---
title: "CDISC USDM Support | Unified Study Definitions Model Protocol Authoring | Trelice"
description: "Trelice supports CDISC USDM, the Unified Study Definitions Model. What USDM is, why it matters, how it relates to ICH M11, and how Trelice produces USDM study definitions."
canonical_url: "https://www.trelice.com/usdm"
---

CDISC USDM

# USDM support: the protocol as data, not just a document

Trelice supports the CDISC Unified Study Definitions Model (USDM). A protocol authored in Trelice is structured data from the first field, so it can be delivered as a USDM study definition that downstream systems read directly, instead of a PDF that people re-type. This page explains what USDM is, why it matters to sponsors, CROs, and sites, how it relates to ICH M11, and how Trelice is extending the same approach to the other documents in the protocol family.

[Book a demo](/contact) [Read the docs](/docs)

app.trelice.com / studies / DEMO-ORGO-THERAPEUTICS-001 / protocol / schedule-of-assessments

![Trelice Schedule of Assessments as a structured visit by assessment matrix, the part of a protocol that USDM represents as a study timeline](/images/app/05-schedule-of-assessments.png)

## What is USDM?

USDM, the Unified Study Definitions Model, is a CDISC standard that defines a clinical study in machine-readable form. It was developed through TransCelerate BioPharma’s Digital Data Flow (DDF) initiative in collaboration with CDISC, and CDISC now publishes and maintains it alongside its other standards. Where a traditional protocol is a long document that humans read, a USDM study definition is a structured dataset that software can read: the study design, its arms, epochs, and elements, the visits and the activities scheduled at each one, the eligibility criteria, the objectives and endpoints, the interventions, and the biomedical concepts that tie planned assessments to the data they will produce.

USDM is delivered as a data model with an implementation guide, controlled terminology, conformance rules, and an API specification in JSON and YAML, so a study definition can be exchanged between systems in a known format. The first version was published in 2022. Version 3.0, released in April 2024, aligned the model with the ICH M11 protocol template, and version 4.0, released in June 2025, extended that alignment to cover the breadth of M11. CDISC followed in 2026 with an implementation handbook on building SDTM trial design datasets directly from USDM metadata.

The purpose is simple to state: define the study once, as data, and let every system that depends on the protocol (EDC build, CTMS, statistical programming, registries, consent, site documents) draw from that single definition rather than from its own transcription of a PDF.

## Why USDM matters

Most of the cost of a protocol is paid after it is approved. The schedule of assessments is re-keyed into the EDC. The eligibility criteria are rewritten into a screening form. The visit structure is rebuilt in the CTMS and again in the site budget. The trial design is re-expressed as SDTM trial design domains at submission. Each transcription is done by a different person, from the same PDF, and each is a chance for the study that runs to differ from the study that was written.

USDM removes the transcription. When the protocol exists as a study definition, the downstream systems read the same arms, visits, and activities the authors approved, and an amendment is a change to data that those systems can detect, rather than a redlined PDF someone has to notice. That is the structural reason the industry’s standards bodies, the large sponsors behind TransCelerate, and the vendors building EDC and study-build automation have organized around it.

There is a regulatory reason as well. ICH M11 was adopted in November 2025 and takes effect in June 2026, and it is explicitly aligned with USDM: the M11 data elements are a subset of the USDM, so a USDM study definition already carries the harmonised protocol content in the structure regulators are converging on. Teams that hold their protocol as USDM data are prepared for an M11 submission without a re-authoring project, and prepared for the next step after that, when health authorities begin to accept and eventually expect structured protocol content alongside the document.

Finally, USDM is what makes protocol automation trustworthy. Generating a consent form, a visit worksheet, or a screening logic from free text means guessing at the structure. Generating them from a study definition means reading it. The quality of every derived artifact is bounded by whether the source is data.

## USDM and ICH M11: how they fit together

USDM and ICH M11 are often mentioned together and sometimes confused. They solve different parts of the same problem. ICH M11 is a harmonised template and technical specification for the clinical protocol document: which sections exist, in what order, with what headings and controlled vocabulary. USDM is the data model that can carry that content, together with the study design details the document describes, in a form systems can exchange.

A protocol can be M11-conformant without being USDM data, in the way a Word file can follow a template. It cannot be a USDM study definition without the underlying structure existing as data. Trelice is native to both: the authoring structure follows ICH M11 (see [native to ICH M11, not retrofitted](/protocol-authoring#native-m11)), and because every section is a structured form rather than text under a heading, the same protocol is available as USDM.

ICH M11 and CDISC USDM compared
| Question | ICH M11 | CDISC USDM |
| --- | --- | --- |
| **What is it?** | A harmonised protocol template and technical specification | A machine-readable data model for the study definition |
| **Who publishes it?** | The International Council for Harmonisation (ICH) | CDISC, developed with TransCelerate’s Digital Data Flow initiative |
| **What does it govern?** | The structure, headings, and content of the protocol document | Study design, timelines, eligibility, endpoints, interventions, and the narrative content, as data |
| **How do they relate?** | M11 data elements are a subset of USDM | USDM carries M11 content plus the design detail downstream systems need |
| **Where does Trelice stand?** | Authoring structure follows M11 natively | The structured protocol is produced as a USDM study definition |

## How Trelice supports USDM

Trelice was built as a structured authoring platform before it was built for any particular exchange format, which is what makes USDM support straightforward. A Trelice protocol section is a form with labeled fields, not a page of prose (see [filling in structured sections](/docs/structured-sections)). The study design, the arms and epochs, the Schedule of Assessments with its visits, windows, and activities, the inclusion and exclusion criteria, the objectives and endpoints, and the interventions are all held as data the platform can read and validate. Those are the entities USDM is made of.

From the approved protocol, Trelice produces a USDM study definition: the structured content mapped onto the USDM model and its controlled terminology, in the JSON form the USDM API specification describes, ready for an EDC build, a CTMS, a statistical programming team, or any other system that consumes USDM. Because the definition is derived from the same source as the Word and PDF exports, the document people read and the data systems read describe the same study.

The protocol lifecycle carries through. Approved protocols are locked, changes travel through a formal [amendment](/protocol-authoring#amendments) with per-field reasons and a frozen, numbered version, and the USDM definition follows the version, so a downstream system can always tell which protocol version it is reading. Golden threads (cross-section consistency checking) keep the facts USDM represents once, such as endpoints and eligibility criteria, consistent across the sections that state them (see [golden threads](/docs/golden-threads)), which is exactly the consistency a study definition needs to be correct.

How Trelice protocol structure maps to USDM
| In Trelice | In USDM |
| --- | --- |
| **Study and protocol record with numbered, frozen versions** | Study, study version, and study definition identifiers |
| **Study Design section: arms, periods, and allocation** | Study design, study arms, epochs, and elements |
| **Schedule of Assessments: visits, timing, windows, and assessments** | Schedule timelines, encounters, timings, and activities |
| **Inclusion and exclusion criteria** | Eligibility criteria with identifiers and categories |
| **Objectives and endpoints** | Objectives, endpoints, and their levels |
| **Interventions and dosing** | Study interventions, administrations, and products |
| **Abbreviations and defined terms** | Controlled terminology and code references |
| **Narrative sections aligned with ICH M11** | Narrative content items mapped to M11 sections |

## Defining USDM for the documents around the protocol

USDM describes the study. It does not yet describe the family of documents every study derives from its protocol: informed consent forms, site visit worksheets, eligibility screening forms, IRB and ethics committee questionnaires, study schematics, site budgets, redacted versions for disclosure. In most organizations those are maintained by hand beside the protocol, and the industry has no shared data model for them.

Trelice is defining USDM-based structured definitions for these other document types, in preparation for a future in which they are exchanged as data the way the protocol is beginning to be. Each generated document in Trelice already starts from the protocol’s structured data rather than from copied text (see [downstream documents](/downstream-documents)). Defining that derivation in USDM terms, so that a consent form’s procedures section, a visit worksheet’s checklist, or a screener’s logic is expressed as a view over the study definition, means the whole document family moves with the protocol by construction, and can be handed to sites, ethics committees, and systems in a form they can read without re-keying.

This is deliberately ahead of where the standards are today. The expectation behind it is the same one behind USDM itself: the protocol stopped being a document first, and the documents that depend on it will follow. Teams that adopt Trelice now get the structured protocol and the USDM study definition today, and the derived-document definitions as they land, without changing how they author.

## What USDM support means for each team

### Sponsors and clinical operations

One approved study definition feeds study build, vendor hand-offs, and the submission, and an amendment is a versioned change to data that every consumer can detect. The protocol development process itself, with writer and reviewer assignment, approval queues, electronic signatures, and the audit trail, stays in one place (see [collaboration](/collaboration)).

### Medical writers

Nothing extra to author. The USDM definition falls out of the structured protocol that M11-aligned authoring already produces, so writers keep writing in forms and fields with regulatory context beside them, and the data model is a result rather than a task.

### CROs and study-build teams

Receive the study as USDM instead of a PDF to transcribe, with the version it came from. EDC build, CTMS configuration, and site budgets start from the sponsor’s approved data, and the derived documents a CRO produces for sites can be generated from the same source (see [downstream documents](/downstream-documents)).

### Biostatisticians and data managers

The trial design, arms, epochs, visits, and endpoints are available as data from the start, which is the input CDISC’s USDM implementation guidance uses to construct SDTM trial design domains, and the consistency checks on endpoints and sample size assumptions run during authoring rather than during analysis.

### Research sites

Visit worksheets, screening forms, and consent documents that agree with the protocol because they were generated from the same study definition, regenerated when it changes, and versioned per site (see [visit instructions](/docs/visit-instructions)).

## USDM questions, answered

### What does USDM stand for?

USDM stands for Unified Study Definitions Model. It is a CDISC standard, developed through TransCelerate’s Digital Data Flow (DDF) initiative, that represents a clinical study definition as machine-readable data rather than as a document.

### Does Trelice support CDISC USDM?

Yes. Protocols authored in Trelice are structured data, and Trelice produces a USDM study definition from the approved protocol, versioned with the protocol, for downstream systems such as EDC build, CTMS, and statistical programming to consume.

### Is USDM the same as ICH M11?

No. ICH M11 is the harmonised template and technical specification for the protocol document. USDM is the data model that carries the study definition, including the M11 content. The M11 data elements are a subset of USDM, so a USDM study definition already contains the M11 structure. Trelice is native to M11 for authoring and produces USDM from the result.

### What is in a USDM study definition?

The study and its versions, the study design with arms, epochs, and elements, the schedule timelines with encounters and activities, the eligibility criteria, the objectives and endpoints, the interventions, controlled terminology, biomedical concepts, and the narrative content of the protocol mapped to its sections.

### Why does USDM matter if regulators still accept a PDF protocol?

Because the cost of a protocol is mostly paid downstream, in the systems and documents that are transcribed from it. USDM lets those consumers read the approved study definition directly, so EDC build, site documents, and submission datasets stay consistent with the protocol and with each other. It is also the data model ICH M11 is aligned with, which is where regulatory expectations are heading.

### Which version of USDM does Trelice target?

Trelice tracks the current CDISC release of the USDM model, implementation guide, and controlled terminology, and updates its mapping as CDISC publishes new versions. USDM 4.0, released in June 2025, is the version aligned to the breadth of ICH M11.

### Does Trelice generate other documents from the USDM study definition?

Trelice generates informed consent forms, visit worksheets, eligibility screeners, IRB and ethics committee questionnaires, study schematics, site budgets, and redacted protocols from the protocol’s structured data, and it is defining USDM-based structured definitions for these document types so that they can eventually be exchanged as data in the same way as the protocol.

### Do medical writers need to learn USDM to use Trelice?

No. Writers author in structured forms with regulatory context beside each field. The USDM study definition is produced from what they write; it is an output of the authoring structure, not an additional authoring task.

**Related:** [Protocol authoring](/protocol-authoring)[Downstream documents](/downstream-documents)[Security and compliance](/security)[How Trelice compares](/compare)[Documentation](/docs)

## See your protocol *as a study definition.*

Bring a synopsis. We will author it in the structured editor, walk through the USDM output, and show how an amendment travels through to the data and the documents.

[Book a demo](/contact)

Still reading?

Fifteen minutes on your own protocol will show the study definition better than the rest of this page. Pick a time that suits you.

[Pick a 15-minute slot](/contact)