Common Child Identifier
Department for Education, pre-Discovery, 2022
Background
Poor information sharing between agencies is one of the most consistent findings in reviews that follow the death of, or serious harm to, a child. Practitioners in health, education, policing and children’s social care each hold part of the picture, and the significance of what one holds often only becomes clear alongside what another holds.
A commitment then going through Parliament, now section 179 of the Health and Care Act 2022, would require the Secretary of State for Education to report on government policy for multi-agency information sharing about children, including the potential role of a consistent child identifier (CCI).
I was product manager on a short cross-government pre-Discovery ahead of that work. The ask was a gap analysis and a proposed roadmap for a cross-government CCI to support safeguarding across children’s social care, health, policing and education, covering user research, technical barriers and requirements, operational considerations such as legal basis and data agreements, and the options for what an identifier could be, including the NHS number and alternatives.
The idea was not new. A 2016 report for the department had found broad support for a consistent identifier, but more work was needed to understand how one could be delivered.
Approach
We ran this as a GDS-aligned pre-Discovery: understand the problem and the landscape well enough to say whether a Discovery is worth funding and what it should look at. No solution.
The team was multidisciplinary: policy, user research, service design, business analysis, data architecture, solution architecture, delivery management and product. I set the scope with the policy lead, ran the working assumptions workshop that framed our hypothesis, and kept a user needs log running throughout so that every recommendation traced back to evidence rather than opinion.
The research approach was:
- desk research across previous reports and studies from 2016 to 2021, more than 900 pages and slides
- a subject matter expert session to build a service blueprint of the children’s social care referral “front door”, showing where identifying a child actually costs time
- a draft survey for social workers, and a proposal for primary research with front door staff, with a research plan to be approved before we contacted any users
- engagement with the Children’s Commissioner’s Office, and attempts to reach existing local initiatives already working on this
We mapped every identifier already issued to children (NHS number, unique pupil number, unique learner number, local authority identifiers, national insurance number, child benefit number) against who it covers and who it excludes. We then tested two options against financial, usability, feasibility, operational, and legal, privacy and ethics barriers, recording a mitigation for each, and closed with a gap analysis.
What we found
Identification is a real cost in real work. Front door staff spend significant time looking for, checking and correcting information, often searching several systems by hand before creating a record, while working to a statutory duty to respond to a referral within one working day. Decisions on risk can be made on partial information.
No existing identifier covers everyone. Each was designed for a different purpose and excludes a group who may need safeguarding: unborn children, children not registered with the NHS, children under 14, children educated privately or at home.
A number on its own shares nothing. This was the finding that mattered most. A consistent identifier gives practitioners confidence they are discussing the same child, but it does not, by itself, give them access to anyone else’s information. Interoperability, data standards and consent do that. A CCI is a prerequisite, not the solution.
Recommendation
We put forward two options: the department generating and owning an identifier built from multiple data sources, or adopting the NHS number as the identifier.
We recommended the first as preferred, with the second to be explored further. The NHS number is the most universal identifier available and has the widest data coverage of any, but it does not reach every child who needs safeguarding, and purpose limitation constrains how non-health agencies could use data held against it. Building from multiple sources covers more children, allows a search on any known identifier where there is a legal basis, and reduces dependence on a single feed over time.
We were also clear about what a Discovery needed to answer next: what information partners actually need to share, what current tools and workflows look like, and what an identifier would have to do to earn its place in them.
Outcome
The pre-Discovery closed with a gap analysis, an options analysis and recommendation, a conceptual data model, a proposed roadmap through Discovery, Alpha and Beta, and a short list of immediate next steps: business case approval, a stakeholder engagement plan, and a research plan signed off before any user contact.
The cross-government work carried on from there. In July 2023 the Secretary of State for Education laid “Improving multi-agency information sharing” before Parliament, meeting the section 179 duty. Its conclusions matched the framing we had set out: adopting the NHS number is a significant part of the picture but only part of it, a consistent identifier cannot improve information sharing on its own, and interoperability between systems is the focus for the next phase.