Cloud Hosting

2020 is the year the NHS was supposed to be paper free. In 2013, then health secretary Jeremy Hunt promised a fully digital health service by 2018. In 2016, a report from NHS England the Five Year Forward View moved that back two years, promising 1.8 billion for the shift away from paper.

But the much-vaunted digital transformation hasnt appeared as quickly as its proponents hoped. In October 2018, 94 per cent of NHS trusts were still using handwritten notes for paper records. In January 2019, the paperless NHS was pushed back to the end of the next decade. What went wrong?

NHS deadlines are always stretch targets, says the CIO of one of NHS Englands 16 digital exemplars a group of digitally strong hospital trusts, who asked not to be named. There is a semblance of a plan, although I think everything has been confused by various ministerial announcements that translate into nonsense.

Others are more optimistic citing last Aprils launch of NHSX, created by Matt Hancock, the Secretary of State for Health, to oversee data, digital and tech procurement for the Department of Health and Social Care and NHS England. The aim is to match countries such as Sweden where, in 2017, everyone over the age of 16 had a Patient Accessible Electronic Health Record, allowing people and clinicians access to individual health data.

Theres finally real national strategy and co-ordination and thats a very good sign, says Natalie Banner, head of the Understanding Patient Data task force based at the Wellcome Trust. Theres a major shift at the central level but the NHS isnt a single organisation. Its made up of many organisations and theres not nearly enough coherence. Individual trusts even individual departments within trusts are entering deals where theres no consistency in the rules or the value of patient data.

At the heart of NHSX is the NHS Artificial Intelligence Laboratory, launched in August 2019 with a budget of 250 million and headed up by Indra Joshi, a former A&E clinician who headed up digital health and AI for NHS England. Joshis role is to streamline the development and deployment of practical applications of artificial intelligence in the NHS, as well as delivering the Artificial Intelligence in Health and Care Award, she explains.

The award will hand out 140 million over three years in a series of bi-annual funding rounds a March 4 deadline currently looms for the first round. The money is intended to back everything from initial feasibility to evaluation of technologies in use in the NHS. Initially, Joshi explains, she will focus on four key areas: screening, diagnosis, decision support and improving system efficiency. These four areas are what the evidence and data shows is where the majority of AI developments are taking place, she explains. We need to improve the basics first, creating efficiencies in the system such as process automation of repetitive tasks or optimising the care pathway. If we can show that these technologies can be successfully implemented and evaluated to show impact, we are immediately freeing up time for people on the frontline.

The award and Joshis previous clinician experience have been cautiously welcomed by many doctors as well as organisations like health privacy body MedConfidential, the Wellcome Trust and the healthcare improvement charity Health Foundation.

Its positive to see NHSX and the AI Lab focus on the basics and what matters to clinician from simple things like one unified login system across the service to understand the burden on staff, says Sarah Deeny, assistant director of data analytics at Health Foundation. Those efforts are really important, but they are happening in the context of outside pressure on the NHS with an underspend in capital projects, a shortage of staff and huge pressure on the workforce. That also underlines the need for the NHS to get the best value for its data.

Health Foundation figures show that NHS capital spend which includes new buildings, equipment and IT, improvements and maintenance of NHS trusts, as well as research and development is very low compared to other, similar western countries. Austria, for instance, spends more than double the share of GDP on health care capital compared to the UK. Matching this in England would mean more than doubling the current capital budget which currently stands at 7.1bn.

Work in AI is dependent on excellent data in health and social care and among its many strengths is the NHSs ability to bring together a comprehensive, longitudinal dataset for 65 million people in the UK, explains James OShaughnessy, formerly secretary of state at the Department of Health under David Cameron, now professor at Imperial Colleges Institute of Global Health Innovation. Right now it's fragmented, not joined up and that needs investment. The entire health service is founded on the assumption of patient trust which can be lost if misused or used for commercial purposes. Patients need to be actively involved in health data strategy at the moment that is not happening either nationally or at any given trust.

The issue of patient trust in the use of data in AI/NHS projects has recently come under scrutiny, thanks to deals between Moorfields Hospital and the Royal Free Hospital with Googles AI subsidiary DeepMind. The agreement with the Royal Free gave DeepMind access to healthcare data on the 1.6 million patients including identifying people who are HIV-positive and giving details of drug overdoses and abortions over the preceding five years. The problem, many doctors argue, is that the needs of Silicon Valley AI health research to provide monetisable software including apps for personal health arent always compatible with the needs of the NHS.

Theres a drive by the secretary of state to get things done quickly and an excitement about digital and apps, but theres a huge gap between that and the NHS world of large scale data systems that support day-to-day healthcare, says Jem Rashbash, director of disease registration and cancer analysis across both NHS Digital and Public Health England. Many of the problems faced are big, difficult and take a long time to address. So while what were seeing now appears exciting, theyre really only scratching the surface. Only very few groups, such as Mihaela van der Schaars Lab in Cambridge and the Turing Institute, are working on machine learning-driven vision and methodologies that will change the future of healthcare.

Joshi is careful to avoid commenting on recent AI/NHS scandals like Google Deepminds deals with Moorfields and the Royal Free Hospital weve got to be cautious when we generalise about AI and Google because we already have partnerships with a range of providers, she explains. In Salford Royal and Bradford Royal Infirmary, for instance, machine learning helps predict how long patients will stay in in hospital, which helps plan and manage patient flow.

She does see a role for diagnostic AI as a junior A&E doctor in the middle of the night being able to read something and decide quickly if its normal or not normal can make a massive difference, she points out. If someone has a possible spine injury they have to be immobilised and put through a CT scan. Helping doctors be confident the scan shows no fractures can make a huge amount of difference but she also sees AI-supported bed management and appointment booking or cancellations systems as immediately useful to overstretched A&E departments, acute wards and outpatient wards alike.

Her final challenge is health inequality. We have centres of excellence in the UK that are focussed in the south, Banner points out. Theres the risk with systems designed for London populations being rolled out nationally which create biased algorithms and unequal access so people in need in Cornwall or Yorkshire are excluded from essential services. We need regional centres alongside a strict regulation and evaluation process which we just dont have right now.

Joshi accepts this. Were very clear that the award, for instance, needs to support NHS staff and private providers wherever they are across the country, she explains. Were active in telling people we would love for your area to make an application so that the less heard voices are included in all of this.

She is also keen to stress that her role includes ensuring reassuring patients through regulation and transparency. We will help people navigate the rules, be clear on what the standards are, help develop policies and make more robust those that are already in place, she explains. The thing is, everybody gets excited by AI and the Lab and the award but this isnt about throwing money at problem and at no point is this about replacing doctors its about supporting the system so its sustainable while maintaining public and workforce trust. She laughs. And thats hard enough in itself

Indra Joshi will be one of the speakers at WIRED Health in London on March 25, 2020. For more details, and to book your ticket, click here

A huge Airbnb scam is taking over London

Thinking of buying an electric car? Read this first

Why do modern tomatoes taste so bad?

How Slack ruined work

Follow WIRED on Twitter, Instagram, Facebook and LinkedIn

Get The Email from WIRED, your no-nonsense briefing on all the biggest stories in technology, business and science. In your inbox every weekday at 12pm sharp.

by entering your email address, you agree to our privacy policy

Thank You. You have successfully subscribed to our newsletter. You will hear from us shortly.

Sorry, you have entered an invalid email. Please refresh and try again.

Excerpt from:
This was meant to be the year the NHS went digital. What happened? - Wired.co.uk