The structured protocol and ICH M11
ICH M11 asks for a protocol built as structured, reusable content rather than as narrative typed into a template. It changes where each piece of a protocol's content is defined, when disclosure is planned for and what compliance with the template can show, and its status and its fit with the TransCelerate common protocol template have each changed over time.
What ICH M11 is
ICH M11 is digital scaffolding for structured, harmonised protocol development, and more than a template to adopt: it is a shift in mindset from the narrative of static text to a data-centric framework. It establishes a common foundation for the structure, content and electronic exchange of protocols across the ICH regions and across the drug development lifecycle. It is not limited to one type of study, and interventional trials of any phase, therapeutic area and product type are in scope. The guidance has three parts: principles, a template and a technical specifications guideline of about 245 pages. The underlying expectation has not changed, since transparent, high-quality research has always been the standard. The underlying principles have not changed either, though the move from static text to a data-centric world is a paradigm shift. What M11 adds is greater consistency and efficiency in preparing and sharing the information.
Why protocols are being restructured
The case for M11 is protocol modernisation. There is no standard way to write a protocol: everyone does it differently and every region has its own expectations, which slows the work and prevents both the easy reuse of information and the introduction of new technology. With M11 everyone uses the same structure, terminology and content placement. Its technical specifications add a structured content model for electronic data exchange and interoperability, which defines content as data, so that it can be pulled automatically into other systems and documents. Protocol content is otherwise copied, pasted, rewritten and tweaked largely by hand, and communicating it electronically across systems reduces both the errors and the time. Fixed definitions and rules allow information to be communicated automatically, and the end point is a unified and automated clinical trial data ecosystem. M11 is designed to enable clarity, alignment, execution and reuse across the document development lifecycle, and a well-designed protocol makes everything easier. The industry has been document-focused, treating the document as the deliverable, and protocols in particular lend themselves to content reuse and data interoperability. ICH M11 reflects this, looking at the reuse of information as well as at structure.
The guideline's status over time
In September 2022 ICH released M11, the Clinical Electronic Structured Harmonised Protocol (CESHARP), as a Step 2 draft guideline outlining a harmonised template for clinical trial protocols to support consistency among sponsors. The draft guideline, its template and the template's technical specifications opened for public consultation on 21 October 2022, and the consultation ended in February 2023. The draft's scope was to establish common instructions for the placement of content and information on technical attributes. ICH's stated aim was clinical trial protocol templates that are complete, free from ambiguity, well organised and aligned with quality by design principles. The template's core set of information extended to the fonts to be used, the numbering of tables and figures and the acceptable abbreviations. In March 2023 the draft had two main themes: a common protocol structure, and technical specifications for the electronic exchange of protocol content, that is, for digital protocols.
In July 2024 Step 4 of ICH M11 was planned for autumn 2025.
In October 2025 the M11 template was still in draft: it was expected to be final that autumn, but that could extend into 2026 and was not yet clear. The TransCelerate template was still the industry standard. Some people were already adapting to the M11 template before its release, and adapting templates was among the concerns and struggles protocol writers named.
In June 2026 a live poll of a webinar audience found that most had not yet started using the template, though more had than expected, and that those who had were divided equally between adopting their own template and using the available one. Many were already implementing the technical specifications.
The Step 2 draft and the TransCelerate template
In March 2023 the draft stood in some conflict with the TransCelerate common protocol template, and the hope was that the two would align as the draft matured. TransCelerate had released a clinical template suite addendum in November 2022, a track-changes version of its common protocol template (version 9) with only limited updates. The addendum said that the next templates would follow in the second half of 2023, to allow alignment with M11 and with EU-PEARL, the EU patient-centric clinical trial platform. TransCelerate was to wait until M11 was more mature before commenting further. In March 2023 the final M11 guidance was expected to set the structure of the clinical trial protocol, on the ground that "ICH trumps everything". As M11 matured, it was expected without doubt to affect a range of study designs, including oncology trials.
On 13 December 2022 the CORE Reference project team published an open-access comparison of the level 2 headings of the TransCelerate template (version 9) and the draft M11 template, to support familiarisation with M11. The comparison gives the draft's level 1 headings in this order: a protocol summary (synopsis, trial schema and schedule of activities); an introduction (purpose of the trial and summary of benefits and risks); trial objectives, endpoints and estimands; trial design; trial population; trial intervention and concomitant therapy; discontinuation of trial intervention and participant withdrawal from the trial; trial assessments and procedures; statistical considerations; general considerations on regulatory, ethical and trial oversight matters; general considerations on risk management and quality assurance; an appendix on adverse events and serious adverse events, with their definitions, severity and causality; an appendix on definitions and supporting operational details; a glossary of terms; and references.
The overall structure of the protocol summary and the level of detail required in the introduction were similar in the two templates. For objectives and endpoints, although no definitive level 2 headings were given, the TransCelerate template proposed more comprehensive guidance than the draft on how endpoints and objectives should be presented. Under trial design the draft required a description of any access to the trial intervention after the end of the trial, which the TransCelerate template had in its section on the study intervention. The draft dealt specifically with the selection of the trial population and its rationale, and the TransCelerate template had a section on criteria for temporarily delaying enrolment, randomisation or administration of the study intervention that the draft did not have. The overall organisation of the trial intervention section differed slightly between the two.
The draft emphasised trial-specific stopping rules, such as guidance on stopping the trial for safety reasons or on when a cohort, a dose escalation or a treatment arm should be ended. The TransCelerate template had participant stopping rules based on particular variables, such as liver chemistry or QTc criteria, as level 3 headings, but no guidance on stopping a trial or a treatment arm. The draft made pregnancy a level 2 heading, where the TransCelerate template had it at level 3. It added an optional section on medical device product complaints for drug and device combination products, where the TransCelerate template had medical device deficiencies as a level 3 heading and further medical device information in an appendix. It had no pharmacodynamics heading, which the TransCelerate template had at level 2.
Although the draft used a detailed level 2 structure for the statistical analyses, the information the statistical section covered was generally similar in the two templates. The draft also had a section on protocol deviations that the TransCelerate template lacked. From section 10 onwards the overall information was generally similar, but the TransCelerate template placed it all in section 10 as a series of appendices, where the draft used separate level 2 headings. Those appendices held the TransCelerate template's regulatory, ethical and study oversight considerations, including data quality assurance and source documents, its definitions and procedures for adverse events and serious adverse events, and its clinical laboratory tests, contraception guidance, country-specific requirements and protocol amendment history. The TransCelerate template put the abbreviations directly after the table of contents and its references in section 11.
Authoring to the template
The M11 template is a PDF, which is of no use for creating a document from the ground up, so a sponsor has to make or adapt its own template or use the TransCelerate common protocol template. The TransCelerate template has been around for years and is now aligned with M11, so it is a good place to start. With the specifications in these templates, the data mapping is already established, and it enables content exchange. Authoring to the template benefits every sponsor, even a large one or a small biotech writing its first protocol.
Automating M11 is costly, and the electronic data exchange it brings calls for a large-scale, highly expensive technical infrastructure. Large pharmaceutical companies and TransCelerate members have already adopted it, because they have the resources and the infrastructure. A company or client without those resources can still use the template, and so be ready to work with the digital and AI-enabled tools that are coming. For now a Word template, TransCelerate's or the sponsor's own, still works, but the thought process behind it needs to change.
In October 2025 M11 was seen as potentially an important change in how protocols are authored: in the process, in who is involved, not only the contributors but also those who sign off and approve, who are best involved as early as possible to avoid late-stage reviewers adding more comments, and in the technology used during development.
From scattered narrative to defined content
Legacy protocols and templates are narrative-heavy, repetitive and redundant. That worked 20 years ago and no longer does, because an industry full of complex adaptive designs and master protocols cannot afford the redundancy and error rates that come with copy-pasting between documents. A deliberately exaggerated example of the old style scatters the information: the objective sits in the primary endpoint section, the drug is named, some values sit somewhere in the middle, the endpoint is buried in the middle, and the end of the study is confused with a time point during it. The downstream effects are inconsistent wording, misaligned analyses, version-control problems, and more quality control and more revisions.
In the structured alternative, the guideline and its technical specifications say where each piece of information is allowed to live, and it is put in that one place and nowhere else. The primary objective goes into what the author sees in the template as a familiar free-text box, filled in according to the specifications in the guidance. Page 74 of the technical specifications carries the guidelines for the primary objective, which many writers will recognise as what they were already supposed to do, except that the guidelines now specify the requirements. The writer moves from writing out a primary objective in a free-text box to the reuse rules the guidance specifies. The objective is defined once, given a unique identifier and used automatically everywhere, with the same definition and the same information. It is then no longer free text buried in a paragraph but a tagged structured data element, an object identifier (OID) populated according to the rules in the guidance. The principle is write once, use many, so that copy-paste errors end.
Structuring a protocol is not converting a word-processor file or a PDF into data. The protocol is built as structured content using the defined meanings, definitions and relationships, so that the content can be reviewed, rendered and reused electronically, and it becomes an electronic source from which information is propagated up and downstream across the document lifecycle. A relationship is being established among CDISC, common terminology and definitions, and digital data flows, and documentation can be structured, traceable and inspection-ready from the beginning. In a fully enabled system a PDF of the protocol can still be printed. Some day a protocol might be configured rather than written, and until then writing must respect the digital data flows that are coming.
In March 2023 Digital Data Flow, a CDISC and TransCelerate effort, was looking to convert the structured content of a digital protocol into downstream activities with no or minimal human intervention, so as to reduce the start-up time of trials.
Asked in October 2025 how protocols had changed, experienced protocol writers described a move away from verbose protocols towards lean, concise documents of shorter, briefer statements, with much less copy-paste and more cross-references to other documents. What used to be an all-in-one document had become a network of related information, a transition from a document form to a content form.
One writer holds that experience of the EU Clinical Trials Regulation may call for revising the process or structure by which protocols and amendments are developed. A protocol needs structure to define clearly the parameters for conducting a trial, but many trials are now global, so how best to accommodate region-specific changes and requests from health authorities quickly needs considering. Under the Regulation, requests for information can come from several member states and need responses, and often protocol amendments, within a couple of weeks. In place of creating and maintaining country-specific protocol amendments for each change, the protocol could perhaps exist in an electronic format with version control, with its country-specific elements tagged and the view switched to the region where the investigator is conducting the trial.
Transparency readiness
Transparency and disclosure are an essential part of the document work. A protocol created prospectively as redaction- and transparency-ready is standardised from the beginning, so that confidential information is much easier to redact without breaking the operational logic of the document, and disclosure is no longer an afterthought.
Aligning the protocol with the report
ICH developed E3 in 1995 and E6, on good clinical practice, in 1996, some 30 years before M11, the guideline for the protocol. So what the output should look like was known before what goes into it, and the older guidelines often led to a mismatch between the reporting requirements of the clinical study report and what the protocol captured. With M11 the protocol can be aligned at the foundational stage with the end-game expectations of the report. M11, E3, E6 and E8 together bring the foundational protocol, the ethical principles, quality by design and transparent reporting into one package.
The link between the protocol, the statistical analysis plan and the clinical study report, backwards and forwards, is critical, because what was not planned for collection, or not collected, cannot be reported. So M11 is the scaffolding for the eventual report and not only a tool for authoring the protocol, and the protocol is best written forwards by thinking backwards from the report. With M11 the data-centric position can be considered at the very start of the process. A test follows: if the protocol is being written now and the report were due tomorrow, would the protocol support the analysis and transparent reporting? An imperfect analogy with school science is to start from the hypothesis, design the experiment and work backwards to prove it: claim it, prove it, design it. Seen this way, the structured component of the protocol becomes undeniable. Strategy and logic come first and the content is crafted afterwards, starting with the end in mind, and the hoped-for gains are fewer review cycles, faster protocol generation and fewer copy-paste errors.
Template compliance and judgement
Template adherence alone never ensured compliance, clarity or traceability. A protocol can be technically correct and fully compliant yet scientifically worthless, and the intellectual framework makes the difference. With M11, data centricity is the goal and compliance, clarity and traceability are built into the process. The M11 framework treats protocols as structured source documents, and it reduces variability, improves quality and lays a foundation for efficient data exchange and reuse. Medical writing judgement is still required, not only adherence to a template. M11 does not make a protocol better; it makes its weaknesses harder to hide. Human critical thinking is a primary hallmark of the profession, and the collective expertise of medical writers remains critical in a structured, data-centric environment. One experienced writer concludes that structured protocol templates do not replace scientific judgement and critical thinking. For medical writers the value of M11 is the harmonisation of thought as well as of structure, along a chain in which structure brings clarity, clarity supports traceability, traceability underpins regulatory trust, and regulatory trust facilitates interactions and brings treatments to patients.