Skip to main content

#OpenDataSavesLives: Session 38 - RAP for Health Inequalities

By Tazmin Chiles, Data & Innovation Consultant, Open Innovations

Our second #OpenDataSavesLives session of the year focused on Reproducible Analytical Pipelines, and offers some crucial context that raises questions on how the cost of living crisis could be worsening health inequalities. It was another busy session with 100+ sign ups and 60 attendees. We live-streamed the event to Youtube this time, which we will continue to do for future sessions so if you’re unable to attend our events via Zoom, then you can still get involved on our Youtube channel.

Thoughts on #OpenDataSavesLives: 

When we set up #OpenDataSavesLives during the pandemic, we offered a convening space for people who were trying to answer important questions about the impact and response to COVID 19. Now the program has evolved into a diverse network of organisations across a number of sectors. Through our sessions and engaging with the community, we keep coming back to the same themes - reproducible analysis, and how we can adopt this to understand and combat inequalities. So, the #OpenDataSavesLives platform now focuses its efforts to support the public sector in addressing these inequalities - in access to healthcare, experience and outcomes.

This speaks to our roots in the pandemic where we saw early identification of unwarranted variation in access to treatment and outcomes after COVID. It now comes in light of NHS guidance, that requires NHS trusts to show they are delivering services fairly, sparking new conversations on the role of data and working in the open to understand and address these inequalities.

The health sector needs to communicate rapidly and effectively to answer questions about their impact and coordinate their response to address inequalities, and #OpenDataSavesLives aims to support them in doing so in the coming months and years.

This session’s topic is therefore a significant one as the public sector moves towards more automated and reproducible methods of data analysis with the goal of addressing inequalities and improving delivery of health and public services.

You can watch the session recording here.

Speakers:

  • Sam Hollings - NHS England - "Reproducible Analytical Pipelines - faster, more efficient analysis - how we're rolling it out in NHS England". View the slides.
  • Molly Broome - Economist, the Resolution Foundation - "Hoping and coping: how the cost of living crisis has affected people's finances and health to date". View the slides.

Sam Hollings - NHS England

A key point that Sam made was that the benefits of RAP - Reproducible Analytical Pipelines (a blog post on which you can read here, if you are not familiar with the term) - are two-fold. For the organisation, and the analysts. Transitioning to RAP makes analysis faster, more robust, reliable, and - the clue is in the name - reproducible. What used to be a time-consuming, manual process now frees up analysts to do more valuable analytical work. “Analysis that took two weeks to do, now takes 40 minutes”.

RAP involves the automation of outdated and expensive legacy processes using open source tools like Python, R and SQL. Publishing well-documented code, methodology and data (unless there is a good reason not to share the data - such as patient privacy), makes the analysis sharable and reusable, so that another analyst in another part of the world can view their repository, understand what they’ve done and replicate it themselves.

RAP is not just confined to the health service and is being adopted across the public service and research sector. The Government Analysis function has published guidance on RAP, there are a number of blogs and example repositories (below) available online for anyone wanting to find out more. And of course, the Goldacre Review sets out a huge wealth of guidance and principles behind the change.



Another key point from the following discussion was the usefulness of a ‘RAP score’ - assessing a project against RAP guidelines. How is a baseline RAP defined and what makes the ‘Gold standard’? NHS England have an approach to define this (below). We think these metrics will be useful to share and develop as it helps others understand where their analysis fits and how they can make it more reproducible.


Scoring RAP... How do we assess whether an existing pipeline is 'RAP ready'?
Credit: NHS England

 Molly Broome - the Resolution Foundation

As we saw with the COVID 19 pandemic, the cost of living crisis is a rapidly changing situation - therefore accurate, timely data and reproducible analysis is important to understand and minimise the impact on the people most at risk.

We invited Molly to present some of their findings on how the cost of living crisis has impacted on the health and finances of different populations.Their analysis was based on the outcome of recent surveys conducted with YouGov with over 10,000 respondents, providing a richer and more timely source of data than ONS surveys. The slides below demonstrate some of their key findings.

Richer and older people have been more insulated from the cost of living crisis.
Credit: The Resolution Foundation



People with a low income have reported eating less or skipping meals due to the cost of living crisis.
Credit: The Resolution Foundation



Poorer and younger people saw worse health outcomes resulting from the cost of living crisis.
Credit: The Resolution Foundation

The inequalities are stark, demonstrating striking patterns of variation in how the crisis impacts on more vulnerable groups.

These findings are fascinating, and prompted a number of questions from the room around how somebody might interact with the data and analysis that informs the report. Perhaps the responses could be further broken down by geography, or paired with open data on regional demand for physical and mental health services? And what might the survey data suggest about ethnic variation in physical, emotional and financial wellbeing during the cost of living crisis?

We’d be keen to explore how the cost of living crisis, on top of the pandemic, could be further contributing to inequalities in health and wellbeing across the UK, and how adopting reproducible analysis can help us coordinate a faster and more informed response to these unwarranted differences.


Conclusion


When we set up #OpenDataSavesLives, we didn’t know that it would evolve into what it is today. We’ve attracted a fantastic network of engaged attendees and sponsors, and we’re looking forward to the next phase of the programme as we narrow down our focus. We're always looking out for collaborators, funders and speakers at our events so please do get in touch if you're interested in what we do.

Thanks to our speakers, Sam and Molly, for sharing their work and to our sponsors and supporters of the event series. Our next event will be on the topic of Virtual Wards on the 21st June. Here is the link to register - see you then!

 

Upcoming events...

Open Data Saves Lives - Session 41