Putting Targets First

Sorry, NHS England, I know your 2013-14 business plan is called “Putting Patients First” but there seems to be a misunderstanding.  We’ve just learned from HSJ that Jeremy Hunt has called the 10 chief execs of hospitals who have the highest percentage of patients failing the 95% target to leave A&E within 4 hours.

What’s really driving the NHS is the Secretary of State kicking an arbitrary 10 hospitals who have “failed” more than an arbitrary 95% discharges within an arbitrary 4 hours. There is no clinical basis for 4 hours, or 95%, or 10.  As a patient I’d be frankly annoyed if I had to wait 4 hours for ten minutes of medical attention – on the other hand, if I had something serious which a hospital could diagnose and treat, then send me home after 6 hours rather than admit me overnight, I’d be delighted.  That is what “Putting Patients First” would be like, hospitals would benefit as they get very little marginal tariff for emergency admissions, and I think staff would like it too when they don’t have a ton of bricks fall on them because a clock went over 4 hours.  We need new measures.

A&E curve 1314

Median discharge time matters more to patients

Before we look at the data let’s agree on a GOOD THING – being treated faster is better, so long as it’s safe and appropriate.  And it’s a BAD THING to wait, to have inappropriate treatment or unnecessary handovers in the hospital.  The charts here attempt to look at performance from a patient point of view, so the time to be treated and discharged matters, and there’s nothing arbitrary about that.  All the data is from HES Online, most recent available April – August 2013, all English type 1 A&E hospitals.  The median time to be discharged is 125 minutes, just over two hours (we’ll come to admitted patients below).  Suspend judgment on whether this is good, bad or indifferent – it is what it is, at present. (15 minutes higher than it was in 2010 by the way).  Click on the chart for a better view.

There is one bar for each of the 137 trusts.  The red bars are the 10 “special ones” who have the highest numbers failing the friendly four hour target.  Amazing isn’t it?  None of the slowest 6 got the call.  They are spread right through the range, including what appears to be the fastest unit which has a median of just 64 minutes, half the average – better or a special cause? [later note:  it turned out to be non-comparable data]

A&E curve 1314 admitted

Median admitted times show effect of target

We move on to the picture for patients who are admitted to the hospital having arrived through A&E, which is about a quarter of them.  They tend to have much more complex problems but they count in exactly the same way towards the 4 hour target.  This is mystifying, as they are going to be in the hospital for much longer than 4 hours, so which department they are sitting in would seem of much less importance.  On the ground I’m told the problem is there are often no beds to put them in, so A&E can do everything right but lack of capacity elsewhere means the clock ticks on.  This is really a measure of the hospital, but thanks to media headlines and even the PM’s personal attention A&E gets all the flak.  (Apparently it’s really hard to find doctors to work in A&E, any connection?)

Compare the shape of the curve with the earlier one.  It’s much flatter at the high end, a telltale sign of a constrained reporting environment, let’s call it gaming, where there is enormous pressure to hit that arbitrary target.  It means that anyone sticking over the line will have CQC/Monitor/DH/NHSE all over them, I forget who’s in charge.  Sorry, JH and DC as well now.  But the really fascinating thing is that our “special ones” are randomly spread through the pack, just as before.  Well well, that surprised me. What this means is that even on the tougher to meet admitted patient pathway, 95% performance has no relationship to median time.

Admittion ratio curve 1314

Admission ratio all English A&Es

We need to go further, because the talk of A&E crises does not mainly concern A&E. With a tariff of £109 per visit attendances are a factor but are not going to break the NHS. The problem is emergency admissions, most of which start from A&E. What matters is the proportion of attendances which convert into admissions, 24% median, but with an astonishing variation between hospitals.  A small health warning on this data: it is straightforward admission ratio whereas a more precise comparison would be age-sex standardised (elderly patients are more likely to be admitted, and there is some variation in age profile of patients by hospital). However this would account for only a small part of the variation. In general, a lower admission rate is less costly and we cannot tell from this data how many are “inappropriate”, though a CQC report claimed these amounted to 500,000 last year.

The range from 7% to 46% at the extremes is truly amazing.  Even the main part of the curve runs from 16% to 33%, a multiple of more than 2.  Hospitals next to each other in the table are north, south, east and west, and case mix differences are not so large.  I want you to notice again that the “special ones” are randomly distributed through the curve.  So on this measure critical to the health of the whole NHS there is no relationship between those getting the attention and those doing a particularly good, bad or even interestingly different job.

In a previous blog A&E has a mountain to climb I looked at the corrosive effects of the four hour target on individual hospitals. Today we are seeing how the overall system management of A&E completely misses the point of what matters for patient experience, staff morale, and economic performance.  Worse, a recent NAO report showed evidence that the target increased admissions (p30).  You are wasting your time Mr Hunt on those phone calls, talking to the wrong people about the wrong things. An example: one hospital keeps patients waiting nearly 3 times as long and admits more than twice as many as one of those you called. But they have ticked all the CQC boxes and stayed off the radar.  All you will get for your efforts is more desperate compliance with the wrong target, while the real opportunity for improvement is lost.  This regime was not by you, it was Labour, an open goal if you want someone to blame.  Before long, one or other party is going to realise the folly of this mismanagement, abolish the target, lead real change and reap the political dividend. I just don’t know which one. Still, I live in hope.


Harry Longman


2 responses to “Putting Targets First”

  1. Paul Brady says:

    Eliminating the target will be a huge backwards step. Despite it’s “unfair”, “subclinical” and sledgehammer approach it has driven an unprecedented amount of service improvement in most departments and the hospitals they feed. It is only one of a balanced set of measures for any department, taking it in isolation is like poking yourself in both eyes and complaining it’s foggy. Failing the 4 hour standard is still is an indicator that something is not right in that emergency department whether it be lack of beds, high admissions, dependency and acuity of patients, clinical staffing, layout, poor management, culture or poor local primary care etc. Often it’s not the fault of the ED itself but it is still an indicator of a failure of the overall acute pathway that ED is part of. Abolishing it will have no effect on the good ED’s, will allow those in the middle to slip but be catastrophic for the ones that are part of a failing acute pathway. There are a few subgroups of patients who are clinically better served under the care of the ED consultants for longer than 4 hrs but most hospitals do not have the consultant capacity to do this.

  2. Harry Longman says:

    Yesterday I heard Dr Clifford Mann, President of the College of Emergency Medicine, describe the triple whammy on the A&E medical workforce: they can’t recruit, they can’t retain and those they have are losing the will to live. A decade of fabulous attention to their 4 hour target has left A&E departments so “improved” that doctors don’t want to work there. It wasn’t always like this.

    Perhaps you would read the blog again because it does address all your arguments. The next piece of evidence will show how waiting times in A&E have increased in an almost straight line over recent years. It’s no story of improvement.

Leave a Reply

Your email address will not be published. Required fields are marked *