Outstanding skill is not always enough to keep a patient safe.
A working resource for surgical, anaesthetic, theatre and ward teams. It gathers the models, tools and evidence from my interest over the years, reading, and delivering HUMAN FACTORS teaching days so you can use them on the next list, the next handover and the next difficult conversation to improve patient safety.
- Three principles we start every course with
- 01Error is normal.
- 02Performance is variable.
- 03People and culture interact with systems and process.
From human factors to team resource management
Human factors thinking explains why capable people get things wrong. Practised consistently, that understanding becomes team skills. Team skills used together make team resource management: the deliberate use of every resource a team has, human and technical.
Step 1
Human factors
Recognising the vulnerability of the human condition. Even excellent knowledge and skill can fail under fatigue, overload or a poor system.
Step 2
Team skills
The behaviours that make team performance reliable: briefing, speaking up, managing workload, giving and receiving feedback.
Step 3
Team resource management
Developing the individual, bringing together the team and getting the job done safely.
Why aviation gets mentioned so often
For decades, accidents were closed as "pilot error". NASA and University of Texas research showed that skilled individuals put together as a crew could still perform badly as a team, and identified the behaviours that made the difference. That work became Crew Resource Management, now mandatory and tested for every commercial pilot.
Healthcare has a similar problem. Around one in ten hospital admissions is linked to an adverse event, and most are judged preventable on review. This is mainly a systems and teamwork problem rather than an individual competence problem.
Two strands of safety
System & process
- Policy and procedures
- Back-up systems
- Regulation
People & culture
- Knowledge, built through training
- Experience, built through practice
- Judgement, built through reflection
Neither strand makes up for weakness in the other.
Choose a topic
Each topic has its own page. Use the menu at the top, or start here.
Performance is variable
Humans are creatures of the day, good for about sixteen hours from waking. After roughly 17 to 19 hours awake, performance is comparable to being at the UK drink-drive limit (Dawson & Reid, Nature 1997). Workload, fatigue and the people around us all move where we sit on the curve.
The arousal curve
What's in your bucket today? (H.A.L.T.)
A fuller bucket lowers your personal best. It doesn't mean stop working. It means stop and think: slow down, double-check, ask for a second pair of eyes.
One brain, one resource
Thinking and doing draw on the same limited capacity (Amalberti). As the demands to act rise, the capacity left to plan, anticipate and notice change falls away. Below the line you are busy but in control. Above it you are still active, but no longer able to catch your own errors.
- When overloaded we all make unforced errors and our judgement goes.
- When one team member is overloaded, the whole team's performance suffers.
- Communication fails under high workload, so an agreed, rehearsed vocabulary matters most then.
- Split cleanly: if there are two of you, the senior thinks and the junior does.
Spot overload in yourself and others
Apathy · dithering · anger · task avoidance · embarrassment · slowed speech or losing words · shaking · jerky, tense movement · sweating · poor decisions
Response: stop to think. Make lists. Pre-think likely events. Rehearse actions until they're automatic. Thinking is usually a better use of capacity than doing, even though we feel compelled to act.
Acute
Overload
Today's list, this crisis.
Chronic
Burnout
Interpersonal communication fails; compassion for co-workers is lost.
A group of people working cooperatively towards a shared goal
Run this with any team: In a great team I've worked in, what enabled its performance? In a difficult one, what inhibited it? Single words. The answers are more useful than any values poster, because they come from the people doing the work.
The eight NASA / University of Texas team skill markers
1
Briefing
Operationally thorough and interesting; covers coordination, planning and likely problems.
2
Leadership & followership
Appropriate leadership and followership, with concern for the task.
3
Communication & decisions
Free and open communication; active participation in decisions.
4 · the gatekeeper
Interpersonal relationships
Quality of relationships and group climate. If this is wrong, none of the rest follow.
5
Team self-feedback
The team recognises the need to give and receive feedback.
6
Preparation & vigilance
Planning ahead, keeping situation awareness, anticipating contingencies.
7
Enquiry & assertion
Advocating the best course with appropriate persistence, even when that means disagreeing.
8
Workload & distractions
Distributing tasks, avoiding overload, managing distraction.
Team shapes (Michael West)
- Hierarchical: the safest shape. Everyone knows their role and responsibility, and performance management is built in.
- Individual: clear roles.
- Pseudoteam: often the greatest capability on paper but no performance management. Theatre teams who have never worked together never learn each other's skills or gain marginal improvements.
Accountable to, responsible for
We are accountable to people, and that relationship runs both ways: to patients through consent, and to employers, professional bodies and society. We are responsible for outcomes. If you delegate a task, the responsibility stays with you.
Train hard, fight easy. Make relationships in peacetime.
Will I be welcomed if I speak up?
Psychological safety is a belief that the team is safe for interpersonal risk: that questions, concerns, ideas and mistakes will be welcomed (Amy Edmondson). It takes a long time to build and can be lost in a moment.
Fostering psychological safety
- Demonstrate engagement
- Show understanding
- Be inclusive in conversation and in decision making
- Show confidence and conviction without seeming inflexible
| It is not… | It is… |
|---|---|
| Niceness | Sharing the truth clearly |
| Avoiding discomfort | Having the tough conversation |
| Avoiding disagreement | Deciding what is right |
When there isn't a culture of safety, people think…
"I don't want to look ignorant"
so I won't ask questions
"I don't want to look incompetent"
so I won't admit mistakes
"I don't want to seem negative"
so I won't question how we do things
"I don't want to intrude"
so I won't offer my ideas
The authority gradient
The perceived difference in power between the decision-maker and the rest of the team. It comes from position, confidence, knowledge, reputation, expertise, culture, gender and experience, real or assumed.
Too steep
Concerns go unspoken; the leader loses information they don't have.
Too flat
Nobody is clearly in charge; decisions drift when time is short.
Leaders lower it with approachability: open, respectful, empathetic. Followers climb it with assertiveness: congruent, concise, clear, confident, cognisant, courteous.
The 10 commandments of a psychologically unsafe system for surgical trainees
Written for a podcast discussion as satire: a list of everything to avoid. Read each as its opposite.
- Promote a culture of fear and intimidation: public humiliation, harsh punishment for minor errors, constant scare tactics.
- Keep a rigid hierarchy: discourage questions, give the best cases to those who conform, show favouritism.
- Encourage unhealthy competition: pit trainees against each other and reward only the "top performer".
- Provide no structured mentorship, and imply that asking for help is weakness.
- Make expectations and assessments unpredictable and arbitrary.
- Undermine confidence through micromanagement and by dismissing achievements.
- Glorify long hours and minimise mental health support.
- Value individual performance over teamwork and discourage peer support.
- Use biased, subjective assessments and vague feedback.
- Limit open communication and keep trainees in the dark about decisions affecting them.
Discussion prompt: is "sink or swim" training designed to toughen people up, or is it simply unsafe?
Rudeness is a patient safety risk
Civility is the sum of the many small sacrifices we make for the sake of working together. Incivility spreads, and it affects everyone who sees it, not only the person on the receiving end (Porath & Pearson, Harvard Business Review 2013).
What can we do?
- Raise awareness
- Engage
- Say sorry
- Praise good work
- Empower others
- Call out incivility
- Say thank you
It could be me
Ask yourself at the end of a shift: What did I notice about my conversations today? Did I do anything that seemed to affect someone badly?
Active bystander: the 4 Ds
Direct · Distract · Delegate · Delay. Decide in advance the words you would use, and rehearse them, so they're there when you need them.
Making sense, then making decisions
Situation awareness is taking in information and using it to decide. Each of us builds our own mental model of what is happening, quickly and unconsciously, filling the gaps from experience. Two people in the same room can hold different pictures and not know it until someone says theirs out loud.
| System 1 | System 2 | |
|---|---|---|
| Speed | Fast | Slow |
| Effort | Effortless | Effortful |
| Style | Instinctive, emotional | Logical, deliberate |
| Risk | Bias, priming | Lost to workload and distraction |
Four habits against bias
- Switch deliberately into System 2 at defined safety-critical points, by pausing.
- Look for what doesn't fit, not only what confirms the picture.
- Bring in different professions and grades: they notice different things.
- Ask for fresh eyes from someone not anchored to the same model.
Local rationality
"It made sense to me at the time." Ask a room to picture an apple and everyone sees a different one. After an event, root cause analysis is prone to hindsight bias and outcome bias. Understand why the decision made sense to the person inside it.
Kahneman, Sibony & Sunstein
Noise: the hidden half of error
Noise is unwanted variation in judgements that should be the same. Pre-assessment and the anaesthetist on the day reach different conclusions about whether a child needs HDU. Error = bias + noise, and both cause harm.
- Aim for accuracy, not individual expression.
- Think statistically; take the outside view.
- Break judgements into separate, independent parts (think Apgar).
- Resist premature intuitions.
- Collect independent judgements before discussion, then aggregate.
- Favour relative judgements and relative scales.
What this means for our MDTs
Anchoring: the first person to speak sets the tone, often the most confident and least familiar with the patient. Substitution: discussing a similar case instead of this one. Deference to seniority over knowledge of the family. Collect views before the discussion starts.
ABCDE against narrative traps
Acceptance · Balancing · Checklists · Diversity · Evaluation
Dolan & Henwood, Front Psychol 2021
Metacognition
Metacognition is the ability to notice and steer your own thinking: knowing what you know, spotting when you might be wrong, and choosing when to slow down. It is the skill that lets us catch our own errors before they reach the patient.
Before
Plan
What am I trying to achieve? What do I already know, and what don't I know? Where is this likely to go wrong?
During
Monitor
Is this going as I expected? Am I running on autopilot? How full is my bucket? Am I still listening?
After
Evaluate
What did I expect, what happened, and why the difference? What would I do differently? This is the debrief, done in your own head.
Confidence is not accuracy
We don't feel it when we're wrong
In a study of over 100 internists working through case vignettes, accuracy fell from 55% on easy cases to under 6% on difficult ones. Confidence barely moved: 7.2 out of 10 on easy cases, 6.4 on hard ones. Doctors also asked for fewer extra tests on the hard cases, when they needed them most.
accuracy confidence (scaled to 100)
Meyer AND et al. JAMA Intern Med 2013;173:1952–8
Moulton et al, Academic Medicine
Slowing down when you should
Expert surgeons aren't on autopilot. They keep a background watch and, at cues such as a critical step, unexpected anatomy or a complication, they shift from automatic to deliberate attention. Sometimes that means stopping to regroup; sometimes removing distractions or simply focusing harder while the conversation carries on.
The danger is drifting: when the background watch quietly lapses. Teams who know each other read the cues and slow down with the surgeon.
Cognitive forcing strategies: Croskerry, Acad Med 2003. Use at handover, before committing to a diagnosis, and before the knife goes in.
Try it
If you overtake the person in second place in a race, what position are you in?
Second. System 1 shouts "first". You only get it right by pausing to check the fast answer. Kahneman calls the people who do this "actively open-minded": they take time to think things through.
Reflection in action, reflection on action
Donald Schön described two kinds of reflection: thinking while you work ("this isn't going as planned, why?") and thinking afterwards. Both can be practised. A debrief trains the second; the habit of a mental pause trains the first.
Never make a decision in theatre
Big decisions made mid-operation are made under load, with System 2 squeezed. Where you can, think them through beforehand: what will I do if I find this? Then in theatre you are choosing a rehearsed plan, not inventing one.
Build it into the team
- Brief the "what could go wrong" out loud
- Invite challenge: "tell me what I'm missing"
- Name a decision observer in MDTs
- Debrief, every time it matters
See also Decisions for System 1 and 2, bias and noise, and Performance for overload.
References
- Flavell JH. Metacognition and cognitive monitoring. Am Psychol 1979;34:906–11.
- Croskerry P. The importance of cognitive errors in diagnosis and strategies to minimize them. Acad Med 2003;78:775–80.
- Meyer AND, Payne VL, Meeks DW et al. Physicians' diagnostic accuracy, confidence, and resource requests: a vignette study. JAMA Intern Med 2013;173:1952–8. doi ↗
- Moulton CA, Regehr G, Mylopoulos M, MacRae HM. Slowing down when you should: a new model of expert judgment. Acad Med 2007;82(10 Suppl):S109–16.
- Moulton CA, Regehr G, Lingard L, Merritt C, MacRae H. Slowing down to stay out of trouble in the operating room: remaining attentive in automaticity. Acad Med 2010;85:1571–7.
- Schön DA. The Reflective Practitioner. Basic Books, 1983.
- Kahneman D, Sibony O, Sunstein CR. Noise. William Collins, 2021.
State what you see, not your interpretation
Listening is hard work. We decide the gist of a message within about two seconds, and once we think we know what's going on we stop listening. Sustained listening lasts around 17 seconds in experimental studies, and bad news makes us zone out fastest.
Register
Social
Relationship-building; informal and imprecise.
Register
Professional
Structured and precise; shared terms used deliberately.
Register
Crisis
Stripped to essentials; scripted and unambiguous.
Hazard
Jargon
"Let's go up to 10,000 feet." What does it mean to everyone in the room?
Attention"
Use it when
- You anticipate a serious incident
- You need the full attention of everyone in theatre
When you hear it
- Stop what you are doing
- Establish silence
- Listen to the person alerting
- Decide as a team what to do
Bullet point thinking
- Prepare: what is relevant?
- Précis: cut the word count.
- Prioritise: most important first.
Avoid "but" before criticism; people stop listening. Signpost to extend attention.
Escalation
- Recognise you've reached the limit of competence or capacity.
- Who to call: chain of command, approachability.
- What to say: think before you call. Use SBAR.
SBAR
Feedback: "When you said this, it made me feel…" Say what you saw and its impact. Escalate line manager to line manager.
Brief, check, hand over, debrief
The brief turns private mental models into a shared one. Checklists are not a barrier in themselves; how the team engages with them is what matters. And a debrief is where the team learns.
A theatre day
Amber = sterile cockpit: no non-essential conversation or activity during brief, induction, time out and debrief.
Dr Liz Shepherd, paediatric anaesthesia
SAFE conversation: think before we do
SAfety
What are the patient safety concerns? What might be difficult? How can the team help?
human Factors
What might go wrong? How will we recognise it? What will we do?
Education
What are the learning opportunities for the team?
Handover
Handover passes on responsibility. It needs a package of data, information and insight, not just a list.
Routine work
Challenge–response
Confirms every step has been done when repetition dulls attention.
Rare or abnormal
Read–do
Better the right thing slowly than the wrong thing quickly.
The evidence for surgical checklists
| Review | Outcome | Risk ratio (95% CI) |
|---|---|---|
| Gillespie et al, Anesthesiology 2014 (7 studies, 37,339 patients) | Any complication | 0.63 (0.58–0.72) |
| Wound infection | 0.54 (0.40–0.72) | |
| Mortality | 0.79 (0.57–1.11) | |
| Bergs et al, BJS 2014 (WHO checklist) | Any complication | 0.59 (0.47–0.74) |
| Mortality | 0.77 (0.60–0.98) | |
| Borchard et al, Ann Surg 2012 | Mortality | 0.57 (0.42–0.76) |
| Lingard et al, BMJ Qual Saf 2011 (team briefing) | Antibiotics on time | 77.6% → 87.6% |
Mostly cohort and before-and-after studies. Benefit tracks adherence to the care the checklist embeds.
Learning from what goes right as well as wrong
Risk management is what we do now to make a good outcome more likely later. It can't remove risk from complex, hazardous work. It can give people under pressure, with competing goals and limited resources, the best chance of a safe result.
Safety I
Learn from failure
Incident investigation and root cause analysis. Still essential when harm occurs.
Safety II
Learn from success
Most cases go well despite the same pressures and workarounds. What is the team getting right, and how can the system support it?
Reason
Swiss cheese
Harm comes from a combination of events, circumstances and decisions lining up, not a single failure. Real healthcare is rarely that linear.
What distinguishes cases that go well
- Dissent is sought and heard
- Discussion of risk stays open after the plan is agreed
- Past success isn't treated as a guarantee
- Anyone can say "stop"
- Deference to expertise, wherever it sits in the hierarchy
- The team notices problems without waiting for audit
- Pride in the work
Normalisation of deviance
Gradual acceptance of departures from standard practice, because nothing bad happened last time. Linked themes: risk normalisation, production pressure, culture and a lack of visible consequences. It featured in the Challenger disaster.
Sedlar et al, J Safety Res 2023 (33 papers)
Risk holding is a learned skill, not an innate skill.Chris Whitty, BMJ / Academy of Medical Sciences lecture, September 2026
Over 30 years, risk has moved up from resident doctors to consultants and then senior consultants, and arrives all at once on day one as a consultant. Trainees asked for more responsibility, not less.
Five proposals
- Teach risk-holding skills early and systematically
- Map expected risk by career stage and specialty
- Increase it gradually, without cliff edges
- Redesign rotas for more stable teams
- Protect continuity with the same patients
Moberly T. BMJ 2026;394:e100924
Look after the people doing the work
Burnout is exhaustion, cynicism and reduced satisfaction: a burnout of caring. Chaotic work processes raise the risk. Solutions work at three levels: individual, team and system.
Tackling burnout at three levels
- Individual: work–life balance, peer support, coaching, counselling.
- Team: relationships and mentorship.
- System: good leadership and investment in culture and wellbeing.
Lozano-Franco et al, JVS-Vascular Insights 2024
Simon Edgar
Joy in work
Thinking, feeling and doing, with people you trust and outcomes you can see. You are not a high performer if you don't make the people around you better.
- Behaviours that build relational trust
- Kindness is contagious
- Do what you do well
- Rehearse your replies to incivility
Ask "How are you, really?"
Check on colleagues before you leave. Your colleagues are there to listen to you too.
Acute vs chronic
From overload to burnout
When overload goes on, with prolonged stress and adrenaline, it becomes burnout: more sustained, and it can lead to loss of engagement and ill health. The stretch zone is where we learn, with the right supervision and support. Beyond it, performance falls.
Spot it in yourself
Spot it in others
- Communication between colleagues breaks down
- Loss of compassion for co-workers
How to avoid overload
Delegate. Form an ad hoc team. Stop to think before doing more.
Try the ball-catching exercise to feel how quickly doing crowds out thinking.
Tackling burnout: individual, team, system
Individual
- Recognise signs of mental ill health in yourself
- Connect with a trusted relative or friend
- Use the support programmes your organisation offers
- Join peer support groups
- Schedule time for hobbies and leisure
- Set boundaries on working hours
- Delegate tasks
Team
- Build relationships outside the workplace
- Seek different kinds of mentorship: mentor, coach, sponsor, connector
- Create mentorship and sponsorship for female trainees
- Promote a culture of wellness
- Keep checking where support is lacking
System
- Leadership and organisational support
- Sensible duty hours and flexible working
- Access to counselling and wellness programmes
- More support staff in theatres, and fairer task distribution
- Fund evidence-based strategies against burnout
- Regular wellbeing checks for trainees
Lozano-Franco et al, JVS-Vascular Insights 2024. For support after an incident, see Coping.
Something will go wrong
Healthcare relies on high human performance, the "human as hero". But performance varies and fails under pressure. When we look at safety failures there is rarely a lack of knowledge or skill; the failure is in connecting what we know and can do to the task in hand. Sooner or later every clinician will be involved in an event that hurts a patient, and it will affect them too.
Second victim
When the clinician is hurt too
A second victim is a healthcare worker who is significantly affected, professionally or personally, by a patient safety incident. Many feel unsupported, or even punished, for an error they never intended.
Common effects include:
Warning signs to watch for
- Re-experiencing the event
- Avoidance and emotional numbing
- Hyperarousal
- Harmful behaviour
- Physical symptoms
- Problems at work and breakdown of relationships
secondvictim.co.uk
Common feelings after a clinical event
It is very common to feel shaken and upset after some clinical events.
Remembering, through dreams and intrusive memories, is common for a while. It usually eases over a few days or up to 4 weeks. Seek help if it isn't improving after a month, or contact your psychology team sooner if you wish.
Balance avoidance. Balance not thinking about it with allowing time to process what you have seen. If it is still distressing to talk about after a month, consider seeking help.
Sustained exposure to repeated intense challenges can cause more distress and fatigue than single events.
Adapted from TIM (Team Immediate Meet), Intensive Care Society
Useful actions
Before you leave
- Don't go home straight away
- Talk to someone you trust, or write a reflection before you go
- When going home, put it to bed before you go to bed
At home
- Treat yourself as you would your best friend
- Do something positive when you get home
- Put non-essential tasks on hold, sleep, avoid excess alcohol, exercise, and talk to people you trust
For the team
- Consider who could be affected after a traumatic event
- Cast your net widely
- Look out for the quieter members of your team
Remember
Look after yourself.
Look out for each other.
Busch et al 2021, adapted from Seys et al
A stepped model of support
Most people need the lower levels. A few need the top.
Managing your wellbeing
Ride the rollercoaster well. Take your breaks, step away for a few minutes to breathe, and allow real down time outside work.
Acknowledge you are only human. Clinical work provokes strong emotions, and they can feel harder when life outside work is difficult too.
Make space to reflect and share. Beware avoidant coping: keeping very busy, avoiding talking, over-eating and drinking to excess are all risk factors.
Beyond healthcare
Aviation already does this
Airlines run pilot peer assistance and support programmes. Second-victim responses are recognised across other high-risk industries too (Conti et al 2024).
Where to get support
BMA · 24 hours a day, 7 days a week
Counselling and peer support
0330 123 1245
Free and confidential, for all doctors and medical students, whether or not you are a BMA member.
BMA counselling and peer support ↗If you are in crisis or worried about your safety, call NHS 111 (option 2 for mental health) or 999, or go to A&E.
References
- Busch IM et al. Promoting the psychological well-being of healthcare providers facing the burden of adverse events: a systematic review of second victim support resources. IJERPH 2021;18:5080.
- Conti A et al. Second victims in industries beyond healthcare: a scoping review. Healthcare 2024;12:1835.
- Lozano-Franco R et al. Addressing burnout in surgery and vascular surgery. JVS-Vascular Insights 2024;2:100062.
- Mira JJ et al. Key elements for designing effective second victim support interventions: a focus group study in European clinical settings. BMJ Open 2025;15:e089923.
Watch the videos
Talks, teaching clips and discussion on human factors, teamwork and patient safety in paediatric surgery.
Open the channelLibrary
Websites, reports and books behind the teaching. Start with one and come back for more.
Organisations
Reports and articles
Books
| Book | Why read it |
|---|---|
| Sidney Dekker, The Field Guide to Understanding Human Error; Foundations of Safety Science | Hindsight bias, local rationality |
| Daniel Kahneman, Thinking, Fast and Slow | System 1 and System 2 |
| Kahneman, Sibony & Sunstein, Noise | Variation in judgement |
| Amy Edmondson, The Fearless Organization | Psychological safety |
| Matthew Syed, Black Box Thinking | Learning from error |
| Chabris & Simons, The Invisible Gorilla | Inattentional blindness |
| Matthew Walker, Why We Sleep | Fatigue and performance |
| Urs Koenig, Radical Humility | Always ask, listen and thank |
| Angela Duckworth, Grit | Effort over the talent myth |
Contact
Questions about the teaching, ideas for a session, or something that worked in your team? I'd like to hear from you.
Only you can see this
Messages received
Loading…
Acknowledgements
This resource grew out of teaching delivered together. Thank you to the colleagues and organisations who shaped it.
Colleagues
Dr Sam Harpur
Consultant Paediatric Anaesthetist and co-faculty on our Human Factors teaching days.
Sanj Gupta
Consultant Paediatric ENT Surgeon.
Rob Hearn
Consultant Anaesthetist and co-presenter of The Human Factor in Patient Safety teaching day.
Dr Liz Shepherd
Consultant Paediatric Anaesthetist and creator of the SAFE Conversation.
Training partner
Terema
For the human factors and team resource management training, models and free resources that underpin much of this site.
Phil Higton & Rachel Broadley
Facilitators of the Terema Human Factors in Risk Management masterclass.