Ensuring everybody benefits from the use of health data

With almost 5 million patients on the waiting list, and over 400,000 waiting over a year for treatment, the NHS faces a huge challenge to get back on track following the pandemic. Meanwhile the level of health need among the population has increased, with reported rates of depression doubling since the pandemic began. The parts of our society with the highest level of need – such as care home residents and the clinically extremely vulnerable – have not had their health needs met, and this will also store up problems for the future. 

Notwithstanding the scale of these challenges, there are reasons for optimism, as my colleague Will Warburton has described. New possibilities are being brought about by advances in analytics, data science and artificial intelligence. The potential is for a health system that continually learns from health data, with the experiences and outcomes of every patient being used to improve care for future patients.  

The potential power of health data 

Health data has played a critical role during the pandemic – from tracking outbreaks, developing treatments, and getting people booked in for their vaccines. It hasn’t all been about NHS data as we are beginning to see the relevance of non-NHS data for improving health too. Yet, despite the great strides made during the pandemic, our health system is a long way from using data optimally.  

The average outpatient consultant, for example, has little information about how patients are doing when they are away from the clinic. Doctors often fall back on scheduling regular, follow-up appointments to manage health conditions, rather than using more tailored approaches that might produce better outcomes and lower cost. Meanwhile managers and policymakers have little information about the impact that major changes to services, such as remote consultations, are having. Still too little is known about the impact of the wider determinants of health, and environmental factors are not routinely factored into analysis.  

So, given this context, what is needed to enable a data-driven approach to health to flourish?   

Identifying the most promising technologies and approaches  

At the Health Foundation, we’re lucky to have the opportunity to try out different approaches to using data, including the following: 

• Rapid cycle evaluation. Our Improvement Analytics Unit has been examining the impact of major changes to heath service delivery, such as integrated care, and providing insights that help NHS leaders improve care over time. It identifies the patients receiving these new approaches from within existing NHS data sets and anonymously tracks their use of services and outcomes over time. Patients are compared against a control group selected using advanced statistical methods.  
   
• Data linkage. By joining data sets together from across a wide range of local services, it is possible to produce exciting insights about the health of a place and how it could be improved. Our Networked Data Lab is working with teams from five regions of the UK and has already produced new analysis of the health needs of people who are clinically extremely vulnerable to COVID-19. 


• New tools at the front line. Through our Advancing Applied Analytics programme, we have tested a range of data-driven approaches at the front line, including the use of discrete event simulation to improve the flow of patients through health care providers, and predictive models to help inform triage decisions. We have now supported over 40 projects.  

None of these approaches are as well established as they could be and we have been making our code freely available online, so that other data teams can also use them. 

Getting the conditions right  

Approaches such as these are never going to take off at scale across the UK unless the conditions are right to enable them to flourish. Our improvement team have been looking at how innovation spreads, and finds that it’s often about properly recognising, resourcing and designing the hard work of adoption. As a result, the Health Foundation has launched a new programme in this area.  

There are also specific changes needed to ensure that advances in analytics and data-driven technologies benefit everyone’s health and health care. The national AI strategy for health and social care, national data strategy, and Goldacre review all provide excellent opportunities to address some of these.  

The priorities include: 

• We need to develop better ways of developing and deploying data-driven solutions to address problems in the health system, which are focused on the problems that matter to patients and the NHS and can operate at scale. We have an exciting partnership in this area with HDR UK and we’ll be sharing more about the progress made soon. 
   
• We need to ensure that our data teams – within all sectors and settings – are flourishing. Our research has found that data analysts tend to be isolated and lack the opportunities to collaborate across organisational boundaries which are so important for learning as well as reducing duplication. The investments that the Health Foundation has made in this area are paying off, with the NHS R Community and APHA providing fantastic opportunities for analysts to connect, linking in with the new AnalystX initiative. There is also potential to use the web to make work more open and shareable, as my colleague Karen Hodgson explains.
   
• We need to act to ensure that data driven technologies have a positive rather than negative impact on health inequalities. Unfortunately, there are real dangers that existing inequalities are entrenched because of the decisions that are made about what data are collected and how it is used. These issues require urgent research, and we are partnering with the Ada Lovelace Institute to research the impact of some of the technologies that were used in the pandemic, and with NHSX to advance the use of artificial intelligence for the benefit of minority ethnic communities.   
   
• We also need to invest in the data itself, particularly with a view to improving the recording of outcomes from the patient perspective, and to ensure that data-driven technology is evaluated and regulated effectively. 

The Health Foundation’s data analytics for better health strategy is tackling these problems. Follow our progress. You can also join our webinar, Tech and NHS recovery – what you should know, on 25 June 2021 for more discussion on these topics. 

Adam Steventon (@ASteventonTHF) is Director of Data Analytics at the Health Foundation.

This content originally featured in our email newsletter, which explores perspectives and expert opinion on a different health or health care topic each month.