There is a solution to all the problems in healthcare – it is a simple one, it has a successful track record in many high risk industries and it is one that is acceptable to all stakeholders in the health transaction – Professionals, Patients, Payers, Politicians and Managers. It also saves money – lots of it…
The solution has been called different things in different industries, it has been adopted in various styles at different points in history by these different industries – but is essentially the same thing. Health is probably the last high risk industry to adopt this solution and is doing so rather slowly, grudgingly if at all – for one simple reason: The harm that healthcare causes does not affect either the payers or the providers of the service.
The airline industry was an early adopter – because it discovered pretty quickly that not adopting it would be fatal to themselves and their customers. There was also pretty effective self selection of non-adopters as heroic, buccaneering individuals and organisations literally crashed and burned. The oil and gas industry followed somewhat later when they discovered that adopting it made them money – lots and lots of it – with safety being a welcome side effect.
The solution has created a world where it is safer to fly than it is to drive to the airport. Where more people are killed by petrol in their own garages than in the entire global petrochemical processing industry. Where we enjoy astonishing improvements in quality, safety and utility of the products we consume at either the same or diminishing cost (think of the mobile phone you use now compared to ten years ago).
The solution is a culture, a state of mind and a way of doing things – it is a committed, system wide and systematic approach to reliability.
The commonest argument used against the proposal to adopt a reliable approach to delivering healthcare goes along the lines of – ‘Patients are not widgets’; ‘Jumbo jets are more reliable than patients’; ‘There is so much uncertainty in medicine’; ‘This constrains my freedom of practice’ etc. Blaming patients for unreliable healthcare is, however, a highly disingenuous argument. Patients are unreliable, they do present us with enormous problems of variance and deviance from the expected, medicine is difficult – but that is their nature, that is their right. Putting them into an unreliable healthcare system produces variance on variance – which, I have said before, is the definition of chaos. Unreliable healthcare results in unsafe medicine, uncertain and poor outcomes, errors that are destined to be repeated (like history) all of which, not withstanding the human misery, costs. It has been estimated that nearly half of all health care costs are related to failures in the delivery of care.
What is Reliable Healthcare Delivery?
The trite (and not very helpful) answer to this question is the familiar ‘Doing the right thing and doing things right’. This particular definition ignores the rather large zone of uncertainty that exists between what we know is the right thing and what we know is the wrong thing. If we were to pause and reflect on the state of medical knowledge and draw a diagram representing each of these three zones – what we know is right (white), what we know is wrong (black), where there is room for argument (grey) – how big would we draw each of the zones? What strategies would a reliable healthcare delivery system adopt in the three different knowledge zones?
The White Zone – Doing the right thing the right way
The truth of the matter is we have a huge amount of medical knowledge – there are very many areas of medicine where we know what the right thing to do is and how best to do it. The medical knowledge base is vast and increasing exponentially (see graph below) and there are swathes of medicine where the important unanswered question is not ‘What is the right thing to do?’ but ‘Why (IGN) are we not doing it?’. One of the main barriers to deploying medical knowledge appears to be the shear volume of new information – over 1 million original medical papers were published in 2010 alone. However we have allies to help us – there are expert groups, royal colleges, specialist societies, NICE, Map of Medicine, Bandolier, national and international consensus bodies who are systematically collecting, sifting and grading the evidence for us and telling us what we should do.
Yet as McGlynn et al. discovered we still don’t do it nearly 50% of the time (see table below). Why? There are healthcare organisations that do take a systematic approach to doing the right thing – Intermountain Health in Utah USA is one of the most outstanding high performing healthcare organisations in the world. Their outcomes for most common medical and surgical diagnoses are way above their peers – their mortality from sepsis is 9.3% compared to a US average of 25% – 40%. They are one of the few organisations in the US that makes a return on Medicare and Medicaid reimbursements. They invented the care process model – of which more later.
There will be doctors reading this (most of them) that are convinced that they are practicing to the highest and most up to date standards, and able account for that standard of care they provide. Yet the outcome from their institutions will come nowhere near those of Intermountain. Whilst they may account for their own practice they will undoubtedly be a little more taciturn on their colleague’s practice, and perhaps a little more vocal about what they perceive the standard of care provided to patients before they arrived in their care and maybe after they left as well. And there is the rub – for to deliver outstanding outcomes we have to do the right thing every step of the way on the patient’s journey – for every patient.
A good outcome is the aggregated marginal gains of multiple inputs by many professionals. Reliable healthcare is a team sport and as Atul Gawande put it in his tour-de-force ‘The Checklist Manifesto’ we are still practicing the medicine of the heroic individual – we are only just emerging from the buccaneering age of medicine, equivalent in the airline industry of when aircraft routinely fell out of the sky. Unlike buccaneering pilots though doctors don’t die with their mistakes.
The problem with doing the right thing is that it is rather mundane. It involves being told what to do, it involves following checklists, care bundles, protocols and pathways. It also involves agreeing with your team ‘how we are going to do things here’; that inevitably involves negotiation and compromise, going along with the consensus because doing so is for the greater good. Heroic doctors are not very good at doing those sorts of things. It also takes a lot of collaborative effort to get there.
There is still plenty of scope for the heroic doctor though. To paraphrase Atul Gawande again ‘Checklists are there to get the 80% of mundane stuff right so that the mind is freed to do the heroic 20%’. It is vital to get the 80% right – otherwise our heroics become expensive futilities.
The Black Zone – when it goes wrong
Understanding and managing medical error is a huge topic in of itself – which I will undoubtedly expand in future posts. I am though in a hurry (for a change) to move on to the bit I am interested today which is the grey zone. Suffice to say for now that you cannot be a highly reliable healthcare organisation if you do not manage medical error well.
The Grey Zone – Learning from uncertainty
This is where it starts getting very interesting. In our daily practice as doctors our patients constantly present us with dilemmas. Situations where doing the right thing seems to be the wrong thing, or where doing the right thing for one problem is definitely the wrong thing for another. As our patients get older and compound multiple pathologies these dilemmas increase all the time. Each time we are presented with these situations it feels as though we are solving the problem for the first time over and again for each patient – we are faced with the huge and overwhelming variance in presentation and response to treatment.
The traditional medical model for dealing with these situations is the ‘iterative care process’ underpinned by the ‘experienced clinician’. This care process involves a combination of medical detective work (history and examination), Diagnostic hypothesis (differential diagnosis), Diagnostic tests or a ‘Diagnostic Therapeutic Trial’ (we’ll give antibiotics and see if they get better…). Experience helps by being able to ask the right questions, come up with a feasible and manageable list of diagnostic hypotheses and also design a diagnostic prescription that does the minimum to confirm or refute the hypothesis. Patients will often go through several cycles of this process (either because it doesn’t solve the problem or because the patient has moved to a different team) – with escalating intensity and invasiveness of investigation – until either a diagnosis is made and correct treatment started or the progress of the condition outpaces the process and the patient succumbs (or they get better despite us).
There are several reasons why this model is problematic and fails to deliver reliable care. The first is that the operating model of the modern hospital (the process by which patients move through the organisation) is not aligned to the iterative care process. The second is we are not very good at it any more – we don’t have enough experienced clinicians to see and review patients progress through the care process in a timely or frequent enough way. Thirdly it is slow, expensive and unreliable. Finally it does not deliver learning at anything other than an individual level – hence the very real sense in which we feel we are re-solving the same problems day-in day-out without seeming to make much progress.
The complexity of modern medicine and modern patients, the dissolution of traditional medical teams and their replacement by the transitory, multi-professional, socially complex, modern alternatives – means we need a new model for delivering care.
This medical model needs to achieve several things:
1) Reliable delivery of care that we know is right that does not depend on the location of the patient, or the presence or absence of a particular professional.
2) Keep up-to-date with the ever expanding body of medical knowledge
3) Generate ‘Team Learning’ from variance in presentation and response
4) Generate ‘Team Learning’ from medical errors
5) Use that learning to modify and adapt the care process
The ‘Care Process Model’ invented by Intermountain achieves all of these things. On first glance it appears to be a set of protocols – and many subsequent implementations have deployed it as such. However seeing it as such is missing the point – the content of the protocol is much less important than the process by which it comes about and the way that it is used.
The diagram below shows the essence of the care process model development and more to the point continuous improvement through organisational learning.
The essential components are:
1) An expert team that crystallise the current state of medical knowledge into detailed guidance
2) Clinical senate that simplify and standardise to a deployable protocol across the whole system
3) An operating system that reliably delivers elements of the CPM at all points of the pathway – the protocol is the record
4) A reporting and monitoring system of deviance from the pathway – an expected deviance of 20% is built in – but all deviance is reported and analysed including medical error
5) Outcomes are monitored
6) Information is fed back to the clinical senate that adjust the CPM
You will see that this generates organisational learning – the system gets smarter – and that it does so using three distinct knowledge types – The global medical knowledge base generated by original research, knowledge of its own patient group through analysis of unexpected response and outcome, knowledge of itself through analysis of medical error and non-compliance.
In my next post I will discuss the pre-requisites in culture, structure and process that are required to create organisational learning and how rare they are in the NHS.