The Clinical Trial Protocol & Regulatory Workbook
Build a SPIRIT-2025-complete, GCP-ready protocol your ethics committee will approve.
The problem it solves
The SPIRIT checklist is free, and it tells you what a protocol must contain — not how to write one. Framing an estimand, laying out a schedule of assessments, justifying a sample size to a committee, running a consent process rather than collecting a signature: none of that is in the checklist.
How it works
Built on the SPIRIT 2025 skeleton with the regulatory-operations layer that pure design books leave out — consent, delegation, safety reporting, monitoring, registration and the trial master file.
Who it is for
- First-time investigators told to write the protocol
- Trial coordinators and research nurses assembling a submission
- Clinical academics preparing an ethics or sponsor package
Who should look elsewhere. This builds a protocol; it does not run the analysis. It supports rather than replaces your trial statistician, your sponsor's SOPs, and your regulatory-affairs and ethics specialists. It is not regulatory or statistical advice, and guideline versions change — verify them at the time you write.
What is inside
- Classifying the trial, and mapping to the SPIRIT 2025 skeleton with version control
- Objectives, and a defined estimand under ICH E9(R1)
- An endpoint hierarchy that distinguishes primary from everything else
- Design, randomisation, allocation concealment and blinding
- Eligibility criteria that recruit rather than exclude everyone
- A TIDieR intervention description another site could deliver
- The schedule-of-assessments figure
- Sample-size justification a committee will accept
- An analysis-plan skeleton, CRF design and data-management plan
- A plain-language consent package, and consent as a process rather than a signature
- GCP safety reporting, monitoring, and the trial master file
- Prospective registration and the submission package
What you will have finished
- A classified trial and a SPIRIT-2025-mapped protocol under version control
- Clear objectives, a defined estimand (ICH E9 R1) and an endpoint hierarchy
- A design, randomisation, allocation-concealment and blinding plan
- Eligibility criteria that actually recruit, and a TIDieR intervention description
- A sample-size justification, an analysis-plan skeleton, a CRF and data-management plan
- A plain-language consent package, a GCP safety and monitoring plan, and a registration and submission package
What this adds to the free guidance
SPIRIT 2025, the NIH-FDA protocol template and ICH-GCP are all freely available, and between them they specify what a protocol must contain. None of them tells a first-time investigator how to frame an estimand, lay out a schedule of assessments, justify a sample size to a committee that will push back, or run a consent process rather than collect a signature. This workbook adds the regulatory-operations layer that pure design books omit — delegation, safety reporting, monitoring, registration, the master file — and carries one worked trial through all of it.
Built on SPIRIT 2025, ICH-GCP, CONSORT 2025, TIDieR, ICH E9(R1). Original tools that operationalise the published standards and point to the free official sources.
Frequently asked questions
What is an estimand and why does it matter?
An estimand states precisely what treatment effect you intend to measure, including how you will handle the events that complicate it — discontinuation, rescue medication, death. ICH E9(R1) made it explicit because trials had long reported effects without saying what question they answered. Defining it before the protocol is written prevents an analysis plan that cannot be defended afterwards.
Does following SPIRIT guarantee approval?
No. SPIRIT is a reporting standard for completeness, not a design review. A protocol can be SPIRIT-complete and still be sent back for an unjustified sample size, eligibility criteria that will not recruit, or a consent process that a committee considers inadequate. Those are what the coaching chapters address.
When do I register the trial?
Prospectively — before the first participant is enrolled. Most journals following ICMJE will not publish a trial registered retrospectively, and this is discovered far too late with dispiriting regularity. The registration chapter comes before submission for that reason.
Do I still need a statistician?
Yes. This workbook makes you arrive at that conversation with a defined estimand, a stated endpoint hierarchy and a draft sample-size justification, which makes the conversation far shorter and more productive. It does not substitute for it.