Writing a Medical Research Proposal

From research question to methods and ethics, the structure supervisors and ethics committees expect from an MD/MS or clinical research proposal.

A research proposal makes the case that your study is worth doing and can be done rigorously and ethically. For MD/MS students and clinical researchers it is the document that secures supervisor sign-off and ethics approval, so it must persuade on both scientific and ethical grounds.

Title and Research Question

A specific, answerable question drives everything. Frame it with PICO where relevant. The title should convey population, intervention/exposure and outcome at a glance.

Background and Rationale

Summarise what is known, drawn from a focused literature review, and identify the gap your study fills. The rationale must answer "why is this worth doing?", clinically and scientifically. This is where you justify the study's existence.

Aim, Objectives and Hypothesis

State one clear aim, break it into measurable objectives, and, for quantitative work, state a testable hypothesis. Ethics committees and supervisors check that the later methods actually address exactly these.

Methodology

Design: RCT, cohort, case-control, cross-sectional, justified by the question.
Population and sampling: inclusion/exclusion criteria and recruitment.
Sample size: a power calculation, not a guess, a frequent point of rejection.
Data collection: variables, instruments, validated where possible.
Analysis plan: the specific statistical tests, pre-specified.

Ethical Considerations

Clinical research demands explicit attention to consent, confidentiality, data protection, risk-benefit balance, and approval from the relevant research ethics committee. Vulnerable populations require additional safeguards. A proposal that treats ethics as an afterthought will not pass.

Timeline and Resources

A realistic Gantt-style timeline and a note of required resources (funding, equipment, access) show the study is feasible, not just desirable.

Common Weaknesses

No power calculation, so sample size is unjustified.
Analysis plan vague or absent.
Ethics section thin.
Objectives and methods not aligned.

Checklist

Is the question focused and answerable?
Is the sample size justified by a power calculation?
Is the analysis plan pre-specified?
Are consent, confidentiality and approval addressed fully?
Have I included patient and public involvement, or explained its absence?
Is the reporting guideline for my design named?

Structuring the Proposal

Section	What reviewers look for
Title and summary	Design, population and outcome visible in one line
Background	Why the question is unanswered, evidenced by a proper search
Aim and objectives	One aim, three to five objectives, matched to the design
Design	Named design, justified against the question
Population	Inclusion and exclusion criteria, recruitment route
Sample size	Power calculation with every input stated
Outcomes	One primary outcome, defined and measurable
Analysis plan	Pre-specified tests, handling of missing data
Ethics and governance	Approval route, consent, confidentiality, data handling
PPI	How patients and the public shaped the design
Timeline and resources	Realistic phases including approval lead times
Dissemination	How findings reach clinicians and participants

One Primary Outcome

Proposals with four primary outcomes have none. The primary outcome determines the sample size, drives the analysis and is what the study is judged on. Everything else is secondary and should be labelled as such.

Define it fully: what is measured, with what instrument, in whom and at what time point. "Improvement in pain" is not a primary outcome. "Change in mean pain score on the 0 to 10 numerical rating scale between baseline and 12 weeks" is one, because two researchers would measure it the same way.

The Power Calculation

Reviewers check this closely because an underpowered study exposes participants to burden without the capacity to answer the question, which is itself an ethical problem rather than merely a statistical one.

State every input. The expected effect size and where it came from, whether a pilot, published data or a minimum clinically important difference. The significance level, conventionally 0.05. The desired power, conventionally 0.80 or 0.90. The expected attrition, and the inflated target that follows from it. And the software used.

The weakest phrase in any proposal is that the sample size was based on what was feasible. If recruitment genuinely constrains you, say so explicitly, report the effect size the study can detect, and frame it as a pilot or feasibility study rather than presenting it as definitive.

Matching the Reporting Guideline to the Design

Design	Guideline
Randomised trial protocol	SPIRIT
Randomised trial report	CONSORT
Observational study	STROBE
Systematic review	PRISMA
Qualitative study	COREQ or SRQR
Case report	CARE

Naming the guideline in the proposal signals that you know how the finished work will be judged, and the EQUATOR Network indexes all of them in one place.

Ethics and Governance Are Not One Paragraph

Approval route: which committee, and whether NHS Research Ethics Committee approval and HRA approval are needed alongside university review.
Consent: how it is taken, by whom, and what the cooling-off period is. Address capacity where relevant.
Confidentiality: how data are pseudonymised, who holds the key, and where the linkage file sits.
Data protection: lawful basis, storage location, retention period and who has access.
Risk: foreseeable burden or distress, and what you will do about it.
Incidental findings: what happens if a scan or test reveals something unexpected. Decide before, not during.
Withdrawal: the process, and whether already-collected data are retained.

Approval timelines are the most underestimated element of any medical research proposal. Build months, not weeks, into the plan, and show the reviewer you have.

Patient and Public Involvement

UK funders and many ethics committees now expect PPI, and its absence attracts comment. It means involving patients in designing the research rather than recruiting them into it. Did patients help choose the outcome that matters to them? Did anyone review the participant information sheet for readability? Was the burden of the visit schedule tested against real lives?

Where PPI has not been possible, say so and say why, and describe what you will do at the next stage. A candid account is treated far better than silence.

Feasibility Is Judged as Hard as Science

Committees see well-designed studies fail on recruitment far more often than on method. Anticipate the question. State how many eligible patients pass through the site in a month, what proportion you expect to consent, and what your contingency is if that proportion turns out lower. Where you have a letter of support from a site or a clinical lead, say so, because it converts an assertion into evidence.

Common Mistakes and Fixes

Mistake	Fix
Several primary outcomes	Choose one. Demote the rest to secondary
Sample size from convenience	Do the calculation, or reframe as a feasibility study
Analysis plan written as "data will be analysed"	Pre-specify the tests and the missing-data approach
Ethics reduced to one sentence	Cover consent, confidentiality, data protection and risk separately
Timeline ignoring approval lead times	Add realistic months for REC and HRA
No PPI section	Include one, or explain the absence
Recruitment route unspecified	Say where participants come from and who approaches them

Where Ethical Support Fits

Research design support, statistical review and PPI groups exist in most institutions and are meant to be used before submission rather than after rejection. Research Design Services and clinical trials units routinely advise on exactly these sections.

What stays yours is the question, the design decisions and the argument that the study is worth doing. Reviewers assess whether you understand your own protocol, and an interview or committee meeting will make that clear quickly.

Frequently Asked Questions

How long should a medical research proposal be?

Follow the funder's or committee's template exactly, since these are usually prescriptive. Where none is given, 2,000 to 3,000 words is typical.

Do I need REC approval for a student project?

It depends on whether it involves NHS patients, staff or data. Many student projects need university ethics approval only, while NHS-facing work usually needs more. Check early, since this drives the timeline.

What if I cannot recruit the calculated sample?

Say so in the proposal, present it as a feasibility or pilot study, and state what the results would inform rather than what they would prove.

Should I register the protocol?

Clinical trials must be registered prospectively, and it is a condition of publication in journals following ICMJE recommendations. Systematic reviews are commonly registered on PROSPERO.

How detailed does the analysis plan need to be?

Detailed enough that someone else could run it. Name the tests, the comparisons and the approach to missing data.

What is the difference between a proposal and a protocol?

A proposal argues that the study should happen and is often for funding or admission. A protocol specifies exactly how it will be conducted. Protocols are longer and more prescriptive, and SPIRIT governs their content for trials.

Your Next Step Today

Write your primary outcome as one sentence naming the measure, the instrument, the population and the time point. Everything else in the proposal, including the sample size, follows from it, so it is the sentence to fix first.

Trusted Sources

Health Research Authority, Research planning. Accessed 11 August 2026.
EQUATOR Network, SPIRIT statement, standard protocol items for clinical trials. Accessed 11 August 2026.
EQUATOR Network, CONSORT statement, reporting randomised trials. Accessed 11 August 2026.
World Medical Association, Declaration of Helsinki, ethical principles for medical research involving human participants. Accessed 11 August 2026.
ICMJE Recommendations, including trial registration and authorship. Accessed 11 August 2026.
Andrade C, Sample Size and its Importance in Research. Open access. Accessed 11 August 2026.

This is general study and writing guidance, not clinical, legal or regulatory advice. Approval requirements are set by your institution and national regulator, so follow theirs.

Related services

Nursing assignment help