1 of 56

Scalable Approach to Vulnerability via Interoperability (SAVVI) Show & Tell

4 November 2021

Shelley Heckman

iNetwork Partnership Director, Tameside Council

2 of 56

�

Welcome & Context

Shelley Heckman, iNetwork

Phil Swan, GMCA

3 of 56

Part 1: The SAVVI Pilots & Outputs

The SAVVI Process: an overview

The SAVVI Pilots

The SAVVI Products

Part 2: Wider Engagement in the National Context

Wider Engagement & the National Context for Vulnerability

Building Relationships: the SAVVI Community

Part 3: What Next for SAVVI

SAVVI Ambitions for Phase 3

Agenda

4 of 56

About SAVVI

There are often fracture points and gaps between services, which people can fall through, so that their needs are not recognised and responded to.
SAVVI is bringing forward data standards, data sharing, and information governance, within a SAVVI model process, that can be applied to ‘find, assess, and support’ people and households over a wide range of vulnerabilities such as

homelessness, loneliness, financial distress, re-offending, school-readiness and so on.

The aim is early identification, and smarter interventions and referrals.

5 of 56

Welcome from Project Sponsor

Welcome from SAVVI’s Project Sponsor

Phil Swan

6 of 56

�

The SAVVI Process: an overview

Paul Davidson, iStandUK

The SAVVI Pilots

Tony Evans, Huntingdonshire District Council

Ruth Little, North Yorkshire County Council

The SAVVI Products

Paul Davidson, iStandUK

PART 1: the SAVVI Pilots & Outputs

7 of 56

�

The SAVVI Process: an overview

Paul Davidson, Sedgemoor District Council/ iStandUK

8 of 56

The SAVVI Process

The steps to find and support vulnerable people.

9 of 56

�

The SAVVI Pilots

Tony Evans, Huntingdonshire District Council

Ruth Little, North Yorkshire County Council

10 of 56

SENA at Huntingdonshire

How we have helped people historically.

Residents start specific application / request processes.

Districts are quite transactional, but handle the crisis of homelessness.

County provides longer running support services.

11 of 56

SENA at Huntingdonshire

How we think we can help people better.

We understand who is vulnerable based on data, residents actions and self service.

Intelligent systems help residents engage with public and third sector services to create a holistic support package.

We provide additional support and intervention to help create this holistic package to high risk households, without waiting for the crisis.

12 of 56

SENA at Huntingdonshire

Maintaining a useful list of services...

...what they are.

...what needs they help.

…"entry" criteria.

… constraints.

An integrated referral process...

...that's not just an email.

...that handles information governance concerns.

… that allows outcome tracking

That is a tool organisations can use themselves...

What we did… (read about it at https://huntingdonshire-project.org)

… user research with vulnerable residents.

… co-design what the service would look like with community groups.

What next…

… we want to progress to an alpha build and rollout of the application.

… we need partners to work with us on that can adopt the solution.

13 of 56

School Readiness at North Yorkshire

Ruth Little, School Readiness Co-ordinator, North Yorkshire County Council

Simon Dickinson, SAVVI Coach, Digital Gaps

Agenda

Context
Identifying Children at Highest Risk of Being Not ‘School Ready’

14 of 56

Context

15 of 56

Context

About 6,000 births per annum across North Yorkshire
Roughly 100 (1.7%) every year identified at highest risk by Health Visitors
300 (5%) per annum at next highest level of risk
2,700 every year eligible for Two Year Old Offer, with about 80% (2,000) take-up
Ages and Stages Health Visitor assessment at age 2.5 identifies that less than 90% of children are hitting development milestones (600 children)
About 1,800 children (30%) fail to reach a ‘Good Level of Development’ at the end of their reception year (EYFSP ‘not ready for school’)
Roughy 500 families per annum (8%) get support through Early Offer

No insight into gaps, overlap between these families

No understanding of what support is offered and whether this has any impact

Also no understanding of the alignment of risk assessment criteria - what data sets used, what decisions are made

16 of 56

Context

17 of 56

SAVVI Process

Purpose - Improve school readiness

identification of families at the highest risk of having children not ready for school

families who are not picked-up by existing assessments

offering the right support to people

in the right way that encourages them to engage
by learning about ‘what works’

increasing take-up of the support offered

by improved insight into which families do and do not take up support

Learning what works - both finding people earlier and which interventions have an impacts

18 of 56

19 of 56

Headline model

Using only 6 data attributes has accuracy of 78% in predicting children who will not be ready for school

Income band / socio-economic
Number of children
Housing type
Ethnicity

Re-use existing approach for data matching / single-view to bring the data together and produce the prioritised lists
Identify highest risk 150 families per annum who are not picked-up by existing assessments
Engage through universal Health Visitor 2-2.5 year old development review
Signpost to the Community Directory

20 of 56

Next steps

Finalise more detailed blueprint
Document the business case
Address the IG

Case study captures the findings

21 of 56

�

The SAVVI Products

Paul Davidson, Sedgemoor District Council/ iStandUK

22 of 56

The SAVVI Playbook

The SAVVI Process
The SAVVI Catalogue
The SAVVI IG Framework
The SAVVI Standards

http://www.savviuk.org

23 of 56

Vulnerability Attributes

Data Attributes	Could be an indicator of ...
UC Benefit Cap / Rent Gap Offending / Anti-Social Behaviour Domestic Violence Police called to domestic incident Mental Health Problems Drug or Alcohol Incident / Support	Potential Homelessness
Living Alone In Arrears Moved Home Recently Bereaved Lost Job Over 70	Loneliness / Financial Distress
Maternal education Household Income band Household number of children Gender	School Readiness
GP Registrations - and who is not registered Multiple GP Appointments Multiple A&E attendance Multiple 999 Calls	New need for Care

24 of 56

The SAVVI Catalogue

which data attributes
from which datasets
have been used to find people with what vulnerability.

25 of 56

The SAVVI Catalogue

8 min video at https://youtu.be/YvQQfatm_VU

26 of 56

SAVVI IG Framework

See the SAVVI IG Framework at the SAVVI Playbook
Embracing ICO Principles

“You must share personal data fairly and transparently.”
“You must identify at least one lawful basis for sharing data before you start any sharing.”
“You must process personal data securely, with appropriate organisational and technical measures in place.”
“You can only use the personal data for a new purpose if either this is compatible with your original purpose, you get consent, or you have a clear obligation or function set out in law.”

Steps to engage with IG professionals to source data for secondary use
Lawfully, Ethically, Transparently
Escalations - e.g. the Digital Economy Act

27 of 56

SAVVI Data Standards

The SAVVI Data Standards

The SAVVI Concept Model
The SAVVI Logical Model
SAVVI message formats
SAVVI Terminology and Taxonomies

Deep-Dive anyone?

28 of 56

The SAVVI Concept Model

Person
Household
Residence
Vulnerability Attribute
Risk Model
Case
Assessment

Need
Event
Delivery Organisation
Plan
Action
Outcome

29 of 56

Deep-Dives to look at the standards

We have done 1-2-1 ‘deep-dive’ sessions with each of

Liquid Logic
ForgeRock
Xantura
Policy in Practice
Genserve
SAP

NHS Digital
CRISIS
Manchester University

Planned

BOLD
GMCA
Scottish Government

30 of 56

Taxonomies

The Local Government Association (LGA) provide code lists, in a data model.
for

Circumstance
Need
Service Type
Organisation Type

see https://standards.esd.org.uk/?

31 of 56

�

PART 2: Wider Engagement in the National Context

Wider Engagement & the National Context for Vulnerability

Patrick Kiernan, Cabinet Office

Toby Hayward-Butcher, Ministry of Justice

Viv Adams, Information Commissioner’s Office

Building Relationships: the SAVVI Community

Georgina Maratheftis, TechUK

John Curtis, Bolton Council

32 of 56

�

Wider Engagement & the National Context for Vulnerability

Patrick Kiernan, Cabinet Office

Toby Hayward-Butcher, Ministry of Justice

Viv Adams, Information Commissioner’s Office

33 of 56

The Digital Economy Act

Patrick Kiernan, Cabinet Office

Finding vulnerable people and households, using existing datasets
Adding a new ‘objective’ to the DeA for ‘Early Help and Prevention’
How to get involved - working with SAVVI

34 of 56

35 of 56

Better Outcomes through Linked Data

Toby Hayward-Butcher, Ministry of Justice

3 year programme from the Shared Outcomes Fund
Four workstreams

Providing support to victims of crime - MoJ leading, signposting to the right services.
Homelessness and Rough Sleeping - DLUHC leading
Combatting the misuse of drugs - OHID (formerly PHE) and PH Wales leading
Reducing re-offending - MoJ leading, a single digital view of an offender

Working with SAVVI

The SAVVI Data Model
The SAVVI Model Process
The SAVVI approach to Information Governance
Building on relationships with other departments, local government, and industry

Toby Hayward-Butcher

Head of Data Strategy

Ministry of Justice

36 of 56

Information Commissioner’s Office: perspective on vulnerability data

Viv Adams

Principal Policy Adviser on the

ICO’s Parliament and Government Affairs Team

37 of 56

�

Building Relationships: the SAVVI community

Georgina Maratheftis, Tech UK

John Curtis, Bolton Council

38 of 56

�

The SAVVI TechUK Group

Georgina Maratheftis, techUK

georgina.maratheftis@techuk.org

39 of 56

Introducing the SAVVI Tech Working Group

Xantura
Gemserv
Liquidlogic
Dedalus
Placecube
SAP
IoT Solutions Group
Civica
Foregrock
Objectivity
Capita
Dignio
RIVIAM Digital Care

A mutual forum for SAVVI and the Tech industry to engage and collaborate on SAVVI data standardS.

Facilitating genuine two way discussion.

A forum for suppliers to share how they propose to use the standards.

40 of 56

�

The SAVVI Information Governance QA group

John L Curtis, Bolton Council

41 of 56

The SAVVI IG QA Group

Information Governance for

using existing data to find and identify people who may be at risk.
collecting and sharing additional data (including secondary use) as people are assessed and supported.

The role for the IG profession during a SAVVI project - capturing the landscape; development of products; propositions to government.
Membership to the IG QA group - dynamic energy and engagement with sectors / subject matter experts.
Workshops e.g.

Health Barriers workshop - listening and understanding local stories - barriers / success and themes.
Feeding back to NHSx - local perspectives.

Products

SAVVI IG Catalogue.
SAVVI IG Framework.

42 of 56

�

PART 3: What Next for SAVVI

SAVVI Ambitions for Phase 3

Paul Davidson, iStandUK

Mike Thacker, Porism Ltd

43 of 56

Let’s Build it

Working with

Councils - e.g. Supported Families at GMCA
Suppliers - e.g. Sharing Data across products and hubs

From Prototypes to Sustainable services

The SAVVI Catalogue
The SAVVI Dashboard
The SAVVI IG Framework

Cataloguing and Unlocking Data

An ecosystem of trusted partners sharing data securely via standards

Collaborating to propose message formats and terminologies

44 of 56

�

SAVVI Platforms for the Standards and the Terminologies

Mike Thacker, Porism

45 of 56

Sharing the standards

Terminology

UC Benefit Cap / Rent Gap

Offending / Anti-Social Behaviour

Domestic Violence

Police called to domestic incident

Mental Health Problems

Drug or Alcohol Incident / Support

Data structures

Terminology server

46 of 56

Identifying vulnerability and relevant types of service

Circumstance

Need

Service type

defines

addressed by

If we start with terminologies. These need to be standardised if different organisations are to understand one another when they describe the circumstances of a person or family, the types of service available, types of organisation, outcomes from interventions, etc. As well as having a joint understanding, we can learn from one another if we’re all referencing the same terminologies.

When addressing vulnerability, definitions of and links between circumstance, need and services type are important.

The Hunts approach (simplified) is: circumstance infers need and need indicates the types of service that might be suitable for a person/family.

LGA lists support this approach. Their standards pages let you inspect terms and mappings, so this shows the “Care at home “ need and the types of services which might address that need.

Lists and mappings can be viewed on screen, downloaded or queried via an API.

47 of 56

Using mapped terms in a service finder

What do you need?

Accessible transport

Adult education

Better health

Care at home

Child care

Child protection

Dealing with chronic illness

Dealing with substance abuse

Digital skills

Disabilities and special needs support

Emotional wellbeing

Employment

Service: Adult Care Plan (1757)

Service: Assisted garden maintenance (147)

Service: Care at home (242)

Service: Community alarms and telecare(313)

Service: Home adaptations and aids (178)

Service

48 of 56

API calls for using mappings

Example mapping

Needs mapped to Bristol services

https://webservices.esd.org.uk/lists/mapping?from=needs&to=services&toTag=Bristol

Example Open Referral UK API query

Services of service type 242

https://bristol.openplace.directory/o/ServiceDirectoryService/v2/services?taxonomy_id=242&vocabulary=esdServiceTypes

49 of 56

Issues

Relies on the LGA’s centralised view
Too much detail for most directories
Designed for just local government
SAVVI considers other types of characteristic
Subject to change for local differences and ongoing learning

50 of 56

The ideal - collaborate & learn from each other

Linked taxonomy (e.g. Needs)

My taxonomy

Taxonomy 1

(eg LGSL)

Selected terms

addressed by

Taxonomy 2

(eg SnomedCT)

Taxonomy 3

(eg Local terms)

Selected terms

Σ

51 of 56

The NHS Digital Terminology Server

“A FHIR^® compliant solution that holds and disseminates assured international terminologies and classifications (such as SNOMED-CT and ICD-10) and national terminologies (such as NHS Data Model and Dictionary codes).”

Code systems

Valueset

Concept map

Linked taxonomy (e.g. Needs)

My taxonomy

Taxonomy 1

(eg LGSL)

Selected terms

addressed by

Taxonomy 2

(eg SnomedCT)

Taxonomy 3

(eg Local terms)

Selected terms

Σ

52 of 56

Local variants code system

Central code system

Value set

53 of 56

SAVVI Ambition - Terminology

Build a “terminology portal” to refine terminologies and mappings through collaboration.

54 of 56

SAVVI Data Models in Mauro

So that’s terminology. When it comes to data structures, the model Paul has described defines different data classes and how they relate to one another. It references existing FHIR data structures where relevant. In practice we need to transfer pieces of information between organisations in a way that everyone understands, however the information is stored by different systems.

There appears to be no government standard repository of standard data structures, but NHS Digital does make use of the Mauro Data Mapper which lets you store definitions of data classes and their elements such that we can define how messages between system conveying information might look.

Again we envisage a collaborative approach letting people record how they work and evolving consensus.

We’ve set up a Mauro server and input part of the SAVVI model to test the principle so we feel ready to put such a server in production.

55 of 56

SAVVI ambition - Data structures

Run a Mauro server for sharing data structures and message formats to make data interoperable.

56 of 56

��Thank you for attending today’s show and tell.

We share at:

http://www.savviuk.org

https://www.istanduk.org/savvi

shelley.heckman@tameside.gov.uk

paul.davidson@sedgemoor.gov.uk�