Statistical Analysis Plan Writing for Sponsors, CROs and Investigators
Pubrica’s biostatisticians write statistical analysis plans (SAPs) for trial sponsors, CROs, academic investigators, device teams and real-world evidence studies, completed and signed before database lock. Each plan defines the estimand under ICH E9(R1), covering population, variable, intercurrent event handling and summary measure, before the analysis methods are chosen. You receive analysis populations defined by rule, statistical methods with their justification, a multiplicity strategy, pre-specified sensitivity analyses and, where requested, table, listing and figure shells.
- The estimand framed before anything else, under ICH E9(R1), which governs how treatment discontinuation and rescue medication are handled
- Analysis populations defined and justified — full analysis set, per-protocol set and safety set, each defined by rule
- Multiplicity handled deliberately, with the testing hierarchy stated and the alpha allocation documented
- Sensitivity and supplementary analyses specified in advance, before unblinding, so that they carry full weight in the report
- Written for direct implementation — a plan a programmer can implement without further clarification
Estimand framing under E9(R1)Included
Populations, endpoints and methodsIncluded
Multiplicity and sensitivity analysesIncluded
TLF shellsIncluded on request
Database build and lockHandled by your data management team
Protocol → estimand → methods → shells → plan
- Estimand settled before methods
- Populations defined and justified
- Multiplicity stated, not assumed
- Sensitivity analyses pre-specified
- Ready for programming
Service Scope
Four Statistical Analysis Plan Service Areas, Available Individually
Pubrica provides each part of statistical analysis plan writing as a separate service, so you can outsource the complete plan or only the part you need, whether you are starting from a protocol or reviewing an existing draft.
Who we serve
Sponsors and Researchers We Support with Statistical Analysis Plans
We support trial sponsors, CROs, academic investigators and other teams whose analysis plans must satisfy regulators, notified bodies, ethics committees, funders and journals.
Deliverables
What Every Statistical Analysis Plan Includes
Each plan is written against your protocol and its primary endpoint. What the plan pre-specifies determines what the study report can later conclude.
From the estimand sectionMissing data will be handled using last observation carried forward.Treatment policy estimand. Intercurrent events: treatment discontinuation (handled by treatment policy) and rescue medication (composite). Summary measure: difference in proportions at 30 days. Sensitivity analyses named.
From the populations sectionThe per-protocol set will exclude subjects with major protocol deviations.Excludes subjects with any of six listed deviation categories, each defined; exclusions decided and documented before unblinding by the blinded data review.
From the multiplicity sectionOne primary and two key secondary endpoints in a fixed sequence, each tested at 0.05 only if the previous one succeeded. Stated in the plan, not decided at analysis.
From the planTipping-point analysis for the primary endpoint, a per-protocol confirmatory analysis and a complier-average causal effect analysis: three analyses, each chosen and justified in advance.
These samples are illustrative and anonymised. Hover over a sample to replay it.
Compliance and Standards
Statistical Analysis Plan Regulatory and Reporting Standards
Where your sponsor template or regulatory agency has a stated position, we follow it.
Statistical principles with estimands (E9(R1), 2019)
The estimand framework. Citing E9 alone is incomplete, because the addendum governs how intercurrent events are defined and handled.
Good Clinical Practice (E6(R3), not E6(R2))
Annex 1 took effect in the EU in July 2025; Annex 2, covering pragmatic and decentralised designs, follows in January 2027.
Structured protocol template (Step 4, November 2025)
The harmonised protocol content and technical specification, known as CeSHarP. TransCelerate has realigned its Common Protocol Template to it; most templates still in circulation predate it.
CONSORT 2025 and SPIRIT 2025 (not the 2010 and 2013 versions)
Where the trial will be published, the analysis plan must support the reporting checklist the journal applies. Both checklists have been updated to 2025 versions.
Responsible Use of AI
Responsible Use of AI in Statistical Analysis Plans
An analysis plan is a commitment made in advance by a named statistician. Its value depends on a qualified person making each decision before the data exist and documenting the reasons.
Why We Do Not Publish an AI Accuracy Figure
Accuracy figures for AI tools are usually reported by the companies that sell them.
For the same reason, Pubrica does not quote an accuracy figure of its own, as it would carry the same conflict of interest. Instead, we tell you which tool was used on your project, at which stage, and what our reviewers changed after it.
Methodology
How Our Statistical Analysis Plan Process Works
We define the estimand first, because every other decision in the plan depends on it.
Protocol read and estimand framing
We read the objectives and endpoints from the protocol and construct the estimand for each one: population, variable, intercurrent events and summary measure.
Methods and populations
We define the analysis populations by rule, select and justify the statistical methods, state the multiplicity strategy and specify how missing data will be handled.
Sensitivity, supplementary and subgroup analyses
We pre-specify, justify and limit these analyses, and mark subgroup analyses as exploratory where appropriate.
Shells and sign-off
We produce table, listing and figure shells where they are in scope, and the plan is finalised and approved before treatment unblinding; in our projects, finalisation before database lock is the usual procedure.
Our experts
Your Statistical Analysis Plan Team
Your project is handled by a named team of subject specialists who stay with it from scoping to delivery. Your project coordinator introduces each team member before work begins.
Reviews the scientific content for accuracy and confirms that each conclusion is supported by the evidence. Matched to your field rather than allocated from a rota.
Checks that statistical methods, tables and conclusions are consistent with the analysis plan and that the analysis can be reproduced.
Prepares, cleans and documents the data so that every step from raw data to final output can be traced and reproduced.
Independently checks code, tables and outputs against the analysis plan. Discrepancies are logged and resolved before delivery.
Packages
Statistical Analysis Plan Service Packages
The cost of a statistical analysis plan depends on the study design, the number of endpoints and whether TLF shells are in scope.
Statistical analysis plan
Sponsors and investigators before database lock
Quoted per projectBased on design complexity and endpoint count
Delivery By scope
- ✓Estimand framed under ICH E9(R1) for each objective
- ✓Analysis populations defined by rule
- ✓Methods selected and justified
- ✓Multiplicity strategy stated
- ✓Sensitivity and supplementary analyses pre-specified
- ✓Two review cycles
Commission the plan early, so that it can be finalised before unblinding and the analysis is genuinely pre-specified.
Deliverables Request a quotePlan with TLF shells
Teams taking the analysis through to report
Quoted per projectBased on the number of tables and figures
Delivery By scope
- ✓Everything in the plan
- ✓Table, listing and figure shells
- ✓Mock outputs for the primary and key secondary endpoints
- ✓Shell-to-plan traceability index
- ✓Handover to programming
Shells written alongside the plan match it; shells written afterwards often diverge from it.
Deliverables Request a quoteSend us your protocol and we will frame the primary estimand, covering population, variable, intercurrent events and summary measure, within five business days at no charge. Every other decision in the plan depends on it.
Benefits
Benefits of a Pubrica Statistical Analysis Plan
The main benefits of commissioning a statistical analysis plan from Pubrica.
Quality Assurance
Our Service Commitments and Limitations
Our service terms are set out in full below so you can review them before you commit.
The Pubrica guarantee
Revision Support Until Agreed Requirements Are Met
We continue revising the work until it meets the requirements agreed within the scope. Rather than offering a refund, which would end the engagement with the work incomplete, we keep working until the agreed requirements are met.
01
Unlimited revisions within the agreed scope
There is no limit on the number or timing of revision rounds within the agreed scope. If the work does not meet the agreed requirements, we revise it. Work outside the agreed scope is quoted separately, and we confirm which applies before any work starts.
02
Corrections at Our Cost
If an error in our analysis or writing needs correcting, we correct it free of charge.
03
Delivered on the date agreed
We quote a delivery date we can meet and work to it. If a milestone is at risk, we inform you in advance.
04
Mutual NDA on request
We sign a mutual non-disclosure agreement before you share any material, at no cost, and we can sign your organisation’s NDA instead of ours.
05
You Retain Final Decisions
We give clear methodological advice and tell you when we disagree with a proposed approach. You make the final decision on everything published in your name.
Average client rating
From 3,000 responses to Pubrica’s post-delivery survey, sent to every client at handover between January 2024 and June 2026. This figure comes from our own records rather than from an external rating service.
Limitations of Our Service We cannot guarantee decisions made by others, such as an editor’s decision, a funding panel’s outcome or a regulator’s assessment. Within the agreed scope, we commit to the work itself: completed to the standards named on this page, delivered on the agreed date and corrected at our cost if we made an error.
FAQs
Statistical Analysis Plan FAQs
If your question is not answered here, please ask an expert. No quote is required.
An estimand is a precise definition of the treatment effect the trial is trying to estimate: the population, the variable, how intercurrent events like discontinuation and rescue medication are handled, and the summary measure.
It comes first because every other choice in the plan depends on it. ICH E9(R1) introduced the framework in 2019 because these decisions were often made after unblinding, and citing E9 without the addendum is now incomplete.
Prespecify the principal analyses in the protocol. Finalise and approve the detailed statistical analysis plan before treatment unblinding and before analyses that could influence analytical choices. Document and justify subsequent changes. In our projects, finalisation before database lock is the usual procedure.
The date matters. In the clinical study report, a change made before unblinding is interpreted very differently from the same change made after it, and section 9.8 must state which applies.
SAP writing is one part of a biostatistician’s work. We offer it as a separate service because many clients, often teams whose analysis will be run elsewhere, need the plan on its own.
Where we also run the analysis, the plan and its execution stay with the same team. Where we do not, we write the plan so that another team can implement it without further clarification.
Yes, provided the database has not been locked and the treatment allocation has not been revealed.
If the data have been seen, we can write an analysis plan marked as post hoc. This is still useful, but it is a different document and the report must identify it as such.
We produce the shells as part of the plan where that is in scope, and we can run the analysis against your locked dataset and produce the outputs.
We do not manage the clinical database. Database build, validation and lock remain with your data management team or CRO.
Related services
Related Research Services
Services connected to the statistical analysis plan, from protocol to study report.
1BEFORE THE PLANdocuments the plan is based on
Clinical study protocolObjectives, endpoints and schedule. Experimental designDesign, sample size and endpoints.2RUNNING THE ANALYSISafter database lock
BiostatisticsAnalysis and TLFs from the locked dataset. Meta-analysisPooled estimates across studies.3WHEN IT REPORTSdocuments that report the analysis
Clinical study reportIncluding every departure from the plan, in section 9.8. Original research articleThe manuscript, to CONSORT 2025.Request a Free Primary Estimand Framing
Send your protocol and we will frame the primary estimand, covering population, variable, intercurrent events and summary measure, within five business days at no charge.