#OpenDataSavesLives - Session 33: Open Data in Health

By Tazmin Chiles, Data & Innovation Consultant/Delivery Manager, Open Innovations

Introduction 

#OpenDataSavesLives has evolved a great deal from its original inception in the midst of the pandemic. Formed as an informal group of collaborators in response to the slow emergence of Covid-19 data and lack of central guidance, it has since become a movement to help people share and publish data quickly on the web. Now, as the availability of Covid-19 data increases and the severity of the pandemic subsides, #OpenDataSavesLives is evolving and focusing on the wider impact of health inequalities and the critical role that data will play in levelling the playing field for all people.

Inequalities in quality of care is a key area of focus within the NHS, and part of its response will require a more robust method of demonstrating how it provides fair and equitable services. In light of this, we will be delivering events later in the year that explore this topic and link with the two excellent speakers who joined us this time to share the work they’ve been doing in this area.

Peter Lacey, Director of the Whole Systems Partnership will give reflections on how his models have aided in monitoring and forecasting during the pandemic. There is also huge potential in connecting NHS data with what are considered as wider determinants of health such as emotional, economic and social environment. David Williams, Director of Inclusion at the Park Academies Trust, will be talking about his work with Pupil Perceptions, a tool that reflects these wider determinants of health in students.

Peter Lacey

Peter has an extensive background in information modelling within the NHS and private sector. The Whole Systems Partnership has emerged out of the need for modelling and ‘what-if’ analysis of complex systems, such as the impact of Covid-19 on cancer services. They start by looking at the system, and he stressed that data is the product of the system you are measuring therefore its reliability and quality is dependent on changes in the system itself. Context matters, and data cannot be considered in isolation.

Peter shared with us the value that system dynamics models can offer, even in the absence of data, using an example developed from a minister’s speech (shown in the slide below), which then can be used to define the data that is most crucial to collect in order to better understand the whole system. Here he demonstrated that their work doesn’t start with the data, but builds a picture of what data would provide the best representation of the system in order to make accurate and actionable models.

The Whole Systems Partnership develop models that represent complex systems such as recovery of referrals through diagnostics and urgent treatment after COVID. We see enormous value in visualising systems in this way to tackle challenges such as health inequalities.
Credit: The Whole Systems Partnership

In covid modeling, this was driven by emerging evidence and generated an expected level of admissions and bed occupancy. It was a constant learning process in which issues with data quality and consistency were identified and smoothed out. He went on to explain how these models might be developed in the future. In order to do that, we need good quality, consistent and often linked data to properly support this learning approach to complex challenges. There have been some drawbacks for the widespread adoption of these models which may be ameliorated through a more open approach. This has sparked the creation of Really Useful Models, a non-profit that aims to inspire a conversational movement and public engagement around system dynamics modelling.

We can see a lot of potential value in adopting these approaches to visualise police, housing and demographics data and linking this with NHS data and beyond. The Whole Systems Partnership have provided models that continue to offer value in the NHS and we are excited to see the future of Really Useful Models.

David Williams

Our next speaker, David Williams is Director of Inclusion at the Park Academies Trust, a chain of academies from nursery to post-16 year education. He told us about some of his previous work on wellbeing data with the University of Oxford, which demonstrated how academic achievement is impacted by the emotional wellbeing of young people.

As part of this work, they contacted Anspear, who developed Pupil Perceptions - a tool that gives a measure of a pupil’s emotional wellbeing and readiness for learning based on answers to an indepth questionnaire spanning key areas of a child’s life including healthy lifestyle, feeling safe, asking for help and opinions towards authority figures. The tool therefore tracks each individual child’s opinion of their own wellbeing and their attitudes towards their teachers, family, friends and professionals over time. Currently, they hold a dataset of over 500,000 students across the country.

Pupil Perceptions is a diagnostic tool that reflects several aspects of a child's wellbeing. Answers to these questions can be broken down by age, gender, demographics and can provide enormous insights to decision makers.
Credit: Pupil Perceptions

Answers to each question can therefore be broken down by gender, economic and demographic group. Some of the data gives fascinating insight into how different parts of the community have different answers, particularly towards the police and healthcare professionals. Comparing this with a deprivation index would be very useful for NHS professionals and local authorities. This raises questions around the potential to link these datasets with medical and police records, where tools like Pupil Perceptions may act as early identifiers for mental health intervention, with the potential to prevent students from reaching a crisis point. From a trend of this data, they are starting to see which students are likely to be using CAMHS services within the next two years.

The fact that we operate in a national context of nervousness around sharing data may act as a challenge to linking these data sources together in the short term. One point was raised during the session that the risk of not sharing this type of data may have a greater impact on a young person’s wellbeing than the risk that comes with sharing it. There have been legitimate GDPR concerns in the past when dealing with this sort of sensitive data, however we know there are numerous altruistic organisations who, with the necessary infrastructure to preserve privacy such as trusted research environments, could use this data to improve services for young people.

Conclusion

The session brought some amazing discussion that will hopefully feed into our work later in the year around health inequalities. Watch this space for future projects relating to both of these topics.

If you’re interested or are working in this space, then please do get in touch with us at hello@opendatasaveslives.org and feel free to reach out to Peter Lacey (peter.lacey@thewholesystem.co.uk) and David Williams (williamsd@theparkacademiestrust.com).

As always, a huge thank you to our sponsors at NHS SCW and TPP - and to the Health Foundation for helping us to deliver this event series.