretation and publication integrity. Independently check AI-generated factual statements and references, and protect confidential manuscripts or participant information from inappropriate external disclosure.

AI systems cannot satisfy human authorship responsibilities merely because they contributed text. The distinction between assistance and accountable authorship should remain explicit. For broader context, see the AI in research section of the Blazon Research Workflow Library. A dedicated guide to the responsible use of AI in research is planned.

12. Edit for scientific clarity and precision

Good scientific editing makes the research easier to understand without changing its meaning or giving uncertainty a more confident appearance. The goal is clear, precise writing, not the most elaborate academic-sounding prose.

Build coherent paragraphs

Give each paragraph a recognisable purpose and link it logically with what comes before and after. Do not combine unrelated methodological details and interpretations in one paragraph, but avoid breaking a connected explanation into several isolated one-sentence paragraphs merely to create visual space.

Prefer direct sentences

We measured blood pressure at baseline and 12 weeks is generally clearer than Measurements of blood pressure were undertaken at the baseline and 12-week assessment points. Passive voice is still useful when the procedure is more important than the person performing it. Choose the wording that communicates the science most naturally.

Preserve precise scientific distinctions

Association is not the same as causation; risk differs from odds; a between-group difference is not a within-group change. A non-significant result does not necessarily show no effect. Keep terms consistent, define unfamiliar language where needed and avoid abbreviations that readers will have to decode repeatedly.

Preserve numbers and uncertainty

Never silently change sample sizes, effect estimates, P values, confidence intervals, units or the interpretation of the analysis. If a value looks wrong, verify it against the original output or source records rather than guessing. Editing should not convert a cautious conclusion into a stronger claim simply because the latter sounds more persuasive.

Write originally and attribute sources

Explain existing research using the authors’ own synthesis and original language, with citations where the evidence informs a claim. A citation does not excuse copying or very close paraphrase of another paper’s wording. Shared technical terminology and official guideline names are often unavoidable; the surrounding explanations should nevertheless be independently written.

Similarity-checking software may help identify overlap but cannot by itself establish whether an article is properly attributed or free from plagiarism. Evaluate the actual wording, source dependence and rights involved, rather than relying on a numerical similarity score.

13. The final manuscript consistency check

Read the completed manuscript as someone who did not conduct the study. Can you trace the primary question, methods, findings and interpretation without guessing? The last review should look beyond formatting to the scientific claims and how transparently they are supported.

Revisit the original aims and hypotheses

Return to the end of the Introduction. Are the objectives clear, and is a prespecified hypothesis stated where relevant? Do the Methods and Results genuinely address them? Does the Discussion explicitly consider whether the primary findings support, contradict or leave the hypotheses unresolved? A paper should not quietly redefine its main purpose around a more attractive secondary finding.

Reconcile numerical information

Compare important participant counts, denominators, outcome names, units, groups, timepoints, estimates, uncertainty and P values across the abstract, main text, displays and supplementary information. Investigate differences before assuming which value is correct. If results differ because of distinct models or analysis populations, explain that distinction.

Check evidence, interpretation and submission requirements

Confirm that influential claims are supported by suitable sources, relevant contradictory evidence has not been omitted, and the conclusions respect the design and limitations. Finally, check the target journal’s current instructions, reporting checklists, declarations, author details and supplementary documents.

Final manuscript quality checklist

  • The primary research question and objectives are explicit, with hypotheses where applicable.
  • The Introduction’s knowledge gap and contribution are accurately represented.
  • The design and Methods are described transparently and are appropriate to the objective.
  • Important deviations from the protocol, registration or analysis plan are disclosed.
  • The Results address the main objectives and report appropriate estimates and uncertainty.
  • Prespecified, secondary and exploratory analyses are distinguishable.
  • Participant or sample numbers and denominators are clear.
  • The abstract, Results, tables, figures and supplements agree on key values.
  • The Discussion returns explicitly to objectives and hypotheses, where relevant.
  • Unexpected, inconclusive and contradictory findings are interpreted fairly.
  • Limitations and causal claimsStatements that one factor causes a change in another, rather than merely being associated with it. are proportionate to the design.
  • The conclusion does not go beyond the evidence.
  • Important citations are verified; the prose is original and properly attributed.
  • Authorship, ethics, registration, funding, competing interests and data statements are accurate.
  • The manuscript follows journal instructions and all authors have approved the final version as required.

A checklist can expose problems worth correcting; it cannot certify that a study is free from bias or that its conclusions are correct. Prioritise issues that affect transparency, scientific interpretation or publication readiness, rather than chasing minor formatting preferences at the expense of substantive quality.

Frequently asked questions

Should a scientific manuscript always include a hypothesis?

No. A manuscript should clearly state its objectives and, when relevant, its prespecified hypothesis. Descriptive, exploratory and qualitative studies may be guided by research questions without directional hypotheses; authors should not invent one simply to satisfy a manuscript template.

Should the Discussion explicitly return to the hypothesis?

Yes, where a hypothesis was specified. Explain how the findings relate to the hypothesis, including their magnitude, uncertainty and limitations. Every Discussion should return to the aims or research questions, even when no formal hypothesis was used.

What is the best order for writing a scientific manuscript?

There is no single best order. Developing the main tables and figures before the Results can help clarify what needs to be communicated, while a protocol may make the Methods a natural starting point. The important goal is consistency between the question, methods, findings and interpretation.

Can AI help with scientific manuscript writing?

AI tools may assist with appropriate language or organisational tasks, subject to journal rules and confidentiality obligations. Human authors must verify scientific claims, numbers, citations and originality, disclose use as required and remain accountable for the submitted work.

How do I know whether a manuscript is ready to submit?

Check the alignment of objectives, methods, results and conclusions; reconcile important numbers; verify consequential citations and declarations; and follow the target journal’s current requirements. An independent scientific or editorial review may reveal ambiguities the authors have become accustomed to overlooking.

References and further guidance

  1. International Committee of Medical Journal Editors. Recommendations for the Conduct, Reporting, Editing, and Publication of Scholarly Work in Medical Journals. Updated January 2026. ICMJE Recommendations.
  2. EQUATOR Network. Reporting guideline library. Browse reporting guidelines.
  3. Hopewell S, Chan AW, Collins GS, et al. CONSORT 2025 statement: updated guideline for reporting randomised trials. BMJ. 2025;389:e081123. https://doi.org/10.1136/bmj-2024-081123.
  4. von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. PLoS Medicine. 2007;4(10):e296. https://doi.org/10.1371/journal.pmed.0040296.
  5. National Information Standards Organization. CRediT Contributor Role Taxonomy. CRediT.
  6. European Society of Clinical Pharmacy. Writing a manuscript for publication in a peer-reviewed scientific journal: Guidance from the European Society of Clinical Pharmacy. International Journal of Clinical Pharmacy. 2024;46:548–554. https://doi.org/10.1007/s11096-023-01695-6. Note: a published correction is available; consult it alongside the paper. Correction.
  7. Cooper ID. How to write an original research paper (and get it published). Journal of the Medical Library Association. 2015;103(2):67–68. https://doi.org/10.3163/1536-5050.103.2.001.
  8. Advice to authors for avoiding flaws in preparation of original research manuscripts. Perspectives in Clinical Research. 2021;12:229–233. https://doi.org/10.4103/picr.PICR_139_21.
Home > Research Resources for Medical Researchers > How to Write a Scientific Manuscript