In life sciences every approved sentence is a regulated artefact. A pharmacovigilance update in the Company Core Datasheet has to land cleanly in every downstream label: EU SmPC, US USPI, every local pack insert. Manage that in documents and the drift only surfaces in an audit.
Most life-sciences organizations manage labeling, study documents, SOPs and CMC content as finished documents: Word files, PDFs, manuscripts. A change to one safety statement updates one file. The same statement in twenty-two other documents is left to find by hand.
That is not a quality problem. It is a structural one. When the regulator's question arrives (who approved this, and when), the answer takes days, because the content was never structured to be retrieved.
None of these standards prescribe a tool. All of them assume that when the question comes (who approved this, and when?), the answer is already structured and can be retrieved without manual reconstruction.
DitaExchange treats each approved fragment as a component with its own identity, version and approval record. The underlying case is older than the audit pressure; we set it out in The Value of Structured Content Management in Life Sciences.
| What changes | How |
|---|---|
| The approved fragment becomes the unit of management | A safety warning, a dosing paragraph, a contraindication, a stability statement: each carries its own identity, version and approval record. |
| Documents are assembled, not maintained | DITA content maps assemble labels, study documents and SOPs on demand from approved components. Nothing is copied forward by hand. |
| One safety update reaches every downstream label | A change made once propagates to every document that references it, with the full audit trail captured at the change rather than at the file. |
| Medical writers stay in Microsoft Word | The SMEs who hold the domain knowledge author in Word. DITA structure is applied behind the scenes; there is no new editor to learn. |
| The seven-day EMA window becomes a query | Component-level version history answers who approved this, and when without reconstructing it from rendered documents. |
The document types DitaExchange has handled in production with life-sciences customers, each assembled from components rather than maintained as a finished file.
Protocols, synopses, study reports, investigator brochures. Reuse across studies in the same program. The interplay between TransCelerate, FDA/NIH and ICH M11 protocol templates matters here. Dx5 holds those template structures as content maps.
Company Core Datasheet (CCDS), Company Core Safety Information (CCSI), Summary of Product Characteristics (SmPC), United States Prescribing Information (USPI). One safety update, one approval, every downstream label refreshed.
Kept current and auditable, with the change history needed to respond to EMA notifications within the seven-day window.
Common task content reused across SOPs, with a single source of truth for shared instructions.
Risk Management Plans (RMP), Periodic Safety Update Reports (PSUR), with intelligent reuse between related documents.
Subject matter experts (the medical writers, regulatory specialists and clinical scientists who hold the domain knowledge) author in Microsoft Word. The reuse and audit posture work without asking them to learn a new tool.
Owns the submission timeline and the audit answer. Component-level approval and fragment-level history make that answer queryable.
Change-control at the fragment rather than the document, with 21 CFR Part 11 and EU GMP Annex 11 alignment and snapshots for inspection.
Assembles labeling, IFUs and patient information from the same approved components, so one wording change reaches every market and language that carries it.
Dx5 runs inside the Microsoft 365 tenant IT already operates. No new identity provider, no new data-residency conversation.
One labeling program usually touches all four. The component model is what lets them work from the same approved source.
Labeling, study documents and SOPs are written by the scientists closest to the work and have to hold up before a regulator years later. Dx5 keeps them as approved components, with the audit trail captured underneath.
See how subject matter experts work
Danish medical-device manufacturer (transcranial magnetic stimulation). Product documentation across regulatory submissions.
QMSR replaced 21 CFR Part 820 and aligned US medical-product quality systems with ISO 13485:2016. The shift is evidentiary rather than editorial. Traceability across clinical, manufacturing and post-market content is expected to exist at the moment content is approved, not to be assembled afterwards from mailboxes and shared drives. In practice that means the approval record, the approver and the applicable regulatory driver sit on each component.
Article 50 sets transparency obligations for AI-generated content, and AI-assisted drafting is now common in protocol and labeling work. The obligation lands on the record rather than on the prose. Where approval is captured on the component, with the person who cleared the wording named and dated, the question of what a model produced and what a human approved has an answer that holds up in an inspection.
No. M11 is the International Council for Harmonisation guideline finalizing a harmonized, structured template for clinical trial protocols. It defines the template, not the system that holds the text. The gain arrives when the standardized sections behave as governed components that carry their own approval record and are reused across studies in the same program, instead of being copied out of the last protocol and edited in place.
IDMP defines the master data for substances, products and packaging. SPL is the FDA format for labeling content that references it, and eCTD is the submission envelope that carries the dossier. At larger sponsors the IDMP master data usually sits in a dedicated regulatory information management system, with the CCMS holding the narrative and labeling components. If what you need is the master-data system of record and the submission gateway, that is a different purchase.
No, and it is worth being plain about it. Part 11 governs electronic records and signatures, and an electronic signature for sign-off is not a claim we make. What the platform records is component-level approval metadata and version history inside your own SharePoint, with snapshots that fix an approved state as a version. If a validated signature step is a hard requirement, scope it separately against your own procedures.
Most life-sciences implementations begin with a single labeling program, a single PSMF, or a single clinical-study suite, and grow from there. The infrastructure scales with the program.
