Mapping the delivery of research
Doing the right research in the right place
Peter Bower, NIHR Clinical Research Network
To do the best research, we need people to take part in research studies.
England is unique in having a specific organisation (the NIHR Clinical Research Network) to help people take part. As a result, many people do – in 2018/2019, over 870,000 (2018/2019 Performance Report). That clearly shows that people are very willing to give time and effort to take part in research
As researchers, we are incredibly grateful to everyone who gives up their time to take part. However, we also need to think carefully about who is taking part and where they live. The leads of the NIHR (Professor Chris Whitty and Louise Wood) raised some concerns in this regard, stating:
"At the more applied, clinical and public health end of the spectrum, there is a strong scientific need for research to be conducted with and in the populations most affected. Research activity should go to the populations who most need it, and we would like to encourage the best researchers, wherever they are based, to undertake clinical and public health research in the areas of England with the greatest health needs."
Why does this matter? There is an argument that as long as people who are recruited to research are representative of the wider population, it should not matter where they are recruited. However, that might not be true for various reasons:
People living in different regions have different risk factors and outcomes, so research done in healthier regions and populations may not generalise
Although lots of people take part in research, recruitment is still difficult. Recruiting in areas where there is lots of people with disease should be easier
Much applied health research evaluates interventions, and their impact may vary by context (including geography). An intervention that works in an area with lots of resources may not work so well in an area that is more deprived
Finally, fairness dictates that research funded by the public should be accessible to all. There is increasing evidence that taking part in research benefits patients and health and care organisations – and it is important that the benefits of taking part are fairly distributed.
When these concerns were raised, we started to wonder: what is the evidence that research is not conducted in areas of greatest need?
Fortunately, the NIHR CRN is in a unique position to answer that, as it collects lots of data on recruitment. Combining this with data that exists in GP systems on disease, we were able to see if research was conducted with and in the populations most affected.
This involved some quite advanced analysis, but the core picture can be presented quite simply.
There are 15 regions in the CRN. Using data from GP systems, we ranked those regions in terms of the prevalence of people with long-term conditions. The graph is shown in Figure 1. Regions with the lowest prevalence of these conditions are on the left (i.e. these are areas with fewer people with disease per head of population). The prevalence increases as we move to the right.
The height of the bar reflects the RATE of recruitment within each region: basically, how many people are recruited to research for every 1000 people who have a disease.
If research followed prevalence, all the bars would be the same height – that would mean that the chance of anyone being in a study would be the same wherever you live.
As can be seen, that is not the case. In fact, the rate of recruitment seems to drop as we move left to right, such that the higher the prevalence of disease, the less likely you are to be recruited. There is some variation, but that is the general pattern
These data are unique, but they are not without problems. The CRN collects data on where people are recruited, not where they live, so our mapping is not 100% accurate. Nevertheless, there is no better data, so this gives the CRN a baseline
How big is the problems? One way of looking at this is to ask the ‘Robin Hood’ question: how many recruits do we need to ‘move’ from regions on the left to regions on the right to make our graphs balanced – robbing the ‘rich’ to feed the ‘poor’. We think it is at least 12%, which represents around one hundred thousand people. It may well be higher.
So how BAD is the problem? This is difficult to judge. These data are unique, and we cannot easily compare between health systems – we simply do not know how other countries compare. Given the national reach of the NIHR CRN, England probably does at least as well as other countries.
Finally, what are we doing about it? Well, there are lots of activities within the NIHR looking to better align recruitment with patient need.
The NIHR is one of the biggest funders of research and is encouraging research teams to think much more carefully about where they will recruit. This will help researchers plan early to ensure their work is aligned with need (YouTube Video)
The CRN has a new Research Targeting Tool which can help teams identify areas with the great prevalence of disease, and plan their recruitment accordingly. If interested, people should talk to their local CRN contacts, or email odp@nihr.ac.uk
The CRN has a project looking to collect and share best practice in recruitment in under-represented groups, to provide advice and support for researchers looking to work in new areas. For more information please visit the INCLUDE page on the NIHR website or the INCLUDE Website.
Working closely with patients on the design of new studies is known to benefit recruitment (BMJ article) – we just need to make our patient partners as diverse as possible to capture the right voices to help must align research with need. The new UK standards for Public Involvement highlight this under ‘Inclusive Opportunities’
Systems like the national Join Dementia Research and Be Part of Research and the local Research for the Future are trying to make access to research easier for patients.
We don’t know how long it will take to better align research with burden, but the data held by NIHR CRN means we can monitor the effects of these changes over time, and work out if we are moving in the right direction – or whether we need more radical change.
Taking part in research is good for the NHS and social care system, it is good for the organisations that take part, and it is good for patients. These data suggest we need to think more carefully about where we do our research and how we engage the populations most in need if we are to maximise those benefits and share them equally with everyone.