Evidence Pack
DCB0129 · version 1.4
DCB0129 clinical safety
Your NHS buyer will ask for a signed clinical safety case. We produce the evidence behind it: the hazard log, the drafted Clinical Safety Case Report, and every control tested on your real product, including what happens when its AI is wrong. Your Clinical Safety Officer reviews and signs.
Evidence Pack
DCB0129 · version 1.4
Illustration. Example values from a fictional project, not a client engagement.
What this route is for
DCB0129 is the NHS England standard for clinical risk management in the manufacture of health IT. It asks you to answer that question with evidence, for every release.
We produce that evidence from testing your real product, including fault conditions such as wrong AI output. Your Clinical Safety Officer makes the clinical judgements and signs the safety case. We do not certify or approve the product.
Each hazard is linked to the control that manages it, the test that proves the control works, and the build it was tested on.
We map who uses your product, in which clinical workflows, and what could go wrong for a patient.
For each hazard, we record the control that manages it and the requirement that describes it.
We test every control on your real product, in normal and fault conditions, including wrong AI output where your product uses AI.
Every result is tied to its hazard, its control and the software version, including defects and re-tests.
What you receive
A fixed 10-day first engagement builds the foundation for the highest-risk area of your product. The complete pack, including the drafted Clinical Safety Case Report, typically follows four weeks later.
Who does what
We run the testing and produce the pack. Your Clinical Safety Officer makes the clinical judgements and signs. Your team gives us the access and input only you can.
| We do | Your Clinical Safety Officer | Your team provides |
|---|---|---|
| Scope and intended use, agreed with you | Evaluates each risk and whether it is acceptable | Clinical and product people for the hazard workshops |
| Hazard workshops and the Hazard Log | Triages every defect for clinical impact | Access to the build and test environment |
| Risk controls, each linked to a test we run | Reviews and signs the Clinical Safety Case Report | Fixes for the defects we find |
| The Clinical Risk Management Plan and safety case, drafted | Carries out the release safety review | The accountable signatory for the release decision |
| A handover pack for your NHS customer's own clinical safety assessment | Running the incident process after go-live |
AI Test Group is not a notified body. We do not certify or approve anything, and we do not determine whether a product is a medical device. The manufacturer's accountable signatory signs the release decision.
DCB0129 applies clinical risk management in the manufacture of Health IT systems. DCB0160 applies clinical risk management in the deployment and use of Health IT systems by health and care organisations. They are related standards with different accountable organisations.
It is the NHS England information standard for applying clinical risk management in the manufacture of Health IT systems used in health and care. It is designed to help manufacturers evidence clinical safety.
No. It is a clinical-risk-management information standard. Evidence work is not a certificate, procurement guarantee, or product approval.
Yes. Where your product uses AI, we supply wrong AI output on purpose and test whether each control catches it. That is the fault-condition evidence DCB0129 asks for, and it is re-run when your model, provider or prompt changes.
No. Your Clinical Safety Officer, or one you appoint, makes the clinical judgements and signs the safety case. We produce the tested evidence it relies on, in a form they can reference directly.
Yes. We map what is useful, establish gaps and assumptions, and scope the evidence work around the product version and clinical-risk-management question in front of you.
A structured clinical-safety evidence record can support the questions raised in procurement and assurance activity. The exact deliverable is agreed around your product and review route.
Each release, and each change to your AI model, provider or prompt, can be re-tested and the evidence re-issued, so your safety case stays current as the product changes.
Typically 10 days · fixed scope
One release and the area of your product with the most clinical risk. You end with an evidence foundation and a costed route to the complete pack.
Typically 4 weeks after
The remaining hazards, controls, tests and documents. You end with the full DCB0129 Evidence Pack, ready for your Clinical Safety Officer to sign.
Every release after that
Each release, and each model, provider or prompt change, re-tested and the pack re-issued, so your safety case stays current instead of ageing.
Timings assume access to your product and test environment, input from your team when we need it, and fixes landing within the window.
A free 45-minute session. We map your highest-risk controls, whether each one has been tested, and the evidence you already have. You keep the map.