How to Write a Research Protocol: SPIRIT Checklist and a Practical Template

A research protocol is the blueprint of your study. It tells reviewers, ethics committees, and your own team exactly what you plan to do, why you plan to do it, and how you will avoid bias. If you want faster approvals, fewer revisions, and cleaner papers, learn how to write a research protocol the right way.

This guide gives you a step by step system to write a strong protocol for clinical, observational, or academic research. You will also get a practical template you can reuse and a checklist of common mistakes to avoid.

Internal reading first: start with the right question

If you are still choosing a topic, begin here: How to Get a Research Idea. A strong protocol starts with a clear, focused research question.

You can also browse more guides on the SciTrack Blog and keep everything organized from idea to publication.

What makes a protocol strong

A strong protocol is:

• Specific: clear question, clear population, clear outcomes.
• Defendable: design choices reduce bias and confounding.
• Feasible: realistic timeline, sample size, and data access.
• Ethical: risk is minimized and consent and privacy are addressed.
• Reproducible: another team could follow the methods and obtain comparable results.

Step 1: Write the research question in one sentence

Use a framework to avoid vague aims. For quantitative studies, use PICO:

• P: Population
• I: Intervention or exposure
• C: Comparison
• O: Outcome

Example: In [Population], does [Exposure] compared with [Comparison] change [Outcome] within [Time]?

Lock this sentence early. Many projects fail because the question shifts after data extraction begins.

Step 2: Define the study design and setting

Write the design clearly: randomized trial, cohort, case control, cross sectional, qualitative, mixed methods, or systematic review. Then specify the setting (hospital, registry, community, lab), timeframe, and data sources.

If you are planning a systematic review, check the PRISMA guidance early so your methods align with reporting expectations: PRISMA Statement.

Step 3: Eligibility criteria that reviewers cannot attack

Eligibility criteria should be precise and measurable. Include:

• Inclusion criteria (age range, diagnosis definition, timeframe, treatment status)
• Exclusion criteria (missing data rules, prior interventions, overlapping conditions if needed)
• How you will confirm eligibility (ICD codes, echo criteria, operative notes, lab thresholds)

A common weakness is vague definitions. If a variable cannot be measured reliably, do not build your primary aim around it.

Step 4: Outcomes and endpoint definitions

Pick one primary outcome. Then choose a small number of secondary outcomes. For each endpoint, write:

• Exact definition (what counts as an event)
• Time window (30 days, 90 days, 1 year)
• Data source and validation (chart review, registry fields, adjudication rules)

Precise endpoint definitions reduce reviewer criticism and improve reproducibility.

Step 5: Sample size and feasibility

Write a short feasibility section. Even if you do not do a full power calculation, you should state expected sample size and why it is realistic. Include:

• Estimated available sample in your timeframe
• Expected event rate if relevant
• Dropout or missingness expectations

If you are planning a trial, protocol item checklists like SPIRIT can help ensure you include the necessary details: SPIRIT checklist on EQUATOR.

Step 6: Statistical analysis plan (simple and defendable)

Your analysis plan should match your question and design. Include:

• Primary analysis method (regression type, time to event model, matching strategy)
• Confounders and covariates chosen a priori (not after seeing results)
• Handling of missing data (complete case, multiple imputation, sensitivity checks)
• Subgroup analyses (limit them and predefine them)
• Sensitivity analyses (alternate definitions, alternate models)

Keep it readable. Reviewers want clarity more than fancy methods.

Step 7: Ethics, privacy, and governance

Write your ethics section as if an IRB will read it tomorrow:

• Consent approach (waiver, written consent, or verbal consent with justification)
• Privacy and data protection (de identification, access control, retention)
• Risk minimization and adverse event monitoring when applicable

Also define authorship and contributor roles early to prevent conflict. Many journals follow the ICMJE recommendations: ICMJE Recommendations.

Step 8: Registration (highly recommended)

Protocol registration reduces duplication, improves transparency, and strengthens credibility. Options include:

• Systematic reviews: PROSPERO
• Clinical trials: ClinicalTrials.gov

Choose the registry that fits your study type and local requirements.

Practical protocol template (copy and fill)

1. Title: clear population and main outcome
2. Background: what is known, what is missing, why it matters
3. Research question: one sentence (PICO)
4. Aim and hypothesis: one primary aim, few secondary aims
5. Design and setting: design, site, timeframe, data source
6. Eligibility criteria: precise inclusion and exclusion rules
7. Variables: exposures, outcomes, confounders, definitions
8. Sample size: expected sample and reasoning
9. Statistical plan: primary method, covariates, missing data
10. Ethics and privacy: consent, confidentiality, governance
11. Timeline: extraction, analysis, writing, submission
12. Dissemination: target journal and conference plan

Common mistakes that cause protocol rejection

• Vague endpoints or inconsistent definitions
• Too many aims for one dataset
• No plan for confounding or missing data
• Feasibility not addressed (sample size and access unclear)
• Ethics and privacy not written clearly

Next step

Once your protocol is written, keep your screening, extraction, and analysis organized in your workspace. If you are working on a systematic review, check your review area here: Systematic Reviews Workspace.

Want more research tips? Explore the blog archive and keep your idea pipeline active.