Lean Service Blog

If you think targets don't influence service design

2010-02-02T15:42:00.003Z

A great programme from The Report on BBC Radio 4 from 17th Dec 2009 on how response times targets for ambulance services are distorting service design to concentrate resources too much in cities and towns and to have too many single responder vehicles that send a single paramedic that can't carry patients rather than a full ambulance to too many cases.

Plus, where does the 8 minutes come from in the first place?

Best,

Rob

PDSA applied to PDSA

2009-12-21T13:00:00.003Z

I was once at a meeting of consultants and we had got to the end of an interesting day learning about new ideas regarding organisations and how to improve them. The volunteer chair of the meeting got to the part of the meeting where we would suggest things that were good and things that we should improve upon for next time. He decided to divide the flip chart into four quarters and label them each Plan, Do, Study, Act.

Now Plan, Do, Study, Act (PDSA) is the Shewhart Cycle (some know it as the Deming Cycle) which is a generalisation of the scientific method as applied to improvement in general.

Plan: make a plan or a hypothesis about what you are going to do.
Do: try it out, run the experiment.
Study (sometimes called Check): see how the plan came out.
Act: implement the findings. Either embed the new way or throw it out.

Then you cycle back to Plan with new information.

Anyway, back to this meeting. We tried our best to fill in the four quarters of the chart with our good and bad points about the meeting but try as the chair could he couldn't explain what he meant in this context by the four quadrants in this context. As far as we could all see we were in Study. We had just Done the meeting and were Studying what we had done for the next Plan for the next meeting. A big row broke out because apparently the chair applied PDSA to everything without problems.

The only new information I gained was that you should never force a technique on a group of people if it is irrelevant and you can't communicate it properly.

Best,

Rob

Targets are all SMART, let's make measures VVAPID

2009-12-18T12:01:00.004Z

For many years I have been told that targets should be SMART. This means they should be Specific, Measurable, Achievable, Realistic and Timed. Well actually, all targets are already SMART:

Simplistic
Meaningless
Arbitrary
Road blocks
Torture

Simplistic

Targets are simplistic. Targets are a sledgehammer to crack a nut and they come with no context or method. Targets are imposed by managers and governments and they walk away thinking by setting a target that their job is done.

Meaningless

Targets don't have any relation to the system they are meant to apply to. They can't have. Targets are not related to purpose. Hitting or missing a target does not give you any new information. Having a target set does not give you any new method to improve it simply gives you something to focus on. And not anything useful. Aiming for a target does not give staff or managers any new understanding of the system. They will be so absorbed trying to hit the target they may well be even more oblivious of what is happening around them.

Arbitrary

An understanding of variation gives the insight that any level at which you set your target is wrong. A study of common cause variation for any measure will give you upper and lower limits within which the system will exhibit predictable variation. Setting a target above or below the limits means the target will always/never be hit (depending on whether higher or lower is better or worse). Setting the target at a level between the limits means that it is virtually random whether the target will be hit in any given period. Therefore there is no reliable way to set a target and hence all targets are arbitrary.

Road Blocks

Targets distract from genuine improvement by sapping energy toward the collecting, analysis and reporting of useless data. Worse, targets also actively get in the way and drag down performance. Think of the 4 hour A&E target that induced one hospital trust to keep patients in ambulances outside A&E, only letting them in once they were sure they could hit the target. Schools are measured and ranked on exam results and so they cheat to get the brightest children in their intake. This is not improving education for all, this is a road block to improving education.

Torture

Targets mean pressure to achieve meaningless, arbitrary, numbers. This is tortuous for staff, managers and inspectors alike.

Staff are judged, rated and rewarded by whether they hit arbitrary targets. They know that they have to apply their minds to hitting the targets when that means they have to degrade overall performance to do so. When staff have to cheat to hit targets that has an effect on morale, self-esteem and respect for their managers who are putting them in the position of having to do these things.

Managers are ravaged by the constant hitting and missing of targets in a seemingly random way. They can never seem to get a good explanation about why a target was missed even though when it is hit they tell everyone they can.

Inspectors often know that what they are doing is ruinous to the performance of the organisation that they are trying to judge. They know that most of the time they are simply assessing compliance to a standard and not looking at an intrinsic ability to give good service. This knowledge must eat away at a person.

Targets are tortuous to users. Using a target driven service will be boring, annoying, frustrating and perhaps dangerous and yet they will be constantly being told that the organisation is hitting or making good progress toward its targets.

So let's get rid of this culture of the target, SMART or otherwise, and move toward measures that are VVAPID.

Value
Variation
Aligned
Purpose
Informative
Deliver

Best,

Rob

Can't get no dissatisfaction

2009-12-15T10:00:00.006Z

The light works in my girlfriend's kitchen.

I bet the light works in your kitchen and you don't even think about it. Well the strip light in this kitchen hasn't worked properly for ages. For the want of 99 pence worth of a new starter, we have both been getting up on a chair to fiddle with the old starter to make the fluorescent bulb flicker and come on and stay on. We just hadn't quite got round to going to an electrical shop to get the starter. So up on a chair every time we wanted to turn on the kitchen light. We even took to leaving the light on all evening when we left the kitchen so that if we popped back to make some tea, we didn't have to get on the chair again to turn the light on again. Talk about a work-around.

The thing is we got used to it. After a while, getting up on a chair to make the light come on didn't seem so much trouble. We forgot that it was a bother. It became the way things were done.

I recall way back when, working for an investment back as an analyst messing with dozens of Access databases and Excel spreadsheets. One day I had to add an extra calculation to the daily work which would have meant a couple of days effort to update the Access database. Instead, because the trader wanted it that day, I exported the data to Excel, wrote a quick and dirty calculation in an hour, ran the calculation and then reimported the results to the database to continue the day's work. This extra procedure added 90 minutes to my day, every day.

I continued that extra step for months. I forgot that it was slow and cumbersome. I got used to it.

It took six months before it started to bug me and I got so bored of the extra step that I spent the two days implementing the calculation in Access. After that the calculation whizzed along in the blink of an eye. Those two days spent right at the start would have saved me 24 working days over the six months I waited. In fact the two days of work would have paid for themselves in only eleven days.

In order to implement change, big or small, you need some negative emotions. You need to be dissatisfied, bored, shocked, appalled, angry and critical. If you are tolerant, accepting, placid and content with your lot, nothing will happen. This is why change agents are always searching for the "burning platform" so they can get people to jump instead of having to push them off a "quite comfortable thank you" platform.

For an evening, the newly fixed light was a revelation. "Wow! We don't have to climb on the chair to turn the light on! Amazing!!" How sad that we take joy in things working as they always should have done. I recall my boss at the bank giving me a pat on the back for rewriting the calculation to save that 90 minutes a day.

A vision of a better way is nice and shiny, but how about a bit of tedium and rage to get us not just to where we should be, but beyond, to where we couldn't dream of? If only we could stop being so accepting of the messy, awful, boring, infuriating status quo.

Get moving. Get some dissatisfaction.

Best,

Rob

Jim Womack Reflects

2009-12-11T11:14:00.001Z

In Jim Womack's latest e-letter he describes a visit to the Arsenale in Venice where they pioneered flow systems in building war ships way back in the 15th century. All this looking backward made him wonder why Lean is not more widespread than it is. Reflecting on the spread of Lean he says,

...we haven't combined all of these tools and management methods in more than a few organizations.

Trouble is that the reasons why Lean hasn't been taken up as much as Jim and I would both like is hidden in that very sentence. Also from the e-letter,

It seems to me that we have already achieved several things of lasting value:

We have transferred and adapted lean process tools for production, product development, supplier management, and customer support to a wide range of industries in a wide range of countries.
We have experimented with all of the management tools - policy deployment, A3 analysis, and standardized management with kaizen - that are needed to introduce and sustain these process tools.

Again, the reasons for the low take up compared to the potential of Lean, are right there in those two very telling paragraphs.

The problem is the tools.

The best thing Womack and Jones ever wrote was the title of the book that came after The Machine That Changed the World. That book was called Lean Thinking. The title emphasised thinking (more than the book, I might add). This is the thing that people need to focus on. The tools are a red herring. It is the way that management and staff think that determines how they see they systems they work in and so how they try to change them.

Looking at work through a filter of a set of tools means that is what you see. If all you know is 5S, kanban, heijunka, poke yoke, work cells, supermarket pull systems, value stream mapping etc., then every problem is seen as an opportunity to apply one of these tools.

Every problem is instead an opportunity to learn. Every thing that is working badly is an opportunity to understand better how to improve.

Taiichi Ohno, the father of the Toyota production system, said, "don't codify method". He meant don't give things names, don't invent tools. When people ask me, "Which tool should we start with?", I ask them to guess which tool Toyota started with. The answer is they didn't start with a tool because they didn't have any. They started to understand their system and to develop solutions to the problems they encountered. These solutions have become codified as the Lean tools. Even the book Learning to See by Mike Rother, which has another promising title, is simply another description of how to apply a set of tools. It should be titled Learning to See Which Tool to Apply.

Jim Womack is in a considerable position of power in the Lean community and the trouble is that instead of reflecting and coming to the useful conclusion that he needs to drop the tools approach instead he is actually trying to extend it by inventing Lean Management Tools to patch up the poor effectiveness of the original Lean tools.

When the tools don't work, using more of them won't help matters.

Best,

Rob

Use and Misuse of Takt Time in Services

2009-11-30T19:39:00.004Z

I had an interesting discussion about takt time not so long ago and I have been mulling over what was said regarding the use of takt time in service organisations.

Firstly, what is takt time? Well 'takt' is a German word meaning 'rhythm' or 'drum-beat'. Takt time describes the average time between every instance of a demand on an organisation. The easiest way to explain it is to describe an example. (The numbers are not realistic. They are dreamt up to make the arithmetic easy.)

Imagine a car manufacturer who gets on average 100 new orders for cars per day, they run the factory for five hours so that is 300 minutes per day. Thus to fill the received orders they need to build 100 cars in 300 minutes each day. If a car popped off the assembly line following a regular drum-beat, that would mean a car finished every 3 minutes. Thus the takt time would be three minutes. This takt time would be used to level the production so that all the processes followed the beat. Now you can't make an engine in 3 minutes so you would have lots of engine assembling stations so you might make 10 engines at the same time that took 30 minutes each to make. That would be an engine every 3 minutes, on average.

The takt time will be used to balance the processes so that they all take a multiple of just under 3 minutes to keep the factory humming along, pushing a finished car out every 3 minutes.

This is informative when you are making things since all the materials to make a car need to arrive at the right time to be attached to the right car and a drum-beat can help organise and regulate things. You also have the option to spread the work over a day or a week. However, in service, takt time should be taken with a pinch of salt. Some might apply takt time to arrivals in A&E, 999 calls to the police or other on-demand services. You need to have a good appreciation of demand in all these services because you need to know how much resource you need to meet the demand. However, managers need to realise that however tempting takt time may be, in a situation where you can't balance demand and delivery processes, takt time doesn't tell you anything useful.

Let's think about the police. Take the same (unrealistic) numbers from the example above. A police service receives 100 emergency calls in a 5 hour shift. Again that gives a takt time of 3 minutes. But this time the managers can't balance the other processes, because there is variety in the demand. One call may be about the discovery of a burglary and another about a mass brawl in a pub. These two calls will not be equivalent in the same way that making two different cars or even a car and a van might be on an assembly production line. The calls will require different response times, need different personnel, varying numbers of officers and all engaged for different amounts of time both in the field and doing different paperwork back at the station.

You may wonder if you have a service process that is more like a assembly line, perhaps processing mortgage applications, then maybe you can find a use for takt time. Perhaps you can, but the point here really is you need to ask yourself what it is telling you. It is only the average amount of time between points of demand and following down the takt time route gives you the feeling that you can treat your office like a factory which may blind you and in turn cause you to break up work even further.

You should always seek to understand demand. It is just that "I want a car." repeated many times a day is much simpler than most of the demands placed on service organisations even if the car itself is more complicated than many services. In most cases, it is the use of Systems Thinking to identify value to the customer and hence the removal of delay, errors, waste and hence failure demand in a process that really brings benefit. Chasing concepts like takt time more often than not take attention off the real problems.

Don't dismiss these tools, but rather apply them with care in your organisations and never use a tool until you understand the problem it is trying to solve and whether you actually suffer from the same problem.

Best,

Rob

Targets in A&E make people cheat

2009-11-24T10:22:00.002Z

A couple of days ago I wrote about an analysis of Accident and Emergency department (A&E) waiting times that the NHS Information Centre released. It showed that a disproportionate number of patients were seen to in the 10 minutes before the 4 hours waiting time target. The Nursing Times ran a survey in response to that analysis which found that across the UK, trusts are cheating to be seen to hit the target. To quote from their article,

Forty per cent of nurses believe their colleagues are involved in helping to meet waiting time targets by underhand means, often referred to as “gaming”.

And one in 10 hospital nurses say they have personally been asked to engage in gaming to help meet waiting times this autumn.

The article goes on give a few examples of the type of gaming that happens:

Discharge times for patients are changed
Patients are temporarily moved from A&E to corridors, observation areas and theatre recovery wards
Patients are unnecessarily admitted to mixed-sex bays or specialist wards
In one trust, part of A&E was re-badged as a "clinical decision unit" and is now longer deemed part of A&E

This is no surprise. Targets make people cheat. It happens all the time and not just in the NHS. Remember the teachers in Bolton who were suspended over allegations of cheating to help students in GCSE language exams?

Targets do not, and never have driven improvement for patients they only drive inventiveness to try to hit the target by any means necessary. There is small cheating like adjusting discharge times and then institutional cheating by relabelling a part of A&E.

It is all wrong.

It is morally wrong but there is also the issue of the opportunity cost. All this effort and inventiveness is going into fiddling the system. Imagine if all that innovation went into improving the service. What about the boost to morale for the staff when they were released from the target and from the need to cheat? Think of the benefit to patients to feel that their care was the most important thing to the staff.

The other hidden problem is that when the figures get changed you no longer have reliable data upon which to base genuine improvement.

Anyone who thinks targets drive improvement has never been moved to a ward in order for the hospital to meet a target, only to be forgotten about because they should never have been transferred there in the first place.

Best,

Rob

Inspectors under fire

2009-11-23T09:33:00.006Z

In The Guardian today, Ofsted (Office for Standards in Education, Children’s Services and Skills) has come under fire from its former Chief Inspector, Sir Mike Tomlinson,

"Inspection systems that rely too heavily on data and tick-box systems is not what we need. I worry we are heading that way."

It was Ofsted who was accused of giving Haringey a top rating before the Baby P incident in its initial report then downgrading it after the case came to light. One tragic case does not make a bad system but then a good review from Ofsted doesn't seem to reliably indicate good system either.

In another incident it was also Ofsted who said that two police officers may not look after each others' children while the other was on duty. This was an entirely reasonable arrangement of two women supporting each other to work and get child care from a trusted friend. Ofsted judged that it constituted "receiving a reward".

If inspection reports cannot be relied on to give a reasonable indication of the state of the organisation being inspected there is a serious failing in either the inspectors, or the whole methodology and assumptions behind the method of assessment and review.

Ofsted was originally only inspecting education and childcare but was given Child Services to inspect two years ago. But a position paper (Nov 2009) from the Association of Directors of Children's Services, which starts by praising some aspects of Ofsted's inspection regime goes on to say,

"But, all too often a reductionist approach is taken to the inspection, moderation and judgements of services, particularly local safeguarding services where risk-averse approaches on the part of inspectors are leading to perverse judgements and unintended consequences. The perceived punitive effects and the impact of judgements on services in terms of the local media and political response are in danger of creating a climate whereby the inspected manage for inspection rather than managing for quality and outcomes for children and young people."

This was Systems Thinkers have been saying for years. Dr. W. Edwards Deming had as point 3 of his famous 14 points: "Cease dependence on inspection." Deming knew that inspection has many unwanted consequences:

People focus on passing inspections rather than doing what is right
Time, money and effort that could be spent on the work is spent on preparing for inspection
The fear of failing an inspection causes stress and worry which in turn reduces performance
The banality of tick-box inspection regimes makes people demoralised and they question why they are doing the job if this is how they are judged

However, Ofsted is not to blame. The problem comes from the widely held belief that quality can be inspected into an organisation. Quality does not come from outside, it is grown from within. That is not to say that outside expertise is not needed to bring new thinking or that some form of audit or inspection is not required, but that audit should be restricted to checking for probity in regards to money and resources, i.e. that no-one is perpetrating any fraud, and inspectors should throw away all the forms with the tick-boxes on them and ask one question,

"What are you doing to understand and improve the work?"

Then if managers are struggling to get understanding or to find or implement a method of improvement, the inspector can pull in assistance. Importantly this would be help and support, not punishment.

Ofsted and its ilk (The Audit Commission etc.) all need to reoriented to focus on helping organisations to improve rather than inspecting them to ensure that they comply with centrally mandated criteria of success. The removal of the current inspection regime would firstly free organisations to do the work without distraction and then the implementation of supportive improvement mechanisms would enable them to far exceed the dreams of any narrow inspection system.

Best,

Rob

More on 4 hour target in A&E

2009-11-19T11:49:00.002Z

The Times Online has an article about the 4 hour target data that I wrote about last time.

In the article, Mark Porter, chairman of the British Medical Association’s consultants’ committee, said that the admission rates were worrying.

“This suggests that when patients have been waiting close to four hours, there is a rush to discharge or admit them so that the hospital meets the four-hour target,” he said.

“Patients must always be treated on the basis of their clinical need, not simply because they have been waiting close to four hours.”

Katherine Murphy, director of the Patients Association said,

“This results in doctors making rushed decisions at three hours and 50 minutes, with patients having to be admitted inappropriately at huge cost to the NHS,” she said. “We have heard instances of ambulance drivers being forced to wait outside A&E with seriously ill patients, until staff have cleared a backlog of people who need to be seen within the four hour target.

“It is unfair to make NHS staff feel like they have to put meeting this target ahead of what’s in the best interests of patients.”

Mike O’Brien, the Health Minister defends the targets in the article. Why is it that ministers can not see the harm they are doing?

Best,

Rob

A&E targets distort behaviour

2009-11-19T10:22:00.004Z

Yesterday the NHS Information Centre published a further analysis of the time that patients wait in Accident and Emergency departments before being seen to. The analysis showed that just before the 4 hour target, there is a huge peak of people being admitted to hospital. See the graph below from the report.

The rising peak just before the 4 hour deadline is very clear.

There is an accompanying Excel spreadsheet which lets you see this distribution for each of the trusts that submitted data. Most of them exhibit this kind of pattern.

The report deals mostly with data analysis but little time is given for the reasons why people are treated and admitted in this pattern. I would postulate that the reason is clear. They have a target to deal with people within a 4 hour window and so either people have something relatively minor and they are treated quickly with no follow up or referral to their GP or they are left to wait. Then as they come closer to the 4 hours someone is there to make sure that the target is hit and people are admitted. There is a further chart from the report that is quite revealing.

This is a chart of the destinations of patients who are dealt in each ten minute slot. The bands dealing with 'referred' (white on top), 'others' (bright green under referred) and 'discharged - referred to GP' (dark green, second from bottom) are all fairly flat. The two categories that seem to cause the peaks are 'discharged - no follow up' (yellowy green) which peaks around the hour mark, and 'admitted' (black, on the bottom) which has a little peak in the first 10 mins and then rises to a very steep peak just before the 4 hour mark.

All the catergories drop-off to almost nothing after the 4 hour point.

It is a pretty safe guess to say that the 'discharged - no follow up' peak at 1 hour because they have minor issues so they can wait a little while, but then they can be quickly dealt with and go home. (A bit like my cut finger!)

It is the admitted patients that are of the most interest. The 4 hour target is skewing behaviour. Why are so many patients being admitted just before the target? It can't be coincidence. How many of those patients are being admitted solely to meet the target and not for clinical reasons? The thing that really gives the game away is the massive cliff-like drop-off after the 4 hour point. If there were a smoother drop-off after 4 hours, it would indicate that the system was behaving more normally, but the fact that almost no-one is left to be dealt with after 4 hours, means that patients must be being treated differently as they come up to the 4 hour point.

Now, isn't it a good thing that people are seen within 4 hours? Well yes and no. It is good if people are seen quickly, of course. Patients with life threatening conditions will surely want to be seen straight away and the rest of us walking wounded don't want to be hanging around A&E for no reason. But the problem is that this data shows clearly that the 4 hour target is skewing behaviour. Patients are being admitted to meet the target and not to give them the best care. Also if this is happening then it must follow that resources are being used to meet the target. There must be some mechanism that is letting staff know that a patient is about to breach the 4 hour target and then staff are diverted away from other patients to get the near-breach patients admitted. This way of managing resources is taking away from treating people as they arrive which otherwise might move the curve as patients are treated sooner.

The irony is this, by diverting resources to ensure that people don't wait more than 4 hours you are ensuring that people are more likely to wait longer and thus more likely to breach the 4 hours. And what of the waste of resources in the rest of the trust when people are admitted to meet the target and then take up beds, nursing time etc. when they don't need it?

This target needs to be dropped and replaced with measures of value to the patient. This will give understanding of demand so the service can be designed to meet the needs of patients, giving maximum value as quickly as possible. If trusts were to be freed from the burden of this target and then given a proper method to improve, they would be able to wipe the floor with the 4 hour target instead of being held back by it.

This data shows that while the Department of Health and central government think that this and similar targets are driving performance, in fact these targets are holding performance back.

Best,

Rob

London Health 09 conference

2009-11-12T09:39:00.003Z

On Monday and Tuesday this week I attended the London Health 09 conference. Day one was focused on local government and social care and day two on improving health outcomes for London.

All in all it was a very useful two days and it highlighted some really good things that are being done and also some fundamental problems.

It was attended by some very senior people including leaders of councils, advisers to the Mayor of London and senior executives in London NHS and local PCTs. They all had a very real need to do something good for the people in need in London and most of them felt a pull to do something different than had been done before. I was very encouraged by this, and if the rest of the health care staff in London are even half as committed then I am sure that some good things will happen.

I could write about many of the interesting talks but I will mention here a few that resonated or said something new.

The first was from Sir Michael Marmot from UCL. He presented research that showed that health inequalities were very strongly to do with social status. The bottom line is that if you are poor and uneducated then you will have a short life with multiple illnesses. His message was that even if social and health care were perfect they won't lift people out of poverty and so you will still have people who need lots of care just because of the start they get in life. It is other policy areas that need to deal with the cause and cure of that societal malady.

The second session I would like to highlight was a breakout session on telehealth. This is simply when you give patients the means (equipment) to monitor things like blood pressure at home with a direct link that sends the data to a nurse or other professional. This demonstrated that you can give people independence, save staff time and reduce the incidences of hospital admissions. This seems very innovative and everyone was very impressed but I don't think it is so clever. They said the units cost £2,000 to install, but you can get weighing scales that tweet your weight now. Having devices connected to the internet should not be so expensive. Or so surprising.

On the second day Dr Robert Chote from the Institute of Fiscal Studies showed us a lot of scary graphs that said that public finances are going down the toilet in the next few years. I think this is known in this community and a lot of the talk at the conference was about money and "doing more with less". Interestingly, I had a conversation over coffee with one panellist who said that he thought that because NHS pay had increased over the past few years and at the same time the mortgage costs had fallen, that meant that a lot of NHS staff were feeling personally a lot better off than they ever had. He thought that this personal feeling of extra wealth was preventing people from digesting the coming budget crash and that their comfort was stopping them doing the things they needed to prepare. One of the speakers, Jim Easton, NHS National Director for Improvement and Efficiency, addressed the audience directly and said, "This [budget] problem is coming and if you are not doing something now, you are burning time."

Lastly, there was a very interesting presentation from Conor Burke, Borough Managing Director, NHS Redbridge about an impressive scheme of polyclincs to join together social and health care services and run them together to remove duplication and budget wrangling between local authority and health services. I will comment further in another blog, but suffice to say there was a lot of talk about commissioning and joint working. Trying to make commissioning function better and putting together joint working between local authorities and PCTs is worthy, but it is really working around a big flaw in the system. The structure of the system shouldn't make it difficult to work together but all the mentions of joint working were mentioned in ways that made it seem like it was an amazing job, how wonderful was it that we can get together and do this. Working together for the purpose of giving the best care to the residents of London should be the number one priority. It shouldn't feel above and beyond to do that over simply delivering the services as mandated.

In the next few blogs I will talk more about commissioning, joint working, improvement methods (or lack thereof) and other topics.

Maybe I will see you at London Health 2010?

Best,

Rob

Total Place - integrated public services or more cost cutting?

2009-11-04T10:34:00.003Z

Total Place is a scheme to join-up public services in an area or a city. Trials have been run in Kent and Birmingham.

It may be useful to read the report of the Total Place Roundtable in The Guardian since I will reference some comments from it there.

The sub-title of the article is:

Total Place is a new initiative to examine how cutting out duplication in public service delivery can improve quality and reduce costs. But is this really a 'magic bullet' solution?

Well nothing is ever a silver bullet. As W. Edwards Deming said, "There is no such thing as instant pudding." But the general idea to join up services to remove duplication and concentrate more on prevention than cure, is a a good one that most people would agree with. There are some encouraging comments, such as:

Hospitals have been trialling specific units for patients with alcohol related problems, where they can be given preventive treatment, with co-ordinated interventions from across a range of agencies.

Figures from the police reveal that a single murder costs around £1.1m in services, from investigation to the legal and social services work, so the scale of preventive cost savings – especially in reducing gang violence, for example – is parallel with the moral gains.

"The focus needs to be on changing the culture and behaviour within public services, rather than fixating on financial outcomes."

These are encouraging but one comment from one of the participants worries me.

The table first heard an account of the Birmingham pilot, where £7.5bn of public sector cash had been mapped out. "The idea was to follow the money, and see where it led us," a participant explained. "Families are facing a range of issues, some not interconnected, but the challenge was to dismiss short-term thinking to analyse the £7.5bn coming into Birmingham every year."

If you follow the money, then you will come to the wrong answer. Budgets and costs are an output. First you should seek to understand the demand on the system and if you do that with the whole system in mind, for example how mental health services, social services and hospitals can work together on chronic alcohol problems, then that is all the better. But we shouldn't be starting with the money. That is using the tail to wag the dog.

Efficiencies in the these type of systems can be made, but we must ask,

What is our purpose?
What is of value to the public?
How can understand demand for that?
How can we design service to meet the demand, designed for the value and removing the waste?

If you do this properly you design public services that the public like and as an output you save money. You also get improvements in results that far outstrip anything that anyone would have dared to set as a target.

One more little niggle from the article:

Kent, a two-tier authority with a £10bn budget, now has a single phone number and single web portal for all local government services.

It goes on to say that "getting even this far was an uphill struggle". Well instituting a call centre isn't providing value but it is certainly following the money. Trouble is the failure demand that will ensue will cause costs to rise, not to fall. Having one place to call does not mean that the services delivered will be joined-up, in fact quite the reverse. Whereas, previously you might have had the call answered by someone who did the work in the department you called, now you get a call centre agent who is, by construction, removed from the work and only connected to it by workflow systems.

I wish the Total Place idea well. I just hope the practice would match the intent.

Best,

Rob

Book review: 'Instruction to Deliver' by Michael Barber

2009-10-27T13:51:00.002Z

I came across this book while browsing the politics section of Waterstones on Goodge Street. It immediately struck me as something interesting to read, a description of the delivery methods of the Blair government from 2001-2005 by the man who set-up and ran the Prime Minster's Delivery Unit. Now note that I fundamentally disagree with centralised targets and top-down imposed processes that this framework employed, but I was still curious to see how the other side tried to do it.

Firstly, overall it was quite a good read. Easy to get along with, clear and nicely paced. It got into detail often enough to make it revealing but it never got bogged down. The style was chatty but serious about the subject and the people involved.

Basically, the Delivery Unit was a programme office for the targets that government focused on. They were in Transport, Health, the Home Office and Education. The Unit reported directly to Tony Blair. They were there to provide method for collecting and reporting data, monitoring progress towards the targets and building capacity in the departments.

The fact that the use of central targets is wrong for me didn't take away from the basic project management messages that anyone trying to get things done need to reiterate on a regular basis.

Focus
Plan
Have method
Monitor
Communicate

The whole myth of "deliverology" that the Delivery Unit spawned, seems to me reading this book, simply effective project management.

Additionally, the parts that deal with the relationships with and between ministers, the prime minister, civil servants and the Unit staff itself are very interesting. There is a description of promises made to the Permanent Secretaries of each department. It was along the lines that we are here to help, we want to pass along the capacity to achieve and in so doing put ourselves out of work and that we are in it together.

I would recommend this book to anyone trying to enact big projects anywhere as a learning tool. I would, of course, warn readers away from the centralised targets and inspection regime they employed, but there is definitely something to learn here. It is just a shame that with the strong project skills described but with better method (e.g. understanding of systems, psychology and motivation) so much more could have been achieved in public services in the UK by now.

Best,

Rob

Worse and worse

2009-10-15T10:53:00.002+01:00

An article on the 'This is Cheshire' web site entitled "VOTE: Is hospital admitting A&E patients so it doesn't breach targets?" is a perfect example of the wider problems that come from targets and incentives based on activity.

Now I don't know the truth of the accusations made by NHS Warrington but I do know that targets do cause people to cheat. NHS Warrington say,

“There has been an increase in admissions and the number of attendances at A&E is staying the same. [...] There has been a 24 per cent increase in admissions this month, of whom the majority then spent less than 24 hours in hospital. [...] It seems driven by the fact we have to deliver four hour targets.”

That wouldn't surprise me in the slightest. There are often reports of people gaming their systems to hit targets to the detriment of the service as a whole.

But also,

The hospital was also accused of getting ‘more effective at coding’ so it can receive more money per procedure.

“In a matter of four months there has been almost a 10 per cent increase in the amount that NHS Warrington paid for hospital treatment,” said NHS Warrington chief executive Andrew Burgess.

Why would they want to get more effective at coding? Surely more effective coding is a good thing. We all want better coding. Don't we?

What we really want is a system designed to meet its purpose. In this case 'to provide the best care to every patient' might be a reasonable purpose.

But the problems with targets and payments by activity (which is what the coding argument is all about) are just the start. With this article we have a dysfunctional relationship between the PCT and a hospital trust. Are they very likely to work together to provide excellent services to their public? Are they likely to work together to study demand? Are they likely to make any improvements at all while they argue in public like this?

Worse, the public bickering is just eroding what confidence the local people have in the service.

So the problems of targets and wrong incentives don't just stop at their immediate effects, they have a profound knock on effect of organisational harmony and the ability to simply get on well enough to work together. This is a vicious circle.

But the other way can be virtuous. Use measures related to purpose, drop all the targets and forget about costs. Then get together with your partners and see how to improve things. Service will improve, costs will fall and managers and staff will enjoy doing a better job, morale will go up and everyone wins.

In short, the environment (the system) is the major cause of behaviour. Change the system and behaviour improves.

Best,

Rob

Now on Twitter

2009-09-29T12:16:00.004+01:00

For those of my readers who like that sort of thing, Worth Solutions has a Twitter account.

You follow us at http://twitter.com/worthsolutions

We will use Twitter to tell you what we are up to, give pointers to interesting news stories, little Lean Service tips and reminders and links back to this blog.

See you there.

Best,

Rob

I hope they don't fail

2009-09-03T10:03:00.004+01:00

thisisleicestershire.co.uk reports that the City Council got a bad report from the Audit Commission and if you believe the article fully it does sound like their is room for improvement.

But to tease out some of the targets for the next 18 months in the article:

Cut phone waiting times from 9 minutes to 1 minute
Average time to complete a claim to fall from 22.5 to 16 days
Stop over paying people to the tune of £6m over four years

For the sake of the people of Leicester I wish them the best of luck, because they are going to need it. I am pessimistic because of the quote right at the end of the article from council leader Ross Wilmot, "This set of targets will put us on the path to excellent performance." But also because "The money will be used to bring in more staff and to open phone lines and buildings for longer, to cut waiting times and speed up claims."

Having targets and getting more staff isn't going to help. I would bet a lot of money that the service has massive failure demand. They will have people calling to chase and complain about claims all day. That is why the phone lines are so jammed. If they would study their demand and have a purpose to pay people the right money as quickly as possible, they could then design their service against that demand and I bet they could get rid of most of the calls, visits and letters that they are currently getting, drop the processing times way below the targeted 16 days, to more like 2 or 3 days, and reduce the mistakes in payment to almost zero. And they could do it in a lot less than 18 months.

Thing is they seem to be stuck in the old way of thinking. "We have too much work so we need to spend more money to get more people and we will set targets and that will get the people to do what we want them to do."

As Deming said, "There is a better way."

Best,

Rob

Poor me

2009-08-20T11:55:00.002+01:00

I had a minor accident the other day. I was washing up a glass and I was forcing my hand down into the bottom to get it properly clean, the glass broke and I cut my little finger. It cut, what seemed to me, a big flap on my finger. There was a bit of blood and I stuck a dressing from my first aid kit on it. I resolved to go to Accident and Emergency to get them to check that there was no glass in the cut.

I arrived at the local A&E and had to queue for reception where I had to give my name and then fill in a basic form with name, address, my GP etc. That took five minutes. I then sat for about twenty minutes until called by a triage nurse who asked me what happened, clarified some things on my details. That took about three minutes. He then asked me to go back and wait. Which I did. For about fifty minutes. Then I was called through and was seen by a nurse practitioner who dressed the cut and gave me a tetanus shot. This took about five minutes. All done. No glass in the cut. Off you go.

Not too bad I thought. I then thought back on the steps and wondered if for my personal journey it might be better to have the nurse practitioner and the triage nurse doing the same job. So that instead of the triage nurse assessing and sending people to sit down again they could dress my wound straight away. And if you took the two people who were doing their respective triage and practitioner jobs and had them do both, lots of people with minor (and it was very minor!) injuries could be out much quicker.

Now we have to be careful here. I am a sample of one and before you redesign a flow you need to take into account the total demand on a system. The story above is one incident in an Accident and Emergency department, the system as a whole would need to be examined.

Interesting though...

The other interesting problem is the queuing to see the receptionist. Would we want our new triage/practitioner nurse to do that job as well? I don't know. But the clever part is that if we understood the process better and worked to understand the type and frequency of demand, the people who worked there including the receptionist and the nurses would be able to tell me better than I can tell you whether that was a good idea or not. The best thing a consultant or manager can do is give people the skills to investigate and understand for themselves how their work works so they can make improvements on their own system. So I am not worried that I don't know, because we would together.

Also while we are adding complexity. Should we just look at A&E in isolation? What about admission to wards? What about ambulance services? How about NHS Direct who tell people to go to A&E? And not forgetting GP services. They all fit together into a wider system of which an A&E department is but a small part.

The finger is healing nicely by the way. I wouldn't want you all to worry.

Best,

Rob

How would you like that?

2009-07-02T15:09:00.001+01:00

Whenever you are assessing a service or task you need to bear in mind the concept of an operational definition. It is the operational definition that spells out what someone means when they ask for something or for something to be done. By way of example imagine I am standing by a table and I ask you to clean it. Before you grab a cloth you need to know the operational definition of what I am asking. Do I mean clean enough to eat dinner at? Clean enough to write a letter on? Clean enough to prepare food on? Clean enough to perform open-heart surgery on? Or even, clean enough to manufacture silicon chips on? Without an operational definition we can't plan how to do something neither can we judge how well we have done it.

We can apply this to our organisations by asking what is the operational definition of the service we provide. Do we even have one? How do we know how to go about planning to provide the service and how to we judge how well it was provided? Note that it is important to write the operational definition from the point of view of the customer or user. Internally generated operational definitions only lead to confusion, bad service and frustration. Customer oriented operational definitions help toward clarity of purpose, good service and satisfaction for customers and staff alike.

Best,

Rob

Start With a Fanfare?

2009-05-14T11:52:00.002+01:00

A fairly accepted way to start a change programme is to announce it, or further to parade it with a big fanfare. "We are starting a new programme!!" Townhalls are held, emails sent and managers and staff briefed. Sometimes mugs and T-shirts are printed too. But I am on the cusp of wondering if that is all a good idea. I read Philip Crosby's book, "Quality is Free" a few years ago. I can remember two things from it. First, that doing things well is cheaper than doing things badly, and I agree with this. The second is that you must have a big unveiling of your new way. But the thing that struck me as odd was that he said - and I hope I remember it right - was that if it doesn't quite go as expected, the thing to do was to have another fanfare and gee everyone up again and keep doing that until it works. But if something is not working why would you keep trying it over and over?

The thing is anything that you can "start" you can "stop". So if you announce a new programme people can further down the line, decide that they don't want to do that any more and revert. So I am wondering if it might not be better not to announce a new programme. Might it be better to just start doing something different without a big to-do? How might that look?

It would have to start with the top. It would come from a Cheif Executive or equivalent. They would start to ask their staff to look at some different things, get some different data, think about things in a different way. They wouldn't be saying that Cheify has a new fad, instead they would simply be asked to go and try a couple of things. These things would be a priority, certainly and there would be no secret about what was happening, no subterfuge. Put simply the top officers would be given 'things to do' to help them see that there might be a better way. Thus, because there is no fanfare, there is nothing for staff to push back on. At first they might not see that the new things they are looking at are ground breaking, but the subtlty of the design of the tasks set are where this approach would live or die.

To clarify, I have not tried this approach yet, but I would look to do so. Instinctually I feel that if you are trying to change the way people think and thus work and do that in a way that sticks, you need to introduce it in a natural way, as part of the work. Not artificially in a big kick-off announcement.

Good luck with your programmes and I will try and report back if I can persuade someone to try this with me.

Best,

Rob

Types of help

2009-03-18T18:15:00.002Z

I sometimes try to think of the levels of help I give to a client - this equally applies to management assisting their staff. For the sake of a simple example, let us think of helping a child to tie their shoelace.

Level 1 - Do it for them.

Simply tie the child's shoelace. They learn that it can be done, but not how to do it.

Level 2 - Show them how to do it.

Tie the shoelace but go slow and describe how to do it. They learn how to do it but by demonstration only.

Level 3 - Teach them how to do it.

Show them, describe it, let them try, correct them when they err until they can do it for themselves. They learn how to do it by demonstration, trial and error and being corrected.

Level 4 - Teach them how to solve similar problems.

Don't teach them to tie the lace but give them skills to approach those type of problems. This could involve analysis (a tied lace has two loops, how might they get there?) and reasoning. They gain skill to enable them to solve similar problems in the future (how to tie a bow tie - which I still can't do right!).

Level 5 - Show them how to apply the above to others.

This is showing them how to be like you. The most powerful level since you have moved on extra level away in that they are passing on the skills that you practised on them.

The point is not that you should always strive to be at Level 5, more that you should know when to use the level that is most appropriate to the situation. Usually you will start at the lower levels and work your way up but it is not a linear progression.

From management point of view (or mine when I work with a client) if we don't get to level 5 at all, then the staff haven't been taught how to pass knowledge on and so any intervention will necessarily wither and die eventually.

Best,

Rob

Unsold cars

2009-01-22T13:46:00.002Z

If anyone thinks that the world's car manufacturers have been adopting lean in the last few years they can think again. The Guardian has some photos of car storage sites. Number six is the best.

The aim of the Toyota Production System is to 'produce cars at the rate of demand'. This means 'get an order, build a car, deliver it, get your money'.

A lean system should pull from the order. Dealers should have one or two examples of each model which are used for test drives and showing off features. When the order is made that should be the trigger to build the car, which should be delivered within a week.

There should be no question of manufacturers having fields full of cars or ever deciding to reduce or increase production. If there are fewer orders then there will be fewer cars made. The system flexes automatically.

In the seven wastes that lean people talk about - they are not applicable to service(!) - there is the waste of over-production. Taiichi Ohno thought that this was the worst waste of them all because it was the cause of many of the other wastes. What he would think of these pictures... (Interesting that there are some Toyota pictures in there. It doesn't say if they are already sold or not.)

Service is usually a bit different - you can't have fields full of mortgage applications pre-assessed before the customers come along - but you can still get bottle necks. You can still design a system that does not serve at the rate of demand. The "inventory" will more typically occur just after the order. There will have mortgage applications submitted by customers that are waiting to be looked at instead.

So, 'serve at the rate of demand' is still the game. Thus you must still have a fast, flexible system, that absorbs variety. More important you must understand your demand, for it is against that information that you design your system.

Best,

Rob

Who are the NHS really cheating?

2009-01-16T14:42:00.002Z

There was a article this week that made me laugh out loud. You can read it in full here. But I will quote from two paragraphs near the end.

The survey found more people than in previous years believed they were listened to by doctors and nurses. But waiting times were longer. The proportion saying they stayed in the emergency department for no more than four hours fell from 77% in 2004 to 73% in 2008.
This might appear to show widespread flouting of the government's target for the waiting time in A&E to be no more than four hours for 98% of patients. But the [Healthcare Commission] said trusts often moved patients from A&E into a nearby "admissions unit". Patients might not be able to tell the difference in location, but time spent in an admissions unit did not count towards the four-hour target.

There is so much wrong with this situation. Error upon, incorrect assumption, upon invalid method.

Let's start at the beginning. Firstly, the practice of using surveys to find out of your service is performing well is deeply suspect. I read a report to say that customers who report that they are satisfied on surveys are just as, if not more likely to defect to another company than those who said they were not satisfied. Why? Well because more people want the survey to be over as quickly as possible and if they had bad service and were thinking of leaving why would they want to give the company more time than they needed to. Obviously with A&E you don't get too much choice about using someone else. But the principle of using surveys is still flawed.

Secondly, the article quotes a statistic that waits more than fours hours fell from 77% in 2004 to 73% in 2008. Where is the understanding of variation? Maybe both those figures are within the predictable range and nothing has changed. The four per cent difference might simply be from common cause variation. We can't know from the figures as presented.

Thirdly, the fact that they have a target of no wait being longer than four hours will be causing problems and in fact longer waits. We know that it is causing them to cheat since the commission itself explained that hospitals more people from A&E to "admission units" to avoid going over the four hours. And the "patients might not be able to tell the difference". Of course they can't. Waiting in one room is the same as waiting in another room and even more annoying if you have to get moved half way through. So even though they should not have the target in the first place, the Healthcare Commission is condoning the cheating that the target causes in order to meet the target. And the big laugh is they are still not meeting the target. We could speculate that the resources need to have an "admissions unit" and move people to it, might be better used in the A and E department.

And if you read the rest of the article you will see that 9% of people who asked for pain relief never got it.

Ouch!

Best,

Rob

Improvement is a Puzzle

2009-01-14T16:49:00.002Z

I was listening to a podcast of a talk given by Russell Ackoff today and he talked about the difference between a problem and a puzzle. It is not about difficulty or anything obvious like that. The difference, Ackoff said, was that unlike a problem, there is something in a puzzle that means you have to change an assumption you have about it before you can solve it. (If you take this literally then they should be called "jigsaw problems" instead!) We've all done those puzzles where in order to solve them you have to change your thinking.

Which made me think that redesigning a system is a puzzle, not a problem because you have to change more than one assumption about how your organisation works.

The other interesting thing he talked about was the difference between analysis and synthesis. Analysis is where you take something that needs improving, fixing or understanding and you take it apart, examine (and perhaps improve) the parts then put it back together. With synthesis you start by understanding the bigger system that the system you are studying sits in.

He said that there are always important things about a system that you can only understand by synthesis and will never understand by analysis. For example, why do cars drive on the left in the UK? You may get an answer by looking at the history of transportation and the transportation system in general, but you will never get an answer by taking a car to pieces.

Similarly, you can juggle the internal processes of an organisation as much as you like, but if the barrier to improvement is in customer demand or in externally imposed targets or regulation, you need to look at the wider system and not just inside the organisation to understand these things.

Best,

Rob

What's In a Name?

2009-01-08T09:00:00.001Z

That which we call a rose
By any other name would smell as sweet;

Shakespeare, Romeo and Juliet

And indeed changing the name of the rose would not change the affect it would have on our olfactory sense, but it may affect our attitudes towards it. Would it be as romantic to buy a dozen red Snozwangers for a loved one on Valentine's Day this year, even if the aroma was as pleasant as it ever was?

I just wonder because I heard recently that one of the big telecommunications companies in the UK has been renaming its entity model. That it is what the things are called in its enterprise wide IT systems. The thing that used to be called a "customer" is now called a "contract partner". Now I will admit that I do have a contract with my phone company, but I wold not call myself a partner. For starters I don't get a share of the profits. But the more worrying thing is how this change might shifts attitudes to customers.

You might argue that this is a very slight argument and you may be right. But just stop to think how we might treat these things. A customer: is served, valued and is "always right". A "contract partner" is something: to negotiate with, to sue for breach of agreement, to work with according to SLAs and service standards.

How would you rather be thought of by this company? And more importantly, when that company comes to discuss issues of purpose and value as defined by their customers, will the fact that they think of them as "contract partners" distort the outcome?

So thank you good ~~reader~~ textual informational absorber, see you next time.

Best,

Rob

Lean London NHS Seminar - 26th Mar 2009

2009-01-07T14:32:00.003Z

There is a new Lean London event happening on the 26th March 2009. It is aimed at those in the NHS who are implementing or thinking of implementing Lean there. There will be case studies, some theory and a chance to meet others who are also interested.

Find out more and register to attend at http://www.leanlondon.org.uk/

Please tell anyone you know who might be interested.

See you there.

Rob