The protocol stops being a documentand becomes data.

That is the Digital Data Flow blueprint the largest sponsors agreed on. AccuraTrials is a full public supporter, and builds on its output rather than around it.

Request a demo See the products
L4Execution

AGE ≥ 18 runs as a check at entry, a randomization rule and an analysis population

Deterministic rules
L3Study model

Eligibility criterion I-01: age at screening ≥ 18 years, cited to p.47 §5.1.2

CDISC USDM v4.0
L2Ontology

Age, as DM.AGE, in years: one concept, whatever the protocol called it

SDTM · CDASH · CT
L1Source

“Subjects must be 18 years or older at screening.” Protocol, p.47 §5.1.2

ICH M11 protocol
One eligibility criterion, from the protocol's sentence to the check that runs. Specimen protocol.

Four layers

Execution rests on semantics,not the other way round.

Choose a layer to see what it stands on. Nothing above can assert what the layer beneath cannot account for.

L4 · Execution

EDC, eConsent, ePRO, labs, safety

Deterministic rules generated from the layer below run the study and capture its data.

Rests on L3, L2, L1

L3 · Study model

One living, versioned model of the trial

Visits, forms, endpoints, eligibility, dosing and windows: machine-readable, and diffable across amendments.

Rests on L2, L1

L2 · Ontology and semantics

CDISC USDM · SDTM · CDASH · controlled terminology

Every concept binds to a standard definition, so “SBP”, “systolic” and “blood pressure (systolic)” are one thing.

Rests on L1

L1 · Source documents

Protocol, amendments, specifications, SOPs

The evidence. Nothing enters the model without a pointer to a line here.

The foundation: everything above points here

What we build on

Five standards,and what each one is actually for.

Foundation · Open study definition repository

Our base layer: an open clinical metadata repository that holds study definitions and terminology as a graph, not a proprietary schema, so a definition stays portable and inspectable. AccuraTrials runs it inside the platform.

// a study, as OpenStudyBuilder holds it
Study ACT-217
  ─ Arm        "Drug 10 mg"  · "Placebo"
  ─ Epoch      Screening → Treatment → Follow-up
  ─ Visit      V2 · Day 28 ± 3
  ─ Activity   Vital signs → CT C25298

Model · Unified Study Definitions Model

The structure our study model conforms to. Born in TransCelerate’s Digital Data Flow and now stewarded by CDISC: a shared way to say what a study is, so it moves between systems without being retyped.

{
  "instanceType": "EligibilityCriterion",
  "category": "Inclusion",
  "text": "Age 18 years or older at screening",
  "source": "Protocol v3.0 · p.47 §5.1.2"
}

Protocol · Structured protocol template

The harmonised electronic protocol template. As protocols follow M11, the document we read and the model we build converge, so extraction gets more certain, not harder.

  1. 1Protocol summary
  2. 3Objectives and estimands
  3. 4Trial design
  4. 5Trial population
  5. 6Trial intervention
  6. 8Assessments and procedures
  7. 9Adverse events and safety
  8. 10Statistical considerations

Data · CDISC data standards

Collection, tabulation and analysis. Fields are mapped at build time, so data comes out submission-shaped instead of needing a mapping project after database lock.

CDASH · collectVSORRES 128SDTM · tabulateVSSTRESN 128 · SYSBPADaM · analyseAVAL 128 · CHG −6

Vocabulary · MedDRA · LOINC · SNOMED CT · UCUM

The vocabularies that make the model mean something. Events, tests, findings and units resolve to codes, not free text, which is what makes automated checking possible at all.

As written“SBP” · “systolic” · “BP (sys)”One conceptLOINC 8480-6 · mm[Hg]

From concept to ADaM, step by step: how the model reaches the analysis.

The base layer

We did not write the foundation.Three of the largest sponsors did.

  1. Novo NordiskInitiated OpenStudyBuilder, released it as open source, and runs it across its own trials.Initiator
  2. Boehringer IngelheimBuilds and maintains it jointly.Joint contributor
  3. BayerBuilds and maintains it jointly.Joint contributor
Open­StudyBuilderOpen source · the base
What AccuraTrials addsThe machine that runs it, and proves it ran.

None of that infrastructure runs a study. The blueprint describes the trial; something has to execute it and keep the evidence.

Digital Data Flow, 2020 on

How the industrygot here.

  1. Digital Data Flow launchesTransCelerate sets one goal: the protocol as data, flowing into every system.
  2. USDM and a reference repositoryThe concept proven end to end; Novo Nordisk open-sources OpenStudyBuilder.
  3. CDISC stewardship, ICH M11USDM becomes a governed CDISC standard; M11 meets it from the regulatory side.
  4. AccuraTrials runs itFull public supporters of DDF, building on its output rather than around it.

FAQ

Standards,answered plainly.

What is CDISC USDM?

The Unified Study Definitions Model is a standard, machine-readable structure for describing a clinical study: its design, arms, visits, activities and eligibility, so the definition can move between systems without being retyped. It came out of TransCelerate’s Digital Data Flow initiative and is now stewarded by CDISC. AccuraTrials’ study model conforms to it.

What is OpenStudyBuilder?

OpenStudyBuilder is an open-source clinical metadata repository that holds study definitions and controlled terminology as a graph. Novo Nordisk initiated it and runs it across its own trials, with Bayer and Boehringer Ingelheim as joint contributors. AccuraTrials builds on it as its base layer and runs it inside its platform.

What is ICH M11?

ICH M11 is the harmonised guideline and template for a structured, electronic clinical protocol. As protocols follow its structure, the document a system reads and the model it builds converge, which makes extracting the study more reliable.

What is TransCelerate’s Digital Data Flow?

Digital Data Flow is a TransCelerate BioPharma initiative, launched in 2020, to move clinical trials from protocols as documents to protocols as data that flows into every downstream system. It produced USDM and a reference Study Definitions Repository. AccuraTrials is a full public supporter.

Look at a real model,on a real protocol.

Thirty minutes. Bring your own protocol and we will compile it while you watch.

OPEN THE EDC Request a demo

Or email us: jgui@accuratrials.com