MAIS Standards Working Group - Meeting 2 - 3rd September 2025

[andjnewman]

Draft Agenda:

1. Review of changes to and outstanding comments on person spec
2. Introducing data exchange
3. Introducing taxonomy/vocabulary

[andjnewman]

Hello,

The next meeting of the MAIS Standards Working Group will be tomorrow, Wednesday 3rd, 3-4:30pm, in Microsoft Teams.

Our aims for the meeting are to finalise version 0.2 of our person specification, introduce our data exchange standard and taxonomy specifications.

To enable the Working Group to focus on outstanding issues this email provides a short context for each specification, an update on decisions we’ve made and issues we have resolved and highlights the outstanding issues we will present for discussion tomorrow. Please give these issues some thought prior to the meeting to enable us to have a rounded conversation.
Person Specification
We’ll be returning to the Person Specification that we shared at our first working group meeting. As a reminder, we are aiming to standardise some of the data collected about individuals in the children’s and adult’s social care system, focussing initially on those fields required to distinguish one person from another, and to identify relationships between people. Its goal is to unblock the use of ‘Single View’ systems, as an aid to multi-agency information sharing.

Decisions we’ve made:
We have updated the spec to accept multiple identifiers (#23 & #34). Not everyone has an NHS number, and we need to facilitate data records where persons are identified by one or more non-NHS identifiers.

We have explored the suggestion that addresses should be separate from core person attributes (#30). Given the usage of address information for the “searching” for a person on a system, we have decided to not do this now, we may revisit this decision in the future.

We have made a change to the specification to enable alternative (or contested) names to be recorded (#29).

More generally, we’ve made some minor changes to the wording of the person spec and the background information provided with it in response to issues #19, #20, and #22. We’ve also answered how we’re aligning (and diverging) from the Common Basic Data Set (CBDS) (#27). We've started compiling this list of standards and initiatives that potentially overlap with or are relevant to this programme.

Issues for decision at the working group:
Our recommendation on the issue around including gender and/or sex in the specification is set out in issue #24 on GitHub. Government policy about the reporting of Gender and Sex characteristics conflicts with our choices about what is and isn’t necessary to describe a person. Importantly, we came to those choices based on analysis of clinical data standards; which are included in our notes on the subject in the thread. We will seek approval of our recommendation at the working group.

Issue (#31) sets out our recommendation for meeting the need to facilitate unknown, estimated, or expected dates of birth. Again, we will seek approval from the working group on this.

We will also share several other issues which we will not be working on imminently but are on our radar.

Data Exchange Specification
We’ll be introducing our proposed data exchange standard. Our initial focus aims to unlock basic data exchange to enable single view systems. It assumes a distributed architecture, which is aligned with DfE’s latest thinking, but will be compatible with the centralised datastores some LAs have already created.

We’ll share some context setting out the assumptions we’re making about what types of exchange we’re enabling in the first instance, and what infrastructure is available to facilitate that, and invite your feedback on those, and see if there is appetite for a deep dive on this specifically.

We’ll then share the draft data exchange specification, asking the group to raise any issues in the session and/or via GitHub as usual.

Controlled Vocabularies
Machine to machine data exchange is dependent on a standardised language, so this is a critical area.

We have commenced the development of controlled vocabularies to support our initial single view use case.

● We need a catalogue or registry of organisations and service providers, so that we can direct requests for information.
● We need to be able to refer to service involvements, which requires an ability to classify services.

We have initial drafts of these vocabularies which we will present for discussion and feedback.
We have also identified a need for a vocabulary to enable data requests to be linked to a Data Processing Agreement, and potentially to specific agreed data sharing use cases which they authorise. We have not commenced development of this yet but will share the problem statement we are working to and invite discussion.
We look forward to seeing you on Wednesday, if you don’t have the meeting invite please reply to this email and we will add you.

Thanks, The Social Care Data Standards Team

[andjnewman]

Minutes of MAIS Standards Development Working Group

Session 2, Microsoft Teams, 3 September 2025, 3pm to 4.30pm

Agenda:

1. Welcome to the working group
2. Intros and expectations
3. Programme Overview
4. Person Spec
5. Introducing data exchange
6. Introducing taxonomy/vocabulary

The meeting was chaired by Andrew Newman (ODI).

Programme Overview

Tom Rintoul (Social Finance) provided a short reminder on the purpose and scope of the programme.

Tom announced that the standards programme, previously focused on children's social care, will now expand to include adult social care, following agreement with the Department of Health and Social Care. This presents a unique and powerful opportunity to transform social care case management.

Tom described that while foundational standards will be relevant to both domains, future iterations may require separate reference or working groups to address domain-specific extensions, such as care coordination for adults or risk assessments for children.

Tom noted that the immediate impact for the current group is minimal, but participants should expect increased involvement from adult social care stakeholders in future sessions, with broader changes anticipated as the roadmap progresses.

Person Specification

Neil Majithia (ODI) presented updates to the person standard, Neil outlined recent changes to the person standard, such as allowing multiple identifiers (e.g., NHS numbers), retaining address for identification purposes, and supporting multiple names.

Neil led a group discussions and votes on the inclusion of gender and sex fields and handling on uncertainty in dates of birth

Gender and phenotypic sex

Neil presented Issue #24 regarding the inclusion of person stated gender versus phenotypic sex, referencing government guidance and sector standards, and proposed making gender mandatory and sex optional fields.

Vote: The group then voted on the proposal with the following result:

Do you accept our recommendation on Gender / Sex?

• Yes - proceed to testing = 14
• Not yet - Proceed to testing = 0
• No - further refinement is required before testing = 6

Group Discussion: Participants shared experiences and concerns about recording gender and sex in discussion and in the meeting chat. This discussion highlighed the emotive nature, system inconsistencies, and the need for consultation with affected groups key points included:

• Need for Clarity: There is confusion about why both gender and sex data are being collected and if both are essential. Some feedback questions the necessity of including this specific data, suggesting it may represent indirect discrimination if other protected characteristics like religion or ethnicity are omitted.

• Mandatory Data: There's a strong argument that sex at birth should be a mandatory data field, not optional. This is seen as essential for aligning with and matching other government datasets, such as school attendance and census data. It's also noted that sex is a fixed, biological data point, whereas a child's understanding of gender may be less defined.

• Complexity and Sensitivity: The feedback acknowledges that gender is a complex and highly emotive topic. It's recognized that gender is not always binary. The suggestion is to approach this topic carefully, with empathy and a recognition of its complexity.

• Importance of Testing: The majority of the feedback highlights the critical need to test the proposed data collection approach. This testing will help to validate or disprove assumptions, expose and resolve potential data-matching conflicts, and provide more information to inform the final decision. The results from testing should be kept under review, with the understanding that the approach can be refined or changed later if needed.

• Need for Further Consultation: A key piece of feedback is that the current group is not the right audience to make this decision. It is strongly recommended to consult with those directly affected by this data collection to understand what truly matters and to avoid making decisions "for" them.

Decision: The group agreed to proceed with testing the proposed approach and to refine it based on the outcomes of this testing and wider policy developments.

Date of Birth Accuracy Handling

Neil and Howard Askew (ODI) discussed approaches for recording estimated or uncertain dates of birth, recommending an accuracy indicator inspired by the Australian FHIR implementation.

Vote: The group then voted on the proposal with the following result:

Do you accept our recommendation on Date of Birth?

• Yes - proceed to testing = 17
• Not yet - proceed to testing = 2
• No - further refinement is required before testing = 2

Group Discussion: Participants discussed the proposal it was noted that the proposed approach may not work with the current data models of some systems. Howard confirmed ongoing supplier engagement to assess feasibility.

Decision: The group agreed to proceed with testing the proposed approach and to refine it based on the outcomes of this testing and wider policy developments.

Future Issue Tracking

Neil encouraged participants to raise further issues or comments via GitHub, non-critical issues will be addressed after the identification use case is tested.

Introducing Data Exchange

Howard, with input from Tom and Andrew, introduced the initial data exchange standard for multi-agency information sharing, demonstrated a prototype API interaction, and led a discussion on matching processes, system interoperability, and the challenges of standardising responses across diverse organisations

• API-Based Data Exchange Approach: Howard described the decision to use RESTful APIs aligned with FHIR protocols for inter-organisational data requests and responses, demonstrating a prototype where a system queries others for information about a person and receives structured responses.

• Matching Process and Validation: Howard and Tom discussed the challenge of ensuring reliable matching across systems, proposing the use of synthetic data for validation and setting minimum standards for matching performance, while acknowledging the limitations of controlling external systems' matching logic.

• Response Structure and Thresholds: Tom elaborated on the need to standardise how systems communicate match confidence, including handling binary and partial matches, and described the importance of describing matching approaches and thresholds in responses for transparency.

• Centralised vs Distributed Matching Debate: The issue of scalability was raised and delegates suggested considering a centralised matching process to overcome the limitations of distributed approaches. Howard inviting further discussion and documentation of this proposal on GitHub.

• Information Governance Considerations: Howard highlighted the need to codify information governance rules in a machine-readable format to ensure organisations can confidently share data, referencing existing guidance and the complexity of local agreements, with plans for a dedicated future session on this topic.

Introducing taxonomy/vocabulary

Howard, supported by Andrew and Tom, introduced the taxonomy work to create shared vocabularies for organisations, services, and data processing agreements, discussed initial scoping and prioritisation challenges, and solicited input from participants

• Taxonomy Purpose and Scope: Howard explained that the taxonomy aims to enable machine-to-machine communication and build trust by establishing controlled vocabularies for organisations, service involvements, and data processing agreements, starting with priority players identified from previous work.

• Organisation and Service Cataloguing: Andrew and Howard described efforts to catalogue organisations and services, referencing existing directories such as LGA and NHS catalogues, and invited participants to help prioritise which organisations to engage, though technical issues with the voting tool delayed this process.

• Service-Provider Distinction: Members discussed the need to distinguish between organisations, services, and providers, noting that in-house and commissioned providers may require different taxonomy levels, and recommended synthesising existing taxonomies from various sectors.

Next Steps

Andrew closed the meeting by outlining the timeline for iterating the person specification, data exchange standard, and taxonomy, encouraged ongoing issue-raising and discussion on GitHub, and invited participants to contribute feedback and existing resources..

Future Sessions: Plans were made for deep dives into taxonomy development and a special session on information governance, dates to follow.

Follow-up tasks:

• Finalise Person Specification ready for testing
• Consider obtaining lived experience input from groups directly affected by gender and sex data recording to inform the person standard.
• GitHub Issue Documentation: Add long-form comments and feedback from the meeting regarding gender and sex data fields and Date of Birth to the relevant GitHub issue thread for further discussion and transparency.
• Assess the impact of implementing the date of birth accuracy indicator on existing case management systems.
• Document and propose a process for verifying matching standards across organisations and publish via GitHub for input.
• Publish taxonomies proposals on GitHub and inform steering group members when these are ready for review.
• Gather and analyse existing taxonomies and catalogues of organisations and services to inform the development of a shared taxonomy for service involvement and organisational registers.