News Story Five

Linda Vernon from Healthier Lancashire and South Cumbria Integrated Care System describes a vision of a single source of reliable information about local preventative services.

Challenges

Healthier Lancashire and South Cumbria is a large Integrated Care System (ICS) covering a population of 1.8 million people, made up of 8 CCGS, 5 NHS Trusts, two County Councils, two unitary councils and 14 District councils. Expenditure across all public sector partners (including the police and fire and rescue) is over £5.5 billion per annum.

Across this complex geography, it became clear during 2019 to 2020 that there was a risk of duplicated collection of information about third sector service provision. There was a growth in social prescribing, council and health initiatives related to the government loneliness agenda and the broad interest in the role of third sector provision in providing support to help divert and prevent people from needing statutory services. All these were driving independent collection and maintenance of information about local support offers.

Process

Through engagement with the Local Government Association (LGA), Department for Culture Media and Sport and Ministry of Housing Communities and Local Government, it became apparent that adopting the Open Referral UK data standard may provide the framework for this information to be Captured Once and Used Numerous Times (COUNT). Early system and data standard trials were delivered during late 2019 and early 2020 by Digital Gaps, as part of the LGA pilot projects to address loneliness.

These early low-cost research projects were critical to create an understanding and buy-in by senior managers to the opportunities of adopting a data standard. In particular senior leaders within Public Health recognised that adopting the standard would help deliver a true ‘market’ for social prescribing referrals, self-care and self-help, and wider community referrals. In many ways it is a re-iteration of a previous set of initiatives related to Making Every Contact Count. Adoption of the data standard and the aggregation of the information into a single Digital Catalogue, provides the digital capability so that frontline professionals can make every contact count by helping people find other support mechanisms or services in their community.

Throughout the programme, since 2019, there has been a focus on 4 challenges of the work. These are evidenced all the way back through the early loneliness projects in 2019:

collection of the information in line with the data standard
assurance of the information, to ensure that it is accurate
technical consumption of the information from the Digital Catalogue into different directories, software and digital processes (through the recent pilots, the information has been surfaced into 5 different end-user systems)
exploitation of that information by frontline workers to support citizens in searching for and identifying the most appropriate services to meet their needs.

Culturally, the ambition has been that frontline workers continue to use their own software tools or access the same directory that they have always used. The programme is making sure the information that they can search or use is the same set of information that is surfaced elsewhere across the widest sets of organisations; and is the same information that is also populating planned digital tools in the future.

The process is certainly not finished; the ICS consists of a very large number of services and organisations and whilst the infrastructure is in place there are many more frontline worker and self-referral uses that can be addressed. The view is that almost every single contact with a citizen within a community setting could be an opportunity to help that individual feel they are able to easily connect with (or can be supported to connect with) appropriate support group in their neighbourhood.

The pilot model

Development of the vision for how the model operates was a key part of the process in creating buy-in. The important elements of this are:

service providers are supported and given the tools to enter and maintain information themselves about their service offers. Information can be collected using a range of systems that adopt the OR UK standard. The pilot focused primarily on implementation of Placecube Open Place Directory
a federated support function has a role to check that the information is accurate and to carry out some light-touch due-diligence to verify key information about the providers that are requesting permission to enter information about their services
members of the community and frontline workers can ‘crowd-source’ information about services that is sent to the assurer for updates and data entry. This crowd-sourcing process is about identifying new services or errors to existing information
after the information has been assured into the Data Lake or Digital Catalogue, it can then be used across a range of uses, including social prescribing, self-care, existing directories, referral processes and even chatbots and call-handling processes
Digital Coproduction took the open source VidaVia Service Finder demonstrator produced by the LGA in the Loneliness project and, working with frontline workers, co-produced a commercial service finder tool
the service finder tool also includes a translator module, which aligns to the unique reference of the services to hold translations of all services. When the English version is changed, this triggers a request for updated translations. This is configured to also hold Easy Read versions as a translation option
the proposal is that service providers will get the service finder tools for free, to incentivise them to keep their information up to date

Benefits

The early trials and pilots have provided the confidence and evidence that technically the elements of the model can work and that this does create the opportunity to deliver the following benefits.

A more efficient model for collecting information about local community services. The baseline work completed indicates there are currently in excess of 500 separate resource lists or directories across statutory, commissioned and third sector providers. There is an average overlap of 15% of services in each of these lists, thus each service is being maintained an average of 75 times. The estimate is that the new approach will save in excess of £5 million resource effort across all partners.
An improved experience and improved productivity for frontline workers. Their focus can be less upon researching for services and verifying the information and solely focused on understanding their clients better and establishing what additional support they might need to attend the potential services that exist.
Therefore, an improved experience for citizens and clients where they can work in partnership with the social prescriber and frontline worker filtering available services to identify the most appropriate services. Citizens and customers offered feedback that they do not trust information currently and hence often don’t attend what they are referred into or signposted towards.
Ultimately, this is about delivering the greatest impact on public health outcomes. Where people have access to the widest set of information about what is available locally to them and can be near-certain of its accuracy, then take-up of appropriate support will increase. The information will also inform analysis of any gaps or duplication in terms of community offers to target the right services to meet local needs. This will all have a direct impact on improved community health and care outcomes.
A real opportunity for wider innovation and improvement, in particular moving towards self-care / self-referral. The use of chatbots, improved call-handling, speedier referrals to the right service and the opportunity for self-care all require a reliable set of information that holds a mapping of how each service meets needs.

Impact

The pilot has provided the evidence of a positive financial business case and that technically the approach can be delivered. In particular, the following were notable successes or offered valuable learning.

The approach has included the ability for frontline workers to be able to search for both traditional services, as well as on-line options as well as web-app and mobile phone-app services. There are a set of 'apps' approved by the ICS and the work has created wider visibility of these apps which provide valuable support to citizens.
Part of the pilots has tested a ‘crowd-sourcing’ capability that enables frontline workers and members of the community to feed back where there are inaccuracies in the information or where there are new services. The feedback from the members of the community, the Parent-Carer Forums and the Maternity Voice Partnership roles was extremely positive. They are very motivated to play a role in helping maintain a set of information that is accurate; this will help shift the culture to value this input and ensure that valued and appropriate services are in place.
The work has engaged with hundreds of frontline workers. Many of these now understand the benefit of internally curated and assured data compared to current reliance on web-based keyword Google searches. It is only when people trial the approach that they realise that the data standard and the taxonomies adopted enable a much more personalised and targeted set of support to be identified.
There is general recognition of the importance of the assurance model that has been trialled. The value of assurance is in providing confidence to users that the information is reliably accurate. As soon as the information is inaccurate, people will start to go back to maintaining their own lists as it feels easier. The assurance model has not been straightforward but is important in gaining credibility, buy-in and the wider behaviour change.
There has been consideration of people who are digitally excluded and a recognition that the Digital Catalogue offers better potential solutions for citizens who have no access digitally (due to skills or technical constraints). So, for example, the opportunity for the data to be used for chatbots, for SMS texting, but also simply that frontline workers have a more reliable set of information with which to support people. The information can be printed out and hence paper records can be shared – but ones that can be re-printed ‘live’ to ensure the information is accurate, rather than resorting to flyers that become quickly out of date.

What we learned

The pilots provide confidence that the model will work and that the move towards sharing and federating the collection and assurance of information is the right direction of travel. There are however some things that have been learned that need to be addressed. Some of these are recognition of additional opportunities, some are about the cultural changes that need to be adopted to deliver a digital future and others are wider still around the real investment, belief by statutory services and consumer appetite for prevention.

Technically, the Open Referral UK documentation is not in-depth enough to enable a robust understanding for technical suppliers. The impact is that currently implementation is time-consuming. Across HLSC, the estimate is that more than 50 software providers will be adopting the standard in the future; the documentation needs to be more comprehensive, which will simplify implementation.
The precise contractual nature of the interface between the ‘back-end’ systems (that collect the data) and the front-end systems (that publish the data) requires further definition. As multiple ‘front-ends’ have started to consume the information, they are reliant on the accuracy of the collection ‘back-end’. If the ‘back-end’ is not working (downtime or inaccurate Application Programme Interface (API)), then the front-end products will perform badly, in terms of the accuracy and access to the data. The exact contract set-up needs further examples to capture more learning and ensure that ongoing local governance is fairly light-touch but still ensures that things can be resolved speedily.
The model will only work where the demand at the front-end is tangible and clear. It is the ‘aggregation of demand’ that will drive further success as providers will want their information to be visible to all of the processes that are linked into the Digital Catalogue. At this stage, the evidence is mixed about the true level of demand, with frontline workers seemingly indifferent to a more streamlined process and happy to do their own in-depth manual web-research for every case they meet. Without significant adoption, then the momentum will be too slow, the information will not be kept up to date and confidence will wane. The true ambition to work to this digital model remains unproven.
The next phase of work must therefore be led operationally, rather than technically, as it is this commitment to work differently in terms of re-using accurate information that will be fundamental to success.
Some service providers do not want to have their offers so readily in competition and competing with others. They would prefer to have people visit their website only, rather than being listed alongside 3 or 4 potential services, with summary information and a link to each website. Again, this is about adoption; where perhaps 80% of all referrals are made through reference to the Digital Catalogue, then providers will have little choice.
There is a need for the data on what is being searched and how often to be more widely available. A model for making this happen is part of the next phase of the work; this data will continue to strengthen the business case for providers to be inside rather than outside the model.
Service quality is an issue that was much discussed. The Open Referral UK standard includes a set of fields that will allow service quality assessments to be included, for example Ofsted or local kitemark for Social Prescribing Ready service. This can be added by a service provider and verified as part of the assurance process, but this was not a focus for the pilot. 0Valuable insight for commissioners about sufficiency, demand and duplication is an expected benefit when there is a critical mass of Open Referral data. Digital Coproduction have developed an example dashboard that provides sufficiency and demand reporting to support commissioner decisions
The work to date has focused solely on surfacing information for frontline workers. The next phase will look at how information is made available to citizens. This has some potential advantages (there are fewer behaviour and culture issues), but a disadvantage as there will be a need for a simple user interface, alongside a complex set of challenges around communication and promotion.
Through the work, it has become apparent that there is a need for better alignment of the taxonomies that are used (the terms or tags within the model). There are no standard taxonomies, and hence health, social care, the police, housing and others are keen that services should be ‘mapped’ to reflect and be compatible with their own sets of language and existing standards (the Open Referral UK standard offers standard fields for the definition of a service, not the tags and taxonomy that should be used). For example, within health there is SNOMED and for local councils, the best example of common terminology is the LGA needs and circumstances list. There needs to be some work undertaken nationally to provide a model for how these can be inter-linked; the value of the work is primarily to deliver better search capability and the aggregation of management information. If the taxonomies are not aligned, then this will not be realised. HLSC will be undertaking their own mapping as it is required, but where this may have broader value (to save others from doing the same thing), then some central co-ordination may be necessary.