What Are the Ethical Considerations in Research Design?
Research ethics is usually presented as a set of procedures that surround a study: a committee application, a consent form, a confidentiality statement. The governing instruments take a different view. They treat the design itself as the primary ethical object, on the ground that a study incapable of answering its question cannot justify any burden at all. That principle has consequences for sample size, for the comparator, and for every procedure a protocol specifies.
- What the three foundational principles govern, and what each one implies
- Why an underpowered study is treated as an ethical failure
- What proportion of participants understand randomisation and placebo
- “Ethics is handled by the committee application and the consent form.” The guidelines locate the justification for research in its scientific and social value, which is a property of the design rather than of the paperwork.
- “A signed consent form establishes informed consent.” Pooled across 103 studies, roughly half of participants understood randomisation, and that proportion had not increased in thirty years.
What this article covers
This article concerns the ethical requirements that bear on how a study is designed, and the instruments that state them. It is not a guide to completing an ethics committee application, which varies by institution and jurisdiction. The focus is instead on the obligations that exist before any form is filled in, because those obligations determine whether the application can succeed. A protocol that is methodologically weak cannot be rescued by careful administration.
Five instruments govern most of the territory, and they overlap deliberately rather than competing. The Declaration of Helsinki supplies the principles that medical research is expected to follow. The Belmont Report supplies the ethical framework underlying much regulation. The CIOMS guidelines elaborate both for health-related research generally. Good clinical practice translates them into operational requirements for trials, and the reporting standards govern what happens at publication.
Specifying the question that the design is meant to answer is covered in why use the PICO framework. Registration before the work begins is covered, for evidence synthesis, in PROSPERO protocol registration. Consent and identifiability for a single patient are covered in writing an effective medical case report. This article takes the design obligations, which those three presuppose.

The instruments and the principles they state
Origins and current status
The Declaration of Helsinki was adopted by the World Medical Association in June 1964 and has been revised repeatedly since. Its current text was adopted by the 75th General Assembly in October 2024.1 That revision is the version the reporting standards now point to, since authors are directed to ensure their research follows the Declaration “as revised in 2024”.4
The Belmont Report was issued in April 1979 by the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research.3 It is shorter than the other instruments and more abstract, and it underlies a great deal of subsequent regulation. The CIOMS guidelines, now in their fourth edition, were prepared in 2016 in collaboration with the World Health Organization.2 Operational requirements for trials sit in the good clinical practice guideline, whose current revision was adopted in January 2025.5
The three principles and their applications
The Belmont framework rests on three principles, each of which generates a specific requirement. Respect for persons holds that individuals should be treated as autonomous agents, and that those with diminished autonomy are entitled to protection.3 Its application is informed consent, which the report analyses into three elements: information, comprehension and voluntariness.3 The second element is the one most often neglected in practice, and it is examined later in this article.
Beneficence requires more than refraining from harm. The report frames it as an obligation expressed in two rules: do not harm, and maximise benefits while minimising possible harms.3 Its application is the systematic assessment of risks and benefits, which the report asks to be a “systematic, nonarbitrary analysis of risks and benefits” wherever that is possible.3 Benefits and risks must be shown to stand “in a favorable ratio”.3
Justice concerns the fair distribution of research benefits and burdens. Equals should be treated equally, and any departure from equal treatment requires justification.3 Its application is the selection of participants, considered at both the individual and the social level.3 The CIOMS guidelines state the same requirement operationally, holding that groups and individuals “must be selected for scientific reasons” and that groups unlikely to benefit “should not bear a disproportionate share of the risks and burdens”.2
Practice aims at the well-being of a particular patient, with a reasonable expectation of success. Research aims to test a hypothesis and contribute to generalisable knowledge.3 A clinician departing from standard care has not thereby conducted research, though major innovations should be studied formally.3 The report resolves mixed cases in one direction: “if there is any element of research in an activity, that activity should undergo review”.3
Exclusion as a harm
One provision runs against the protective instinct and deserves separate statement. Groups that are more vulnerable receive specific protections, but the Declaration also treats exclusion from research as a harm in its own right.1 Research with such groups is justified where “it is responsive to their health needs and priorities”.1 The CIOMS commentary makes the corresponding warning about exclusion, cautioning against excluding groups in ways that worsen health disparities.2 A protocol that excludes pregnant participants, older people or those with comorbidity by default has made an ethical choice, not avoided one.
Figures: from ethical principles to protocol decisions
Figure 1 connects each ethical principle to the protocol decision it governs. Figure 2 summarises the evidence on what participants understand after giving consent. Authors can use them to show reviewers and ethics committees how ethical requirements were built into the study design.
Each principle generates one application. Each application lands on a decision taken while the protocol is being written.
What to look for. The right-hand column contains no administrative tasks. Sample size, choice of comparator, eligibility criteria and the procedures a protocol specifies are all methodological decisions, and each is the point at which a principle becomes operative. An ethics application documents those decisions; it does not make them, and it cannot repair them.
Schematic. The three-principle structure follows the Belmont Report, and the right-hand column is drawn from the design requirements in the other instruments cited here.
Pooled proportions from 103 studies across 135 cohorts. The two components that define a randomised trial are the two least understood.
What to look for. The ordering is the finding. Participants understood that they could leave, and broadly what the study was for, at around three-quarters. Understanding of randomisation and of placebo sat near a coin toss. Those two components are what distinguishes a trial from treatment. Roughly half of consenting participants did not grasp the feature that made their participation research rather than care.
Pooled proportions from a random-effects meta-analysis of literature to October 2013. Bar widths are to scale.
Design quality as an ethical requirement
The strongest claim in the governing instruments is also the one least often repeated in summaries of them. Methodological quality is not a scientific virtue that sits alongside ethical conduct. It is a precondition of ethical conduct, and a study that cannot answer its question has no ethical justification for the burdens it imposes.
Social value and scientific validity
The CIOMS guidelines locate the justification for research in exactly this place. The first guideline states that “The ethical justification for undertaking health-related research involving humans is its scientific and social value”.2 Scientific value means the study is capable of producing “reliable, valid information”.2 The relationship between the two is asymmetric, since “scientific value is a necessary but not a sufficient condition for the social value of health research”.2
The consequence for a weak design is direct. A study “cannot be socially valuable without appropriate and rigorous research methods”, so methodological failure removes the justification rather than merely reducing the quality of the output.2 The commentary notes that “methodological shortcomings can derail promising avenues of research”, and identifies seeding trials, designed to promote a product rather than produce knowledge, as violating the requirement outright.2
The Declaration states the same obligation as a property of the protocol. Research “must have a scientifically sound and rigorous design and execution”, and the protocol must clearly describe and justify the design and methods.1 Risk is permissible only on terms: research “may only be conducted if the importance of the objective outweighs the risks and burdens”.1 An objective that the design cannot reach weighs nothing on that scale.
Statistical power
The argument has a specific and uncomfortable application to sample size. A trial too small to detect a clinically meaningful effect exposes participants to burden while being unlikely to produce the knowledge that justifies it. One analysis concluded that such trials are unethical except in narrowly defined circumstances, holding that “underpowered trials are ethical in only 2 situations”.6
The two exceptions are specific. The first covers small trials of interventions for rare diseases, where investigators commit in advance to pooling their results in a prospective meta-analysis.6 The second covers early-phase trials that are adequately powered for purposes other than randomised treatment comparison, such as establishing tolerability.6 In both cases the authors hold that participants must be told that “their participation may only indirectly contribute to future health care benefits”.6
The authors also reject the two defences usually offered for small trials. Neither defence is held to supply the missing justification. One is the prospect of later inclusion in a meta-analysis that nobody has arranged; the other is the production of a wide confidence interval.6 A power calculation is therefore an ethical document as much as a statistical one, and its absence from a protocol is a substantive omission rather than a formatting lapse.
Proportionality and unnecessary burden
The third design obligation concerns everything a protocol asks of participants beyond the minimum. Good clinical practice requires that processes be proportionate to the risks to participants and to the importance of the data collected.5 Its wording on complexity is unusually direct: trial processes “should be operationally feasible and avoid unnecessary complexity, procedures and data collection”.5
That principle has teeth at the design stage. A blood draw added in case it proves interesting, a questionnaire included because it was in the last protocol, an extra visit retained without a stated purpose: each imposes burden that nothing justifies. The same guideline asks that factors critical to trial quality be identified in advance, with quality by design focusing on those factors rather than on uniform intensity across everything.5 The Belmont requirement points the same way. It directs that risks be reduced to what the research objective actually requires, and that investigators check whether human participants are needed at all.3
Consent as a process rather than a document
Informed consent is the application of respect for persons, and the Belmont analysis separates it into information, comprehension and voluntariness.3 A signature evidences the first and the third. It evidences the second only by assumption, and the empirical literature does not support the assumption.
Requirements in the instruments
The Declaration requires that each potential participant “must be adequately informed in plain language of the aims, methods” and the remaining elements of the study.1 Consent must be voluntary, documented, and revocable at any time without reprisal.1 Where a potential participant stands in a dependent relationship to the investigator, consent must be sought by an appropriately qualified individual who is independent of that relationship.1 Additional safeguards apply to those unable to give consent for themselves.1
Good clinical practice states the requirement as its second principle, directing that “Freely given informed consent should be obtained and documented from every participant” before participation begins.5 It adds a requirement about the function of the process rather than its documentation. The process should enable participants to weigh the benefits, risks and burden, and then to decide whether to take part.5 A process that produces a signature without producing that understanding has not met the stated purpose.
Comprehension measured after consent
A meta-analysis pooled comprehension data from 103 studies across 135 cohorts of trial participants, covering literature to October 2013.7 Its results are set out in the second figure above, and the ordering is the point. Three-quarters understood that they were free to withdraw at any time, at 75.8 per cent.7 Comparable proportions understood the nature of the study and the potential benefits, at 74.7 and 74.0 per cent.7
The two components that define a randomised trial fared worst. Understanding of placebo reached 53.3 per cent and understanding of randomisation 52.1 per cent.7 Potential risks and side effects were understood by 67.0 per cent, while only 54.9 per cent could name at least one risk.7 Therapeutic misconception, the belief that a trial is a form of individualised treatment, was absent in 62.4 per cent, which means it was present in roughly a third.7
The review covered three decades of research on this question. The proportion of participants who understood informed consent “had not increased over 30 years”.7 Thirty years of longer forms, clearer templates and stricter documentation produced no measurable gain in comprehension. The implication is that the form is not the instrument through which comprehension is achieved.
Consequences for the design
Three consequences follow for a protocol rather than for a form. The first is that comprehension should be assessed rather than assumed, which means the protocol specifies how understanding of randomisation and of the comparator will be checked before enrolment. The second is that the components with the lowest comprehension deserve disproportionate attention in the consent discussion, since they are the ones carrying the ethical weight.
The third concerns who conducts the conversation. The meta-analysis identified age, education, critical illness, trial phase and location as factors associated with understanding.7 A protocol recruiting acutely unwell participants, or participants with limited formal education, has a foreseeable comprehension problem and should say how it will be addressed. The CIOMS guidelines point in a related direction for research in low-resource settings, calling for “full participation of communities in all steps of the project”.2
Review, registration and reporting
Three obligations bracket the study itself. Each is stated in the Declaration, each has a counterpart in the operational guidance, and the third is the one most often left undischarged.
Independent review
Before research begins, the protocol “must be submitted for consideration, comment, guidance, and approval” to a research ethics committee.1 The committee must be independent and transparent, and it must retain the power to monitor ongoing studies and to amend or halt them.1 Good clinical practice states the same requirement as a principle, holding that trials “should be subject to an independent review by an IRB/IEC”, whose purpose is to safeguard the rights, safety and well-being of participants.5 Approval must precede enrolment.5
Review is necessary but not sufficient, and the reporting standards say so explicitly. Approval by a review body does not prevent editors from forming their own judgement about whether research was conducted appropriately.4 Where there is doubt about compliance with the Declaration, authors must explain their rationale and demonstrate that the review body explicitly approved the doubtful aspects.4 A committee's signature is therefore evidence rather than immunity.
Prospective registration
The Declaration requires that research “must be registered in a publicly accessible database” before the first participant is recruited.1 The timing is the substance of the requirement. Registration after recruitment has begun records what was done; registration before it records what was intended, which is what makes selective reporting detectable afterwards.
Registration therefore functions as a design commitment rather than an administrative step. The outcomes named in the register are the outcomes the study undertook to report, and any divergence between those and the published ones becomes visible to a reader. That visibility is the mechanism by which the obligation in the following section can be enforced at all.
The obligation to publish
The Declaration places a reporting duty on researchers that extends well beyond favourable findings. “Negative and inconclusive as well as positive results must be published” or otherwise made publicly available.1 Funding, affiliations and conflicts of interest must be declared, and the Declaration states that reports not conforming to it should not be accepted for publication.1
The evidence on compliance is poor. Methodological guidance on missing results describes “convincing evidence for several types of non-reporting bias” and characterises the overall evidence as overwhelming.8 Registry data give the clearest measure. As of May 2024, one major registry held almost half a million studies. Only about 13 per cent of them had summary results posted.8
Participants accepted burden on the understanding that their participation would contribute to knowledge. A study that is conducted and never reported withdraws that contribution after the fact. The burden was incurred, and the justification for it was not delivered. The same reasoning that makes an underpowered study unethical applies with equal force to a completed study left unpublished.
Identifiability at publication
A final obligation attaches at the point of writing. Identifying details such as names, initials or hospital numbers should not be published unless the information is essential and consent has been obtained.4 Where an identifiable patient appears, that patient must be shown the manuscript, and consent should be written and archived according to local law.4 Authors should also tell patients whether identifiable material may appear online as well as in print.4
Two further provisions are routinely misunderstood. Covering the eyes in a photograph is described as “inadequate protection of anonymity”.4 Where identifying details are altered to protect anonymity, authors must give an assurance, and editors should note, that the scientific meaning has not been changed.4 Where anonymity is in any doubt, consent should be obtained.4
Obligations that outlast the study
Two requirements in the current Declaration concern what happens to participants after their involvement ends. Both are design decisions in the sense that matters, because both must be settled before recruitment begins and both constrain what the protocol can propose.
The control arm
The choice of comparator is an ethical decision before it is a methodological one. The Declaration's position is that new interventions must be tested against the best proven intervention, with defined exceptions.1 Placebo or no intervention is acceptable where no proven intervention exists. It is also acceptable where compelling and scientifically sound methodological reasons apply, and where participants face no additional risk of serious or irreversible harm.1
The Declaration attaches an unusual warning to that second exception. “Extreme care must be taken to avoid abuse of this option”.1 The wording acknowledges that methodological convenience and ethical justification are easily confused here. A placebo arm is cleaner to analyse than an active comparator, and that cleanliness is not among the reasons the Declaration accepts.
Post-trial provisions
The second obligation concerns continued access. Before a trial begins, sponsors and researchers must make arrangements for participants who still need an intervention found to be beneficial and reasonably safe to continue receiving it.1 Exceptions require approval from a research ethics committee, and the provisions must be disclosed to participants during the consent process.1
The timing is what makes this a design requirement. An arrangement made before recruitment can be described in the consent discussion, which is where the Declaration places it. An arrangement considered after results are available is a different thing, negotiated under pressure and in the absence of any commitment to the people who generated the evidence.
Risk monitoring during the study
A third obligation runs throughout rather than afterwards. Risks must be assessed in advance, minimised, and then monitored continuously as the study proceeds.1 Researchers must consider stopping or modifying a study where the risks come to outweigh the benefits.1 The ethics committee retains a parallel power to monitor, amend or halt.1
Two things follow for the protocol. The first is that stopping rules belong in it, stated in advance rather than improvised. The second is that monitoring arrangements must be specified with enough precision that a committee can see how an emerging harm would actually be detected. The CIOMS guidelines add a point about whose judgement counts in that assessment, holding that risks and benefits “must be evaluated in consultation with the communities” concerned, because community values shape what counts as a benefit or an acceptable risk.2
Common mistakes
| Fault | Stage | What it costs | The remedy |
|---|---|---|---|
| The study is too small to answer its question | Design | Burden imposed with no prospect of the knowledge that justifies it | Power the study, or meet one of the two stated exceptions explicitly |
| Procedures retained without a stated purpose | Design | Unjustified burden on participants and unused data | Trace every procedure to a critical-to-quality factor, or remove it |
| Placebo chosen because it analyses more cleanly | Design | A justification the Declaration does not recognise | Test against the best proven intervention, or meet a stated exception |
| Comprehension assumed from a signature | Consent | Around half of participants misunderstand randomisation | Specify how understanding will be checked before enrolment |
| Consent sought by someone the participant depends on | Consent | Voluntariness cannot be established | Have an appropriately qualified independent person seek consent |
| Vulnerable groups excluded by default | Eligibility | Exclusion is itself a harm, and disparities widen | Justify each exclusion, or widen eligibility with added safeguards |
| Registration after recruitment has begun | Registration | Selective reporting becomes undetectable | Register before the first participant, with the outcomes named |
| No post-trial access arrangement before recruitment | Protocol | Nothing can be promised during the consent discussion | Settle continued access in advance, and disclose it in consent |
| Stopping rules left unstated | Protocol | An emerging harm has no specified route to detection | State monitoring arrangements and stopping rules in the protocol |
| Negative results never reported | Reporting | The burden was incurred and the justification never delivered | Publish or otherwise make public, as the Declaration requires |
Nine of these ten occur before a participant is enrolled, and the tenth occurs after the last one leaves. None of them is reachable by a consent form, and none is addressed by the committee application considered as a document. They are design decisions with ethical content, which is the argument this article has been making throughout.
Research ethics is a property of the design before it is a property of the paperwork. The governing guidelines locate the justification for research in its scientific and social value, and hold that a study cannot be socially valuable without rigorous methods. That principle makes sample size, choice of comparator, eligibility criteria and every specified procedure into ethical decisions. Consent is the application of respect for persons, and the evidence shows that signatures substantially overstate comprehension, with roughly half of participants understanding randomisation across three decades of study. Registration before recruitment and publication of negative results are obligations in the Declaration, and the second is widely unmet. An ethics committee documents these decisions. It does not make them, and it cannot repair a design that has already failed.
How to apply this to your work
- Establish that the design can answer the question before considering any other ethical requirement.
- Power the study properly, or state explicitly which of the two recognised exceptions applies.
- Justify every procedure against a critical-to-quality factor, and remove those that fail the test.
- Choose the comparator on ethical grounds, treating analytical convenience as insufficient.
- Specify how comprehension will be checked, concentrating on randomisation and the comparator.
- Justify each exclusion, recognising that exclusion is itself a harm.
- Register before the first participant, with the outcomes named in advance.
- Settle post-trial access and stopping rules in the protocol, and disclose the former in consent.
Common questions
Is a weak study design an ethical problem or only a scientific one?
It is treated as an ethical problem. The guidelines state that “The ethical justification for undertaking health-related research involving humans is its scientific and social value”, and that a study “cannot be socially valuable without appropriate and rigorous research methods”.2 The Declaration requires a “scientifically sound and rigorous design and execution”, and permits research only where the importance of the objective outweighs the risks.1
Are underpowered trials ever ethically acceptable?
In two circumstances only, according to the analysis most often cited on the question. Its authors conclude that “underpowered trials are ethical in only 2 situations”: small trials in rare diseases where pooling in a prospective meta-analysis is arranged in advance, and early-phase trials adequately powered for purposes other than treatment comparison.6 Participants must be told that their participation may contribute only indirectly.6
How well do participants understand what they consent to?
Less well than signatures suggest. Pooling 103 studies across 135 cohorts, 75.8 per cent understood their freedom to withdraw and 74.7 per cent the nature of the study.7 Understanding of placebo reached 53.3 per cent and of randomisation 52.1 per cent.7 Only 54.9 per cent could name at least one risk, and the overall proportion “had not increased over 30 years”.7
When is a placebo control acceptable?
Where no proven intervention exists, or where compelling and scientifically sound methodological reasons apply and participants face no additional risk of serious or irreversible harm.1 New interventions must otherwise be tested against the best proven intervention.1 The Declaration adds a warning to the second exception: “Extreme care must be taken to avoid abuse of this option”.1
What must be arranged before a trial begins regarding post-trial access?
Sponsors and researchers must make arrangements for participants who still need an intervention found beneficial and reasonably safe to continue receiving it.1 The arrangement must be in place before the trial starts. Exceptions require approval from a research ethics committee, and the provisions must be disclosed to participants during the consent process.1
When must a study be registered?
Before recruitment starts. The Declaration requires that research “must be registered in a publicly accessible database” before the first participant is recruited.1 Registering afterwards records what was done rather than what was intended, which is what makes any later divergence between planned and reported outcomes detectable.
Is there an obligation to publish negative findings?
Yes, and it is stated directly. “Negative and inconclusive as well as positive results must be published” or otherwise made publicly available.1 Compliance is poor: as of May 2024, one major registry held almost half a million studies, and only about 13 per cent had summary results posted.8
Does ethics committee approval settle the question of whether research was conducted appropriately?
No. Approval by a review body does not prevent editors from forming their own judgement about whether the research was conducted appropriately.4 Where there is doubt about compliance with the Declaration, authors must explain their rationale and demonstrate that the review body explicitly approved the doubtful aspects.4
Declared interest: Pubrica provides research and medical writing support commercially. Nothing on this page should be read as suggesting that editorial assistance can supply what a design lacks. Where a study is underpowered, or where the comparator cannot be justified, the appropriate response is to change the design rather than to describe it more persuasively.
References and sources
- World Medical Association. WMA Declaration of Helsinki: ethical principles for medical research involving human participants. Adopted June 1964; current revision adopted by the 75th WMA General Assembly, Helsinki, October 2024. read 21 October 2026, in full.View the Declaration
- Council for International Organizations of Medical Sciences, in collaboration with the World Health Organization. International ethical guidelines for health-related research involving humans. Fourth edition. Geneva: CIOMS; 2016. read 21 October 2026; the guidelines cited here are 1 to 4, and guideline 7 was not reachable in the portion returned.View the full text
- National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. The Belmont Report: ethical principles and guidelines for the protection of human subjects of research. Washington: Department of Health, Education, and Welfare; 18 April 1979. read 21 October 2026, in full.View the full text
- International Committee of Medical Journal Editors. Recommendations for the Conduct, Reporting, Editing, and Publication of Scholarly Work in Medical Journals: protection of research participants. read 21 October 2026, in full.View the section
- International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use. ICH harmonised guideline: guideline for good clinical practice E6(R3). Step 4 version, adopted 6 January 2025. read 21 October 2026; the text returned covers the principles and the opening sections of Annex 1, and the citations here are confined to that portion.View the guideline
- Halpern SD, Karlawish JHT, Berlin JA. The continuing unethical conduct of underpowered clinical trials. JAMA. 2002;288(3):358–362. doi:10.1001/jama.288.3.358 read 21 October 2026 at the journal; the abstract and article record were available rather than the full text, and the statements cited here are those the abstract makes.View the record
- Tam NT, Huy NT, Thoa LTB, Long NP, Trang NTH, Hirayama K, Karbwang J. Participants’ understanding of informed consent in clinical trials over three decades: systematic review and meta-analysis. Bulletin of the World Health Organization. 2015;93(3):186–198H. doi:10.2471/BLT.14.141390 read 21 October 2026 from the Directory of Open Access Journals record, which carries the pooled proportions cited here.View the record
- Chapter 13: assessing risk of bias due to missing results in a synthesis. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editors. Cochrane Handbook for Systematic Reviews of Interventions, current version. Cochrane. read 21 October 2026; the registry figure cited here is dated May 2024 in the chapter itself.View the chapter
One further source was read and is deliberately not cited. An updated meta-analysis of consent comprehension, published in 2024, was retracted by its journal in June 2025 over concerns about manipulation of the publication process. Its figures are not used here, and the 2015 review above is cited in its place.
How to cite this article
Pick your referencing style and copy it. Your university or target journal may vary the punctuation — check its guide if it has one.
This article is written to be learned from. Cite it if you are referring to what it says; if you are using the evidence behind it, cite the sources listed above instead.
Work through these before the protocol goes to a committee.
DESIGN [ ] The design is capable of answering the stated question [ ] Sample size calculation present, with its assumptions stated [ ] If underpowered, one of the two recognised exceptions applies and is named [ ] Every procedure traced to a factor critical to the trial's quality [ ] Procedures that serve no stated purpose have been removed [ ] Data collection limited to what the objective requires [ ] The question could not be answered without human participants COMPARATOR [ ] The comparator is the best proven intervention, or an exception applies [ ] Any placebo arm justified on grounds the Declaration recognises [ ] NO placebo chosen for analytical convenience alone [ ] Participants face no added risk of serious or irreversible harm PARTICIPANTS [ ] Groups selected for scientific reasons, not accessibility [ ] Each exclusion justified, since exclusion is itself a harm [ ] Burdens not concentrated on groups unlikely to benefit [ ] Specific protections recorded for participants with reduced autonomy [ ] Research responsive to the health needs of any vulnerable group included CONSENT [ ] Information given in plain language [ ] How comprehension will be checked is specified in the protocol [ ] Randomisation and the comparator given extra attention in the discussion [ ] Consent sought by someone independent of any dependent relationship [ ] Right to withdraw without reprisal stated explicitly [ ] Post-trial provisions disclosed during consent [ ] Additional safeguards recorded for those unable to consent REVIEW AND REGISTRATION [ ] Protocol submitted to an independent ethics committee before enrolment [ ] Committee has the power to monitor, amend or halt the study [ ] Study registered in a public database before the first participant [ ] Outcomes named in the registration record DURING AND AFTER [ ] Risks assessed in advance, minimised, and monitored continuously [ ] Stopping rules stated in the protocol [ ] Post-trial access arranged before the trial begins [ ] Plan in place to publish negative and inconclusive results [ ] Funding, affiliations and conflicts of interest declared AT PUBLICATION [ ] No identifying details published without consent [ ] Identifiable patients shown the manuscript before submission [ ] Any altered details accompanied by an assurance that meaning is unchanged [ ] Ethics approval and consent documented in the manuscript
We help you test a protocol against the instruments that govern it. We check that the design can answer its question, and that each burden it imposes is one the design requires.