Skip to content

What is a Document

Status: Request for collaboration, Decision
Date: 2025-05-07
Governance: Collaboration to be confirmed, TDA, TDAG?

Situation - Context and Problem Statement

There is inconsistency and confusion across teams and systems regarding the definition of a "document" within digital health services. This impacts communication, integration, and data governance.

Background - Decision Drivers (Optional)

Clinical documents are a fundamental part of a patient’s record, capturing key information about care events, decisions, and outcomes. They are used for communication between professionals, legal recordkeeping, and continuity of care.

Documents can vary in formats—from scanned paper records and PDFs to structured digital notes generated by clinical systems. Despite their varied formats, these share common traits: they are authored, versioned, time-stamped, and intended for long-term reference in the clinical record.

There is ongoing ambiguity across systems, teams, and stakeholders about what constitutes a clinical document, leading to inconsistent handling, impact on storage, misclassification, and poor user experience in accessing patient records.

By clarifying what is a document makes them distinguishable from other clinical artefacts such as, but not limited to, clinical data, images and results. This definition is essential for effective governance, maintenance, storage, metadata tagging, information lifecycle management, and ensuring the right information is available at the right time for care decisions.

Assessment - Considered Options

A document should be defined not just by its file format (e.g., PDF, CDA, DOCX), or by its type (clinic letter, discharge letter, clinic note) but by its function, intent, and context within clinical and administrative workflows. A true document represents a discrete, authored unit of information that:

  • Has clear authorship and creation metadata
  • Is version-controlled and auditable
  • Is bound to a specific clinical or administrative event
  • Is intended for long-term storage and retrieval
  • Is readable, shareable, and legally defensible as a snapshot of care or decision-making
  • Can be indexed and discovered via metadata and search

Documents serve a critical role in continuity of care, legal accountability, information governance, and clinical safety. They offer a formal, persistent record designed to be understood by others (clinicians, patients, auditors), often across organisational boundaries.

However, it is equally important to define what a document is NOT:

  • It is not a discrete data point: individual measurements like heart rate, lab results, or blood pressure readings are data—not documents.
  • It is not a real-time stream: telemetry feeds or continuously updated dashboards are transient, not persistent.
  • It is not an image: diagnostic images (e.g., X-rays, CT scans, echocardiograms, ultrasound) or media files that may be referenced by documents but are not themselves documents.
  • It is not raw, unstructured notes without context or authorship (e.g., free-text comments in forms or internal system logs).
  • It is not UI output or screen printouts unless preserved in an authored, versioned, and clinically contextual format.
  • It is not temporary or incomplete drafts: unless formally authored, signed off, and stored as part of the health record.
  • It is not an active form, that is structured clinical data and should be standardised and maintained in a clinical data repository. When a form is finalised and made into a pdf it can be a classed as a document and stored in a document repository.

Without this functional definition and boundary-setting, we risk storing unnecessary or incomplete information, leading to:

  • Information overload in the patient record
  • Search and classification issues
  • Legal and governance gaps (e.g., storing unapproved or draft content as part of the official record)
  • Inconsistent clinician experience and trust in the system

A shared understanding of what a document is, and is not, will help inform system architecture, clinical workflows, information governance, and retention strategies across NHS Wales and beyond.

Recommendation - Decision Outcome

Adopt and operationalise a definition of a clinical or administrative document across systems, teams, and governance processes to ensure consistency, safety, and interoperability. This definition should guide system design, information governance, user training, and record-keeping practices.

A document should be defined as:

A discrete, authored, version-controlled unit of clinical or administrative information that is contextually bound to a care event or activity, intended for long-term storage and reference. It should be retrievable, legally defensible, and include sufficient metadata (e.g., author, creation time, document type, subject, version control and encounter context).

Examples include:

Discharge summaries

  • Clinic letters
  • Consent forms
  • Care plans

To ensure clarity and prevent misclassification, it must also be stated that a document is NOT:

  • A single data point (e.g., a lab value or BP reading)
  • A dynamic feed or real-time telemetry
  • A medical image (e.g., X-ray, scan, ultrasound, echocardiograms)
  • Raw or un-contextualised text notes
  • Temporary drafts or internal UI outputs
  • A file lacking authorship, provenance, or clinical context
  • An active form, as that is structured clinical data.

Action Steps:

Embed this definition into technical specifications, governance frameworks, and clinical safety cases

Align with national and international standards (e.g., Document metadata standard, IHE XDS, FHIR DocumentReference)

Train system designers, product teams, and clinical staff on appropriate classification and usage

Ensure search, lifecycle management, and metadata tagging are aligned with the above definition

Exclude non-document artefacts from document repositories (e.g., data points, logs, raw media, images, active forms, etc)