Targets are, oddly perhaps, seen as generally a good thing in the NHS. After all, who doesn't remember the horrendous waiting times that people experienced before they were enforced both for people waiting for operations and for A& E. A well constructed target, well implemented is necessary and works, goes the logic. Without them we would continue to live with the outrageous waits with, in some cases, people dying before they could get treatment. It is poor implementation that is responsible for the tragic consequences we hear about not the target itself, is the argument.
I don't agree. In fact, I would go so far as to say I agree with Andrew Lansley. We should drop targets. My point of departure with him is what you replace it with. My view is that in place of targets we should have measures derived from the work.
For some this might sound completely bonkers. After all, without a target, how would we be able to judge how well we are doing? And how else would you motivate people to perform?
For others, it might seem like a semantic debate. What's the difference between a target and a measure derived from the work? For me the difference is profound and goes to the very heart of how we think about what works. Targets always distort the performance of a system. The whole notion of targets is flawed. Their use in hierarchical systems engages people's ingenuity in managing the numbers instead of improving their methods.
For the people who think I am bonkers, let me give some examples. I vividly remember in my first job as a general manager looking at the waiting list for outpatient plastic surgery. The list was in good shape. No one was breaching, at least not according to the Korner waits (as was). These were the statistical waiting times which meant that the clock could be re-set to zero and your wait started all over again on the most flimsy of reasons including being offered appointments that were completely unsuitable for your circumstances.
What I wanted to know was how long people had been waiting end-to-end. From the start of their wait to the current date. I discovered one person had been waiting for over 5 years, just for their outpatient appointment. Their Korner wait? 8 weeks and well inside the maximum waiting time. It was unbelievable, perhaps more so because the person concerned still wanted the appointment!
So what, you might say, that happened years ago. We have come a long way since then. Have we? Consider this. In Rob Findlays and Anthony McKeevers summary of the waiting list position (here - £?) they point to real and terrible consequences of making people, the forgotten few, wait. They say; "Real examples include an infant listed for an operation because of a haemangiona (malformation of the blood vessels) on her wrist, who was still waiting in her teens. There was also a man whose operation was cancelled because of “no theatre time”. He was found, still waiting to be offered another date, 15 years later"
Then there was the shocking review by the Royal College of Surgeons which I previously posted about here. In that post I talk about those people hidden by the target, as if wearing a cloaking device, invisible to the system. In the specific examples given, highlighted by the Royal College of Surgeons report, the result of delivering a target was significant with much impaired outcomes and even avoidable deaths.
We know that there are some Trusts who have allowed their backlog to build up in order to continue to deliver the target. They are allowed to do this because there is a statistical quirk they can take advantage of. The target is measured against those who have been treated, not those who are waiting. So if I am in charge of a waiting list I can drip feed the list with patients who I know are going to maintain my target delivery position, instead of taking people (clinical priorities aside) in the order they are listed. It's a risky tactic because at some point the balance of the waiting list changes and the risk of reporting failure to deliver the target increases. We also know that in times of financial hardship, more lucrative cases will be listed ahead of others in some Trusts, irrespective of length of time waited. Shocking isn't it.
Then there is the issue of people being operated on unnecessarily. Evidence shows that involving people in decision making about undergoing an operation can result in a decision not to go ahead. Weighing up the balance of risks and the impact on lifestyle individuals can and do opt for different approaches. But systemically we don't do this. We list and treat or validate off the list, not understand and reach decisions together in the context of how people live their lives.
The other popular target is A&E. Again, I remember the war zone that was A&E during my time as a general manager. Some people even died on trolleys in corridors, much to my enduring shame. So getting people through in 4 hours has to be a good thing, right? The war zone and 12 hour + waits are now consigned to history - aren't they?
But take a closer look at the numbers and a different pattern emerges. I have found, and reports from the Information Centre support this finding, that as time spent in A&E increases, the percentage of patients discharged with no follow up required falls, while the percentage admitted increases. If you look at the graph they produce, which I have replicated, activity drops off a cliff at the 4 hour mark. It is a very unusual thing to see and worth taking a look at. The numbers will vary but the national data shows that, of those leaving A&E, the percentage of patients admitted to hospital peaks in the last 10 minutes before the 4 hour target is reached - over 60% in this time cohort will be admitted. Funny that.
In other words, what we think of as good performance, deliver the 4 hour target, is not when you view from the users perspective. And this fact is acknowledged because in countless publications there is a recognition that large numbers of people in hospital beds do not need to be there. But large numbers of them are there because of the four hour target. And they are then put at increased risk as a result.
And for the unlucky ones, it is the start of a flightpath to dependency and avoidable cost to the system and the catalyst for many of the stories we hear reported in the press. What I have seen is a process where people are passed from service to service, professional to professional with labels like confused or incontinent when they are neither.
Of course no-one wants to wait unnecessarily long periods of time for their treatment. But because we approach it from a command and control logic, the solutions are based on a resource management logic, where capacity needs to be increased. More doctors, nurses, other stuff to get people treated. So a four hour target is, in part at least, a response to a system wide inability to respond to peoples needs for urgent care in a way that solves their problem when they put their hand up for help.
It's not much better for staff either, as I discussed here. As the Mid Staffs enquiry is hearing, the pressure of delivering targets has devastating consequences for staff as well as patients and their loved ones. And, of course, targets limit innovation and creativity as highlighted by Bevan amongst others.
Is there a better way?
The short answer is yes. The lingua franca of targets might mean they are considered normal but they are in fact counterproductive and should be changed. They should be replaced by a lingua franca about how the work works and how measures should be designed that support the delivery of purpose from a users perspective.
The better way to improve capacity is to remove waste. Creating more capacity in and for A&E for example - more doctors, nurses, beds etc, simply adds more resource to a wasteful system, compounding inefficiency. For elective care the level of waste created by this kind of thinking has been enormous. Only 85% of the elective work that had been paid for in ISTCs had actually been done according to the DH in 2008. For diagnostics the position is even worse with only 25% of the work that had been paid for actually being completed.
Better is to put the following principles at the heart of all strategy and operational management:
I realise that this is quite a long post. In fact I have plenty more to say. I have, after all, only touched on a fraction of the targets that are imposed arbitrarily from people who are not in the work. There will be those that, so far as the 18 week wait is concerned at least, will not be persuaded. After all, the longest waits are occurring in relatively few hospitals. Let's focus on them. Leave the target alone, to do otherwise is sheer madness they might say.
But however counterintuitive this sounds, all I have said is based on empirical evidence. And I am seeing a better way through my experiments, the results of some of which I have already posted about but there will be more to follow. Results which are profound for the system as a whole and the people who rely on it, not just bits of it.
I don't agree. In fact, I would go so far as to say I agree with Andrew Lansley. We should drop targets. My point of departure with him is what you replace it with. My view is that in place of targets we should have measures derived from the work.
For some this might sound completely bonkers. After all, without a target, how would we be able to judge how well we are doing? And how else would you motivate people to perform?
For others, it might seem like a semantic debate. What's the difference between a target and a measure derived from the work? For me the difference is profound and goes to the very heart of how we think about what works. Targets always distort the performance of a system. The whole notion of targets is flawed. Their use in hierarchical systems engages people's ingenuity in managing the numbers instead of improving their methods.
For the people who think I am bonkers, let me give some examples. I vividly remember in my first job as a general manager looking at the waiting list for outpatient plastic surgery. The list was in good shape. No one was breaching, at least not according to the Korner waits (as was). These were the statistical waiting times which meant that the clock could be re-set to zero and your wait started all over again on the most flimsy of reasons including being offered appointments that were completely unsuitable for your circumstances.
What I wanted to know was how long people had been waiting end-to-end. From the start of their wait to the current date. I discovered one person had been waiting for over 5 years, just for their outpatient appointment. Their Korner wait? 8 weeks and well inside the maximum waiting time. It was unbelievable, perhaps more so because the person concerned still wanted the appointment!
So what, you might say, that happened years ago. We have come a long way since then. Have we? Consider this. In Rob Findlays and Anthony McKeevers summary of the waiting list position (here - £?) they point to real and terrible consequences of making people, the forgotten few, wait. They say; "Real examples include an infant listed for an operation because of a haemangiona (malformation of the blood vessels) on her wrist, who was still waiting in her teens. There was also a man whose operation was cancelled because of “no theatre time”. He was found, still waiting to be offered another date, 15 years later"
Then there was the shocking review by the Royal College of Surgeons which I previously posted about here. In that post I talk about those people hidden by the target, as if wearing a cloaking device, invisible to the system. In the specific examples given, highlighted by the Royal College of Surgeons report, the result of delivering a target was significant with much impaired outcomes and even avoidable deaths.
We know that there are some Trusts who have allowed their backlog to build up in order to continue to deliver the target. They are allowed to do this because there is a statistical quirk they can take advantage of. The target is measured against those who have been treated, not those who are waiting. So if I am in charge of a waiting list I can drip feed the list with patients who I know are going to maintain my target delivery position, instead of taking people (clinical priorities aside) in the order they are listed. It's a risky tactic because at some point the balance of the waiting list changes and the risk of reporting failure to deliver the target increases. We also know that in times of financial hardship, more lucrative cases will be listed ahead of others in some Trusts, irrespective of length of time waited. Shocking isn't it.
Then there is the issue of people being operated on unnecessarily. Evidence shows that involving people in decision making about undergoing an operation can result in a decision not to go ahead. Weighing up the balance of risks and the impact on lifestyle individuals can and do opt for different approaches. But systemically we don't do this. We list and treat or validate off the list, not understand and reach decisions together in the context of how people live their lives.
So, those who support the imposition of targets and do so comparing what it used to be like, may be surprised to find that many of the problems that individuals faced then are still occurring now. People are still waiting too long, potentially being over treated and, in some cases, having impaired outcomes or even dying as a result of the target. And people are still cheating or gaming the system. Facts that are hidden from view. It's just that the numbers we are talking about (around 20,000 compared to nearly 600,000 a decade ago) are smaller. So that's OK then.
But take a closer look at the numbers and a different pattern emerges. I have found, and reports from the Information Centre support this finding, that as time spent in A&E increases, the percentage of patients discharged with no follow up required falls, while the percentage admitted increases. If you look at the graph they produce, which I have replicated, activity drops off a cliff at the 4 hour mark. It is a very unusual thing to see and worth taking a look at. The numbers will vary but the national data shows that, of those leaving A&E, the percentage of patients admitted to hospital peaks in the last 10 minutes before the 4 hour target is reached - over 60% in this time cohort will be admitted. Funny that.
In other words, what we think of as good performance, deliver the 4 hour target, is not when you view from the users perspective. And this fact is acknowledged because in countless publications there is a recognition that large numbers of people in hospital beds do not need to be there. But large numbers of them are there because of the four hour target. And they are then put at increased risk as a result.
And for the unlucky ones, it is the start of a flightpath to dependency and avoidable cost to the system and the catalyst for many of the stories we hear reported in the press. What I have seen is a process where people are passed from service to service, professional to professional with labels like confused or incontinent when they are neither.
Of course no-one wants to wait unnecessarily long periods of time for their treatment. But because we approach it from a command and control logic, the solutions are based on a resource management logic, where capacity needs to be increased. More doctors, nurses, other stuff to get people treated. So a four hour target is, in part at least, a response to a system wide inability to respond to peoples needs for urgent care in a way that solves their problem when they put their hand up for help.
It's not much better for staff either, as I discussed here. As the Mid Staffs enquiry is hearing, the pressure of delivering targets has devastating consequences for staff as well as patients and their loved ones. And, of course, targets limit innovation and creativity as highlighted by Bevan amongst others.
Is there a better way?
The short answer is yes. The lingua franca of targets might mean they are considered normal but they are in fact counterproductive and should be changed. They should be replaced by a lingua franca about how the work works and how measures should be designed that support the delivery of purpose from a users perspective.
The better way to improve capacity is to remove waste. Creating more capacity in and for A&E for example - more doctors, nurses, beds etc, simply adds more resource to a wasteful system, compounding inefficiency. For elective care the level of waste created by this kind of thinking has been enormous. Only 85% of the elective work that had been paid for in ISTCs had actually been done according to the DH in 2008. For diagnostics the position is even worse with only 25% of the work that had been paid for actually being completed.
Better is to put the following principles at the heart of all strategy and operational management:
- Purpose – measures – method
- Design against demand
- Partnership through collaboration and mutuality
I realise that this is quite a long post. In fact I have plenty more to say. I have, after all, only touched on a fraction of the targets that are imposed arbitrarily from people who are not in the work. There will be those that, so far as the 18 week wait is concerned at least, will not be persuaded. After all, the longest waits are occurring in relatively few hospitals. Let's focus on them. Leave the target alone, to do otherwise is sheer madness they might say.
But however counterintuitive this sounds, all I have said is based on empirical evidence. And I am seeing a better way through my experiments, the results of some of which I have already posted about but there will be more to follow. Results which are profound for the system as a whole and the people who rely on it, not just bits of it.
Having read the Conservative pre-election policy document "Outcomes not Targets" very carefully when it was first published, I get a sense of deja vu. That's not meant to imply anything about you, just to say that politically there were plans. However, in practice, the targets could not be dropped.
ReplyDeleteI know from talking to managers at my local hospital how perverse the targets can be. However, having been on a waiting list - in discomfort and with some fear - I can also attest to how popular targets are with patients. There has to be a middle way.
Here's a suggestion, well, two. First, achievement (or not) of targets should be informative. The problem is all too often that it is the providers problem without a recognition that perhaps that are other issues (lack of investment, demographics, affect of neighbouring providers etc). DH should attempt to help providers to achieve the targets, rather than assuming it is solely the provider's responsibility.
Second, patient choice. One of the most unsettling thing about being on a list is not knowing *when* you'll be treated. If you are given an absolute date - and guaranteed as much as possible that you will be treated on that date - then the uncertainty goes. I would argue that with an absolute guarantee of a date many patients will not mind if the date is more than 18 weeks away. And Dare I also suggest that if electives were carried out at *weekends* patients would be happier to have an appointment more than 18 weeks away if it meant that they didn't have to lose a day's work.
Get patients to choose: guarantee within 18 weeks, but at a date the hospital chooses; or your date, as long as it is more than 18 weeks away.
Spreading the treatments over a longer period will help hospital managers schedule treatments.
There's another huge problem in the way, and that is the capability of the system and people in the system to take on board operational management principles, whether they are badged Vanguard method, Lean, Six Sigma or whatever.
ReplyDeleteExperiments are required. And that means learning. And that means failures from which to learn. And the current macho, Theory X paradigm is that failure is not an option.
Leadership is required, not merely management.
Bruce Gray.