Connected care: how AI is improving outcomes

Artificial intelligence is healthcare’s best chance of genuine innovation. 

NS

Sign Up

Get the New Statesman's Morning Call email.

“In the 21st century NHS, it might not be the sound of a bedpan dropping that is heard in Whitehall, but that of a robot picking it up. The NHS turns 70 this year but we must turn our sights to the future. We should not accept an analogue NHS in a digital decade.” – Lord Darzi, former health minister.

The use of data and artificial intelligence (AI) forms part of a new transition to connected care, a care model tasked with realising the promise of a value-based healthcare system – in other words, creating a health service that focuses on improved patient and healthcare professional (HCP) outcomes, at a sustainable cost. The data revolution, connected care and AI, presents us with exciting new opportunities to improve the accuracy and speed of diagnosis, to better manage complex diseases.

This emphasis on building exciting digital solutions for better outcomes is the focus of Philips’ most recent global market survey of HCPs, the Future Health Index (FHI). The report reveals that digital innovations are currently pushing towards two considerable gains: improved data collection and analytics, including electronic health records (EHRs); and care delivery, such as telehealth, diagnosis and treatment solutions. The report concludes that an integrated approach delivers the most value. 

AI is something that we all come into contact with on a day to day basis, for example when selecting a “recommended” video on YouTube or ordering an Uber. However, within the context of healthcare, the term AI can alarm people due to the perception that this will involve the transfer of tasks from human hands to robots. In reality, this is not the case.

EHRs can lay the groundwork for data-driven health systems, while AI can be used to support the valuable analysis of this data, aiding diagnosis and offering insights for HCPs and patients. By extracting insights from multiple sources of patient data, for example from radiology, pathology and genomics, that are almost impossible to analyse independently, AI-enabled solutions can facilitate the delivery of more personalised medicine and support clinicians to optimise their work flows, including planning, procedure times, and, vitally, diagnostics so clinicians select the right exam for the right patient. 

Take mammograms, for example. Unfortunately, it is common for mammogram results to be misinterpreted due to differing perspectives by the human and digital eye. This has the potential to cause great distress and worry. However, AI visual recognition software is estimated to be five to ten per cent more accurate than the average physician, thus using this tool could decrease the false-positive rate and prevent numerous cases of unnecessary concern

In addition, the ability of AI and the latest software to sift through large amounts of data can help hospital administrators to identify trends to improve the efficiency, sustainability and overall standard of their hospital. Who would say no to shorter waiting times in A&E? 

The need for increased efficiencies in pathology labs was highlighted recently in a survey by the Royal College of Pathologists, which found that 97 per cent of NHS histopathology labs were understaffed. This has led to some patients experiencing distressing delays in receiving appointments and results. Digital pathology provides the means to speed up and simplify access to histopathology information across the laboratory, helping pathologists to work more efficiently and accurately.

Philips is currently working with the NHS to bring next-level cancer care to the UK through data and AI, including in the field of pathology. The iCAIRD (Industrial Centre for Artificial Intelligence Research in Digital Diagnostics) project with NHS National Services Scotland, aims to create one of Europe’s largest fully-digitised pathology labs. iCAIRD will involve the development of a system whereby AI can assess digital diagnostics, pathology and radiology data within Scottish NHS pathology data, and consolidate the information in one digital location. The initial focus is on gynaecological cancers, to increase speed and accuracy of diagnosis of this complex area, so HCPs can provide patients with the best care possible. 

Philips and University Hospitals Coventry and Warwickshire NHS Trust, along with four digitised NHS trusts, are also working on project PathLAKE to create a centre of excellence for AI. Alongside accelerating the provision of world-class training and education to pathology and computer-science communities, PathLAKE will assess diagnostic efficiencies and develop novel AI tools to track potential response to treatments. 

Through ongoing education around the world, we can build belief in the power of data and AI to drive connected, personalised and better care. To build trust in this new era of digitalised healthcare, governments, businesses and healthcare institutions should continue to invest in developing digital solutions that address the needs of the population.

We need to educate about AI and underscore the benefits of its use in the healthcare setting. We do not want AI to be misconceived as a replacement for physicians, but acknowledged as another tool in the treatment and care arsenal. One which can free up HCPs from certain tasks and assist them in accurate decision-making, allowing them to dedicate more time to provide quality care to more patients.

This year is a monumental year for healthcare in the UK, with the 70th anniversary of the NHS. Whilst the NHS of 1948 may have evolved considerably, the operating principle of providing free care to all still remains. An ageing population and tight budgets put pressure on the men and women of the NHS to deliver high-quality care. However, with data and AI we have the tools to continue to support and accelerate the innovations and medical excellence of this country.

For more information, please click here

Neil Mesher is CEO of Philips UK and Ireland (UKI).