Patient-centred care – how to make it a reality
How can a user-centred approach help make patient-centred healthcare in the NHS a reality? That was the key question that innovators in digital health...
Laura Cameron – former Operations Technology Programme Manager, Raptor Data Management for Yorkshire Water (recently appointed Director of Technology, The Travel Corporation)
Stuart Edmondson – Director, Bluesmith
Andy McMurray – Head of Product Delivery, Opencast
Lee Foster – Chief Technology Officer, Opencast (moderator).
Stuart Edmondson, Director at Bluesmith, said that, while there is data everywhere in every business, the key to unlocking its value is getting that data in front of people who have the power to use it to inform their decision-making.
“It’s great to create reports and have these available, but unless it’s actually going to a consumer who can use it to make a real change – do something more efficiently or create insight for instance - then what’s the point?” he said.
Andy McMurray, Head of Product Delivery at Opencast, and who works on client data projects, said that although data was inherently valuable, it was vital to recognise its potential and focus on how it can help in both public and private sectors.
He said there were huge amounts of data about us as individuals, and less valuable generic data. “We can see much more tailored to you personally – you surf the web and your history goes with you.
“People wearing medical devices or fitness devices can access this more tailored interaction. We have a much more valuable personal relationship with it as we start to access our own data and use it for our own benefit,” he said.
Laura Cameron, who previously worked on data projects for Yorkshire Water, said data quality was often an issue, particularly when organisations felt their data was of excellent quality, but in reality was very poor.
“Companies acknowledge their data is poor, but don’t have the appetite to change how the data is captured. As we all know, if data quality is poor, no amount of technology is going to help you,” she said.
“We don’t need excellent data for every use, but you have to understand the quality of data then address how it is collected. Too often organisations are focused on reporting, when the priority should be producing quality data in the first place,” she added.
Lee agreed – adding that a stress point can be getting an organisation’s executive aligned on the need to collect higher-quality data. He said that could be a challenge in trying to deliver value at pace while working to ensure data collection is secure and compliant through governance teams.
Laura felt the solution was starting small with a business challenge for a particular department and then building from there. “Try not just to boil the ocean,” she advised.
Stuart agreed, explaining that Bluesmith used an “agile data methodology” to build out cloud-first modern data stacks for its clients. Instead of spending months building a perfect enterprise data model with polished enterprise features that don't deliver business value, he urged organisations to start with the business problem that needs solving for a specific area or department.
Taking that approach, they would be able to build a horizontal slice of the architecture, at the same time delivering business value. From that point organisations can iterate and add enterprise features as they add more data and deliver to more parts of the business.
“Get some data in front of people, start getting them excited about it,” he suggested. “You then get a bit of a snowball effect and people start coming to you and asking for their area to be included in this new thing you’re delivering.”
Laura said Yorkshire Water had started its project with a 12-week assignment to deliver for its first use case. It then iterated over that and productionised the platform over the following months, and now had 40-50 data products in different areas of the business. That first case ultimately led to something much bigger.
She said it was very difficult to deliver a programme that has a purely strategic use. It was difficult to justify a project that would take 18 months and cost millions, yet delivered no immediate value. Delivering projects in “slices” would allow you to justify the cost in individual increments, with each data product delivering additional value.
Iterative delivery was beneficial, ideally when combined with an approach that includes business change from day one. Laura asked: “Team A really want to be able to do this with their data that they can't because of technology blockers…so how can we help Team A reorganise themselves to use this data more efficiently, so that instead of being a reactive team, they’re a proactive team? That switch isn't just about technology, it's also about people.”
Stuart described a previous project at the Co-Op, where the documentation for the data platform was captured in a wiki, and users were heavily involved and were finding their own insights in the data, to the point that it effectively built its own community documentation.
“We should always go and speak to people and really validate what the data is showing us, bring humans to the centre of it and not just make an assumption based on ‘we saw this in the data, therefore X.’”
Laura stressed the importance of responding to feedback over time, so that the ultimate objective of a project might change several times depending on what users want.
Lee said data professionals could learn from examples of a quantitative and data-driven mindset being brought into non-data roles with positive results.
Andy said a data-driven approach could inform teams working in different disciplines. User research, for example, could take a more quantitative and data-driven approach to research; “On a government project we were recently working on we were able to use the data we collected from the system to then focus our user research and be positive around what the data was suggesting,” he said. “We had some hypotheses being created from the data. It is also vital to get into the human side of it.”
Highlighting the Co-op project as an example, Stuart said Bluesmith had considered the issue of teams of people with varying levels of capabilities and expertise, all working with data in different ways.
Bluesmith had built a platform that worked for all different kinds of users, from those using technical tool sets and vast quantities of raw data, to the executive who just wanted a quick report on his iPad to take into a meeting.
Andy said that upskilling people in digital literacy would empower them to question the narratives they’re presented with and work with data in new ways. He felt that all projects involving digital transformation should involve educating and upskilling stakeholders, beyond simply delivering a project.
Stuart agreed, adding that those delivering need to be open about what products can do and cannot do, and to show their working wherever possible. Honest conversations can build trust in a product, he argued.
“Taking local government as an example, you have multiple different systems all capturing very similar data, but they don't talk to each other. So you've got your revenues and benefit systems, your housing system, maybe your parking system. They don’t talk to each other at all. Having an MDM solution would help. So before you start investigating data, creating that master data solution is key,” she said.
Stuart pointed out that, with a centralised data platform, teams can bring together more than one data source or more than one data domain, which would then allow them to unlock things that they wouldn't have seen by looking in a single operational system.
Laura highlighted the difference between operational reporting and analytical reporting, with the analytical bringing multiple different data sets together so users could understand data in a different way.
Our panel agreed that AI and machine learning could transform the way we analyse and leverage data, Stuart said that Bluesmith had received a few requests for ‘one of those AI things’, and senior stakeholders were keen to see the use of AI even when it might not be appropriate.
Andy said that while AI has popularly become associated with tools such as ChatGPT, it was important to take a broader view: “Anything that is artificial and provides some form of intelligence and does something for us as humans, is a form of artificial intelligence.”
What it ultimately goes back to is the quality of the data, he said. Good quality data was needed to train those models, because biased data would produce poor results.
Andy predicted that, over the next few years we will see more people taking ownership of their own data, and becoming more analytical with their own data. This was already happening in healthcare, he said, but said it was starting to change in other sectors as well.
Laura said: “My dog has a little pet app, which captures all of his movements. I absolutely love consuming that data. Now, the fact that the product exists tells me that it's not just the data geeks of the world that really want to consume data.
“I don't know if my one-year-old knows that she creates data, but she will when she's older. And let's not be naive, companies stand to benefit hugely from being able to sell things in a different way.”
Andy said that a key challenge for all of us was to equip the next generation of citizens to be data literate enough to take ownership of their own data.
With the rise in sensitive personal data and analytics, strategies to protect that personal data would remain paramount, the panel agreed, based on an understanding up front of how that data would be used. where it would need to be shared, and its sensitivity overall.
Opencast's talk on ‘unlocking the power of data' took place on 18 June at our Newcastle HQ, as part of TechNExt 2024. Watch a full video of the session.
We can see much more data tailored to you personally – you surf the web and your history goes with you...people wearing medical devices or fitness devices can access this more tailored interaction. We will have a more valuable personal relationship with data as we start to access and use it for our own benefit
Loading...