ICH M11 Protocol Template 2026: Which Fields a Study Builder Can Load Into EDC, RTSM and eTMF, and Which Stay Text

TL;DR

Takeaway: M11 is now final at ICH, FDA and EMA, and most of what it standardizes is placement and text rather than loadable values. The ICH Assembly's regulatory members adopted the guideline, template and technical specification under Step 4 on 19 November 2025. [1] [2] FDA's guidance page labels the set Final, May 2026. [3] The EMA Step 5 guideline and technical specification print 11 June 2026 as the date for coming into effect. [4] [5]

Our reading of Appendix 1 counts 575 terms. Only 43 are Value, the pick-list rows, against 244 Data and 284 Heading. Four further cells name more than one class. [5] The schedule of activities is one required component of type table, text or image, not a structured visit model. [5] [2] USDM v4.0 states that it was aligned to the 14 March 2025 public-review text, with alignment to the final specification expected in version 4.1, so check any v4.0 code list against DDF-RA issue 703. [6] [7]

The build pass applies that split to a received protocol. Shell parameters come from the controlled values and the numeric synopsis fields. Visits, contacts and windows come from that one schedule component. Headings and required text stay in the protocol.

Protocol complexity and study-build workload (30 August 2026) sizes a build from registry multipliers. The first-trial module map (18 September 2026) is a purchasing path across data capture, the trial master file, ODM, Define-XML and the TMF Reference Model. This article starts from the M11 template.

What is final, and what legal effect that status has

Takeaway: Step 4, the FDA final label and the EMA effect date are publication status. The legal boundary is the notice and the guideline scope, stated once below.

M11 is three documents. The guideline states the design principles. The template is the format, the headers and the instructions. The technical specification lists the data elements and the exchange attributes, including definition, conformance and cardinality. [1] [2] [8] The explainer of 5 May 2026 says the guideline was adopted for implementation on 19 November 2025. [8]

FDA announced a draft on 22 December 2022 (87 FR 78696) and a revised draft of the technical specification on 6 June 2025 (90 FR 24146). The May 2026 notice says a final draft was endorsed by the regulatory agencies in November 2025 and that the guidance finalizes those two drafts. [9] The guidance page, docket FDA-2022-D-3054, labels it Final, May 2026, and describes the technical specification as harmonized terminologies and standardized data fields for electronic exchange. [3] Final, in that heading, is publication status. The notice, issued under 21 CFR 10.115, states the legal character of the document:

FDA's guidance documents do not establish legally enforceable responsibilities. Instead, they describe the Agency's current thinking on a topic and should be viewed only as recommendations, unless specific regulatory or statutory requirements are cited. [9]

The same notice says this guidance "does not establish any rights for any person and is not binding on FDA or the public," and that an alternative approach may be used if it satisfies applicable statutes and regulations. It says the guidance contains no collection of information of its own and refers to existing collections in 21 CFR part 312 under OMB control number 0910-0014. [9]

The guideline's scope is interventional clinical trials of medicinal products across phases and therapeutic areas: pharmaceuticals, biologics, vaccines, drug-device combination products when being developed as a drug, and, when applicable, cell or gene therapy. [1] Section 1.3 then limits the role of the three documents. Neither the guideline nor the template or the technical specification is intended to specify processes for developing or maintaining a protocol. On content requirements, the guideline says:

They do not supersede or negate other guidelines that establish requirements for protocol content. They neither provide instruction on the development of a well-designed trial nor do they characterise a well-crafted final protocol. Rather, the M11 Guideline, Template, and Technical Specification establish common instructions for placement of content, as reflected in other prevailing guidelines, as well as the technical attributes for interoperable electronic exchange of that content. [1]

Both saved EMA Step 5 covers print final CHMP adoption on 11 December 2025 and a date for coming into effect of 11 June 2026. [4] [5] The cover phrase is "date for coming into effect." Those two files are silent on whether a clinical-trial application lacks validity without an M11 protocol. The row reading below uses the saved technical specification. The EMA Step 5 template PDF sits outside that reading.

A Center for Drug Evaluation notice dated 12 June 2026 was identified and left unread. This article is silent on implementation in China.

How the specification marks a row

The technical specification defines the class cell for every term. Heading is a "section heading including table heading, non-numbered title." Data is "Content such as text, image, equation, table." Value applies "if there is a pick list for the data or when a universal text is required or conditionally required or when optional text is suggested." Conformance is the rule for how the header, the data element or the text will conform. The same page stores the concepts in NCI Thesaurus subset C217023. [5]

The template is the author's view of that cell. Square brackets with grey shading are "variable text modelled as a field with pre-defined valid values (i.e., a pick list) in the electronic manifestation of the protocol." Chevrons mark where text is inserted. Level 1 headings stay in the document: the template says to retain them and to indicate "Not applicable." [2] A bracket can still be conditional or optional. Conformance is what tells the builder whether the protocol has to contain the row.

Appendix 1 of the EMA Step 5 technical specification was read term by term. Page furniture was removed. A term name repeated across a page break was merged until the row closed. A class token wrapped onto an "or" line was read as that letter. Four cells that name more than one class were kept out of the single-class totals. The extract has 575 conformance lines and 575 terms. [5]

Class by conformance in Appendix 1 of the EMA Step 5 technical specification. EClinCloud reading of that file, 28 September 2026. The class cell is a required attribute of each term. The count is that column, and sponsor practice sits outside it. [5]

The four multi-class cells are why a single bucket fails on a minority of rows. Population Diagnosis or Condition, in section 1.1.2, is marked "V or D" and Required. The guidance says to use a MedDRA preferred term or to indicate "other" and describe it. The value cell allows a coded term or text, including "N/A – Healthy" for healthy volunteers, and the row repeats per condition. [5] The planned duration of trial intervention and the planned duration of trial participation are marked "D, V" and Conditional: an integer plus a unit of time, with a short explanation when duration varies. The guidance's examples of that explanation are "event-driven" and "adaptive design." [5] The overview of changes from each prior protocol amendment is marked "Universal text and V, D" and Conditional. [5]

Where the visit grid sits

Takeaway: The Schedule of Activities content row is one Required Data element whose data type is Table, Text or Image. The visit grid is the content of that component.

Section 1.3 of the template has one insertion point, "Enter Schedule of Activities." The instruction, repeated as the row's user guidance, says:

The schedule of activities must capture the procedures that will be accomplished at each trial visit, and all contact with participants (e.g., telephone contacts). This includes any tests that are used for eligibility, participant randomisation or stratification, or decisions on trial intervention discontinuation. Allowable windows should be stated for all visits and procedures. A tabular format is recommended. [2] [5]

The content row is class Data, data type Table; Text; Image, conformance Required, concept C132349. Cardinality is one to one, and the repeating rule allows another copy for each schedule if needed. The template names the case: a separate table may be added for extensive sampling, "e.g., serial PK sampling." [5] [2] One schedule in the protocol is one component. A serial-sampling table is a second copy of the same component. Windows are written inside the table.

Those procedures are collection timepoints. The schedule says when a procedure happens. The inclusion sentence in section 5.2 says who may enter, and an estimand in section 3 says which effect the objective is about.

Trial Schema, section 1.2, is the other single component at the front of the protocol. It is Required Data, type Image or Text, concept C93682, cardinality one to one, repeatable within the section. The template says the figure shows the arms and the participant's path through periods or epochs, including screening, washout or run-in, intervention, and milestones such as randomisation, cross-over, end of treatment, end of study and post-treatment follow-up. Further schemas may be added for a complex trial. [5] [2] Epoch names used as system visits are taken from the schedule table. The figure shows the path.

Which synopsis rows can configure the shell and randomization

The values that identify the study, and the numbers beside them, sit on the title page, in the amendment block, and in section 1.1.2. Treating a Value row as a shell parameter, and leaving a Text row in the protocol, is a reading of class, conformance and the template role of the row. Terms below are printed on the specification cells. The NCI directory dated 19 December 2025 is the production package for this terminology. [5] [10]

Original Protocol Indicator is a Required Value, list C217046, Yes (C49488) or No (C49487). Amendment Scope is a Conditional Value when there is an amendment, list C217047, Global (C68846) or Not Global (C217026). An amendment that applies to all sites takes Global, and the country, region and site identifier fields are then deleted. A single-country trial also takes Global. The country, region or site identifier is a Conditional Value, used for the term Not Global. Sponsor approval is a Required Value of type Date or Text: the date, or where the approval information can be found. The guidance says every version should be uniquely identifiable. [5]

Primary and secondary reasons for amendment are Conditional Values on list C217276. Two of the printed terms are Regulatory Agency Request To Amend (C218490) and Change In Strategy (C218496). The guidance says to pick the closest match, to record a change of primary estimand, endpoints or related measures as a change of strategy, to use Other with a description when nothing fits, and to mark the secondary reason "Not applicable" when there is none. More than one value may be selected, except on the original protocol. [5]

Three further conditional amendment values sit beside that list. Substantial impact, whether the amendment is likely to affect the safety or rights of participants, is answered on C217046, with a short explanation when the answer is Yes. Amendment Scope Enrollment Definition, list C217275, is Globally (C68846), Locally (C41065) or By Cohort (C218489), with a conditional number or percent enrolled. The section a change touches is list C217272, limited to sections 1 through 14, the title page and amendment details. Section 12.3 uses that list for a prior amendment, and the prior-amendment table itself is optional. [5]

Trial Phase is a Required Value on the title page, concept C48281, list C217045. The explainer's identifier example is that list, ICH OID 2.16.840.1.113883.3.989.2.3.1.18. Examples on the row are Early Phase 1 (C54721) and Phase 3 (C15602). A drug or vaccine combined with a device is classified by the phase of drug development. [5] [8]

The section 1.1.2 values a shell can take as coded parameters are these. Lists longer than three terms continue on the same specification row. [5]

Four numbers next to those lists are Data, type Number. Minimum Age and Maximum Age are required integers. A trial with more than one eligible age band records the overall minimum and maximum, and a comment for any excluded range. Number of Arms is a required integer. Where periods have different numbers of arms, the field is the total. Number of Participants is a required integer. The two Value rows say whether that integer is a target or a maximum, and whether it counts people assigned to intervention or people enrolled. [5]

The randomisation-type row

{<Randomisation Type>} is Conditional Text, class Data, concept C223137, present when Randomisation is the selected method. The guidance's examples are simple randomisation and block randomisation, written as a description. A further text row appears when Other is selected. [5] Stratification Indicator remains its own required Yes or No on C217046. Its conformance stays Required when the assignment method is something else. The coded parameter is the assignment method. When that value is Randomisation, the type is the description.

The nonproprietary name of an active control is a Conditional Value, concept C97054. The cell gives WHO INN, USAN, JAN and XEVMPD as examples, repeats for each name used as a control, and says to indicate "Not applicable" when the name is unused. [5] Where the blind differs by period, the schema row is answered for the portion of the trial with the highest number of blinded roles, and the rest goes in section 6.7.3. The open-label case on the blinded-roles guidance is "Not applicable (no blinding)." The committee description belongs in section 11.4. [5]

Which product rows are optional values

The section 6 intervention table carries Optional Value rows, conformance "if the table used." An optional row may be left out. An empty table gives the build no product record. Where the protocol fills the table, the coded cells can seed an intervention record at the cardinality on the row. In this reading, those rows and the section 9.2 event type are the Optional Value cells. [5]

Arm name, dosage strength and dosage level on that table are optional text. A strength written as "10 mg" stays text unless the study build parses it. [5]

Event Type in section 9.2 is an Optional Value, list C217287, if that table is used, repeatable per event type. The first three printed terms are Adverse Event (C41331), Serious Adverse Event (C41335) and Trial Intervention Complaint (C218508). [5] The row classifies events for that protocol section. A safety-database case form is a separate build.

Which required text stays in the protocol

On a Text row, Required means the protocol has to contain the component. The component is stored as text.

In section 3, <Primary Objective> is Required Data, type Text, concept C85826, one per numbered primary objective. The heading tells the author to state each primary objective in natural, nontechnical language. The reuse rule sends that text to the table in section 1.1.1, and the same pattern sends each endpoint from section 3.1, and each secondary objective and endpoint from section 3.2, into the synopsis. [5] Secondary objective text is required. Exploratory objective text is conditional. Population, population-level summary and intercurrent-event strategy are conditional text. A conditional table of estimand characteristics, with the endpoint at a minimum, sits in the secondary and exploratory blocks. [5]

The synopsis table points at sentences whose home is section 3. Copying section 1.1.1 into a second endpoint parameter duplicates a reuse the specification already defines. An estimand attribute left conditional stays empty when the protocol leaves it empty. Turning the endpoint sentence into an analysis dataset, or into a coded endpoint id, is a further choice. The cell is text.

Inclusion Criterion, section 5.2, is Required Data, type Text, concept C25532, one to one on the row and repeatable, numbered consecutively. The number is described as a unique integer and not replaceable. If a criterion is deleted for an amendment, the repeating rule leaves that number unused and calls for no second copy. Exclusion Criterion, section 5.3, is Required Data, type Text, concept C25370, with the same amendment rule. [5] A screening checkbox is a further operationalization of one of those sentences. The visit at which an eligibility test is performed stays a line in the section 1.3 table.

Level 1 headings remain in the protocol, including a heading marked "Not applicable." [2] The specification defines protocol content, and the file that holds the protocol is the trial-master-file object in the pass below. A separate document type for each heading is outside the specification.

Section 2.1 of the guideline says protocol content can be reused in trial management and review, in registry postings, and "for standardised clinical trial data capture." [1] The class cell separates a pick list the shell can load from text the protocol keeps. The explainer says structured elements support automated processing, and it describes an M2 ESTRI objective "to recommend technologies and standards that create measurable time and resource savings for both industry and regulators." [8] That is the stated aim. The explainer gives no measured time or resource figure, and none of the documents read for this article report one.

What USDM v4.0 and the Common Protocol Template already give you

Takeaway: USDM v4.0, released 3 June 2025, states that it was aligned to the CeSHarP public-review text of 14 March 2025 and that alignment to the final specification was expected in version 4.1. The newest DDF-RA tag retrieved on 28 September 2026 is still v4.0.0.

Keeping a pick list, a narrative block and a heading apart is already the working practice. The template's brackets, chevrons and level 1 rules are that practice on the page the author just signed. [2] USDM v4.0 stores narrative as HTML in NarrativeContentItem, and the section order can follow a document structure such as ICH M11. [11] The CDISC page describes the implementation guide as intended for upstream protocol authoring and for downstream consumers of systems such as electronic data capture, clinical trial management and the trial master file, and of documents such as the protocol. [12] TransCelerate's 2026 materials, as retrieved, say member companies aligned the CPT Basic Word edition with the finalized M11 template and technical specification and with USDM, that CPT version 11 aligns with key elements of M11 while keeping additional CPT detail, and that the CPT V011 to M11 Mapping Reference maps CPT version 10 headings to CPT version 11 headings. [13] A direct request for that page returned HTTP 403. These sentences are from a search retrieval of the same URL. The retrieved description is a heading crosswalk between CPT editions.

What those sources leave for the builder is the final-row comparison: which 19 November 2025 rows are the Value cells, which single row carries the schedule, and which terminology differences were still listed after v4.0.

The same fundamentals page says a USDM study design can continue to be represented in the CeSHarP template. [6] The released-version table lists v4.0 on 3 June 2025 and has no v4.1 row. [12] Release v4.0.0, published 2025-06-03T12:22:29Z, calls itself "the final release of phase 4, USDM version 4." [14] The release note closes phase 4. Version 4.1 is the alignment the fundamentals page still names. The tag check is the 20 newest releases in that repository.

DDF-RA issue 703, "M11 Adopted," was opened on 19 December 2025, the date on the NCI M11 readme. [10] [7] On 28 September 2026 the issue was still open, locked, labeled "MOVED TO UGG JIRA," and last updated on 14 April 2026. It is a project record of terminology differences. The load check is that list beside the cell. Compatibility of any single USDM attribute with its matching M11 row is outside what the issue decides.

Against "the previously publicly available version (used to align M11 with USDM v4)," the issue lists the differences a load of v4.0 lists will meet. Preferred terms for primary and secondary amendment reasons "appear to be the same, but the C codes differ (new terms have been created)." Section 1.1.2 has "additional information regarding Randomisation and the new Stratification Indicator element." Event-type terminology in section 9.2 "would be supported by NarrativeContent and NarrativeContentItem" in the current model. The section code list in section 12.3 has been "shortened to just to the top level sections." "M11 uses Yes/No text for Adaptive Trial Design Indicator but USDM stores such information as Characteristics (Adaptive Design present) from which the Yes/No status is derived." Trial-phase terms "are a subset of the CDISC terms with additions," and the issue says that change may reach SDTM trial-design domains. [7]

Those names match the cells already quoted: C217046 for the adaptive Yes/No, C217276 for amendment reasons, C217045 for trial phase, and the top-level entries in C217272. [5] Narrative content remains how a block of protocol text follows an M11 section order. [11]

A build pass on one illustrative protocol

Editorial reading, on an illustrative protocol that is a registered trial in none of the files read here. A form for every Heading row would treat 284 headings, table headings included, as capture structure. The pick lists that can seed the shell are the 43 Value rows, beside four cells that accept a code or text. A USDM v4.0 code list is checked against issue 703, and against the terms printed on the row, before it is treated as the final M11 list. The right-hand column is that reading.

Two mis-loads follow directly from the cells. A form built from the section 5.2 heading has turned required sentences into data-capture items. The visit for an eligibility test is already a line in the section 1.3 table. A randomization setup that expects a code list named randomisation type has read conditional text as a Value. The coded row is the assignment method. When that value is Randomisation, the type is a description.

EClinCloud provides electronic data capture and study build (EDC), randomization and trial supply management (RTSM), and an electronic trial master file (eTMF) as separate products. This pass is an order of work across those three: shell parameters in the study build, assignment method, stratification indicator and the randomisation-type text in randomization, and the protocol document, headings included, in the trial master file. The M11 documents reviewed here do not measure that configuration.

Frequently asked questions

Is USDM v4.0 already the model of the final M11 specification?

The v4.0 implementation guide states alignment to the public-review text of 14 March 2025, with alignment to the final specification expected in version 4.1. [6] The CDISC table lists v4.0 on 3 June 2025 and has no v4.1 row. The newest of the 20 DDF-RA tags retrieved on 28 September 2026 is v4.0.0, and its release note calls that tag the final release of phase 4. [12] [14] Issue 703, opened 19 December 2025 and still open on 28 September 2026, lists terminology differences against the text used to align USDM v4: amendment-reason C-codes, the stratification indicator, trial phase, and Yes/No adaptive-design text stored in USDM as a characteristic. [7] A USDM study design can still be represented in the CeSHarP template, and narrative content can follow an M11 section order. [6] [11]

Sources

1. International Council for Harmonisation. "ICH M11 Clinical electronic Structured Harmonised Protocol (CeSHarP)." Step 4 guideline, adopted 19 November 2025. Sections 1.2, 1.3, 2.1 and 2.2. https://database.ich.org/sites/default/files/ICH_Step4_M11_Final_Guideline_2025_1119.pdf

2. International Council for Harmonisation. "ICH M11 protocol template." Step 4, 19 November 2025. Section 0.3 conventions; sections 1.2 and 1.3. https://database.ich.org/sites/default/files/ICH_Step4_M11_Final_Template_2025_1119.pdf

3. U.S. Food and Drug Administration. "M11 Clinical Electronic Structured Harmonised Protocol." Guidance page, status Final, May 2026. Docket FDA-2022-D-3054. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/m11-clinical-electronic-structured-harmonised-protocol

4. European Medicines Agency. "ICH M11 guideline on clinical electronic structured harmonised protocol (CeSHarP) - Step 5." EMA/CHMP/ICH/778799/2022. Cover: final CHMP adoption 11 December 2025; date for coming into effect 11 June 2026. https://www.ema.europa.eu/en/documents/regulatory-procedural-guideline/ich-m11-guideline-clinical-electronic-structured-harmonised-protocol-cesharp-step-5_en.pdf

5. European Medicines Agency. "ICH M11 technical specification on clinical electronic structured harmonised protocol (CeSHarP) - Step 5." EMA/CHMP/ICH/778800/2022. Cover dates; description of table elements; Appendix 1 rows for the schedule of activities, the synopsis, the title page, amendments, objectives, eligibility, interventions and event type. https://www.ema.europa.eu/en/documents/regulatory-procedural-guideline/ich-m11-clinical-electronic-structured-harmonised-protocol-cesharp-technical-specification-step-5_en.pdf

6. CDISC. "Fundamentals of the USDM." USDM Implementation Guide v4.0, section "Representation of ICH M11 CeSHarP in USDM." https://wiki.cdisc.org/spaces/USDMIGv4/pages/234980516/Fundamentals+of+the+USDM

7. CDISC DDF-RA. Issue 703, "M11 Adopted." Opened 19 December 2025. State open when read on 28 September 2026. https://github.com/cdisc-org/DDF-RA/issues/703

8. International Council for Harmonisation. "ICH M11 explainer." Published 5 May 2026. https://database.ich.org/sites/default/files/ICH%20M11%20Explainer_May2026.pdf

9. U.S. Food and Drug Administration. "M11 Clinical Electronic Structured Harmonised Protocol (CeSHarP)." Federal Register, 22 May 2026, 91 FR 30310. Document 2026-10295. https://www.govinfo.gov/content/pkg/FR-2026-05-22/html/2026-10295.htm

10. National Cancer Institute, Enterprise Vocabulary Services. "ICH M11 terminology ReadMe." Directory dated 19 December 2025. https://evs.nci.nih.gov/ftp1/ICH/M11/ReadMe.txt

11. CDISC. "Unstructured Content." USDM Implementation Guide v4.0, Narrative Content. https://wiki.cdisc.org/spaces/USDMIGv4/pages/234980628/Unstructured+Content

12. CDISC. "Digital Data Flow." Released versions of the Unified Study Definitions Model, and the implementation-guide description. Accessed 28 September 2026. https://www.cdisc.org/ddf

13. TransCelerate Biopharma. "Clinical Content and Reuse Solutions." 2026 CPT materials. Page text retrieved by search on 28 September 2026 after a direct request returned HTTP 403. https://www.transceleratebiopharmainc.com/assets/clinical-content-reuse-solutions/

14. CDISC. "DDF-RA release v4.0.0." Published 3 June 2025. Newest of the 20 most recent release tags retrieved 28 September 2026. https://github.com/cdisc-org/DDF-RA/releases/tag/v4.0.0