Ergodicity and its application to organisational management

It is extremely interesting the link between human biology, mathematics and management.

It is often true that poor performers continue to perform poorly – for instance most hospitals with a high standardised mortality rate continue to have it for very long times, difficult to shift. In case of human body, obese people often continue to be obese. This phenomenon is true for average performers and for high performers as well.

In the human body, this phenomenon is called homeostasis.

Mathematical explanation of performance can partly be understood by the law of large numbers. According to this law, if initial seemingly random results are observed over a longer period of time, they converge to the expected average results i.e. it "guarantees" stable long-term results for the averages of some random events

Often what looks like an improvement or worsening of performance if observed over a period of time essentially reverts to the mean, which means over a period of time performance remains unchanged; this phenomenon is explained by the normal distribution of most random events.

Well, that is how nature works. We can accept that if our organisation (or our health) is a top performing one and so mostly it will stay top performing over longer period of time. What if our organisation was a poor performer; the chances are it will continue to perform poorly; that would not be acceptable for a variety of reasons, especially if our organisation is in the business of healthcare. So when an organisation wants to become better, it embraces change in the hope that change will lead to an improvement. Often this change takes the form and language of transformation, reorganisation, change and other optimism inducing terms.

Here is when the concept of ergodicity becomes useful in leadership and management.

ERGODICITY

Ergodicity is the concept which states that:

·      The time average is the same as the space average

·      A system is Ergodic when the time average is the same as the space average

As an example suppose you’re trying to figure out the most popular park in London is. One method (time average) is to follow one person over a long period of time and see which parks he visits. Alternatively you can obtain a snapshot (spacial/statistical) average by seeing how many people are in a given park at a given point in time. The degree to which the time average equals the spatial average is called ergodicity and when the time and space average are the same then it is ergodic. 

Concepts of Ergodicity applied to Organisational Management

Applying this to organisational management:
 

1)    If you take the whole organisation average on any given day vs whole organisation average over period of time (say 3 months) – if they are the same, then organisation is erogodic

2)  If you take the average of one department on a given day and the average of the whole organisation on the that given day – if they are the same, then organisation could arguably be ergodic but may not be truly so by definition.

3)    If you take the best of one department on a given day vs the best of the whole organisation on on that given day – if they are the same, then the organisation could, arguably, be erogodic but may not be truly by definition. 

The three conditions provided above can be simplified into one to arrive at a fourth condition to demonstrate that a system is ergodic (note that conditions 2 & 3 are independent of time and hence alone are not sufficient to prove that a system is ergodic):

4) If you take the whole organisation average on any given day vs the average of one department over a period of time (say 3 months) – if they are the same, then the organisation is erogodic

·      This essentially could mean that a process even if random is or could be, stable

·      This is why organisations do plenty of activity, calling it transformation, change programmes, etc but yet do not improve as it is the nature of systems to show ergodicity.

·      It is okay for good performing organisations that are high performing to be ergodic.

That is very well, but what should organisations do to enable a higher level of performance and results irrespective of whether they are high performing or poor performing organisations do to improve?

Ergodic Transformation

(https://en.wikipedia.org/wiki/Ergodic_theory#Ergodic_transformations)

Here is the example from wikipedia 

·      if the set is a quantity of hot oatmeal in a bowl, and if a spoon of syrup is dropped into the bowl, then iterations of the inverse of an ergodic transformation of the oatmeal will not allow the syrup to remain in a local subregion of the oatmeal, but will distribute the syrup evenly throughout. At the same time, these iterations will not compress or dilate any portion of the oatmeal: they preserve the measure that is density. 

·      Ergodic transformation does a thorough job of ‘stirring’ the system without disturbing the fundamental nature of the system (due to the inverse of ergodic transformation seen in oatmeal example above)

·      This becomes a Measure Preserving Transformation – which means though there has been an addition (in the oatmeal + syrup example) the system remains ergodic (space average = time average)

Ergodic Transformation applied to organisational management

This is good for systems that are already high performing and want to become even higher performing.

For systems and organisations that are poorly performing a measure preserving transformation is of no use; hence the ergodicity of the system must be broken to see if better results can be obtained.

Ergodicity breaking

(https://en.wikipedia.org/wiki/Spontaneous_symmetry_breaking)

·      Spontaneous symmetry breaking: This is what wikipedia says “Is a spontaneous process It is a spontaneous process by which a system in a symmetrical state ends up in an asymmetrical state. It thus describes systems where the equations of motion or the Lagrangian obey certain symmetries, but the lowest-energy solutions do not exhibit that symmetry.”

Spontaneous Symmetry Breaking applied to organisational management

My lay application of this to organisational management is that when organisational symmetry breaks spontaneously though the overall organisation would be stable the processes and people within the organisations have changed so much that there could be significant improvement (there may be significant worsening as well, which is what the organisation has to monitor and prevent)

·      Explicit symmetry breaking (https://en.wikipedia.org/wiki/Explicit_symmetry_breaking) : This is what wikipedia states “this term is used in situations where these symmetry-breaking terms are small, so that the symmetry is approximately respected by the theory”

Explicit Symmetry Breaking applied to organisational management

·      By demanding explicit transformation we seem to end up (at least according to mathematical, physics theories) with changes that are very apparent, visible, but it does not transform the whole system. There is obviously cautiousness resulting only in very small changes, subject to resistance, these changes are planned, defined and delivered – yet the organisation does not change

Do we need to have spontaneous symmetry breaking when managing organisations?

·      Not always

·      If you have a high performing stable system you may not want this (you may want explicit symmetry breaking or measure preserving transformation so that you can get improvement without making the system unstable, improvement without the trauma of conventional transformation)

·      Yes – if you have a poor performing organisation you want spontaneous symmetry breaking so that we can have improvement with transformation – there is risk of asymmetry which may be beneficial (explicit symmetry breaking may actually be harmful by ensuring status quo while putting out an image/impression of change/improvement)

In summary

Ergodicity could be a useful principle when applied to operational management. Organisations, their systems and people have their own stability irrespective of whether they are high or low performing. To improve the performance of an organisation or to transform an organisation, it may be relevant to consider whether different approaches apply to high and low performing organisations. For high performing organisations it may be relevant to consider the concept of measure preserving transformation where there can be explicit induced changes which are absorbed as a part of good process measures which are maintained as the average increases.  For poorly performing organisations, it may be relevant to consider the concept of spontaneous symmetry breaking where beneficial asymmetry within the organisation is sought out to enable transformation; this means looking for areas within the organisation where people are attempting to or doing things differently and when they are beneficial to capture and systematise them even though the changes may not be compatible to what was externally mandated or top down defined for them; in poor performing organisations demanding explicit changes which are externally mandated or defined top down may result in status quo if lucky or could result in poorer performance making the organisation worse.

PS: Any mathematician, statistician, probability expert or physicist can educate me on this topic, especially enlighten me where I am wrong, I would be very grateful.

Acknowledgement: The above writing was advised and supported by Mr B. Patel

©M HEMADRI

Follow me M HEMADRI on Twitter @HemadriTweets

M Hemadri’s mini e-book 'Standardised Management Conversation' is available - click http://www.amazon.co.uk/Standardised-Management-Conversation-Hemadri-ebook/dp/B018AWBJTU 

MSc in Healthcare Improvement Leadership

Applications are now open for the second cohort of the MSc in Healthcare Improvement Leadership starting in October 2015.  This is a bespoke programme that has been developed jointly between University of Hull (Faculty of Health Sciences) Hull University Business School and Hull and East Yorkshire Hospitals and is open to all Healthcare Professionals.  

The first cohort (a mixture of clinical and non-clinical staff) have reported that the programme is “interesting, enjoyable and thought-provoking” as well as “blowing my mind with different perspectives of quality”

If you are interested in being a part of the second cohort and for further information please contact:-

Tracey Heath – Director of Enterprise

University of Hull

T.N.Heath@hull.ac.uk 01482464519

Further information:

MSc in Healthcare Improvement Leadership

A unique opportunity has arisen to undertake a Masters in Healthcare Improvement
Leadership at Hull University. This MSc programme is developed jointly between Hull
University, Hull Business School and Hull and East Yorkshire NHS Trust. This programme is
open to all healthcare professionals from primary, secondary and other health care sectors.

Aims

- To provide the participants with theoretical and practical understanding of the
concepts of quality improvement in healthcare delivery

- To equip the participants with practical tools to enable quality improvement

- To equip the participants with the attitude and ability to be a leader of healthcare
quality improvement

Objectives

- To explore the links between evidence, experts, experience, policy and practice.

- To understand the relationship between quality and cost

- To understand the concepts of shared baselines, local clinical protocols and the
improvement method

- To understand the modelling of the process of quality improvement

- To review tools available for healthcare delivery improvement

- To understand the relevance of measurement in improvement and to learn about
the tools to do so

- To understand the various kinds of leadership that brings about the preferred
response from colleagues using a selection of human factors and communication
methods thus defining the human face of quality improvement leadership

- To appreciate the importance of learning from immediate peers and colleagues.

Attendance Requirement
There will be 10 contact classes in the first year which participants are required to attend.
Other aspects of the course will be delivered by a combination of e-learning and support as
needed.

Modular progression
At the successful completion of first year there will be the option to take the qualification of
Certificate in Healthcare Improvement Leadership and to progress to a Diploma and Masters

©M HEMADRI

Follow me on Twitter @HemadriTweets

 

Declaration of interest: I teach some parts of the course. No current financial interest.

The Problem with PDSA in healthcare

In healthcare it is now generally understood that using the PDSA cycle is a good and valid method to try to achieve improvement. The PDSA is very widely known in healthcare and often used, though it is thought it ought to be used even more. Yet, when we look at health organisations that are using the PDSA we do not find the improvement at a range or scale or impact that is very often found in other industries who use PDSA.

Why?

To resolve the angst around this we need to know what comes before and after any particular individual PDSA cycle.

Before a PDSA

How is the specific individual PDSA cycle conceived? Why was this particular PDSA chosen over many possible PDSAs that could have been done?

Before choosing to do any one particular PDSA there are at least five prior major detailed outlining steps to be completed that involves objective and subjective methods, data analysis, prioritisation, setting aims, measures and interventions. Only after this a PDSA ought to be done by a very small team which has mostly understood the prior steps as a matter of overall context – i.e. the how and the why, the logic that validates your activity, the reason that requires your engagement and the rationale that demands your time and energy.

If you are currently doing a PDSA or soon planning to do a PDSA it is important for you to consider how it was chosen. If you chose it out of an impulse, hunch, suggestion, obligation, instruction that is great for your personal learning of the tool which is of course very important. It may (or may not) show an improvement on that particular cycle or cycles, but you or your organisation should not be under the illusion that this PDSA effort is going to contribute to sustained or widespread improvement. It is important to prove to yourself on where the PDSA fits in within a broader department, division, directorate, organisation context.

One of the ways to identify whether there is any link to anywhere other than to you is to observe if your boss or your boss’ boss is as keen and enthusiastic about your PDSA not because they support you but because your PDSA has an important link to moving the dots in the right direction that they are supposed to move and they can prove it. They should be able to stop further PDSAs that is not working and you should be happy with it.

After a PDSA

What happens to your PDSA after you have completed and you think it shows some positive result? Are you in a position to pilot it further in repeatedly larger areas/scales? Do you have the support for it? Have your bosses confirmed your PDSA cycles have proved as shown by a series of linked organisation wide data that it has led to wider improvement? Eventually after a series of such PDSAs does your intervention, process and outcome become official standard protocol for the area?

The problem with PDSAs as we do it in healthcare right now

Every empowered person does PDSA based improvement activity but there is usually no one to track all of these, guide the people doing PDSA projects, help them do the run or SPC charts, identify where these projects are in the overall organisational improvement effort (say by using a driver diagram), capture and roll out good ideas for the whole organisation's benefit.
I would say that we should stop healthcare employees from doing unsupported PDSAs for at least two reasons a) it wastes individual staff time which could be usefully spent on something more useful b) if the unsupported PDSAs are successful then it leads to small individual areas shining which is usually a drain on resources and general emotion (technically known as sub-optimisation). In theory it is possible to even cause harm by such poorly designed activity.

The issue is Tools vs Philosophy

PDSA has great history and comes from the times of superior masters like Shewhart Juran and Deming. It is a part of an overall philosophy that can be called the QI movement or which after adaptation now more familiarly known as the ‘Lean’ (though some purists, even non-purists will be able to differentiate between the two).

To understand this better, we need to ask ourselves whether the PDSA is used as a tool for individuals or as a part of a philosophy for organisations. Similar to the issue whether Lean is used as a method or philosophy. If you or your organisation are using PDSA (or Lean) as a mere tool or a method – you are designing is poor and destined to fail.

We are at a point in history of improvement healthcare that we are training a large number of people on ‘quality improvement’ and letting them do unsupported PDSAs. We do that under the guise that we do not want to interfere with the freedom of senior and experienced healthcare staff. When these ‘trained’ ‘senior’ people do not see the improvement that the lean system claims that it offers, they then become committed disbelievers in the philosophy while at the same time being obliged to follow the tools and the methods.

We are at the risk of defiling and debunking a well established validated healthcare improvement philosophy because of our unwillingness to adopt it as a philosophy. It will be to the eternal shame of us in healthcare. We are creating proof that lean healthcare does not work, instead of accepting that we do not know how to do lean healthcare properly and we are not doing it as we are supposed to do. We need to act swiftly to avoid this - there is life and limb at risk.

  

©M HEMADRI 

Follow me on twitter @HemadriTweets

'Nakamura invented the light bulb'

Would Nakamura invented the light bulb if he was working in the NHS?

Nakamura invented the light bulb, that is what we might probably say one day, that Nakamura invented the light bulb or to put it correctly that Nakamura re-invented the light bulb. Shuji Nakamura's inspiring story has been told before but here is an ultra-short version of it

Nakamura gets a masters from a relatively small university in a small city in Japan, goes off to work in a small company in a lab competes against the big companies, discovers many right things and makes products that would not sell.

Times get difficult, his department shrinks. He goes to his boss and wants to make a product that the big boys have tried to make and failed; with his record, he gets turned down. He goes to his boss's boss and gets some support to make it, despite his record of making nothing that sold, his company chairman gives him money, $2mil actually and he reinvents the light bulb. Well, he actually gets the blue component of the LED to work and the rest is history.

What is interesting is that Nakamura was a non-PhD working in the industry as a lab scientist who then gets a doctorate from his local university and within 5 years is head hunted by University of California and becomes a professor.

Nakamuras in NHS?

Let us imagine a scenario of a doctor who becomes a consultant in a DGH in the NHS and wants to do something that the big boys tried and failed.Then the DGH consultant fails as well, fails repeatedly - what are the chances that he will not be performance managed out of his/her activity and driven to the end of his wits.

What are the chances that the medical director or CD will be over-ruled by the CEO or Chairman and a doctor provided funding to carry on despite a record of 'failure'? What are the chances that even after this doctor discovered something interesting a big place will head hunt and make him/her an 'academic'? In fact he/she should be grateful if the GMC and the rest of the regulation did not land on him/her and crushed him/her out of existence.

Getting real

Now a lot of you are going to say that reinventing a light bulb while surely profound is unlikely to involve any damage to real human beings. You might say that any lurking Nakamuras in the NHS if supported could end up hurting patients. Good logical argument. Is that what is really hurting patients? Probably not. It is not any innovation by enthusiastic people that harms patients, it is the bureaucratic nay sayers who use the language of clinical governance and risk yet know very little about process capabilities, refuse to learn shared baselines, practice unimaginatively poor leadership who perpetuate harm in healthcare. They refuse to fix the system instead try to 'fix' the people in the system. Of course the medical profession does not do itself any favours by its ego, jealousy and macho attitude which will aim to shoot down anything that arises outside its hierarchical constraints by treating them as bad apples and recommending the use of evidence the origins of which  can probably attributed to the Abilene paradox.

In healthcare especially in the NHS it is pretty much impossible these days to take an extra breath without CD, CG, R&D, GCP, LREC, NREC, NICE, and every other alphabet in the soup wanting to spoil it for you, while claiming to support you. It is when people who are typically NHS managers and every other hierarchical bureaucrat stops behaving like researchers and most doctors who are not researchers begin to look at operational evidence as a valid method of creating a new practices, innovation and improvement that healthcare will be truly successful.

Allowing and managing 'Nakamuras' in healthcare is not easy but will be rewarding; eliminating the healthcare 'Nakamuras' will allow the managers to sleep peacefully but might push true healthcare innovation into a coma.

If you know of any 'Nakamuras' in the NHS please let me know by leaving a comment below.

©M HEMADRI 

Follow me on twitter @HemadriTweets

Links & reference
The dream of the blue laser diode
http://engphys.mcmaster.ca/undergraduate/outlines/4e03/Nichia%20%27s%20Shuji%20Nakamura%20Dream%20of%20the%20Blue%20Laser%20Diode.htm
Time magazine short feature on Nakamura
http://www.time.com/time/magazine/article/0,9171,1604891,00.html

Human Error: Does not exist

Human Error. Does it really exist?

We have discussed wrong site surgery/procedure (http://successinhealthcare.blogspot.co.uk/2012/10/mark-site-campaign.html)

The equivalent for this in histopathology would probably be labeling errors. Labeling errors could at the best lead to rework/reprocess and at the worst result in wrong report with potentially catastrophic effects on patients which can be as grim as wrong site surgery. In most laboratories there are multiple checking steps within the process to detect errors and prevent them leading to errors in reports that could harm patients. In a busy pathology laboratory in England in 2007 there were 113 slide and block labeling errors. By 2009 after a series of Kaizen events it dropped down to just 2 labeling errors which would be a 98% improvement giving a short term six sigma score of 5.8.

 What is interesting are the results of the root cause analysis of the 113 pre-Kaizen. Most of them showed that human error as one of the root causes. What is remarkable was the post-Kaizen improvement was achieved with the same people. The root cause analysis of the 2 post-Kaizen errors showed further opportunities for system improvement.

If system improvement can reduce or eliminate (well, nearly eliminate in this example) human errors, the immediate logical obvious question to ask is 'Does human error exist?'

Deming says that 80% of quality problems are caused by management and 20% by employees. It is further thought that since the employees are essentially a part of a system for which the management is responsible, almost all quality problems are caused by management. Deming seems to have taken the view that the focus and emphasis on quality has to be top down and the creation and delivery of quality should be bottom up.

There are a number of areas where zero errors or quality problems (or virtually zero errors) are possible. In the same pathology lab the number of endoscopic biopsy request clarifications (which used to happen due to doctors illegible handwriting) are now down to zero since the lab started asking for a copy of the printed endoscopy report to accompany the specimen. Previously it was thought that poor handwriting and not putting enough information was a part of human error due to human fallibility, in practice it caused arguments, distress and wasted time.

We are now beginning to question whether there is anything called human error at all. As realists and practical professionals we realise that there will be some areas where perhaps human error does exist and possibly cannot be avoided but we believe that for people working within well organised systems this should be a rare thing. We wonder if people with poor training and no experience in quality methods who nonetheless think they are capable of understanding quality improvement are unable to analyse with an aim of system improvement and hence blame human error as a reason by default. After all everyone has recognition and sympathy for the phrase 'to err is human'

We are having an emerging view that 'Human Error' as an attribution for quality problems is a cop out clause used by poor managers and weak leadership. It need not be so. However it requires managers and leaders to shoulder the responsibility for building continuous quality improvement into their work and the way their teams function. CQI systems are already available and they have to be applied with patience and persistence - those who do that will find  the path of continuous improvement and will eventually share this view of ours that 'Human Errors does not exist in organised systems' though it may sound very radical right now.

M Hemadri & David Clark

Co-authors

David Clark is a Consultant Pathologist and National Clinical Lead, NHS Improvement. David's thoughts expressed in this blog post are his own personal views.

PS: Regular readers of this blog would have read a previous post about how we find designs which set us up for failure (not deliberately) and we then blame it on 'human error' http://successinhealthcare.blogspot.co.uk/2012/12/blame-it-on-human-error.html

©M HEMADRI 

Follow me on twitter @HemadriTweets

My mini e-book 'Standardised Management Conversation' is available - click http://www.amazon.co.uk/Standardised-Management-Conversation-Hemadri-ebook/dp/B018AWBJTU 
till 31 December 2016 all my earnings from the sale of this book will be donated to charity  http://successinhealthcare.blogspot.co.uk/2015/11/standardised-management-conversation.html