Simulation and Learner Safety

Primarily when we talk about safety in simulation we are referring to patient safety. Patient safety in two senses. The first is that one of the main reasons for carrying out simulation is to improve patient safety by looking for latent errors, improving teamwork, testing equipment, etc. The second is that "no patient is harmed" during simulation exercises.

In the brief before the simulation event, safety is also often mentioned in the establishment of a "safe learning environment (SLE)" and, in this context, it refers to Learner Safety. A recent clinical experience reinforced my appreciation of the SLE.

It was 10pm and I was resident on-call when my phone went off to tell me that a poly-trauma was on its way in. 2 adults and 3 children had life-threatening injuries after a collision on the motorway. Although I have been an anaesthetist for 13 years, a consultant for 5 of those, my clinical experience of polytrauma in adults is minimal and in children is essentially nil. I have looked after a man who had major injuries and 95% burns after an industrial explosion, another man who suffered severe injuries after he ran his car underneath a flatbed truck and the occasional stabbing and shooting victims. In children I have intubated a  2-week-old "shaken baby" and anaesthetised a large number of children on the trauma list for broken wrists, arms, ankles, etc. 

When faced with infrequent events it is not unusual to carry out a memory scan to draw on previously obtained knowledge relevant to the situation at hand. I remembered the above patients and I also remembered a simulation course I had been on at the SCSCHF: Managing Emergencies in Paediatric Anaesthesia  for Consultants (MEPA-FC). My scenario involved a boy who had been run down by a car, he had a number of injuries including a closed intracranial bleed. My first thought when I remembered this scenario was "I did okay". Then I mentally went through the scenario again, thought about what had gone well and what, with input from the debrief, I should have done better. This then was the knowledge I had front-loaded and the emotional state I was in when the patients arrived in the ED.

When I talked through the above with David Rowney, the facilitator on the MEPA-FC course, he expressed surprise that my first thought was "I did okay" rather than remembering the Take Home Messages for my scenario. But there it is. It may be that I am very different from other people but I think it is not unusual to have an emotive reaction to a memory before a logical one.

This then made me think about the simulation participant who might not have had the SLE I had. The participant who, after their paediatric trauma scenario, had been dragged over the coals and made to feel incompetent. What would the emotional state of that doctor be as they walked down to the ED? And how would that affect their performance?

This blogpost is not a plea to "take it easy" or "be gentle" with participants. Poor performance must be addressed, but it must be addressed in a constructive manner. Help the participant understand their performance gaps and how to bridge them, while at the same time remembering "I'm okay. You're okay." Very few of us come to work (or to the simulation centre) to perform poorly. In fact most people in a simulation are trying to perform at the peak of their ability. When they fall short it is important to help them figure out why that is, while re-assuring them that they are not "bad".

Book of the month: Resilient Health Care (Hollnagel, Braithwaite and Wears (eds))

About the editors

Erik Hollnagel has a PhD in Psychology and is a Professor at the University of Southern Denmark and Chief Consultant at the Centre for Quality Improvement, Region of Southern Denmark. He is the chief proponent of the Safety-II paradigm and helped to coin the term "resilience engineering".
Jeffrey Braithwaite, PhD, is the director and a professor of the Australian Institute of Health Innovation and the Centre for Health Care Resilience and Implementation Science, both based in the Faculty of Medicine and Health Sciences at Macquarie University, Australia. He is also an Adjunct Professor at the University of Southern Denmark.
Robert Wears, MD, PhD, is an emergency physician and professor of emergency medicine at the University of Florida and visiting professor at the Clinical Safety Research Unit, Imperial College London.

About the contributors

There are 27 other contributors, including well-known names such as Charles Vincent and Terry Fairbanks. The contributors are a world-wide selection, encompassing the US, Europe and Australasia. The majority are from a sociological/psychological research background rather than front-line clinical. 

Who should read this book?

This book will be of interest to those who are tasked with improving patient safety within their organisation, whether this is by collecting and analysing incident reports or "teaching" healthcare workers. It would be useful reading for board members, healthcare leaders and politicians involved in healthcare.

In summary

The book is divided into 3 parts (18 chapters), as well as a preface and epilogue by the editors

1. Health care as a multiple stakeholder, multiple systems enterprise
1. Making Health Care Resilient: From Safety-I to Safety-II
2. Resilience, the Second Story, and Progress on Patient Safety
3. Resilience and Safety in Health Care: Marriage or Divorce?
4. What Safety-II Might Learn from the Socio-Cultural Critique of Safety-I
5. Looking at Success versus Looking at Failure: Is Quality Safety? Is Safety Quality?
6. Health Care as a Complex Adaptive System
2. The locus of resilience - individuals, groups, systems
7. Resilience in Intensive Care Units: The HUG Case
8. Investigating Expertise, Flexibility and Resilience in Socio-technical Environments: A Case Study in Robotic Surgery
9. Reconciling Regulation and Resilience in Health Care
10. Re-structuring and the Resilient Organisation: Implications for Heath Care
11. Relying on Resilience: Too Much of a Good Thing?
12. Mindful Organising and Resilient Health Care
3. The nature and practice of resilient health care
13. Separating Resilience from Success
14. Adaptation versus Standardisation in Patient Safety
15. The Use of PROMs to Promote Patient Empowerment and Improve Resilience in Health Care Systems
16. Resilient Health Care
17. Safety-II Thinking in Action: 'Just in Time' Information to Support Everyday Activities
18. Mrs Jones Can't Breathe: Can a Resilience Framework Help?

I haven't got the time to read 238 pages...

For the time-poor, the preface and epilogue are worth reading. Chapter 3 on the challenges resilience poses to safety, Chapter 5 on quality versus safety and Chapter 11, co-authored by Charles Vincent, on the downsides of resilience, are also worth reading.

What's good about this book?

This book makes it clear that "resilience" can mean different things to different people. The authors identify resilience as part of the defining core of a system, something a system does rather than something that it has (p.73, p.146, p.230). This is in contrast to some who call for more resilient healthcare workers, with the implication that if they were "tougher" then they would make fewer mistakes. Resilience is also not just about an ability to continue to function but an ability to minimise losses and maximise recovery (p.128).

The authors also make it clear that resilience is not a self-evident positive attribute. More resilience in a system does not come without cost including, for example, a system which may resist "positive" change, such as some of the changes that the patient safety movement is trying to embed. Safety may focus on standardisation and supervision while resilience focuses on innovation, personalisation and autonomy (p.29). In Chapter 3, René Amalberti argues that "it is not a priority to increase resilience in health care. The ultimate priority is probably to maintain natural resilience for difficult situations, and abandon some for the standard" (p.35).

The book helps to explain the lack of rapid advance in patient safety because of the "economic, social, organisational, professional, and political forces that surround healthcare" (p.21). Healthcare may be unique in the diversity and strength of these influences. In addition the authors argue that there is a gap between the front-line and those who manage "safety" (p.42), a finding echoed by Reason and Hobbs in their book on maintenance error.

The book makes a good critique of the "measure and manage" approach of Safety-I (p.41) which:

• is retrospective
• focuses on the 10%
• misses learning to be found in safe practice
• focuses on the clinical microsystem rather than the wider socio-cultural, organisational, political system 

Lastly, much work is currently focused on standardisation, however the authors argue that we should  acknowledge the inevitability of performance variability, the need to monitor it and to control it (by dampening it when it's going in the wrong direction and amplifying it when it's going in the right direction). (p.13) The standardisation that does improve resilience is the type that decreases the requirements for effortful attention or the need to memorise (e.g. checklists, layout of workplaces).

What's bad about this book?

Throughout this book, resilience is linked with the Safety-II concept (e.g. "Chapter 1: Making Health Care Resilient: From Safety-I to Safety-II"). The argument for Safety-II can be a nuanced one, therefore a good book on resilience would use simple language and provide specific examples. This book fails on the former and performs poorly on the latter. In particular, how Safety-II can be put into practice now is only vaguely referred to. Even the chapters which purport to show resilience in action do not make this very clear. Exceptions include Chapter 12 "Mindful Organising and Resilient Health Care" which suggests that people should be shown their inter-relations, i.e. how their actions affect those who interact with a patient upstream and downstream. 

At times, the championing of Safety-II gives its proponents the appearance of a cult, e.g. "Enlightened thinkers in both industry and academia began to appreciate..." (p.xxiv) while one must imagine that unenlightened thinkers continued to live in their caves. There are also attacks on the PDCA/PDSA cycle (p.177) and the use of barriers (p. 131) as Safety-I thinking. In addition Safety-I, as a term and paradigm, has been created by Safety-II advocates, and in fact "pure" Safety-I probably does not exist. For example: "In contrast to Safety-I, Safety-II acknowledges that systems are incompletely understood...", however very few people working in healthcare, even within a Safety-I system, would argue that they fully understand the system.

One of the examples in the book of proactive safety management is the stockpiling of H1N1 drugs and vaccines in 2009. This was later deplored by a number of sources as the mild epidemic killed fewer people than seasonal flu and millions of pounds of stockpiles had to be destroyed. 

Lastly one of the arguments the authors use against Safety-I thinking is that focusing on the small number of adverse events means we miss the opportunity to look at all the times things went well. However, with 10% of patients admitted to UK hospitals being subjected to iatrogenic harm (Vincent et al 2008), the number of times things go wrong is still a large chunk of the total work.

Final thoughts

This book makes a strong argument that we must stop looking purely at what has gone wrong in order to find out how to prevent mistakes. It also makes it clear that healthcare, as a complex adaptive system, will not be "fixed" by silver bullets, and that all solutions to problems create their own problems.

The concepts underpinning Safety-II, which include an urge to focus less on incidents and accidents and more on things that go well, are antithetical to much current thinking within healthcare. In addition patients and their families would not accept "I'm sorry you were harmed but we're focusing on things that go right" as an apology. This means that rather than pushing Safety-II, it may be more effective to advocate Safety-III. In Chapter 12 this is defined as: 

"... enactive safety - embodies the reactive [Safety-I] and proactive [Safety-II] and therefore both bridges the past and future, and synthesises their lessons and prospects into current action." (p.155)

Hollnagel himself says "...the way ahead does not lie in a wholesale replacement of Safety-I by Safety-II, but rather in a combination of the two ways of thinking" (p.16). Safety-III may turn out to be a quixotic Theory of Everything. Or it may mature into an accepted, practical and applied paradigm, with "a degree of autonomy at the interface with the patient, yet predictability and effectiveness at the level of the organisation" (p.132). Its adherents still have much work to do.

Further reading:

Vincent, C., et al. (2008) Is health care getting safer? British Medical Journal, 2008;337:a2426.

Book of the month: A life in error: from little slips to big disasters by James Reason

About the author

James Reason is one of the greats in human factors research. English Wikipedia does not have an entry for him (the French site does). Instead we have to content ourselves with a page on perhaps his major contribution to broadening the appeal and understanding of human factors, the Swiss cheese model of accident causation. Reason is Professor Emeritus of Psychology at the University of Manchester and has authored numerous papers and books on human factors, including: Human error, The Human Contribution and Managing Maintenance Error (A Practical Guide)

Who should read this book?

Anybody with an interest in human factors and patient safety (see below for why).

In summary

The book consists of 14 chapters:

1. A Bizarre Beginning
2. Plans, Actions and Consequences
3. Three Performance Levels
4. Absent-minded slips and lapses
5. Individual differences
6. A Courtroom Application of the SIML (Short Inventory of Mental Lapses)
7. The Freudian Slip Revisited
8. Planning Failures
9. Violations
10. Organizational accidents
11. Organizational Culture: Resisting Change 
12. Medical Error
13. Disclosing Error
14. Reviewing the Journey

What’s good about this book?

The book is very well written and easy to read. Reason takes us on an humorous, insightful, autobiographical journey from his first encounter with "human error" to his later theories. The book explains a number of concepts. For example, Reason argues that some familiar objects develop local control zones (p.3). In healthcare, an IV cannula may exhibit this property. If one finds oneself with a syringe in hand, distracted and near a cannula there is a strong possibility that one will inject the contents of the syringe into the cannula. When the syringe contains local anaesthetic or 1:1000 adrenaline this may result in adverse consequences.

Reason talks about differences between novices and experts. The former show a lack of competence, while the latter are much more likely to commit absent-minded slips, i.e. misapplied competence (p.21). Reason argues that, in absent-mindedness, it is the suppressive function which goes absent. Pre-programmed, habitual actions are normally actively suppressed, but in "strong habit intrusions" they are carried out by the distracted person.

Reason discusses the "Stress-vulnerability hypothesis", people under chronic stress are more likely to have cognitive failures such as absent-minded slips and lapses (p.33). However he argues that association is not causation, and it may be that people who are more likely to complain of chronic stress may also me more likely to be absent-minded, i.e. that the same poor cognitive resource management is responsible for both.

In his discussion of planning/decision-making, Reason describes the planning process and the sources of bias which lead to failure, grouping them by planning stage (p.56):

1. Working database (e.g. recency, successes better recalled than failures)
2. Mental operations (e.g. covariation, "halo", hindsight)
3. Knowledge schema (e.g. confirmation, resistance to change, "effort after meaning")

For those interested in groups and organisations, Reason discusses "satisficing", i.e. groups will tend to select the first satisfactory outcome rather than an optimal one. He also looks at the heuristics of group decision-making, such as avoidance of uncertainty and selective organisational learning (p.59).  In terms of accidents, Reason contrasts "individual" (frequent, limited) and "organisational" (rare, devastating) accidents. He therefore agrees with Steven Shorrock that having a sign which says e.g. "135 days since our last accident" does not tell you how safe the system is. Why? Because they have different causal sets (p.79).

"Turning a blind eye" (Nelson commits a violation, p.68)

In terms of violations, that is conscious decisions to ignore or circumvent a rule, Reason argues that it is better to focus on decreasing the benefits of violations rather than trying to increase the costs of doing so. This means that one should look at why the system is promoting violations rather than punishing individuals for committing them.

Reason also covers latent conditions, active failures and how they combine with local triggers into an accident trajectory (p.75).

What’s bad about this book?

At 124 pages, this is a short book, however it is probably too short. A lack of explanation may leave some readers puzzled. For example, on p.30 Reason states: "The correlation between [two independent samples of the Short Inventory of Mental Lapses] over the 15 items was 0.879." It would probably have been better to leave out the numbers or to explain them. His coverage of the planning process and its biases is too short and superficial, he mentions "groupthink" (p.61) and provides 8 main symptoms but does not explain these in sufficient detail to allow one to use this knowledge in practice.

Final thoughts

This book spans the whole gamut of human factors science and touches on a great number of subjects including all the above, as well as a typology of safety cultures, vulnerable system syndrome (blame, deny, pursue wrong goals), why and how organisations resist change, models of medical error (plague, legal, person, system) and more. And if you would like an easy-to-read, broad introduction to human factors and healthcare then this book is a must-read.

On the use and abuse of "human factors"

Words shape our world

The words we use, and how we use them, not only allow people to know what we are thinking but also shape the way we think. As a car mechanic, for example, knowing what all the components of an engine are called will make it easier for her to talk to a fellow mechanic and think about what the problem might be and how to fix it.

"Human factors"

In the podcast "Human factors, non-technical skills and professionalism", Liz Chan, a specialist in Veterinary Anaesthesia and Analgesia at the Royal Veterinary College, University of London tells us:

"Human factors were defined by a guy called Martin Bromiley, who set up the Clinical Human Factors Group [CHFG]... He defines them in such an excellent way I always steal his definition because it is, basically to paraphrase: 'Everything that makes us different from predictable machines.'"

Although it is possible that Martin Bromiley used that definition, it is extremely unlikely and, in trying to paraphrase, Liz Chan has changed the meaning of the term "human factors". This means that the podcast listeners are also likely not to use the term appropriately and when they read human factors literature they may wonder how this fits with their definition.

On their website, under "What is human factors?" the CHFG uses the Chartered Institute of Ergonomics and Human Factors (CIEHF) definition of:

"Ergonomics (or Human Factors) is the scientific discipline concerned with the understanding of interactions among humans and other elements of a system, and the profession that applies theory, principles, data and methods to design in order to optimise human well-being and overall system performance."

An easier definition is provided by Martin Bromiley in a Health Foundation blog.

"I often talk about human factors making it easy to do the right things with reliability of outcome..."

So human factors is a science whose aim is to make it easy for us to do the right thing, and difficult to do the wrong thing.

The abuse of "human factors"

A lack of clarity around the use of the term "human factors" means that when it is used in the press, for example, it is almost always in a pejorative manner. This reinforces the (false) idea that if we could remove the humans from the system then things would be much safer.

In his book, The Human Contribution, James Reason argues that the predominant view of humans in complex systems is as "hazards" when they are often "heroes".

Still one of the predominant examples of the latter view is the "miracle on the Hudson", when Captain Chesley B. "Sully" Sullenberger landed an Airbus A320 on the Hudson river in New York after both engines had failed due to bird-strike. But we didn't see headlines like this:

We should use the term human factors to refer to the science of ergonomics and avoid using it to mean "human error" (itself a poor choice of words). This will help us and others to have more meaningful discussions and clearer thinking on the causes of, and remedies for, incidents and accidents.