Pharmaceutical Market Europe • April 2024 • 30-31

AI AND HEALTHCARE

Data integration and the AI healthcare revolution

Looking at some of the chief opportunities and challenges for digital health providers in implementing and scaling technological solutions in the UK

By Thomas Colomer and Chris Eastham

Image

Digital health applications have transformed the way healthcare is delivered and understood.

Its potential to increase access to treatment, improve patient outcomes and reduce the cost of healthcare delivery has made digital health one of the fastest-growing technology subsectors globally.

In the UK, pent-up demand for healthcare post-COVID, coupled with the pressures of an ageing population have exacerbated the need to find ways to manage this demand.

Given the NHS’ budgetary and staffing constraints, attention has turned to making operational processes more efficient, removing friction in clinical and enterprise service delivery, and using clinical data to make smarter decisions.

This has created an environment in which digital health companies can flourish, albeit with more limitations on investment and growth today than there were immediately post pandemic lockdowns.

Data integration

The UK’s COVID-19 vaccination programme accelerated long-overdue integration of vast amounts of NHS data, as data sets from across the UK’s 200+ NHS Trusts were brought together in the ‘COVID-19 data platform’ in the space of five weeks.

While largely successful, this exercise laid bare many of the issues the NHS has with collection, storage and sharing of data.

One of the most common problems digital healthcare providers face coming into NHS care settings is the absence of a single source of accurate data.

In the majority of NHS care settings, data will be stored in multiple databases and electronic patient record (EPR) systems or in unstructured formats, such as long-form digital (or even handwritten) notes.

For digital health solutions to make a positive impact in their respective target areas, these disparate sources of data need to be integrated into a single source and encircled by strict data quality rules.

As well as clinical data, there are also opportunities for NHS care providers to become more efficient by integrating enterprise data, such as workforce data, finance data and operational data.

This can help with budgeting, inventory management and staffing – for example, optimising theatre lists in hospitals by clearly showing which staff are available when and matching free slots with waiting lists.

‘Attention has turned to making operational processes more efficient... and using clinical data to make smarter decisions’

Evidence suggests that integrating waiting list data, for example, can significantly reduce the size of lists, simply by providing greater visibility of patient information.

Data quality

One of the biggest challenges for digital health companies is the problem of poor data quality.

The quality of data generally depends on how the data is structured and arranged – for example, whether it is contained in a database or unstructured formats – and how comprehensive it is, ie, whether key datapoints are missing, and of course accuracy.

New data extraction technologies can compensate for poor quality data to some extent, but ideally the problem should be fixed at source – ie, when the data is entered, rather than trying to fix issues once problems have occurred.

Much attention is being paid to user experience (UX) of data systems and workflows, and using feedback to improve data interfaces for software programmes, stripping out non-essential features and making them as simple as possible for NHS staff to use.

Co-development of solutions with NHS clinical and operational teams is now common practice, with flexibility built in to accommodate differences between care settings rather than trying to impose ‘one size fits all’ pathways.

Process automation to remove manual entry of data, particularly in primary care settings, should reduce human error and help address staff capacity issues.

Interoperability issues

Innovation diffusion is a major challenge in the UK’s publicly funded, federalised healthcare system.

Due to the autonomous nature of individual NHS Trusts, there are significant differences in the way different care settings operate.

This means that technology developed for one hospital or NHS Trust will not necessarily work in another hospital or Trust.

One solution to reduce these interoperability issues is to build open data models and use cloud architecture to make data easily shareable, and workflows and applications available to multiple users across different NHS care settings.

Modelling and trials by digital health innovators suggest sharing data and systems can enhance the efficiency of services and improve patient outcomes, as well as increasing the speed at which digital health providers can replicate innovations.

Building on the good work of the COVID-19 vaccination roll out, the NHS plans to replace the COVID-19 data platform with a ‘federated data platform’ (FDP), allowing NHS organisations to bring together operational data to support staff to coordinate, plan and deliver care.

To address concerns about the privacy and security of sensitive data, in 2023 NHS England set out ‘five data promises’ covering among other things use cases for data, training provision, transparency and accountability, to ensure it provides a ‘data safe haven’.

The NHS has also been clear that the FDP is not a research environment, for which there is a separate initiative to establish a secure data pool for research purposes.

Both the FDP and the NHS’ pooled data research initiative are extremely promising developments for digital health providers, who will be able to use this data to develop and trial solutions.

Other practical challenges for digital health providers

Timelines
When a care provider invests in a new digital health solution, clinicians are understandably impatient for value. Given constraints around data, as well as funding restrictions and resistance to change, it can take years to see benefits from these investments – timelines that will hopefully speed up with the introduction of new NHS data initiatives.

Cost inflation
The NHS is in the midst of a perfect storm of funding and staffing shortages, increased expectations on what the system can deliver, cost inflation and the advent of promising but very expensive new treatments.

These pressures mean NHS providers have less time and money to spend on new digital health solutions.

The digital health sector itself has also experienced cost inflation pressures, partly as a consequence of the post-COVID digital health funding boom of 2021, which raised salary expectations.

Change management
Encouraging NHS organisations to fully embrace the potential of digital health solutions is a key challenge for innovators.

‘Companies that proactively align their AI solutions with evolving regulations can gain a competitive advantage, showcasing their commitment to ethical practices and patient safety’

Many existing systems are not being used to their full potential – a limitation that will almost certainly apply to any new systems that are introduced, even when they have been co-developed with users.

Clearly defining processes for recording information and mandatory systems training for all users can help close this gap.

Duplication
A common inefficiency in the digital health ecosystem is the tendency to try to develop technologies from scratch that have already been shown to work elsewhere, rather than focusing on building on existing systems to develop specialised or enhanced versions of these technologies.

Integrating AI

In recent years we’ve seen how increasing integration of AI into the healthcare sector is revolutionising the way medical technologies operate. However, among all the optimism, there’s a cloud that comes with a lack of regulatory clarity and questions over safety and effectiveness. Regulators are closely scrutinising the implementation of AI in healthcare, recognising the need to establish clear guidelines and standards. Non-compliance with these regulations can carry severe consequences and organisations should prioritise accordingly. While the prospect of increasingly stringent regulation may be daunting, it is not to be feared. Well-developed regulatory frameworks will bring more certainty to the MedTech space and provide a more stable foundation for innovation. Companies that proactively align their AI solutions with evolving regulations can gain a competitive advantage, showcasing their commitment to ethical practices and patient safety.

This article is based on a seminar hosted by Fieldfisher featuring contributions from Chris Barker, CEO of Spirit Health; John Bull, Technology Partner at MBI Technologies; and Karri Vuori, Managing Director of IMAP UK.


Thomas Colomer is Corporate partner and Chris Eastham is Tech partner, both at Fieldfisher

0