NHS trust’s Pioneer data platform joins up care delivery and improves outcomes

University Hospitals Birmingham NHS Foundation Trust has implemented a data platform with near real-time data to improve care across its hospitals.

The Pioneer data platform, which the country’s largest NHS trust built together with Ensono, is essentially a research hub, linking up patient data from different healthcare settings to provide a full picture of a patient’s care journey.

The aim is to build a more effective healthcare system, allowing clinicians to have access to data to see a joined-up picture of what is going on.

Elizabeth Sapey, acute and respiratory medicine consultant at the trust and lead for acute medicine research, said acute care is provided across several different healthcare organisations, whether it be community, ambulance or hospitals.

“Patients often use a number of these settings to get help for a single problem – for example, perhaps seeing their primary care doctor and then calling an ambulance or presenting to an emergency department,” she said, “and sometimes this healthcare journey can take months or even years to gain a diagnosis and have the medical condition fully treated.

“To really improve emergency, unplanned healthcare for patients, you need to understand every part of this complex healthcare system. You can only do this by bringing together anonymised [de-identified] routinely collected healthcare data from each of these clinical settings, so you can understand the entire patient journey, and identify places where new processes, treatments or devices could have improved outcomes for people.”

To achieve this, the data, which was previously stored in different locations on different systems with little or no data-sharing between them, needed to be brought together.

“We needed a way to bring together and harmonise de-identified patient data from a number of different acute healthcare services so that the data could be used more effectively to improve patient care, through research and innovation,” said Sapey.

The system, which Sapey believes is the first of its kind in acute care, aims to help researchers and clinicians make better, smarter decisions.

“The data within Pioneer is not static, it is refreshed frequently so that the data is as up-to-date as possible – we call this ‘near real time’,” she said. “It is also longitudinal, meaning it captures events as they occur over time.

“Because the data is anonymised, it cannot be used to provide care to an individual, named patient. But by bringing data together from millions of de-identified patients, it can describe clinical problems, highlight where diagnostic pathways are delayed, and identify which types of patients respond best – and worst – to different treatment regimes. And as the data is already in one place, and linked, we can answer urgent questions very quickly indeed.”

The data hub also allows clinical teams and researchers to access and learn from near real-time data from across the West Midlands, speeding up diagnosis processes and improving care choices for patients.

One group of hospital doctors, pharmacists and IT staff have already used Pioneer data to design an automatic prescribing calculator for a complex drug regime used during overdoses, which has led to prescription error rates dropping from a national average of 20% to 0%, said Sapey.

For patients at risk of deep vein thrombosis, using de-identified health data, the trust has been able to test methods to ensure patients are assessed appropriately and find that a prescribing rule that doctors had to perform when seeing a new patient was the most effective way of ensuring safe prescribing, rather than simply individual training or feedback.

“I have cared for a number of patients where complex diseases could have been diagnosed earlier,” said Sapey. “But because people go to different healthcare providers for different problems, the dots are never joined, and a full diagnosis can often be missed.”

Overall, the system helps ensure that care is delivered to the highest standard possible, she added. “This is true for conditions like myeloma, where people might present with infections, kidney damage or bone fractures. Pioneer should be able to change this by joining the dots and building ‘diagnostic prompts’ to get health professionals thinking about the different conditions that are suggested by this longitudinal data – sometimes collected over decades.”

When the trust began the project, it launched a series of workshops with more than 400 members of the public and patients to build up a picture of what they wanted from the data, and how they felt about it.

This, in turn, became a framework to guide health data use within the Pioneer platform, and the trust partnered with Ensono, Microsoft and HDR UK to build the Microsoft Azure cloud-based infrastructure, and test its cyber security performance.

“We had to apply for ethical approvals to ensure our processes for bringing data together and using it for research and innovation had all the necessary legal approvals,” said Sapey. “We also brought together and trained our data trust committee, a group of public members who review and provide a public voice on all data access decisions.

“We then started to work with healthcare providers to start sending their de-identified data to us, and built our processes for ensuring the data – often stored using different systems – could be linked, and the quality of the data was high.”

Even in the middle of the Covid-19 pandemic, the processes for the platform were developed, the systems were build and approvals were put in place, to allow the platform to go live.

In fact, Sapey said the pandemic probably increased the progress of the work, because data was badly needed to help healthcare workers understand how to care for people with Covid-19.

During the pandemic, the Pioneer platform also facilitated data collection and sharing across four regional hospitals and a real-time Covid-19 dashboard, which provided a fast-track screening tool for nationally prioritised clinical trials.