How to Prepare for ICH M11 CeSHarP: An Implementation Guide for Sponsors and Medical Writers

Introduction

On May 22, 2026, the FDA published its final guidance adopting ICH M11, the first internationally harmonized clinical trial protocol template, following the European Medicines Agency's own Step 5 adoption, which came into effect June 11, 2026 ^[3][4]. Before M11, ICH had not adopted a harmonized standard for protocol format and content, so a sponsor's internal template, a CRO's house style, or a legacy TransCelerate Common Protocol Template build could all place the same information in different sections. For scale, the FDA's Center for Drug Evaluation and Research and Center for Biologics Evaluation and Research together received an estimated 4,900 unique clinical trial protocols a year, according to a 2025 conference estimate built on 2018 to 2019 data; the person who presented it was speaking in a personal capacity and specifically not on behalf of EMA or ICH, so treat it as an illustration of scale rather than an official figure ^[7]. For sponsors and medical writers who have spent years working from whatever template a given program happened to inherit, M11 changes the baseline document many clinical-trial and development submissions depend on.

This ICH M11 CeSHarP implementation guide covers what the Clinical Electronic Structured Harmonised Protocol actually requires in its final form, what the evidence says about the problem it is trying to solve, and what sponsors and medical writers should do differently starting with their next protocol cycle.

Why This Topic Matters in Clinical Trials

Protocol inconsistency is not a cosmetic problem. It shows up as delay and cost. A 2016 analysis of Phase II and Phase III protocols found that 57% required at least one substantial amendment, with Phase II protocols averaging 2.2 amendments and Phase III protocols averaging 2.3, and nearly half of those changes were judged avoidable ^[8]. An updated 2024 benchmarking study from the same research group found the picture had shifted further: amendment prevalence across Phases I through IV rose to 76%, and the mean number of amendments per protocol climbed 60%, from 2.1 to 3.3 ^[9]. The same study measured how long the process now takes: an average of 260 days from identifying the need to amend a protocol to the last required oversight approval, with sites operating under inconsistent protocol versions for roughly 215 days in the interim ^[9]. Any one of those amendments may also require consequential updates to the documents built on top of the protocol, from the informed consent form to the statistical analysis plan.

Cost follows time. The Tufts Center for the Study of Drug Development estimated in 2024, using budget data from 447 Phase II and III protocols, that a day of direct trial expense averages approximately $40,000 ^[10]. That figure is a mean drawn from a specific protocol-budget dataset rather than a guaranteed incremental cost for every delayed day, and it says nothing about lost product revenue on top of it. None of this delay is caused solely by inconsistent formatting, and M11 does not claim to change the scientific decisions that drive most amendments. What the template targets is the layer underneath those decisions: making sure everyone reading a protocol, human or machine, knows where to look for a given piece of information.

Current Evidence and Research Landscape

ICH M11 is not a single document. It consists of three parts: a guideline explaining the rationale and design principles, a standardized protocol template with a fixed table of contents and section headers, and a technical specification defining the data elements, conformance rules, and cardinality needed for electronic exchange ^[1][2]. The ICH Assembly endorsed the package at Step 4 on November 19, 2025 ^[2]. Regional adoption, Step 5 in ICH's process, follows its own schedule in each jurisdiction. The EMA's version came into effect June 11, 2026, and its Step 5 document does not describe use of the template as mandatory; it states that the template is "intended to assist" stakeholders and does not set a transition period beyond that effective date ^[3]. The FDA's guidance, published May 22, 2026, is explicit on this point: FDA guidance documents "do not establish legally enforceable responsibilities" and "should be viewed only as recommendations," a status that applies to CeSHarP as much as to any other FDA guidance ^[4]. Sponsors running trials in other ICH regions, including Japan's PMDA, should check that authority's own notice directly rather than assume the EU or US position applies.

The scope is narrower than "all clinical research." The EMA's Step 5 text specifies that the template and technical specification apply to interventional clinical trials of medicinal products across all phases and therapeutic areas, including drug-device combination products when they are being developed as a drug, but not to standalone medical device studies ^[3].

The finalized technical specification itself is narrower than a lot of industry commentary implies. It defines protocol data elements, controlled terminology, conformance attributes (required, optional, or conditional), cardinality, and how elements relate to each other within the protocol hierarchy, describing this as "a technical representation of the ICH M11 protocol template" that leaves flexibility in how data exchange is actually implemented ^[19]. It does not name FHIR, and it does not name CDISC's Unified Study Definitions Model, or USDM, as normative content ^[19]. Those connections exist, but they live in separate, ecosystem-level work: CDISC and its partner organizations have mapped USDM to M11's data elements, and HL7's Vulcan UDP project has built a Clinical Study Protocol Implementation Guide that maps both M11 and USDM content into FHIR, specifically because regulatory agencies asked for a FHIR-based exchange path. That FHIR guide is still in ballot, trial-use status as of this writing, not a finalized standard ^[20].

The USDM and Digital Data Flow (DDF) ecosystem built around M11 is further along than that trial-use status might suggest. USDM version 4.0 shipped in mid-2025 with conformance rule specifications and an implementation guide, an open-source Study Definitions Repository reference implementation is publicly available, and a 2026 CDISC handbook describes a target operating model for generating CDISC SDTM Trial Design domains directly from structured study definitions ^[13][14]. What remains uneven is production use of that model. DDF project materials present the reference implementation as one possible implementation approach rather than a required platform, and the handbook's workflow still depends on a sponsor or vendor building or buying a conformant tool, so it is not yet evidence that automated domain generation is routine across the industry ^[13][15]. A sponsor authoring a protocol using the final M11 template today has implemented the template's human-readable component. Structured, machine-readable exchange requires an additional implementation layer on top of that. USDM and DDF, and the developing HL7/Vulcan FHIR guide described above, are prominent and partly complementary approaches to building that layer, not three separate requirements M11 obliges a sponsor to satisfy.

Industry groups had a hand in shaping the template before it reached Step 4. The Clinical Trials Transformation Initiative ran stakeholder webinars during the draft consultation period, walking sponsors through the template's main body, general considerations, and appendices structure well ahead of the technical specification's 2025 update ^[12]. PROMETRIKA, a contract research organization, reported running mock protocol exercises against the draft structure to surface authoring issues before the template was finalized ^[11]. That early testing gave participating sponsors an opportunity to identify authoring issues before the template was finalized, rather than after.

Operational Impact for Sponsors, CROs, and Sites

The final template moves familiar content into specific, numbered places. Trial design sits in Section 4, with trial stopping rules in Section 4.3. Eligibility, previously often a single combined section, splits into Section 5.2 for inclusion criteria and Section 5.3 for exclusion criteria. Safety assessments and procedures are addressed in Section 8.4, while the timing and procedures for collecting and reporting adverse and serious adverse events sit in Section 9.2, with a dedicated Section 9.3 for pregnancy and postpartum information ^[5]. Appendices carrying supporting detail are consolidated in Section 12 ^[5]. The cover page now calls for a plain-language trial title alongside the scientific one, along with the applicable regulatory identifiers, using the field label "EU CT Number" rather than the EudraCT number sponsors may be accustomed to entering, plus amendment history with sponsor approval dates ^[5].

The synopsis carries a new, but conditional, requirement. For trials intended to estimate or test a treatment effect, the template instructs authors to "summarise the primary and secondary objectives and any associated estimands in natural, nontechnical (layperson) language" ^[5]. Trials not designed to estimate a treatment effect follow different synopsis instructions. Sponsors whose statistical planning has not yet built estimand language into early protocol drafts, for the trial types where it applies, will need to change that process before their next M11-formatted synopsis goes out for review.

Some of the changes create friction that sponsors should plan for rather than discover mid-cycle. The introduction section is now expected to focus on the study-specific benefit-risk conclusion rather than restating investigational product background that belongs in the Investigator's Brochure ^[5]. The template does not prescribe a specific drafting sequence, but that instruction gives sponsors a practical reason to keep the IB current before protocol drafting begins rather than patching new information into the protocol's introduction later. Drug-device combination products, when the investigational product is being developed as a drug, bring in conditional fields distributed across the relevant sections rather than a single standalone medical-device section ^[3].

Whether the new structure reduces the version-control confusion behind the 215-day period of inconsistent protocol versions across sites is not yet demonstrated ^[9]. A fixed template with defined cardinality rules does not resolve the scientific disagreements that trigger most amendments. What it can do, in principle, is make it easier for a compatible clinical trial management or EDC system to isolate exactly which data element changed between two protocol versions, since those elements now have consistent locations and definitions across sponsors. Realizing that benefit still depends on the platform doing the comparison, not on template adoption by itself.

A Practical Adoption Sequence for Sponsors

The cited EMA and FDA documents do not state that use of the template is mandatory and set no transition deadline, so sponsors have room to test it deliberately rather than adopt it under submission pressure. A workable sequence looks like this:

1. Inventory current templates (regulatory operations, medical writing). Catalog whatever a program uses today, whether an internal format, a CRO house style, or TransCelerate's Common Protocol Template, and select a small number of upcoming, not-yet-amended protocols as pilots rather than retrofitting protocols already deep into amendment cycles.
2. Build a section-by-section crosswalk (medical writing, clinical data standards). Map the existing template to the final M11 table of contents, flagging both relocations, such as eligibility moving to Sections 5.2 and 5.3 and stopping rules moving to Section 4.3, and genuinely new content, such as the conditional estimand statement and the EU CT Number field. TransCelerate's own CPT V011 mapping reference already covers this ground for organizations using the Common Protocol Template ^[6].
3. Confirm vendor conformance (clinical data standards, IT, regulatory operations). Where USDM is the sponsor's selected implementation target, ask protocol-authoring, study-design, and metadata-repository vendors, the parties who actually own structured protocol data, not eTMF vendors, whether their tools can produce a USDM-conformant export, not merely a document whose headings visually match M11.
4. Update governance (regulatory operations, quality). Assign clear ownership of the new template version and define change-control procedures for amendments made under the new section structure.
5. Brief internal and investigator-facing teams (clinical operations, medical writing). Walk medical writers, biostatisticians, CRO teams, investigators, and site staff through the specific relocations before a pilot protocol reaches review. An IRB or independent ethics committee cannot be trained by the sponsor, but a plain-language note describing what moved and why, submitted through whatever channel that board normally accepts, may reduce format-navigation questions.
6. Check regional status directly (regulatory affairs). Confirm the applicable guidance for each jurisdiction in a program rather than assuming one region's status carries over: the EMA's version has been in effect since June 11, 2026, the FDA's May 2026 guidance remains nonbinding, and PMDA and other ICH regulators set their own timelines ^[3][4].

A pilot is reasonably complete once a protocol has cleared internal review under the new structure without new format-driven queries, and, where a structured export was attempted, it has cleared two separate checks rather than one: content coverage against the M11 template and technical specification, and, separately, conformance against whichever USDM, API, JSON, or FHIR implementation artifact the sponsor's tooling actually targets. Sites or CROs involved should also be able to confirm they can trace which sections changed between versions.

Regulatory and Documentation Considerations

M11 governs structure, not substantive content. The underlying requirements for what a protocol must actually say remain set by ICH E6(R3), whose Principles and Annex 1 reached Step 4 in January 2025 and whose Annex 2, covering decentralized elements, pragmatic trial design, and real-world data, separately reached Step 4 in June 2026, and by ICH E8(R1), the general considerations guideline finalized in 2021 ^[16][17][21]. Depending on the trial, ICH E9(R1)'s estimand framework, finalized in 2019, along with regional regulations and therapeutic-area-specific guidance, also shapes what has to go into the sections M11 now standardizes ^[18]. A protocol built on the CeSHarP template still has to satisfy all of that content guidance; the template only fixes where the content lives and how it is labeled for exchange between systems.

Organizations currently using TransCelerate's Common Protocol Template have less of a gap to close than they might expect, and less than earlier CPT versions would have implied. TransCelerate released CPT V011 in 2026, realigned specifically to the finalized M11 template and technical specification, with its own CPT V011-to-M11 mapping reference for converting legacy CPT-based protocols ^[6]. Sponsors still running an older CPT version, or a heavily customized internal variant, face more of a reorganization exercise: mapping existing content into M11's table of contents, plus fresh drafting for the genuinely new elements, the estimand statement in the synopsis where it applies and the updated cover-page identifiers, rather than a wholesale rebuild ^[5]. Sponsors have room to run the new template on a subset of upcoming protocols, most practically new studies rather than protocols already deep into amendment cycles, before committing every program to it.

AI and Automation Perspective

Automation here depends on layers the M11 technical specification itself does not build. The specification defines conformance rules and cardinality for each data element in the protocol, but it does not prescribe an exchange format on its own ^[19]. That job falls to the surrounding ecosystem: CDISC's USDM gives those data elements a common model, and the current HL7/Vulcan implementation guide defines a FHIR representation for M11-aligned protocols and maps the M11-relevant overlap with USDM. Its own documentation is explicit that this present mapping "does not cover the whole of USDM," and the guide remains a continuous, changeable build at ballot/trial-use maturity rather than a finalized standard ^[20]. Once a sponsor has implemented those available USDM and DDF artifacts through a conformant authoring or metadata-repository tool, a protocol can feed a trial registry entry or the CDISC SDTM Trial Design domains that downstream systems already expect; CDISC's own 2026 implementation handbook walks through that mapping as a target workflow ^[13][14]. Feeding a statistical analysis plan or a data management plan the same way is a further step that depends on how a given vendor or sponsor has built its own systems, and is not yet the standardized, out-of-the-box outcome that registry and trial-design population is closer to becoming. That is a real efficiency gain where it has been implemented, not a guarantee that comes with adopting the paper template on its own.

Automation here has clear limits, and FDA's own description of its guidance is a useful check on overstated claims. FDA guidance "represents the current thinking of FDA" on the topic, "does not establish any rights for any person," and "is not binding on FDA or the public" ^[4]. Structured formatting cannot, by itself, fix a poorly reasoned inclusion criterion or a badly designed endpoint, whatever the tooling built around it. Any AI-assisted drafting or consistency-checking tool built around the M11 structure still needs a human reviewer validating that extracted data elements match what the protocol actually intends, particularly around estimands and stopping rules, where a subtle misstatement carries real consequences for how a trial is interpreted or amended later.

How Kitsa Fits Into This Problem

The problem M11 addresses at the format level, that a protocol, an Investigator's Brochure, and an informed consent form drafted independently tend to drift out of sync with every amendment, is one Kitsa's KScribe was built around. KScribe is designed to generate and update these documents from a shared set of trial data rather than have separate teams retype the same eligibility criteria in three places, which is the same underlying goal behind CeSHarP's push toward structured, machine-readable protocol content. How far that alignment goes in practice depends on the same tooling and mapping work described above, and Kitsa's own product documentation, rather than this article, is the right place to evaluate specific claims about it.

Key Takeaways

FAQ

References

1. [1]International Council for Harmonisation. "M11: Clinical Electronic Structured Harmonised Protocol (CeSHarP) Explainer." May 2026. https://database.ich.org/sites/default/files/ICH%20M11%20Explainer_May2026.pdf
2. [2]International Council for Harmonisation. "M11 Guideline: Clinical Electronic Structured Harmonised Protocol (CeSHarP)." Step 4 Final Guideline, November 19, 2025. https://database.ich.org/sites/default/files/ICH_Step4_M11_Final_Guideline_2025_1119.pdf
3. [3]European Medicines Agency. "ICH M11 Guideline: Clinical Electronic Structured Harmonised Protocol (CeSHarP)." Step 5, effective June 11, 2026. https://www.ema.europa.eu/en/documents/regulatory-procedural-guideline/ich-m11-guideline-clinical-electronic-structured-harmonised-protocol-cesharp-step-5_en.pdf
4. [4]Federal Register. "M11 Clinical Electronic Structured Harmonised Protocol (CeSHarP); International Council for Harmonisation; Guidance for Industry; Availability." 91 FR 30310, May 22, 2026. https://www.federalregister.gov/documents/2026/05/22/2026-10295/m11-clinical-electronic-structured-harmonised-protocol-cesharp-international-council-for
5. [5]U.S. Food and Drug Administration. "M11 Clinical Electronic Structured Harmonised Protocol (CeSHarP) Template." Final template. https://www.fda.gov/media/192647/download?attachment=
6. [6]TransCelerate BioPharma Inc. "Clinical Content & Reuse Assets: Common Protocol Template (CPT) V011." 2026. https://www.transceleratebiopharmainc.com/assets/clinical-content-reuse-solutions/
7. [7]Halsey, N. "ICH M11 Protocols." CDISC Europe Interchange, 2025. https://www.cdisc.org/sites/default/files/2025-05/ICH%20M11%20Protocols%20-%20CDISC%20Europe%202025%20-%20Nick%20Halsey.pdf
8. [8]Getz, K.A., Stergiopoulos, S., Short, M., Surgeon, L., Krauss, R., Pretorius, S., Desmond, J., Dunn, D. "The Impact of Protocol Amendments on Clinical Trial Performance and Cost." Therapeutic Innovation & Regulatory Science, 2016;50(4):436-441. https://pubmed.ncbi.nlm.nih.gov/30227022/
9. [9]Getz, K., Smith, Z., Botto, E., Murphy, E., Dauchy, A. "New Benchmarks on Protocol Amendment Practices, Trends and their Impact on Clinical Trial Performance." Therapeutic Innovation & Regulatory Science, 2024;58(3):539-548. https://pubmed.ncbi.nlm.nih.gov/38438658/
10. [10]Smith, Z., DiMasi, J., Getz, K. "Day of Delay White Paper." Tufts Center for the Study of Drug Development, August 2024. https://csdd.tufts.edu/sites/default/files/2025-02/Aug2024%20Day%20of%20Delay%20White%20Paper%20Final.pdf
11. [11]PROMETRIKA. "The M11 Template: Clinical Electronic Structured Harmonised Protocol (CeSHarP)." Regulatory Corner. https://www.prometrika.com/thought-leadership/regulatory-corner/the-m11-template-clinical-electronic-structured-harmonised-protocol-cesharp/
12. [12]Clinical Trials Transformation Initiative. "ICH M11: Clinical Electronic Structured Harmonised Protocol Webinar." January 26, 2023. https://ctti-clinicaltrials.org/wp-content/uploads/2023/01/ICH-M11_CTTI-Webinar_01-26-23.pdf
13. [13]CDISC. "Digital Data Flow (DDF) for Clinical Trial Protocols." https://www.cdisc.org/ddf
14. [14]CDISC. "USDM Handbook 1: Leveraging USDM Metadata to Automatically Construct the CDISC SDTM Trial Design Domains." June 2026. https://www.cdisc.org/sites/default/files/2026-06/USDM-HB1%20v1.0_FINAL.pdf
15. [15]TransCelerate BioPharma Inc. "Digital Data Flow (DDF): Getting Started." https://transcelerate.github.io/ddf-home/getting-started.html
16. [16]International Council for Harmonisation / European Medicines Agency. "ICH E6(R3) Guideline on Good Clinical Practice (GCP): Principles and Annex 1, Step 5." Finalized January 2025; effective at EMA July 23, 2025. https://www.ema.europa.eu/en/documents/scientific-guideline/ich-e6-r3-guideline-good-clinical-practice-gcp-step-5_en.pdf
17. [17]International Council for Harmonisation. "E8(R1) General Considerations for Clinical Studies." Finalized 2021. https://database.ich.org/sites/default/files/ICH_E8-R1_Guideline_Step4_2021_1006.pdf
18. [18]International Council for Harmonisation. "E9(R1): Statistical Principles for Clinical Trials, Addendum on Estimands and Sensitivity Analysis in Clinical Trials." Step 4, finalized 2019. https://database.ich.org/sites/default/files/E9-R1_Step4_Guideline_2019_1203.pdf
19. [19]International Council for Harmonisation. "M11 Technical Specification: Clinical Electronic Structured Harmonised Protocol (CeSHarP)." Step 4 Final, November 19, 2025. https://database.ich.org/sites/default/files/ICH_Step4_M11_Final_TechnicalSpecification_2025_1119.pdf
20. [20]HL7 Vulcan Accelerator. "Clinical Study Protocol Implementation Guide." Version 1.0.0-ballot2, trial-use; continuous build, subject to change. Accessed September 2026. https://build.fhir.org/ig/HL7/vulcan-udp-ig/en/
21. [21]International Council for Harmonisation. "ICH E6(R3) Guideline on Good Clinical Practice (GCP): Annex 2." Step 4, adopted June 3, 2026. https://database.ich.org/sites/default/files/ICH_E6%28R3%29_Annex%202_Guideline_Step%204_2026_0603_0.pdf