Developing an integration scorecard

A model for understanding and measuring progress towards health and social care integration

Published: November 2017

In April 2017 the Department of Health commissioned SCIE to undertake research and stakeholder engagement to:

  1. identify an overarching framework for what good health and social care integration looks like, and
  2. propose a set of metrics for measuring progress towards the government’s ambition of full integration.

This takes forward SCIE’s earlier research about the Integration Standard, and creates a framework for action.

Executive summary

A national policy objective over many years has been the integration of health, social care and related services. Expected to deliver improvements in care and support, along with better experiences and outcomes, in recent times this objective has been pursued through initiatives such as integration pilot schemes, the New Models of Care Programme and the Better Care Fund. Underpinned by a government intention that by 2020 integration should be achieved across the country, discussions began on how progress could be best measured, monitored and shared to help encourage and embed integrated approaches in very different local environments.

Research and consultation carried out in 2016 by the Social Care Institute for Excellence (SCIE) for the Department of Health (DH) concluded that an Integration Standard and associated metrics should be driven by people’s experiences and outcomes, and that it should include a focus on leadership and partnership approaches as well as the processes leading to integration of services and care. In its Integration and Better Care Fund Policy Framework (March 2017), the Government confirmed its plans to develop a wider integration scorecard incorporating these points. Other recent developments in respect of the integration agenda have been the use of a performance dashboard with metrics centred on the health/social care interface (July 2017), and the start of DH commissioned Care Quality Commission system reviews in areas defined as having particular issues around Delayed Transfers of Care (September 2017).

In April 2017 SCIE was asked to carry out further research. Comprising additional desk and case study-based research and engagement with service users, carers and other stakeholders through interviews and workshops, the aims were to:

Logic model for integrated care

We created a logic model to illustrate what good health and social care integration looks like. The model depicts visually how a fully integrated system might be structured and how it might function, what interventions and services it might consist of, and the outcomes and benefits it would be expected to produce, especially for people who use services and their carers.

The logic model has four main components:

Stakeholders who contributed to developing the logic model suggested it would be a useful tool for local planning and performance monitoring. We encourage you to use the logic model for this purpose, and to provide us with your feedback. We plan to create a suite of support materials and links to integration resources in line with the logic model's content. We will continue to refine the metrics for measuring the logic model's outcomes through pilot testing and evaluation. Details about the metrics under consideration are included in our full report.

Conclusions and recommendations

This research has shown that despite the enormous challenges local areas face in integrating health and social care, a strong appetite exists for tools that help local areas think about how they deliver and measure better integrated care and support.

This research sought feedback on a framework for understanding what good integrated care looks like, in the form of a logic model, and a set of aligned metrics for measuring progress on integration. We received positive and encouraging feedback about the logic model. Stakeholders commented that the model presents an understandable visualisation of what enables and supports good integrated services, and that it would enable local areas to develop plans that are supported by evidence. By presenting the whole system in this way, the logic model moves firmly away from the limitations of the original Integration Standard, which emphasised structural changes. The grouping of outcomes into three sets – people’s experience of care, service integration and whole system – also received stakeholder support. Logic models, by their very nature of being a concise visual aid, can never be fully comprehensive and reflect the huge variations of approaches in different local areas; but most stakeholders felt that this model provided a useful starting point for discussions on local plans. There are many ways in which this model could be used, including as a planning tool to help local areas prioritise where they need to focus efforts.

The proposed metrics received much more mixed feedback. The measurement of progress on integration has been a goal for many years, and there is some frustration in the sector that greater progress has not been made, especially in developing measures that work with the Think Local Act Personal and National Voices ‘I’ Statements. Because of the limitations associated with current datasets, the metrics seem overly focused on acute care, and they inadequately capture the quality or outcomes of integrated services in primary and community settings. The greatest gap that needs to be filled is measuring how people experience integrated care. Policymakers’ plans for using the proposed metrics – as an integration scorecard for judging progress towards integration or supporting improvement or for both – need to be clarified and resolved, as this will facilitate their acceptance and create support for further research and development.

Sector feedback also confirmed widespread support for the goals of integration. However, the achievement of fully joined-up care was viewed as a lengthy journey, characterised by a process of complex changes and best judged by better care outcomes rather than a fixed end point, such as that expressed in the ambition of Integration 2020.

Three sets of recommendations and next steps are proposed:

  1. In recognition of its positive reception to date, publish and disseminate the integration logic model as a tool for supporting health and social care integration efforts in local areas. Policymakers should no longer use the previous Integration Standard as a reference point, although it was a useful starting point for discussions about integration. A welcome clarification includes recognising that the model focuses on older people, but that it can be adapted to fit for other groups of service users and local circumstances. Additional actions are suggested:

    • Policymakers need to clarify how the logic model (and metrics) will be used – for performance measurement or improvement or both.
    • Policymakers need to clarify the geographic footprint for a local area to support accountability for delivery of whole system integration, e.g. the footprint of a Health and Wellbeing Board.
    • Further work on the logic model is suggested, allowing adaptations and modifications to ‘future proof’ it and for its use with different population groups, from people with learning disabilities to mental health users, etc., to including different sectors, such as housing.
    • The logic model, and indeed the metrics, might be revised further to take account of relevant learning from the current CQC reviews.
  2. Further research, testing and evaluation of the metrics are recommended before the metrics are used nationally

    • Using existing performance data, we recommend the metrics be piloted with a sample of local areas, selecting a range of areas considered good, average or performing less well. We would suggest the following evaluation questions: Do the metrics align with or confirm what we know about each area? Or do they present a different picture and, if so, what does this tell us about measuring progress towards integration?
    • Additionally, it would be helpful to hold discussions with a small set of local area leaders, probing both the utility and acceptability of the metrics: Do the metrics enable local leaders to self-assess their progress and to identify what’s working and where attention and support are required?
    • Similarly, it would be helpful to establish: In what ways can the proposed metrics be combined with local metrics to provide a more rounded picture of local progress? Are local metrics better able to capture the complexity of integration? Are there consistent sets of local metrics being used?
    • Other metrics need developing. For instance, should there be metrics for other parts of the logic model, such as measuring the quality of service provision, and not just the outcomes? How could some of the enablers or components of care be measured?
    • National investment is necessary for the development of user experience metrics, which measure the experiences of integrated care.
  3. The logic model and metrics should be used as a framework for improvement support and resources .

    • At the stakeholder events in particular, we asked for feedback on how the logic model and integration metrics could be used to support improvement.

      Participants commented on how the model could be used as a prioritisation tool; a signposting tool; a self-assessment and planning tool; and even a benchmarking tool. To advance the model as a roadmap, the following actions are suggested:

      • The development of guidelines or a checklist that sits alongside the model to help people apply and use it appropriately.
      • Embedding hyperlinks in an interactive version of the logic model, e.g. to research evidence, good practice case studies, available support, checklists, next steps, and so forth.
      • Offering guidance and support on how to link health and social care data, as well as data from allied sectors, to understand how well a system is integrating over time.