Skip to content

Meaningful measures of integration


The NHS long-term plan confirmed, once again, that collaboration is now the organising principle of the NHS, and that integrated care is the idea on which national leaders have pinned their hopes.

There have been a number of initiatives to integrate services over the past decade, but the pace picked up considerably with the appointment of Simon Stevens as the Chief Executive of NHS England in 2014 and the publication of the NHS five year forward view. If anything will define his legacy when he eventually steps down, it will be the success or otherwise of this.

Up to now, efforts to measure the success of integrated care have tended to focus on whether demand for acute services is reduced as a result of integration and whether costs for acute care are reduced. The evidence for these is inconclusive but focusing on these as the desired outcomes misses the point. NHS England says that for health, care and support to be ‘integrated’, it must be person-centred, co-ordinated and tailored to the needs and preferences of the individual, their carer and family. So, who is best placed to tell us whether integration is working? It is users themselves. What impact does integration have on users’ experience of services? Is their quality of life improved? The problem is we don’t have the right measures in place to answer these questions. This is a significant gap.

The health and social care system has an extensive survey programme, measuring people’s experiences of a range of services from the GP patient survey and the inpatient survey to the adult social care survey. The programme hears from over 1 million people a year, but while they measure experiences of individual services, they provide limited understanding of how well (or otherwise) these services are working together. Given we know that many of the problems people face are because services are not always joined up, we need to come up with new ways of measuring and collecting this type of data. If the system is transforming, then so should the ways we measure it.

This issue is recognised in the NHS long-term plan which outlines the need for a new ‘Integration index’, the aim of which is to ‘measure from patients’, carers’ and the public’s point of view, the extent to which the local health service and its partners are genuinely providing joined up, personalised and anticipatory care.’ It seems there is a real opportunity to put this right.

With this in mind, The King’s Fund, together with Kaleidoscope, held a workshop earlier this year bringing together a group of people to come up with potential solutions to this problem. It certainly felt like we had the right people in the room; representatives of national bodies, patient groups, academics, survey experts and people from local systems (both local government and NHS). As I was sending out invites, it became clear to me not only how many people were working on aspects of this question, but also how long people had been trying to come up with ways of doing this, myself included.

Several potential solutions were developed, tested, challenged and refined. There were a few ideas that might just work with the right backing and further testing. There will be pitfalls and barriers on the way; those of us who have worked on this agenda for some time are under no illusions about how challenging this will be to get right. The danger of this type of work is that it looks far easier at first glance than it is. To produce measures that are valid and reliable, allowing us to understand both variation across areas and between different patient groups, is no mean feat. There are lots of technical questions that need answering, including how to achieve the right sample – but the solutions developed in the workshop were promising, as was the consensus around them. There was a real optimism that this can be done with the right backing.

The participants were then asked to come up with reasons why the development of a new measure might not succeed, and the number one reason cited was that there would no ownership of the need to find a solution. It can be done, but it will not be easy and will need support from the very top of the NHS and local government to overcome the inevitable challenges. Too many efforts in the past have not had the necessary backing from those in the right places.

It needs to be done correctly, providing robust data that is fit for purpose. There is science behind these surveys in the way they are sampled, the way the data is analysed and how the results are interpreted. What is measured should start with what matters to people and it must be user led, but there should also be expertise in how the collection is designed.

Importantly, this work will require investment and time to understand the feasibility of the ideas developed in the workshop. There is a real risk here that something will be developed quickly that does not work. This is too important to get wrong. If done right it will be a significant achievement. If not, as Richard Taunt from Kaleidoscope said, we could be sitting in the same workshops in ten years’ time, still asking whether integration ‘works’ and still wondering why not enough has been done to put the right measures in place to find out.