Some time ago I got a call on a Friday evening from my footballing husband. Apparently a ball he considered to be a header was thought by someone else to be at kicking height and they had a head-boot interaction, resulting in a good Harry Potter imitation.
He said, “It’s bleeding a lot and the boys have called an ambulance…. they’ll make me go to X hospital”. So I got in the car and took him to Y ED, where it was kindly sutured by one of the consultants.
But later I wondered about where that left us regarding X ED. Why was he so keen to avoid it? What was wrong with it? How can we expect the public and policymakers to know which EDs they might and might not want to attend?
Well, this is what the “Friends and family test” is supposed to address. Since April 2013, patients discharged from the ED have (when we remember) been given a card on which they can comment on their care. But does a good F&F report guarantee a good ED? To quote the MPS
“Understandably patients experience difficulties in assessing the technical competency of a doctor, so will frequently judge the quality of clinical competence by their interactions with a particular doctor” (page 8).
In other words, patients may use the quality of your communication as a proxy for the quality of your care – often, but not always, true. Might this not be the same of EDs? Departments may have extensive local support and loyalty despite (for example) lack of supporting services or failure to be accredited for specialist training.
So what else might we look at? For many years the only metric applied to UK EDs was the “4 hour target”. The problem with this was the gaming it led to – for a summary of the distorting effect of targets even a 10-year-old can understand, see Inspector Guilfoyle. – Tom Locker and Sue Mason demonstrated nicely the peak in admissions just before 4 hours, with evidence that it was due to frank manipulation of performance data in some cases, and it persisted. Anecdotes exist of majors cubicles with doors being designated “Decision Units” so the clock could stop ticking, and we have to ask whether a patient is better treated spending 5 hours in the ED then being admitted to a bed on the appropriate ward or 3.5 hours in the ED and a further 8 hours in the waiting room on MAU (no stopwatch there).
Which of these graphs does your department resemble? And which would you want it to resemble? [Hint – the clue’s in the colours]
For a more up-to-date exploration of this and a great geographical analogy, see Harry Longman’s mountains.
This was recognised and in 2010 there was a move, supported by CEM and the RCN, to develop a wider set of quality indicators. Suggested parameters included adherence to relevant NICE guidance, mortality rates for specific conditions, time to specific interventions (eg door to balloon for AMI), proportion of patients with ambulatory care sensitive conditions admitted to hospital, senior review of high risk patients and staff experience.
Some of these are now being published on the Health and Social Care Information Centre website: left without being seen rates, unplanned reattendance within 7 days, time to initial assessment of ambulance cases, time to treatment and total time.
So what should we measure?
Back in 1966 Avedis Donabedian addressed this problem in relation to the whole of healthcare; he identified three facets of care that could be quantified: structure, process and outcome.
In terms of the ED, structure might be number of resus cubicles or whole time consultants; process door to balloon time or number of patients receiving analgesia; outcomes tend to be measured in terms of mortality (not as straightforward as it sounds – more of that in another post).
Now this is where I start to have trouble with our current setup; it’s all very well having lots of space and having a lovely short time to triage if the patients don’t benefit – after all, the patients are who we’re there for. So all these process and structural measures are proxies because we think they’ll improve outcomes. But we only know that for a very few: reperfusion time in AMI, prompt defibrillation in VF, and care in a specialist stroke unit being some. Processes and structures are easy to measure, but that doesn’t mean they’re the right thing to measure. I’ll let you think about that.