The Centre Cannot Hold For Electronic Medical Recording
By Professor Kevin Moore
“Turning and turning in the widening gyre
The falcon cannot hear the falconer;
Things fall apart; the centre cannot hold;
Mere anarchy is loosed upon the world”
-W.B. Yeats
I am referred to as a Liver Expert by the Courts for whom I write medicolegal reports. I am often critical of the care provided by other doctors. As a Professor of Hepatology and a Clinical Pharmacologist, I see cases where clinicians have failed to monitor drug safety according to national guidelines. I frequently see the same problem in the management of long-term medical conditions.
Healthcare systems monitor medicines and diseases poorly
It’s late Friday afternoon, and I’m on my own holding a monster of a liver clinic in a peripheral hospital that will finish way past 6pm. As usual, I have spent much of the afternoon advising people not to drink alcohol, and I will arrive home late, frazzled, and desperate for a drink myself.
I have a patient with autoimmune hepatitis who I need to start on azathioprine. It is a drug that requires lots of baseline tests to check that azathioprine is safe for the patient. Once started it requires weekly blood tests for safety monitoring for four weeks, and then monthly for two months, and then quarterly. A recent medicolegal case I have completed is still fresh in my mind. A nearly fatal neutropenic sepsis in a young woman whose doctors had failed to monitor the bone marrow toxicity of azathioprine.
Why? Manually monitoring is not safe
Using a State-of-the-Art electronic patient record system, I manually request, change and forward-date each test request for the next three months, a duty I discharge with resigned diligence. But here at this hospital (“a paperless hospital”) a paper form is needed for each test, and so each form must be printed. I dutifully click ‘print.’ Nothing prints. I try again, and I try again. I check the paper draw. I open and close the front of the printer. I turn it on and off at the wall, the sure-fire solution to all technical riddles. Still nothing prints. I am staring at the printer now, boring holes in it with my eyes, willing it to print, trying to bend it to my will. It just sits there silently, determined not to flinch. The software offers me a selection of alternative printers, hundreds of them, spread across the hospital. The queue outside my door, like my agitation, is growing. I take a sheet of paper from the printer, hand-write the list of dates the tests are needed and hand it to the patient.
Half an hour later she returns to say that the phlebotomy department have told her ‘No form, no test’ Now I am stressed, not mildly but acutely, visibly stressed – stressed enough to switch into Dr Angry mode and march down to speak to phlebotomy reception. They reluctantly agree they can look up the patient’s details on the EPR system and print the bar code labels. It is, they remind me, ‘much easier with the paper form.’ ‘Thank you,’ I say, suppressing the less generous responses bubbling up in my head. The problem is nearly solved.
Patients get lost in the system all the time
But how would I know if the patient has had the blood test, and ensure I check the results? I give the patient my email address and ask her to contact me as a prompt until I see her next time. Not the most reliable safety-net. Most doctors won’t give their email address to patients, but in my experience this safety net is welcomed and rarely gets abused. Most NHS employees get lots of emails. I receive over 200 per day. Have you ever missed an email?
This is not an uncommon vignette. Many clinical professionals will recognize that these moments of minor farce are moments of major risk. The end-to-end process of practicing medicine in the current NHS is riddled with potential points of failure, risks whose realization or mitigation too often turn on the vagaries of human agency or blind luck. A faulty printer may serve to raise one’s blood pressure and exacerbate the danger, but it is hardly at the root of it. There is an understandable, if somewhat nebulous, clamour around AI and the way it will revolutionize healthcare. In time it will prove transformative, but significant parts of the NHS remain paper-dependent. The current digital infrastructure, such as it is, remains patchy and unnetworked. It is a landscape entirely at odds with the institution’s ambition to coordinate and deliver healthcare across ever more geographically disparate and complex populations. The risk is of a vision of a gleaming future which takes no heed of the insufficiencies of the present, the conceit that we can run before we walk.
The EPR Gap: Twenty Years Behind in Digital Adoption in Healthcare
Where they do exist, digital processes are often punctuated by, and dependent upon, more analogue ones – the slip of paper, the email which enters a deluge of hundreds of others like it – and, along with the goodwill of clinical professionals and the persistence of patients. This is the fabric holding the NHS together, a gossamer thread which will not hold indefinitely. Where “total” digital solutions are in place, unlike the patients they serve who may be seen across a variety of locations in primary, secondary and community care, they tend to serve only one, a single site or group of sites. The cloud-based solutions which could address this siloing of digital practice remain bafflingly under-adopted.
Moreover, too often the systems in place seem to have been designed without the user in mind, their inadequacies more or less guaranteeing that corners will be cut and workarounds found, that, despite the associated risks, time will be saved with minimal rather than maximal digital engagement or simply wasted altogether. One system used widely across the Health Service for the registration of cancer patients does not acknowledge that 3-7% of breast cancers are bilateral (1). Whatever the nature of the cancer.
Each breast must be registered separately, a fact at which the mind simply boggles. To produce game-changing outputs, AI requires good inputs, and by and large the software currently in place doesn’t cut the mustard. To this extent, the inadequacies of the present are already defining the limits of our future.
Digital adoption variation in 2022 vs 2021 (percentage point)
Source: McKinsey Global Digital Sentiment Insights Survey
Successful monitoring is about automation for and collaboration across healthcare staff
In a 2022 paper entitled 'A plan for digital health and social care', The Department for Health and Social Care announced (2);
A brighter future depends on a stream of transformative technologies being developed and spreading fast through the health and social care system. – good technology spreads faster through the system when innovators and frontline teams collaborate on its development. We are supporting best practice partnerships between tech innovators and frontline teams. We are also helping NHS organisations in commercial negotiations with industry and funders to align interests behind products that make a real difference to people’s health, staff workload and system productivity.
As a vision, it is all there: an appetite for collaborative innovation; an understanding, at least strongly implied, that co-development is an empirical process, a case of trial, error, adjustment and refinement in which risk can always be intelligently and transparently managed, but never eradicated; and, in light of this, an appreciation of the acumen and autonomy required to negotiate the terms of such partnerships.
What the NHS calls risk aversion is highly risky
The reality, however, is that many NHS Hospitals are reluctant to take risks of this nature. Sometimes they lack the necessary commercial expertise and confidence to do so, but more often their innovational ambitions are defeated by the risk-aversion of Information Governance teams who find themselves in the unenviable position of going unrewarded for capable facilitation but being roundly criticised when things go awry. Caught in a lose-lose situation, their standards for approval can be implausibly high and, as a result, self-defeating. Their role is to keep patient data safe, an imperative no reasonably impartial voice would dispute, but seeing innovation and safety as zero-sum game sets up a false opposition, one wilfully blind to the profound risks of inaction.
No software can be fully developed and safely launched without a real-world pilot. We need a system which recognizes risk as an integral part of the kind of innovation that can help solve the problems which beset our health service. Regrettably, we have built a bastion of guardians for whom it is often easier and safer to their careers to do nothing, and who remain ignorant, selectively or otherwise, of the anarchy which presides over the status quo. Until the NHS wakes up to the severity of its own predicament, it will remain stuck in a perpetual yesterland, and the “gyre”, of which Yeats wrote, will continue to widen.
References
Bilateral primary breast cancer: A prospective study of disease incidence Get access . M A Chaudary, R R Millis, E O L Hoskins, M Halder, R D Bulbrook, J Cuzick, J L Hayward. British Journal of Surgery, Volume 71, Issue 9, September 1984, Pages 711–714, https://doi.org/10.1002/bjs.1800710924.
About the Author
Dr. Kevin Moore is a Liver Expert for the Courts, where he provides an independent opinion on the care provided by other doctors. As a Professor of Hepatology and Clinical Pharmacologist at University College London, his expertise in drug safety monitoring and adherence to national guidelines provides a unique perspective on the healthcare system's successes and failures. With over two decades of specialized experience in liver transplantation and general hepatology, he brings substantial clinical and academic authority to both patient care and medical-legal evaluations, along with a keen understanding of the practical challenges facing modern healthcare delivery.
Kevin Moore at UCL, Royal Free has authored over 100 research papers with an H index of 64 and is the author of the Oxford Handbook of Acute Medicine.
It is time to be Healthcareful®