Artificial Intelligence in the NHS – transformative benefits requiring open mindedness and a well-considered plan

In June’s budget announcement, new healthcare minister Matt Hancock announced the NHS funding headlines:  real-terms funding growth for the NHS of 3.4% per year over the next five years, amounting to an extra £20.5 billion by 2023/24.   The Government’s intention is to marginally ‘frontload’ the extra money, meaning 3.6% in each of the first two years. The NHS budget for the year 2018/19 is set at £113.8 billion.   Nonetheless, whatever these headlines, for NHS organisations to calibrate themselves correctly to be able to transform their IT services in 2020 and beyond – they must be IT fit.

Examples where healthcare AI is starting to make a difference
• Precision AI – AI is revolutionising healthcare tech using graphics software in areas like radiology.  Here radiologists can use AI-based video analytics to find brain bleeds or tumours in MRIs and another creates 3D colour representations of organs from a chest x-ray.
• Predictive AI – using muscle:fat ratio from CT scans, AI analytics is being developed to create predictive clinical outcomes for illness or disease.
• Natural Language Processing (NLP) – layering this onto data could enable AI algorithms to have access to all patient information vs a small part ie. making sense of badly written disparate medical notes.

Artificial intelligence impact on business
By deploying the right AI technology, organisations can gain the ability to:
• Save time and money by automating routine processes and tasks
• Increase productivity and operational efficiencies
• Make faster business decisions and greater strategic direction, based on outputs from cognitive technologies.
• Avoid mistakes and ‘human error’, provided that smart systems are set up properly
• Use insight to predict customer preferences and offer them better, personalised experience
• Achieve cost savings, by optimising your business, your workforce or your products
• Mine vast amounts of data to generate quality leads in Sales and grow an organisation’s customer base in business.
• Increase revenue by identifying and maximising sales opportunities
• Grow expertise by enabling analysis and offering intelligent advice and support

Some of the challenges around AI
With so much rapid innovation taking place in technology there are enormous digital capabilities on offer especially around AI.  One challenge for healthcare ICT leaders may rest with how they prioritise their spend to evidence best return on improvement and experience for both staff and patients in future years.   On top of this, AI needs access to good quality data.  Not something commonly associated with the NHS.   There is a lack of data standardisation and centralisation which continues to hamper progress (despite healthcare attracting plenty of AI start-ups fostering innovation in the marketplace).

Could personal ‘health clouds’ be the answer?
One answer leading to digital transformation could be ‘personal health clouds’?  Currently data is often stored in separate silos on local hospital servers.  There are no data standards nor centralisation.  The physical dislocation of NHS data is coupled with the many governance issues around Personal Health Information (PHI) which create an inhibiting mix.  If patient records, test results etc. could be stored centrally, with the patient confirming who has access to the info, AI could view all data versus only a part of the data.  It could be transformational for the efficiency of the overburdened hospitals – as well as the patients.  A main barrier to conversion is the necessity of patient buy-in, agreeing to the use of their data and trusting that all healthcare trust compliance measures are met and protecting their data 24×7.

Clearly, the opportunities for the NHS to seize are massive.  But just how hospitals and healthcare organisations go about seizing AI opportunity to transform the sector and how we might receive very different treatment in future may largely centre on the following elements:
• All public sector organisations having an ICT digital lead and visionary who is connected strategically to the Board – and fully conversant of the complexities of the public sector environment.
• Preparedness of the project teams to spend the right length of time at discovery to understand and map out the needs and most desirable outcomes for all stakeholders to ensure the touch of digital feels like a light wand not a heavy fist for implementation or users.
• Starting with getting one project right rather than seeking to fix the whole system in one go.
• Working closely with innovative software companies focused on the sector.
• A highly skilled IT team or partner (internal or outsourced) to safely walk the journey together from infrastructure transformation to a better digital future.

Rome was not built in a day, but those straight roads from 2000 years ago are still much admired today.  The transformative opportunities facing the NHS through AI are immense – and since IBM Watson beat Ken Jennings in 2011 the progress of AI has marched on into our consciousness.   AI requires an open-minded attitude and a willingness to embrace new opportunities when they arise.  It’s baby steps to bring about large scale progressive change, but having advanced technological understanding, drive and support to champion and deliver change and connect is a true game changer.    Let us know what you think to add to this article?

Hospital and Council start to test AI in undertaking tasks, as ‘virtual workers’

Ipswich Hospital, part of East Suffolk and North Essex NHS Foundation Trust (ESNEFT) has engaged three robots over the past three months to mimic keyboard and mouse clicks, to assist with routine admin referrals, using ‘virtual workers’.  Handling 2,000 referrals on average per week, this has saved the Trust considerable costs and increased efficiency amongst frontline staff.

• Tasks have included sending scan and blood test results from Primary Care GPs to Secondary Care hospital consultants, with intelligent automation supplied by British software company Thoughtonomy. This works over the existing applications and systems as part of the GP Electronic Referral Service (where patient data is collected and sent to the relevant consultant).

• Initial results claim the new system is eight times more productive than using humans, releasing 500 hours of medical secretaries’ time, working 24×7 and anticipated to save £220,000 over the next nine months. The referral process has been seen to successfully reduce down the referral process from 15-20 minutes, to 5 minutes.

Darren Atkins Deputy Director of ICT, commented, “At the Trust we have a philosophy of making time matter”.   Mr Atkins continued: “Done to scale, intelligent automation has the capacity to massively drive transformation on a huge scale within the NHS.”

Neurology department medical secretary, Christine Harvey added:  “We used to extract information from (patient records) from one bit of software and put it into another bit of software. This was time consuming and carried the risk of mistakes from the patient notes system”.

The technology has been received positively at the hospital, offering flexibility from ‘virtual buddies’, maximising the benefit of automation and enabling frontline staff to spend more quality time with patients and on workload, be more productive and enabled the Trust to cut down on short term agency use.  Technically, the queues and productivity can also be monitored very easily on the move through smart devices.  The scheme has the capability to be extended to using AI ‘staff’ to make appointments and do accounting and patient queries.

A report by the Institute of Public Policy Research (IPPR) states that automation could save the NHS up to £12.5bn a year, or 10 per cent of its annual budget.  The Department for Health identified that the NHS’ infrastructure was “not currently fit for purpose for AI” and would require standardising to fully harness the technology’s potential.  A move in this direction however is clearly signalled as a route to improvement and cost savings for the NHS.

Dr. Bernhard Kainz at the Department of Computing at Imperial College London said: “At the moment the NHS has still an IT problem rather than a data science or AI opportunity.. and rather, it marks an important step towards clinical decision making supported by AI because it makes sparse, unstructured data accessible for automated data processing”.

Meanwhile, in a bid to improve transformation of social care through technology, Bradford Council is in the process of developing a proof of concept for an AI tool in adult social care, funded by NHS Digital and the Local Government Association.  Bradford, working with Rescon Technologies has designed the relevant service requirements which take on board matters as diverse as dental care, financial matters and watching football.  The findings of the project will be known in December and decisions on how to take matters further determined at that time with stakeholders.

For an understanding of the core distinctions between IT transformation and Digital Transformation and routes of developmental interest for the public sector, read Amicus ITS’ latest White Paper here.

If you would like to discuss this with a member of the Sales department, please contact Amicus ITS on 02380 429429

‘Orangeworm’ the new superworm hacking group that’s targeting healthcare

Hacking activity targeting the healthcare sector continues to rise.  New security research just released by Symantec has identified a global hacking group called ‘Orangeworm’.  Though its targeted victims accounted for a small number of organisations in 2016 and 2017 (mostly in the USA and Asia), some were identified as being based in Europe.  Analysis by industry has revealed that the healthcare sector is Orangeworm’s primary target, with 39% of hacking outcomes manifesting themselves in this data rich sector which includes hospitals and pharmacies.

Symantec said, “Based on the list of known victims, Orangeworm does not select its targets randomly or conduct opportunistic hacking. Rather, the group appears to choose its targets carefully and deliberately, conducting a good amount of planning before launching an attack”.

Orangeworm’s wormable trojan, named ‘Kwampirs’ is able to vet the data to determine if the computer is used for research, or contains high value data targets eg. patient information.  The Kwampirs then create a backdoor on compromised computers, enabling the hackers to remotely access equipment and steal sensitive data – and Orangeworm survives reboots.

The trojan worm has a penchant for machine software on critical hospital equipment which includes kit like x-ray machines and MRI scanners, as well as machines used to assist patients in completing consent forms.  If the ‘victim’ computer is of interest, the malware then “aggressively” spreads itself across open network shares to infect other computers within the same organisation and uses built-in commands to grab data. This includes “any information pertaining to recently accessed computers, network adapter information, available network shares, mapped drives, and files present on the compromised computer.”

The supply chain is a key part of this vulnerability funnel, with targets including manufacturers providing medical devices and technology companies offering services to clinics, plus logistics firms delivering healthcare products.

Director of Technology, Security & Governance, JP Norman advises:  “Ensure your anti-malware provider can detect Kwampirs activity and to prevent and detect an infection, ensure that:

•        A robust program of education and awareness training is delivered to users to ensure they don’t open attachments or follow links within unsolicited emails.
•        All operating systems, anti-virus and other security products are kept up-to-date.
•        All day-to-day computer activities such as email and internet are performed using non-administrative accounts.
•        Strong password policies are in place and password reuse is discouraged.
•        Network, proxy and firewall logs should be monitored for suspicious activity.
•        User accounts accessed from affected devices should be reset on a clean computer.”

Sales Director, Les Keen added, “Where there is the option for healthcare / supply chain organisations to prioritise IT funding, updating the Operating Systems is a primary, as is ensuring a strong and regular policy on Patch Management.  Our Sales and Security teams  are always on hand to review and audit organisational IT infrastructure and offer holistic remediation advice as part of our security readiness programmes.  Just call us on +44 2380 429429”.

 

Take Up For GP Online Services Hits 42% rise YOY in 2018

With 1 million patients now using the NHS every 36 hours, the pressure is firmly on the nation’s healthcare system to cope with an increased, ageing population, more complex medical conditions being treated, increased waiting time for treatment and more ingenious medicine which is keeping people alive for longer.  Tie that to Brexit and the drive for a 7 day service against today’s staffing pressures and you can see a perfect storm brewing.

Enter then, Health Minister Jeremy Hunt and reflect on his 5 Year Forward View for the NHS. Published in 2015, he laid out his vision for a gradual but persistent transition to patient power – in which digital technology would play a central role.  This was expanded on with his appointment of digital guru Martha Lane Fox to identify four key changes to map out a digital NHS for everyone.  Her stated targets included the following:

o  To reach the furthest first and leave nobody behind
o  To provide free WiFi for all throughout the NHS
o  To build the skills of NHS staff to support people’s needs in the digital age
o  To boost take up of online GP services

So it’s exciting three years on to hear good news coming from NHS Digital’s Leeds HQ. Their latest figures show patient registration take up for secure GP online services in England has risen sharply.  Nearly 14 million patients are now going to their GP’s online for a variety of NHS services without the need to visit a surgery or phoning the practice. These include:

•  Booking appointments
•  Ordering repeat prescriptions
•  Patients view their own records

The figure of 14 million patients is up 42% on February 2017 and amounts to a total of 24% of patients in England now being registered.

For the GP surgeries in local communities who have taken the plunge to embrace technology and overcome initial reservations in parts, these digital pioneers are now reaping the benefits from a variety of online GP cloud service providers.  The results are significant time savings for both the staff and practice GPs, fewer ‘no shows’ and improved patient awareness as patients become more knowledgeable about long term conditions.  The net benefit is derived because GPS have integrated the online public service with single source information clinical systems like EMIS and SystmOne.

There will always rightly need to be a hawkish attitude around protecting sensitive data, however if these new online systems are well governed and securely managed, the public that take up this offer can enjoy a degree of ownership of their healthcare data in intelligent partnership with their GPs and healthcare providers – and in so doing, feel good too they are contributing to improving NHS service efficiencies in the 21st Century.

Not Much Deep Thinking Evident Behind NHS Trust’s Data Share with Google DeepMind

Not for the first time, the NHS has come under fire from patients, patient groups and the scrutiny of the UK’s National Data Guardian (NDG), Dame Fiona Caldicott – and the ICO’s chief Elizabeth Denham.

The Royal Free Hospital in London commissioned Google’s DeepMind division in 2015 to help develop a Streams app to detect acute kidney injury through a blood test to identify deterioration. They provided DeepMind with 1.6 million patient records in the process to enabling ‘real time’ testing.

• Patients at the Royal Free Hospital in London were mainly unaware that their details were being used by a third party, nor how it was being used.
• No details on the financial terms of the deal have been disclosed publicly.

To Dame Fiona Caldicott, whose letter to the Royal Free was recently leaked, laid out her  concern that the data had been transferred on a ‘legally inappropriate’ (read ‘unlawful’) basis.  The app being developed was not ‘central’ to patient clinical care.  Caldicott shared her concerns with the ICO.

Caldicott does not dispute the app’s ability to help clinicians save lives today, but added in her letter: “Given that Streams was going through testing and therefore could not be relied upon for patient care, any role the application may have played in supporting the provision of direct care would have been limited and secondary to the purpose of the data transfer.  My considered opinion therefore remains that it would not have been within this reasonable expectation of patients that their records would have been shared for this purpose.”

Google DeepMind’s clinical lead Dominic King, was swift to distance any cross-use of the patient data with other Google products or services, or use for commercial purposes.

The ICO’s Elizabeth Denham has yet to give her judgement on misuse under the Data Protection Act, but the issue underlines the importance of individual consent.  This will be evermore intensely examined with the forthcoming GDPR regulations in 2018.  As it stands though, the ICO nonetheless has powers to fine a company up to £500,000 for the misuse of personal data as well as seek individual criminal prosecution.

Irrespective of the worthiness and potential benefit to patients in the longer term from the app, Dominic King agrees: “I think one thing that we do recognise that we could have done better is make sure that the public are really informed about how their data is used.”

It may prove a costly oversight to the Royal Free at a time of increasing NHS budget constraints, as well as prompting an ignominious slap in the face to the Trust from its patient body through damage reputation.

Amicus ITS is continuing its series of thought leadership events, this time on GDPR through 2017 for its customers and invited guests.  Further information on the programme can be found by contact Marketing (email) or calling Lindsay Burden on 02380 429475.

3D printing gets smarter in healthcare

Since we last reported an amazing 3D printing story in January 2015, the technology continues to demonstrate its extraordinary enabling powers in the operating theatre for the NHS, with another life transformed as reported this week.

Surgeons were able to use 3D printers to replicate body parts in a kidney transplant from father to daughter at Guy’s and St Thomas’ NHS Foundation Trust in London recently.  With the contrast in size of the organs, 3D printers were used to make models of the daughter’s abdomen and father’s kidney from CT and MRI scans. This enabled the surgeons to accurately plan and rehearse the complex operation.

Hard printouts created the girl’s pelvis, whilst her liver was made softer in a liquid plastic model to enable the doctors to practice pushing it out of the way to make way for the new kidney. Happily, the little girl can now run around and eat normally and enjoy a very different outcome and normal childhood, whilst her parents have the simple joy of planning for her nursery integration in the Autumn.

Unlike in medical robotics where there have been more than two million operations since 2000 the robotics arena still carries challenges in winning over patient confidence.  Here however, the winning smiles of father and daughter amply reflect the achievement of partnership between the human hand and advanced printing technology that shows there is plenty more in store in the future of 3D printing.

first-kidney-transplant-3dprinting-changes-life-northern-ireland-toddler-lucy-1

How on target is the NHS to going paperless in 2018?

Health Secretary Jeremy Hunt has been seeking a paperless records target of 2018 for the NHS since 2013.   In a recent focused healthcare survey of 67 members of the Health CIO Network and CCIO Leaders Network of clinicians and digital health IT leaders, there remains a mixed response:

•  67% stated they were ‘quite confident’ or ‘extremely confident’ their organisation will be paper-light by 2020.
•  14% stated they are ‘not at all confident’ or ‘not very confident’ of achieving the target.

However, on the question of having “integrated health and care records, enabling effective co-ordination of health and social care, by 2020” there was less certainty:

•  56% said they were ‘extremely confident’ or ‘quite confident’ of achieving this, but a quarter (24%), said they were ‘not at all confident’ or ‘not very confident’.
•  28% said they were confident of giving patients read/write access to their records, while 53% said they were not confident.

The top priorities for most of those involved focused on:

  • moving to paperless working – 73%
  • improving quality of services – 68%
  • supporting new models of care – 67%

When asked about their next major IT project, these were reported as:

1.  Top ranking for personal health records and patient portals, to give patients access to their medical record and test results, plus services such as appointment booking and email consultations.
2.  Next were shared record initiatives
3.  Third were e-prescribing and medicines management.
4.  In fourth place finally, one-third of respondents said Electronic Patient Records (EPR) – suggesting many are perhaps already some way down the line with this already?

Not surprisingly, with all the other cutbacks facing the NHS, this drive to go paperless might have a lot of goodwill in the sector to deliver, but the barriers facing NHS providers can be summarised by two principle points of feedback:

  • lack of adequate resource (73% affirmed that their IT budget was insufficient)
  • lack of staffing resource

With the breakup of the NHS from a truly national health service to a regional health service, primary and secondary healthcare organisations around the country will need to start showing they are making this work and that we are benefitting.  Then, we may wonder why it took so long when other major data institutions such as banks and industries such as insurance groups have managed to do this.  After all a 100-1 shot just won the Melbourne Cup.

NHS