Final report from the Seminar and Workshop – 3 December 2020

In December 2020 Southern Policy Centre (SPC) and Nquiringminds (NQM) completed a 21-month project – Care Analytics. The project, supported by Innovate UK, explored how social care authorities can use Artificial Intelligence techniques and a wide range of data to guide service improvement, support decision-making and allow for more efficient use of resources. An important output was an analytic tool which used local authority date to provide detailed assessments of the local care users and services. To mark its completion SPC hosted a seminar and workshop with care practitioners and data experts.

The event included speakers from Southampton City Council, NQM and social care think-tank Future Care Capital (FCC). Nick Allott, Chief Executive of NQM spoke about work on domiciliary care his team have undertaken with Portsmouth City Council, as well as the Care Analytics tool. The workshop provided the opportunity for a more detailed discussion about the issues raised by the morning’s seminar.

We had five broad objectives, the event aimed to:

  • Build a bridge between data and practitioner communities;
  • Highlight some examples of data-driven policy making;
  • Understand the challenges in using data to improve services;
  • Showcase examples of service improvement, efficiency gains and financial savings; and
  • Identify fertile areas for further work.

This report summarises our discussions.

The case for data

Local authority colleagues were clear about the importance of using data on the local population and their needs to plan and deliver services. As one speaker put it, “good data helps you take intelligent decisions”.

But information needs to be of good quality and clearly presented if it is to help shape policies and design services. It must be accurate, timely and accessible to decision-makers and those at the service front-line, and paint a meaningful picture of what is happening to gain the trust of Councillors and social care teams. Perhaps most important to service providers is the importance of ‘real time’ data which then helps understand and respond to changing demand or need.

Elected Members, the key decision-makers when it comes to allocating resources and shaping services, must be able to see how their interventions can improve the quality of people’s lives. Work in Portsmouth, for example, had used data on domiciliary care visits to ensure visits by care providers were better managed, addressing a demand failure through a more dynamic approach to commissioning and more accurate charging of self-funding service users. This improved the service, with greater user satisfaction, and gave the Council the ability to charge users accurately for care visits received.

The best councils will use data on their population’s care needs alongside other data on matters as diverse as educational outcomes or economic inequalities to shape services which reflect local priorities. We heard about the example of the Southampton data observatory which draws together a broad range of information on the city’s community. Importantly, it is open to all to see and use that information.

Unfortunately, there are too many examples of where data is not integrated or used well. Policy making based on dry performance data, which may be six months out of date, will not lead to effective responses to ever-changing challenges and needs. Nor will service commissioners or social workers have faith in information which is patchy or inaccurate. The challenge is to bring data and care specialists together to shape systems which provide the timely, accurate data needed.

The problems with data

The sad truth is that most local authority data is, in the harsh words of one speaker, “rubbish”. Data sets can be inaccurate, patchy in coverage and often out of data. Experience shows that those data sets need a lot of ‘cleaning’ before they are usable. Locally the problem often lies with the fragmentary nature of data, when information may be held by several provider agencies in different forms. The fragmentary nature of services often makes it difficult to capture a simple overview of key information.

Nor is there any consistency in how data are collected at a local level: one council’s practices may differ significantly from those of a neighbour. Only national reporting requirements offer a consistent set of information, although that is rarely sufficiently up to date to be of much value in day-to-day management. The problem is exacerbated by the lack of integration between health and social care data, making it hard to shape data-driven policy solutions to cross-sector challenges.

Our seminar discussed the case for national standards on health and social care data, to provide for a more consistent approach to collection and a better guarantee of quality. Colleagues from FCC have been arguing for a ‘digital duty of care’ to set shared ground rules on data standards, interoperability, access and transparency.

One challenge to overcome is achieving the right balance between a centrally imposed national data set and local data sets which reflect local needs/priorities. Complete centralisation will stymie innovation, which is essential if local authorities are to tackle local priorities. On the other hand, absolute local autonomy hampers the creation of robust, reliable data sets which allow a more collaborative approach to social care.

Best practice

The seminar heard about the work Portsmouth City Council had undertaken to tackle demand failures in domiciliary care due to fragmented and inflexible systems. Their approach, which had the backing of the Council’s political leadership, made full use of a wide range of data sets to review systems and processes. It resulted in a better understanding of demand, better use of service provider resources and more accurate charging to users, which in turn increased satisfaction. We heard how social workers were able to take a more dynamic approach to match provision with demand and provide a more user-focused service.

Nick Allott introduced the NQM Care Analytics tool, describing it as a mechanism for gathering wellbeing and outcome data. Building on a wealth of local data drawn from local councils NQM have worked with, the tool gathers the evidence necessary to support change. It offers a means of testing the quality of data, and can provide analysis of both user cohorts and individuals. Importantly, the tool allows councils to examine the impact of interventions on service users’ wellbeing. Nick gave practical examples of how, using councils’ own data, the Care Analytics tool could shape informed, targeted interventions and help optimise delivery to give a high impact whilst remaining cost-effective. He showed how data could suggest new ways of delivering a service, giving significant financial savings.

FCC spoke about their proposals for a Community of Practice in social care, bringing together analysts, researchers, policy makers and practitioners to share ideas, build links and exchange best practice. They suggest that can help to stimulate more effective use of data and analytics in social care.

Challenges for the future

Our discussions highlighted a number of inter-linked issues we need address to ensure social care providers make full use of data to inform policy making and shape efficient and responsive services:

  • Making the case for investment in good data – helping decision-makers appreciate the value that cutting edge data analytics making full use of AI techniques can add by encouraging a greater understanding of the power of good data. Too often their experience is of poorly presented, inaccurate and out-of-date information, but there are examples of best practice others can learn from. The FCC proposal for a Community of Practice offers a good vehicle for building that understanding.
  • Improving quality and consistency of data – the necessity of accurate, timely and accessible data was a theme our speakers returned to again and again. Ideas such as a digital duty of care could assist in the quality of data, but that will require sector-wide leadership, perhaps from organisations such as ADASS or the LGA. We will also need to resolve the tension between sector-wide standardisation and local autonomy.
  • Ensuring councils have the right data skills – councils need the support of good data analysts if they are to realise the potential of information they hold or can access. Those skills are not easy to find and are costly, so they may need to explore options such as shared posts. But they also need to make sure the social care professionals are comfortable working alongside those analysts and can interpret data and so guide a service response. And most of all, the sector needs leadership, from politicians and senior managers, to ensure the necessary aptitude and attitude are within their organisations.
  • Measuring impact, what does good look like? – data is only a tool for achieving better outcomes, allowing more dynamic management and more robust forecasting. It is important that authorities seeking to work with data have a clear understanding of what they want to achieve: more targeted interventions leading to better outcomes.
  • Identifying a priority area we should focus our effort on – discussions at the seminar and workshop identified some key areas where effective use of data could have the greatest impact. Many agreed that preventative interventions, working ‘upstream’ to spot problems and design early interventions, would offer real benefits. Colleagues also suggested we should explore gathering better real-time data from provider agencies and interventions which avoid hospital admissions/readmissions.

What next?

The care analytics project highlighted the potential that gains from using data to inform how we plan and manage social care. There are a wealth of good examples of what can be done, but the sector also faces some challenges in realising that potential.

SPC will continue to work to foster debate and dialogue to help us tackle these challenges. We will look for opportunities to address the challenges this seminar and workshop have highlighted. Please feel free to contact us (Simon Eden – edens@southernpolicycentre.co.uk) if you want to discuss further what can be done. You may also want to contact Nick Allott at Nquiringminds (nick@nquiringminds.com) to find out more about the use of AI techniques, or Josefine Magnusson at Future Care Capital (Josefine.magnusson@futurecarecapital.org.uk) to hear about their ideas on a ‘digital duty of care’ and work to build a community of practice.

Follow this link for more information about how we can use data to find creative solutions.