There is a phrase which goes around the wards and departments of many NHS hospitals:

Inappropriate attendance

This is most often used when talking about patients who pitch up to A+E with conditions which could usually be managed elsewhere.  These are the patients who are thought to cost the NHS a lot of money and are the target of various schemes to stream them to more appropriate settings.

So what are these ‘inappropriate attendances’?

I am very lucky to have benefited from a great education, am lucky enough to work in a pretty comprehensive health service, and because of my day job, have become pretty adept at navigating it, and getting help where and when I need it.

However, imagine if you haven’t had that benefit, and don’t know what a drop-in centre staffed by nurse practitioners can offer, or that you can get good advice over the phone from a GP out of hours, or that a pharmacist at the local chemists could deal with your ailment?  Well, what would you do then?

The decisions patients make about where to go for help are not simply random and unthinking, but they are made when patients are distressed, and searching for answers, and quickly at that.

The NHS is very good at taking a problem and designing a solution to it which works perfectly in a committee room, on paper and in a consultation document.  However, as Helmuth von Moltke the Elder once said, “no plan survives contact with the enemy”.  Similarly, no treatment pathway, referral criteria, single point of access or similar will survive in its original form, and the consequences are very often unintended rather than those set out by their architects.

Once we have set out a plan though, we often don’t recognise that patients (including ourselves) will follow the path of least resistance, and seek help where they will get it.  So if one sets up a system where the most efficient way to get a diagnosis for a funny rash “which isn’t a huge problem, but I don’t really want to take a day off work for it” is to go to A+E, then go to A+E the patients will.

We must recall also that “every system is designed perfectly to achieve the results it yields” (Paul Batalden)  In this case, we must recognise that we cannot force people to make choices which fit with our ideal, but instead that they will make choices which seem to them to make the most sense, and offer them the help the want as quickly as possible.

To change the way in which patients behave we must either match their behaviour (put urgent care centres staffed by GPs in A+E departments and hive off those we think are “inappropriate for A+E” to the GPs next door, or we must improve the alternative offer – and improve community services, awareness of community services such that they can compete with the A+E service to offer reassurance, diagnosis and therapy for those patients who seek it outwith office hours – however the promise of swift treatment and diagnosis at a hospital may prove too much – and the draw of A+E too strong 

There are positive points on both sides of the fence on this one, but one thing is clear – there really are very few cases where an attendance at A+E is “inappropriate”:

It may be that the patient didn’t want to wait and was playing the system – but then the system may be inappropriate, or it might just be that the patient was anxious, tired, scared and wanted some help. Equally, the patient might be lacking the skills and knowledge to manage and requires some additional information on how to navigate the complex health economies we have generated.

And I seem to remember that that falls fairly squarely into the lap of the caring profession of which I am proud to be a member.  It does not become us to castigate our patients for their lack of understanding or anxiety.


I am guilty, guilty as sin.

I think I commit this particular crime on a fairly regular basis, but I am trying to get better.

No, not speeding, not gluttony (although some might disagree), certainly not anything as serious as murder, but still, this crime has the potential to cause harm, and long-lasting harm at that.

So what crime is it, some sort of fraud?  I guess so… it is:

The Crime of Procrustes.

Procrustes was a robber (and wayward son of Posiedon) who lived in Attica and had a particularly nasty way of ‘helping’ his victims.  He would invite travellers on the road to Athens (or Eleusis) to come in and stay the night in his fort, promising them a comfy bed which had amazin properties – namely that it was a perfect fit for everyone.

The slight issue was how this bed managed such a seemingly impossible task.

Well, if you happened to be too long for Procrustes bed, you would be cut down to size, and if you happened to be a bit short, then there were handy winches at each end of the bed to ensure that you would fit just right (after a bit of stretching.) To catch out those lucky ones who fitted the bed without adjustment – he had two beds!

So, a nasty chap – and not an invitation you would want to accept – and thankfully, Theseus put an end to this practice by getting Procrustes to ‘fit’ his own bed when he stayed the night on his way to Athens.

So – what manner of torture chamber do you imagine I keep at home? Thankfully not – only a toddler bed which feels like torture if you are ever unfortunate enough to have to kip on it.

And what does this have to do with me?

The crime I am guilty of is of fitting information to the case I wish to make, rather than using it without prejudice to form an accurate picture of what is going on.

This is an easy trap to fall into, and causes real problems. The trap is often set unwittingly by those who are taking the initial details from the patient – be that the GP, the ambulance crew, or the A+E staff.  The initial phase of a patients admission is often a confused/confusing time, and the diagnosis is often attempted on incomplete information, or without the benefit of the increasing volume of data we generate about patients from the time they arrive in the hospital.

There is a desire amongst the medical profession to be right (it is, after all what we are drilled to do for 5 years or more at medical school) and there is a particular satisfaction in being able to tell a patient what is wrong with them.

Once enough information is gathered to have a reasonably firm diagnosis, it is usually written at the top of a differential – and becomes the working diagnosis.

The difficulty is that this suggestion, when a new pair of eyes comes to pick up the thread of the story, influences their thinking and sets them in the ways of Procrustes.  Further information as it is gained seems to reinforce the original diagnosis, and thereafter it is incredibly difficult to challenge it.

Now, there are a few caveats – the initial diagnosis has to hold some weight for this process to play out.  The medical tribes and hierarchies which exist mean that if the working diagnosis was made by someone that you consider to be less good than you, you will have little difficulty overturning it.  However, if someone of stature makes the initial diagnosis and plan – then it is increasingly difficult to turn the thinking around – and it somehow seems easier to fit the information to what we want to believe.

The key to avoiding this is being aware of the influences on you when you approach a case – especially if you are junior.

The data will be what it is, no matter what spin one tries to put on it. If things are not adding up, or there are inconsistencies in what is going on – go back to the beginning, and re-examine.  It has never ceased to amaze me how many diagnoses are challenged by the juniors on ward cover at night – and I am certain that it is because there is less distraction, more time to focus, and less fear of looking stupid if one questions the status quo.

So please, next time you are faced by a slightly ropey diagnosis, or are trying to explain away anomalies in what you are seeing – think again and try to avoid being guilty.

Clinical experience is paid for by patients

Misdiagnosis has been described as “the greatest threat to patient safety in the UK” and at a recent acute medicine conference was highlighted again as one of the major threats to patient safety which hasn’t yet been the central target of a big safety campaign.

Missing a diagnosis is relatively easy; the information we gain from patients of often through the lens of an individual who is frightened, delerious, or on occasion completely unconscious. The story very often changes overnight- a classic trap for a junior doctor (this song (lyrics) (iTunes) (YouTube) by The Amateur Transplants helps explain what a frustration that can be.)  The problem with the crime of procrustes is that we frame the information available to us in such a way that it supports our position, rather than using it to test our hypotheses and whittle down our differential to a definitive (or at least confident) diagnosis.

So how do we combat this, and in a system which is trying hard to make attendance at hospital ever more efficient and less time consuming?

In the ‘good old days’ there was three-layer clerking where the houseman would take the history, the senior houseman repeat it, and finally the registrar review it before the final, glorious history (now practised, refined and distilled into a thing of wonder) was presented to the consultant.

Nowadays there is less of this time consuming repetition taking place, and decisions are made more swiftly, on incomplete, less accurate information, and management plans enacted accordingly.

This has advantages and disadvantages.

The patient often has a, quicker path through the system, gets treated quicker, and one would hope, gets better sooner.  The turnover of patients in a department can be greater, and so the system is happy.

But what about the patient, and the doctors? If patients can avoid repeating their story a thousand times, and still get the right treatment, great. But, the reason stories change, and management changes so often in the first few hours is because new information changes the differential diagnosis, a different slant in thinking produces a new hunch and eventually, the voice of experience can declare on what the final, or at least working diagnosis should be.

Unfortunately this uncertainty over diagnosis when patients are admitted can lead to a “nebulised co-amilomoxifruse” approach where all differentials are treated with equal importance, and therapy is rather scatter-gun like rather than perhaps the pinpoint accuracy that the popular media present as the pinnacle of medical endeavour.

So how can one get that vital experience (a posteriori knowledge)?

Few junior doctors these days get to regularly present their patients to a specialist in the field of the patients problem, and then have a useful conversation about management and further investigation of the problem. Acute medicine is more often a game where one hopes to get on the fairway in the first 24 hours, on the green on the next ward round, and finally sink the ball into the diagnosis hole on the second day – without the continuity of a single firm then it is difficult to follow a patient through this seive of investigations and opinions before arriving at a final definitive diagnosis.

In general practice this can be amplified in the case of patients with acute illnesses – who may go on to have a prolonged and complicated hospital stay – and that initial diagnosis may never be confirmed or modified by the information sent back by the hospital.

To combat this, better feedback loops are required.  The learning cycle proposed by Kolb demands a period of reflection, gained from concrete experience to allow effective learning.  But reflection – without accurate outside information to ensure that the concrete experience is valid – will lead to a self-propelling myth that the trainee is always right. There is value in understanding our failings, and planning to rectify them in the future. (Tim Harford on this)

As Kierkegaard puts it:  life can only be understood backward, but it must be lived foward.

One way that used to be employed in a hospital where I worked was that the discharge summary of each patient seen by medical trainees was sent to them in the internal post so that they could reflect on what the ultimate outcome of the admission was, and what the final diagnosis was.

In these days of electronic patient records, demands that doctors keep a record of the numbers of patients they see (word document explaining how to calculate how many you might have seen) it should be relatively simple to provide such feedback and allow trainees to develop true experience in diagnosis and management of medical problems, rather than using a medicine by numbers approach which may see a patient through the first 24 hours, but may not influence their progress over the longer term.

As members of a profession which relies so much on the experience of our elders to filter and clarify the importance of information when making decisions on how to treat patients, we have a duty to develop better ways to make every encounter count for our junior colleagues – experience is hard won, and on the cost of every missed diagnosis is borne by a patient who may suffer as a consequence.