What the Signature Certifies: Four asks for the revision of DCB0129 and DCB0160

NHS England is revising the clinical risk management standards DCB0129 and DCB0160 for the first time since 2018. The review’s own focus groups have identified artificial intelligence governance, post-implementation monitoring and compliance mechanisms as gaps. This paper accepts those three findings and proposes a specific mechanism for each. It adds a fourth ask the focus groups did not reach: that the pre-deployment signature should certify the receiving environment as well as the system.

Ask 1. Redefine what the signature certifies — the system and the environment receiving it. Deployment risk is treated as capability × culture, and a multiplicative risk cannot be certified one factor at a time; the automation bias and SEIPS literature already measures the environmental term the standards leave unassessed.

Ask 2. Post-deployment monitoring obligations fit for systems that change after sign-off, with a rule on what counts as evidence: dataset shift, measured performance decay, deskilling of the humans using the system, and the limits of the system’s account of itself — including the structural inability of deviation-from-baseline monitoring to flag a person already at high risk at first contact.

Ask 3. A compliance architecture using the statutory powers that arrived in 2025–26: a central register of safety case existence and sign-off status, a named checker, and continuous tamper-evident compliance evidence. Seventy per cent non-assurance under a mandatory standard is a verification failure, not a drafting failure.

Ask 4. Name the qualifying professional registers and state what the signature makes the signatory accountable for. The double-checking evidence shows that mandated verification without independence collapses into a single check with no measurable safety benefit; a revised standard should not reproduce that at organisational scale.

Asks 2–4 propose mechanisms for gaps NHS England has already named. Ask 1 is a proposal for a gap it has not. The requirement-level drafting submitted through the consultation is reproduced in Appendix A so that it can be cited, since consultation responses are otherwise anonymised. Companion to consultation response ANON-4Z5E-5X5H-J, submitted to NHS England 11 September 2026. Contributor attributions in §4, §5 and §6 are scoped exactly as stated in the text. Part of the Heartbeat Framework series.