Covid-19 shows that we must overcome organisational barriers to deliver clean, realtime, standardised data in support of direct care, system planning, and urgent response
Covid-19 is now a pandemic, and rapidly imposing new challenges on the health service. New ways of working with software and data could help. But the NHS is being held back by a long legacy of closed working models, reluctance to embrace open standards, lack of systemic design for systems and data, and a tendency to regard clinical informatics as a low-status backroom activity rather than an applied science on a par with other medical specialties.
In the Oxford DataLab we informally assembled a small team of clinicians and digital health experts to think through urgent actions to improve our use of digital technologies during the current crisis, and actions to start implementing in preparation for future pandemics. This post briefly summarises our suggestions and links to the full document: but its main purpose is to flag the urgent need for swift progress on digital in the NHS, and more sophisticated open discussion around the tools, systems, and culture towards digital technology across the health service.
Firstly, on data: there are numerous sources of data that can be better exploited from primary and secondary care, each with their own attendant barriers. There are also some relatively neglected sources such as workforce login data from the NHS Spine information exchange service (which could give a view on workforce absence), and eDischarge data passing through the MESH service. We describe how current data collection arrangements could be modified; how templates and coding extensions can be used to enhance data quality; and how neglected analytic expertise across diverse sectors could be harnessed to generate operational insights with these data. Covid-19 shows more clearly than ever that we can and must deliver clean, realtime, standardised data to support direct care and all aspects of system planning and response. This is not a “back office expense” to be minimised, but a core part of delivery.
Secondly, remote working: access to healthcare, especially primary care services, will be challenging when many patients, but also clinicians, are self-isolating. Teleconsultations offer opportunities to overcome this, via video consultations or simple telephone calls. However communications covers only part of the challenge: there is also an urgent need to get clinicians remote access to the Electronic Health Record systems which are now used, almost universally, to access and update patients’ notes, so that they can provide safer care outside a normal workplace. It is extraordinary that almost every clinician in the country can easily open their NHS email account from home, or from their phone, accessing vast amounts of confidential patient information; but many cannot easily access the electronic health record systems—across the NHS, not just in their own organisation—used to keep notes and manage patient care. This will prevent self-isolating clinicians from offering advice and telephone consultations from home. Such technology has been secure, partially adopted, and fit for purpose, for many years. However, in many care settings it has not been supported by the necessary commissioning mechanisms or tariffs, which has slowed progress on implementing change across the system.
Thirdly, there is progress to be made on how we disseminate information. We suggest SMS and login screens to rapidly disseminate updates. More technically, it is disappointing that we are still distributing treatment pathways only in the form of visual flow-charts on printed documents. 96% of GP practices were computerised by 1996, and in 2019 three quarters of hospitals finally had an EHR in place. By 2020, we should be able to deploy updated computerised clinical decision support pathways at scale, immediately, to clinicians’ EHR systems, using open standards.
Shortages could be worked around by crowdsourcing stock information, and sharing it through popups to prescribing clinicians, which would in turn help patients avoid wasted trips to the pharmacy, and give early warning on bottlenecks to those coordinating supply. In terms of research, we discuss opportunities for rapid evaluation of effectiveness of treatment options; the need for good open structures around documenting current uncertainties; and the need for rapid open surveillance of research activity with an R&D observatory. It is good that the Health Research Authority are offering expedited reviews for covid-19 research: this has been challenging before, and routine audit is not just for clinical work; it may be advisable to create dashboards on turnaround times at HRA.
Some of our suggestions can be implemented rapidly during the current covid-19 episode. Others will take longer to address. We have deliberately not separated the two, with good reason. Some of the perceived barriers to implementing digital technologies are movable, and a function of regulatory barriers or interpretation, rather than absolute blockages.
In addition, much of what we discuss relates to information flows in the NHS that are either poorly documented, or whose true nature is only understood by a small number of individuals, which can block good estimates around implementation challenges. This reflects a dangerous historic tendency towards closed ways of working in NHS informatics, rather than more open approaches; and a broader longstanding failure to invest in creating a “commons of knowledge” around this applied science. Any other technical medical specialty creates library shelves filled with textbooks, papers and lecture notes, for diverse audiences, at diverse levels of specialisation. This rich pool of knowledge reduces access barriers for new thinkers, and helps innovators in the private and public sector to see more easily where they can help. We hope that one day we will see the same rich commons of knowledge for clinical informatics that we take for granted in renal medicine, virology, or radiology.
The Secretary of State has set out a clear and positive Tech Vision built around open standards: these are long battles, and we feel optimistic about the long-term prospects. But the urgency of a pandemic should help us all see more clearly what we must change, despite the barriers posed. We hope the challenges around responding to covid-19 will help the system appreciate the value and urgency of building a truly digital NHS.
This working document was produced pragmatically on a very short timescale; we recognise there will be omissions and other shortcomings; we are keen to hear additions and critical feedback below, or via email@example.com.
Ben Goldacre, Director, DataLab, Nuffield Dept of Primary Care Health Sciences, University of Oxford.
Amir Mehrkar, GP, St Mary’s and Telephone House Surgery.
Rachel Dunscombe, CEO NHS Digital Academy.
Jessica Morley, Policy Lead, DataLab, Nuffield Dept of Primary Care Health Sciences, University of Oxford.
BG has received research funding from the Laura and John Arnold Foundation, the Wellcome Trust, the NIHR Oxford Biomedical Research Centre, the NHS National Institute for Health Research School of Primary Care Research, the Mohn-Westlake Foundation, the Good Thinking Foundation, the Health Foundation, Health Data Research UK (HDRUK), and the World Health Organisation. He is Chair of the HealthTech Advisory Board reporting to Secretary of State for Health and Social Care. He receives personal income from speaking and writing for lay audiences on the misuse of science.
JM is employed on BG’s grants to work on better uses of data in healthcare and has previously been an employee of the Department of Health and Social Care and NHSX. She is also a member of the Oxford Internet Institute’s Digital Ethics Lab through which she has received research funding from the Digital Catapult.
AM is an NHS GP; he has recently stepped down as acting Chief Medical Officer of NHS Digital and co-founded INTEROPen. He is a member of the RCGP Health Informatics Group. He holds a small number of share options in Dr Morton’s – the medical helpline.
RD is CEO of the NHS Digital Academy, Visiting Professor at Imperial College and International Arch Collaborative lead for KLAS research. She is an unpaid Non Exec at the Digital Health Society and is a Non Exec Board member at Hybridstat and Malinko.
Contributors on this document so far include: Ben Goldacre (clinical academic, director DataLab, Oxford), Rachel Dunscombe (CEO NHS Digital Academy), Amir Mehrkar (GP, previously acting CMO NHS Digital), Tom Lewis (NHS consultant microbiologist), Brian MacKenna (pharmacist, DataLab and NHS England), Helen Curtis (researcher, DataLab), Caroline Morton (GP Registrar, engineer at DataLab), Merlin Dunlop (GP, CCG GP IT Lead, Director at Ardens), Seb Bacon (CTO DataLab), Richard Hobbs (Head of Dept, Primary Care, Oxford), Carl Heneghan (GP, Director Centre for Evidence Based Medicine, Oxford), Jeff Aronson (clinical pharmacologist, CEBM Oxford), Jess Morley (policy lead, DataLab).