Significant & Learning Event Analysis
Because every near-miss is a free lesson — and ignoring it is the expensive option.
SEA and LEA are two of the most powerful tools in GP training — and two of the most misunderstood. This page will demystify them, show you how to write one that actually matters, and explain why the Swiss cheese model is your best friend when something goes wrong.
⚡ Jump to section
DownloadsHandouts, templates & teaching resources
Handouts, summaries, and teaching extras — ready when you are. Includes SEA templates, risk assessment grids, worked examples, and more.
path: SIGNIFICANT EVENTS
- risk-assessment-in-significant-events
- significant-event-analysis-templates
- critical incidents and critical error - scenarios.doc
- critical incidents and critical error.doc
- critical incidents in primary care.ppt
- how doctors think - types of bias.pdf
- human factors and critical incidents.docx
- learning event activity and reflection.pptx
- patient safety - three bucket model.pdf
- risk assessment - profiling grids.pdf
- risk assessment and management.doc
- risk management and assessment - pascals wager.doc
- sea - cancer diagnosis template.doc
- sea - types of sea and more sea guidance.doc
- sea example in hospital - good example of looking at all the aligned holes in swiss cheese model.pdf
- significant event analysis by ram.ppt
- significant event analysis in general practice.ppt
- significant event example (1).pdf
- significant event example (2).pdf
- significant event example to work through.doc
- significant events - a practical approach.doc
- significant events by rcgp.pdf
- significant events guidance notes.pdf
- significant events in GP.ppt
Web ResourcesA hand-picked mix of official & real-world GP training resources
Because sometimes the best pearls are not hiding in the official documents.
Core Guidance⚡ If You Only Read One Thing…
In general practice, things go wrong. Not because GPs are careless or incompetent — but because general practice is a complex, high-volume, under-resourced system with many moving parts. A prescription going to the wrong patient. A test result that slipped through. A home visit that was logged but never actioned.
SEA and LEA exist to turn these moments into learning — for you, for your team, and for your patients. Done well, they are some of the most powerful quality improvement tools available to a GP.
Done well, SEA/LEA helps you…
- Identify system weaknesses before harm occurs
- Build a culture of openness in your practice
- Demonstrate reflective practice for ARCP & revalidation
- Genuinely improve patient safety
- Show professional growth and self-awareness
Done poorly, it becomes…
- A box-ticking exercise with no real change
- An individual blame exercise (completely wrong approach)
- A brief description of what happened with no analysis
- Evidence that is rejected or flagged at ARCP
- A missed learning opportunity for your whole team
Insider Tip: Why trainees find this hard
Most trainees struggle not with the process, but with the mindset shift. You've been trained to think about what you did wrong. SEA asks you to think about what the system made possible. That is a genuinely different — and more useful — way of thinking.
SEA vs LEA — What's the Difference?And when do you use which?
Important RCGP Update
The RCGP now uses the term Learning Event Analysis (LEA) as the broader category for most reflective practice on events. Significant Event Analysis (SEA) is retained specifically for events that reach the GMC threshold for harm. Both are logged separately on the FourteenFish ePortfolio.
| Feature | LEA — Learning Event Analysis | SEA — Significant Event Analysis |
|---|---|---|
| What happened? | Something notable occurred — good or bad — but harm did NOT reach the GMC threshold | An event occurred that did or could have caused harm meeting the GMC threshold |
| Examples | Near-miss prescription error; unexpected good outcome; unusual presentation; communication challenge | Wrong medication administered to patient; delayed cancer diagnosis; failure to act on abnormal result |
| On FourteenFish | Logged as LEA entry — does NOT need to appear on Form R | Must be reflected on AND declared on Form R before ARCP panel |
| Team meeting? | Ideally shared with team — but can be individual reflection | Should be discussed in a team SEA meeting where possible |
| Minimum per year? | At least 1 per training year (ST1, ST2, ST3) | Required only if an event reaches GMC harm threshold |
| Tone | Curious, exploratory, systems-focused | Honest, analytical, focused on prevention & change |
Quick Rule of Thumb
Ask yourself: "Did this, or could this, have caused significant harm to a patient?" If yes → SEA. If not, but it was still a notable learning event → LEA. When in doubt, discuss with your trainer — and err on the side of the more thorough process.
Decision Flowchart — Which Log Entry Should You Use?Never be unsure again
One of the most common practical confusions among trainees — especially in hospital posts — is which type of entry to create. This flowchart answers that in under 30 seconds.
Still not sure? Default upward
If you're genuinely unsure whether something meets the SEA threshold, err on the side of treating it as an SEA. It's better to over-document than under-document. Talk to your trainer if you're not certain — that conversation itself is a learning opportunity.
What Counts as an Event?Far more than you'd think
One of the most common questions from trainees is: "Is this worth writing up?" The answer is almost always yes — if it made you pause and think, it's worth exploring.
🔴 Classic SEA Events
- Wrong medication prescribed or dispensed
- Failure to act on abnormal blood result
- Delayed diagnosis with harm
- Incorrect patient record — prescription or referral
- Missed home visit with clinical consequences
- Safeguarding concern not escalated
- Medication overdose or interaction causing harm
🟡 Near-Miss / LEA Events
- Prescription sent to wrong patient (spotted in time)
- Blood result not reviewed until follow-up chase
- Referral to wrong specialty (corrected before appointment)
- Test ordered but not chased up
- Confidentiality breach avoided at last moment
- Miscommunication between GP and pharmacist
- Home visit to wrong address (patient was fine)
🟢 Positive Event LEAs
- Unexpected early cancer diagnosis in a routine consultation
- Excellent management of a complex, ambiguous presentation
- Patient expressed gratitude for unusually thorough care
- Effective team communication preventing a serious error
- Brilliant colleague intervention that you observed
- Creative safety-netting that turned out to be crucial
- Unusually effective communication with a distressed patient
Small events matter too
Significant Events don't have to be dramatic. A receptionist discussing a patient's medication at the front desk in front of other patients — that's a confidentiality event. A prescription printed in the wrong record — that's a near-miss. These are exactly the small, low-harm events that reveal systemic weaknesses. If left unexamined, they're the cracks through which bigger failures later fall.
The Swiss Cheese ModelThe most important patient safety concept you'll ever learn
The Swiss Cheese Model was developed by Professor James Reason and is now used everywhere — from aviation to nuclear power to the NHS. It explains why bad things happen even when everyone is trying hard to prevent them.
🧀 How Harm Gets Through the System
Each layer of cheese = a safety barrier. The holes = weaknesses in that barrier. When holes line up — harm reaches the patient.
Based on Reason's Swiss Cheese Model (2000). Adapted for primary care by Bradford VTS.
✅ When the System PROTECTS the Patient
Even if one safety layer fails (a hole is penetrated), the next layer catches it. The holes don't line up. No harm reaches the patient. This is why most errors are caught before they cause harm — but only if the barriers are all doing their job.
Example: A wrong dose is prescribed → the pharmacist spots it → patient is safe. The prescription layer had a hole, but the dispensing layer didn't.
⚠️ When the System FAILS the Patient
When holes align across multiple layers simultaneously, harm gets through. This is rare, but it doesn't require every layer to fail — just enough of them, in the wrong way, at the same time.
Example: Wrong dose prescribed → pharmacist is busy and misses it → patient doesn't read leaflet → harm results. Multiple layers failed together.
Active Failures vs Latent Conditions
This is a key distinction in the Swiss Cheese Model — and often missed in trainee SEAs:
| Type | What is it? | GP Example | Who creates it? |
|---|---|---|---|
| Active Failure | An unsafe act by the person at the "sharp end" — immediately obvious when it goes wrong | GP prescribes the wrong drug; receptionist books appointment in wrong patient record | Individual at the frontline — but almost always influenced by latent conditions |
| Latent Condition | A pre-existing weakness in the system, often invisible until an active failure triggers it | Two patients with similar names on the same list; outdated prescribing protocol; no double-check system | System design, management decisions, organisational culture — often set up long before the event |
The Golden Insight
"System flaws, not character flaws." The Swiss Cheese Model teaches us that most medical errors are the result of systems that make it easy to do the wrong thing — not bad doctors doing bad things. Your SEA should explore the latent conditions, not just the active failure on the surface.
What Types of Events Get Written Up?A visual breakdown — and what that means for you
Research reviewing SEA reports from UK primary care shows a consistent pattern of which event categories come up most often. Understanding these categories helps you recognise them in your own practice — before they become serious.
Common SEA Categories in UK General Practice
Based on published UK primary care SEA research. See legend below.
What the Categories Mean for Trainees
Wrong drug, wrong dose, allergy not checked, interaction missed
Test result not communicated, miscommunication between team members
Delayed diagnosis, missed diagnosis, atypical presentations
Booking errors, record errors, IT failures, filing mistakes
Incorrect specialty, delayed referral, referral lost in system
Safeguarding, confidentiality, equipment, facilities
What this means practically
The most common SEA categories — prescribing, communication, and diagnosis — are also the ones where the Swiss Cheese Model applies most clearly. Each one involves multiple layers: clinical decision-making, system design, protocol existence, communication channels, safety-netting. Next time you're choosing what to write up, look for events that fall in these categories. They're the richest learning opportunities — and the most likely to reveal genuine systemic issues.
LFPSE — The New National Reporting SystemWhat changed in 2024, and why it matters for trainees
Important Update — 2024
The old National Reporting and Learning System (NRLS) was decommissioned in June 2024. It has been replaced by the Learn from Patient Safety Events (LFPSE) service. GP practices are now required to register with and use LFPSE. This is a contractual requirement from the 2025/26 GP contract.
How the Reporting System Works — From SEA to National Learning
What is LFPSE?
LFPSE is the NHS's new national system for recording patient safety events. It replaced the NRLS (decommissioned June 2024). All GP practices must register and submit applicable events. It is designed to be simpler to use, faster to submit, and better at generating national learning from safety data.
What does this mean for GP trainees?
As a trainee, you are unlikely to submit to LFPSE directly — your practice manager or Clinical Lead usually does this. But you should know it exists. Your SEA documentation in FourteenFish and at the practice level feeds into this wider system. Your learning contributes to national patient safety improvement.
📊 The Healthy Reporting Culture — What Good Looks Like
A healthy reporting culture produces more near-miss reports than harm reports. A ratio of 3:1 (near-misses to harm events) or higher shows a team is catching problems before they reach patients. Low near-miss reporting usually means people aren't reporting — not that nothing is happening.
How to Write a Good SEA or LEAStep-by-step, with what assessors want to see
🧠 The RAM Framework for Writing Your SEA/LEA
The 4 W's. Simple. Memorable. Assessors genuinely look for all four.
Step-by-Step: Doing Your SEA or LEA
Identify the event
Notice something worth exploring — not just dramatic events. Near-misses, unexpected good outcomes, and system wobbles all count. The best SEAs often come from small, quiet events that most people would ignore.
Immediate action first
Before reflecting, fix any ongoing harm. Contact the patient if needed. Alert your supervisor. Notify the relevant team. Then document. Don't write the SEA before you've dealt with the patient.
Describe what happened (factually and clearly)
Chronological narrative. Anonymised patient (use initials or "a 60-year-old man"). What happened, when, who was involved. Be precise. Do not write a vague summary — write the story.
Analyse WHY it happened (this is the most important bit)
Identify contributing factors — clinical, organisational, human, environmental. Apply the Swiss Cheese Model: what were the active failures? What latent conditions allowed this? Explore each layer of the system. This section should be the longest and most insightful part of your SEA.
What did you learn?
Personal reflection — what did this teach you about yourself, your practice, or the system? Be honest. Assessors love genuine insight. They can tell when it's formulaic or rehearsed.
Discuss with the team (for SEA — essential; for LEA — strongly encouraged)
Bring it to an SEA meeting, tutorial, or team discussion. More perspectives = richer analysis = better action plans. Document who was present and what was said.
Action plan — specific, assigned, time-bound
NOT "I will read up on X." Something tangible: "The practice will update the repeat prescribing protocol by [date]. Dr X will lead this. Review at SEA meeting in 3 months." Assessors want to see SMART actions.
Review and close the loop
Did the change happen? Did it make a difference? A SEA that never gets followed up is incomplete. You should be able to add a note: "Reviewed 3 months later — protocol updated, no further events."
Risk Assessment: How Bad Was It? (and Could It Be?)
A good SEA includes a risk assessment. This means scoring the event on two dimensions:
| Dimension | 1 — Minor | 2 — Moderate | 3 — Severe |
|---|---|---|---|
| Likelihood How likely is this to happen again? |
Rare — unlikely to recur | Possible — might happen again | Likely — will happen again without change |
| Severity How much harm could result? |
Minimal harm — no significant impact | Moderate harm — temporary impact | Serious harm — permanent impact or death |
The Risk Matrix
Multiply likelihood × severity to get a risk score. High-scoring events need urgent action plans. Low-scoring events may just need documentation and monitoring. This shows you've thought systematically about the event — and assessors love it.
Reflective Models Made VisualUse these to structure your write-up — pick the one that fits you
All SEA and LEA write-ups need good reflection. These models give you a structure to hang your thinking on. You don't need to follow any of them rigidly — but having one in mind helps you avoid the most common mistake: describing what happened without actually reflecting on it.
1. Gibbs' Reflective Cycle (1988) — The Most Popular Choice
Widely recommended for GP ePortfolio entries. Structured, cyclical, and thorough. Works especially well for SEA/LEA write-ups.
The cycle then repeats — learning is ongoing, not a one-off event. Gibbs G (1988). Learning by Doing.
2. Driscoll's "What? So What? Now What?" — The Quick Option
Perfect for near-miss events where you want a fast, clear structure. Especially useful for LEAs. Three questions. That's it.
WHAT?
What exactly happened? Stick to the facts. Brief, clear, chronological. No analysis yet.
Description only
SO WHAT?
Why does it matter? What did you learn? What system factors were involved? How did it make you feel?
Analysis + reflection
NOW WHAT?
What will change? Specific actions — for you, for the team, for the system. Who? What? By when?
Action plan
Which model should you use?
For a full SEA (harm event) — Gibbs gives you the depth you need. For a LEA (near-miss or positive event) — Driscoll's "What, So What, Now What" is perfectly adequate and much quicker to complete. Either way, the most important part is stages 4 and 6 of Gibbs (or the "So What / Now What" of Driscoll) — the analysis and the action plan.
Good vs Weak SEA WritingThe difference between 'adequate' and 'impressive'
❌ Weak SEA Analysis Section
"I prescribed the wrong antibiotic. I should have been more careful. I will make sure I check the allergy box before prescribing in future."
Why it fails: Individual blame, no system analysis, vague action plan, no learning.
✅ Strong SEA Analysis Section
"Applying the Swiss Cheese Model: at the clinical layer, the allergy alert was dismissed without being read. At the system layer, the alert was small, amber-coloured, and appeared alongside routine warnings — making it easy to overlook. At the organisational layer, there was no second-check protocol for patients with known drug allergies. Three layers failed simultaneously. The latent condition was the poor visual salience of allergy alerts in our clinical system — a known design issue. The immediate action was to contact the patient and check for adverse effects. The team action plan is to request an IT change to make allergy alerts visually prominent, and to add a verbal allergy check to the prescribing protocol."
Why it works: Systems thinking, Swiss Cheese applied, specific latent conditions identified, SMART action plan.
What Gets You High Marks
- Using the Swiss Cheese Model explicitly — naming the layers and the holes
- Distinguishing active failures from latent conditions
- Honest personal reflection: "I felt…", "I realised…", "This challenged my assumption that…"
- A team discussion with named participants
- Specific, time-bound, assignable action points
- Evidence that the action happened — a follow-up note
Common Pitfalls & Trainee TrapsMistakes that weaken your SEA or LEA
⚠️ Common Mistakes Trainees Make
- Individual blame: "I should have been more careful" — this misses the point entirely
- Only describing what happened — with no analysis of why
- Vague action plans: "I will read up on X" is not an action
- Choosing easy, low-risk events that don't generate real learning
- Not involving the team in the discussion
- Never following up on whether the action happened
- Confusing LEA and SEA — putting harm-level events as LEAs
🎯 What Candidates Often Forget
- SEAs must appear on Form R — not just in the learning log
- The analysis section is where assessors judge the quality
- Positive events are just as valid as adverse ones — and often produce better insight
- You need the event anonymised — initials, age category, no identifying details
- Human factors matter — tiredness, cognitive load, time pressure are all valid contributors
- Emotional impact is valid too — "how this made me feel" is a legitimate part of reflection
Insider Tip: The "Blame Instinct" Problem
When something goes wrong, the first instinct — especially for doctors — is to blame themselves. "I should have known better." This instinct is understandable but unhelpful in SEA. The practice manager who set up the booking system, the IT team who designed the allergy alert, the staffing pressure that meant no double-check happened — these are the factors that need addressing. Blaming yourself changes nothing in the system. System analysis changes everything.
🚩 Red Flags in Your SEA Write-Up
- The word "I" appears in every sentence of the analysis — individual blame, not systems thinking
- No mention of organisational or system factors
- Action plan says "be more careful" or "read the guidelines" — too vague and personal
- No team discussion section at all
- The event is clearly not significant — "I saw an interesting case" without any near-miss or learning value
- No follow-up to check whether the action plan was enacted
What Trainees Actually SayRecurring insights from real trainee experience across the UK
These patterns come from trainee discussions, GP training forums, and educational communities across the UK. Only insights that align with official guidance are included here.
"I wrote three paragraphs describing what happened. My trainer said — where's the analysis?"
This is the single most common piece of feedback trainees receive. Description is just the starting point. The whole point of the exercise is the WHY — what in the system allowed this? Without that, the write-up is just a diary entry.
"I didn't know SEAs had to go on Form R. Found out the week before my ARCP."
This trips up trainees every year. LEAs stay in the FourteenFish learning log only. SEAs (harm events that meet the GMC threshold) must be declared on Form R. Missing this can lead to an ARCP panel requesting you resubmit.
"My best SEA was a near-miss I nearly didn't bother writing up. It turned out to be the most useful one."
Near-misses are where the real learning lives. A prescription nearly sent to the wrong patient. A referral to the wrong specialty spotted before it was sent. These reveal systemic cracks — and they're much easier to write about honestly because no harm occurred.
"I asked my trainer to look at my SEA BEFORE I wrote it up. Best thing I ever did."
Talking it through first makes the write-up far richer. Your trainer has seen this type of event before. They know which system factors to look for, which layers of the Swiss Cheese are relevant, and how to frame the action plan to make it genuinely meaningful.
"Don't write all your log entries in the last two weeks. The ARCP panel can see exactly when you added them."
The FourteenFish ePortfolio timestamps every entry. Reviewing panels notice when all your SEAs and logs appear in a rush at the end of the review period. Spread them out. Write them close to when events happen — that also means they're more honest and more detailed.
"Quality matters more than length. My trainer said 'concise and insightful beats long and descriptive every time.'"
A tight, focused SEA of 400 words with genuine systems analysis is worth more than 1,500 words of narrative. Trainers and panels are reading a lot of these. Get to the point. Focus on the analysis and the action — not the story.
"When a patient complained about me, I didn't realise that also needed to be documented as an SEA."
All complaints where you have personal involvement should be documented as SEAs in your FourteenFish ePortfolio. This is an RCGP requirement — not optional. If you've been involved in a complaint, discuss with your trainer and document it promptly.
"I kept choosing 'easy' events that weren't really significant. My trainer pulled me up on it."
Choosing minor events to tick the box is a commonly spotted pattern. Trainers know when a trainee has played it safe. The purpose of SEA is genuine learning, not portfolio decoration. Pick events that actually challenged you — even if they feel uncomfortable to write about.
🎓 What Trainees Consistently Wish They Had Known Earlier
Insider Pearls & Real-World WisdomWhat trainees who've done this well actually say
The Near-Miss is Gold
The most powerful SEAs often come from near-misses — events where harm was avoided. These are easier to write about honestly (there's less defensiveness) and they often reveal more systemic issues than actual adverse events. Don't wait for something to go badly wrong before you write one.
Human Factors Count
Tiredness, hunger, cognitive overload, a difficult previous consultation, an interruption mid-task — these are not excuses. They are legitimate contributing factors in the Swiss Cheese Model. Your SEA should include them. Identifying them helps the system be redesigned to account for human limitations.
Positive SEAs Are Underused
Most trainees write SEAs about things that went wrong. But a really insightful positive event LEA — "I diagnosed a rare condition early because of X" — can be just as powerful. It shows you're not just avoiding mistakes, but actively seeking out what makes excellent care happen and replicating it.
Your Trainer Is a Resource, Not Just an Assessor
Many trainees bring SEAs to their trainer only after writing them alone. Better: discuss the event with your trainer first, before you write. They've seen this kind of error before. They know what analysis layers to explore. Use them as a thinking partner — the write-up will be much richer.
What Actually Gets You Good Marks
- An analysis that clearly uses the Swiss Cheese Model — even if you don't name it explicitly
- Honest emotional reflection — "I felt anxious", "I was embarrassed", "I was frustrated" — these are real and valid
- Naming systemic rather than individual causes as the primary analysis
- A team meeting that changed something real — documented with names and dates
- A follow-up: "Three months later, the protocol was updated and no further incidents occurred"
- Showing that this event changed how you or your practice operates — not just what you know
Primary Care Shortcut: The 5 Whys Technique
When analysing a significant event, try asking "Why?" five times in a row. Each answer reveals a deeper cause. By the fifth Why, you're almost always at a systemic or organisational factor rather than an individual one. This simple technique transforms surface-level blame into genuine root-cause analysis — and makes your SEA analysis substantially stronger.
Example: "The wrong medication was prescribed." Why? "The allergy alert was missed." Why? "Allergy alerts look similar to routine warnings." Why? "Our clinical system doesn't prioritise serious alerts visually." Why? "Nobody configured this when the system was installed." Why? "There was no protocol for system configuration at go-live." → Systemic, latent condition identified.
Practical Tips From UK GP Training CommunitiesThe small things that make a big difference
📝 On Writing
- Write it within 48 hours. Your memory fades fast — and so does your emotional honesty.
- Keep the description to 3–4 sentences. Save your word count for the analysis.
- Use plain, direct language. Avoid clinical jargon in the reflection sections — write as you think.
- Don't hedge your analysis. "I think this might possibly have contributed" is weaker than "This was a contributory factor because…"
- A good title makes a difference: "Near-miss prescription error in a penicillin-allergic patient" tells the assessor far more than "Prescribing event."
🤝 On Team Discussion
- Bring the whole team — the receptionist who made the booking error is often the person with the most insight into the system flaw.
- The meeting should feel safe, not disciplinary. "What allowed this to happen?" not "Who did this wrong?"
- Name the people who attended in your write-up — this shows the team discussion actually happened.
- If you can't get a full team meeting, a conversation with your trainer still counts — document it.
- For hospital posts — a Morbidity & Mortality (M&M) meeting or Datix report can feed directly into an LEA entry on your ePortfolio.
✅ On Action Plans
- Every action needs an owner (a named person), a task (specific, not vague), and a date (by when).
- At least one action should be systemic — not just personal. Show you've thought beyond yourself.
- Come back three months later and write a follow-up note: "Protocol updated as agreed — confirmed by practice manager April 2025."
- Include a PDP entry linked to your SEA — your trainer loves seeing the loop close properly.
- If the practice says "no change needed" — document that too, and explain why. That is also a valid outcome.
Primary Care Shortcut: The "Repeat Tomorrow" Test
After writing your action plan, ask: "If this exact event happened again tomorrow, with the same staff and the same system, would my action plan prevent it?" If the answer is no — your actions are not specific enough. Keep revising until the answer is clearly yes. This one question transforms vague intentions into genuine system change.
Deeper Knowledge — Click to ExploreFor those who want to go further
Enhanced SEA (or Enhanced Learning Event Analysis) is a development of the standard process, created with a human factors approach. It was piloted by NHS Education for Scotland (NES) and is increasingly used across UK GP training.
It adds a structured analysis of contributing factors under four domains:
- Patient factors — patient characteristics, behaviour, complexity
- Staff and provider factors — knowledge gaps, workload, human error types
- Environment factors — physical space, staffing levels, technology
- Organisation factors — policies, culture, protocols, leadership
Enhanced SEA also specifically addresses the personal emotional impact on the healthcare professional involved — acknowledging that being involved in a significant event is stressful, and that wellbeing matters in reflection.
See the NES Enhanced SEA resources in the Web Resources section above.
Reason's model identifies four key types of human error, all of which appear in GP significant events:
| Error Type | What it is | GP Example |
|---|---|---|
| Slips | Automatic actions done incorrectly — attention failure | Selecting the wrong patient from a dropdown list of similar names |
| Lapses | Memory failures — forgetting to do something planned | Forgetting to check the results of a blood test you ordered |
| Mistakes | Rule-based or knowledge-based errors — doing the wrong thing deliberately but for the wrong reasons | Prescribing a drug contraindicated in renal failure because the protocol is outdated |
| Violations | Deliberate deviations from safe practice | Skipping the allergy check because "it takes too long" — common under time pressure |
Identifying the error type in your SEA shows sophisticated analysis. Most GP significant events involve slips or lapses under time pressure — rarely violations. Understanding this shifts the response from blame to system design.
The Three Bucket Model is a simpler companion to the Swiss Cheese Model, used in patient safety education to show how conditions accumulate before an adverse event:
🪣 Bucket 1: Self
- Tiredness, illness
- Personal stress
- Emotional state
- Unfamiliar task
🪣 Bucket 2: Context
- Busy clinic
- Interruptions
- Unusual circumstances
- Time pressure
🪣 Bucket 3: Task
- Complex/unfamiliar task
- Poorly designed system
- Missing information
- Ambiguous protocol
When all three buckets are "full" — i.e., you are tired, the clinic is hectic, and the task is complex — the risk of error is highest. This model is particularly useful for explaining to trainees why errors happen at specific moments rather than randomly throughout a day.
For qualified GPs, SEA evidence is a mandatory component of both annual appraisal and the 5-yearly GMC revalidation process.
- GPs must declare all significant events at appraisal — particularly those where they were named
- The GMC's professional duty of candour requires GPs to be open and honest about adverse events
- CQC inspectors actively look for a practice's SEA culture — practices rated "Outstanding" typically have robust, regular, team-wide SEA processes
- Individual SEAs can be used as appraisal evidence even if you were not directly involved — observing or hearing about a team SEA counts
- SEA can be used to demonstrate learning in the "quality improvement" domain of appraisal
For Trainees: Start the Habit Now
The trainees who find SEA easiest as a qualified GP are those who genuinely engage with it during training. Treat it as a learning tool — not a documentation chore — and the habit will serve you well for your entire career.
For Trainers & TPDsHow to teach SEA well — and what to look for
Core Teaching Challenge
Most trainees approach SEA with a mixture of anxiety and defensiveness. They've been trained in a culture of individual accountability and high standards — and significant events feel like admissions of failure. Your job as a trainer is to shift this to genuine systems thinking. That shift is the most important thing you can teach them about patient safety.
Common trainee blind spots on this topic:
- Confusing SEA with personal criticism or disciplinary process
- Identifying only the "active failure" and missing latent conditions entirely
- Writing a description of events with no genuine analysis layer
- Producing vague, unenforceable action plans
- Not following up on whether the action plan was actually implemented
- Only writing up negative events — missing the value of positive ones
- Not understanding the difference between LEA and SEA, and misfiling on FourteenFish
Tutorial ideas and approaches:
- Draw the Swiss Cheese Model together before the trainee shows you their SEA — then apply it to their event
- Use the "5 Whys" technique together in the tutorial
- Role-play the team SEA meeting — what would you say? Who would you involve?
- Review an anonymised SEA example and identify what's weak — build the habit of critical analysis
- Ask the trainee: "If this event happened again tomorrow, what would stop it?" — forces systems thinking
- Discuss the emotional impact explicitly: "How did this make you feel?" — normalises human factors
Reflective questions to use with trainees:
- "If this happened to a colleague, what system change would you recommend?"
- "What was the latent condition that made this possible?"
- "Who else in the system played a role in this event — even if they didn't cause it?"
- "What would need to change so this couldn't happen again — even on your worst day?"
- "How did this event affect your confidence? Your wellbeing?"
- "What does this event tell you about how our practice works — not just about you?"
What Trainers & TPDs Love to See
- A trainee who can distinguish active failures from latent conditions without prompting
- Genuine emotional honesty about the impact of the event on them personally
- An action plan that involves the whole team — not just personal learning goals
- Evidence that the trainee followed up and confirmed the change was made
- A positive event written up with the same rigour as an adverse one
- A trainee who sees SEA as a learning opportunity, not a punishment
Quick FAQsQuestions trainees actually ask
Yes — near-misses (where harm was avoided) are some of the most valuable events to analyse. They reveal system weaknesses before harm occurs. The correct form for these is usually LEA, not SEA — unless the event could have caused serious harm even if it didn't.
Absolutely — and it's encouraged. A positive event LEA might analyse why an unexpectedly good clinical outcome occurred, or why a particularly effective team communication prevented an error. The aim is to understand what made it work so it can be replicated.
Yes — all SEAs (events reaching GMC harm threshold) must be declared on Form R before your ARCP panel. LEAs do NOT need to appear on Form R. Getting this wrong is a common ARCP pitfall.
The culture of SEA should be supportive, not punitive. If you feel unsafe reporting an event, this is itself a significant concern about your learning environment — speak to your trainer or TPD. You can also contact your deanery or the BMA for confidential advice. A culture that punishes self-reporting is itself a safety risk.
Yes. Events from hospital posts can absolutely be documented as LEAs or SEAs in your FourteenFish ePortfolio. You may want to discuss with your educational supervisor whether the hospital team should also be informed so a parallel process can happen at their end.
Quality matters far more than length. A thoughtful, specific analysis of 400–600 words is worth far more than a detailed description of 1,500 words with a weak analysis section. Focus on depth of analysis and quality of the action plan. If the event is complex, it can be longer — but never pad for length.
Final Take-Home PointsThe bits to remember tomorrow
When Not to Panic
If you've had a near-miss event and you're worried about how to document it — don't panic. Near-misses are not evidence that you're a bad doctor. They're evidence that you notice things, reflect on them, and want to do better. That's exactly what an assessor wants to see. Write it up honestly, apply the Swiss Cheese, and propose a real action. You'll be fine.
Watch this little SEA video...
The predecessor term to Significant Event Review was Critical Incident Analysis. This was a process to look at those negative outcomes which may have been triggered by negative events. The idea was based on the notion that if you could identify what were the causes which led to a bad undesirable outcome (for example, giving the wrong prescription to a patient) then you might be able to change things around to stop them happening in the future and thus improving patient safety and care.
The term was changed to significant event because sometimes you can get a really great outcome that was not anticipated. By reviewing the events which led to that “great” outcome might just be able to help you duplicate it and do it more often.
So, significant events can be good or bad outcomes. Significant Event Review is a process to help you look at such events to identify factors which might need changing or promoting with the ultimate aim of improving patient safety and care. Promoting patient safety has other ramifications: it protects the practice, its staff and the doctors too. In other words, everyone’s a winner.