Key takeaways
- A significant event is anything that could have caused, or did cause, harm, or that shows unusually good practice worth learning from.
- Errors are estimated to occur in 2% to 3% of primary care encounters, so every practice has events to learn from; a quiet log is a warning sign, not a success.
- Keep the review blame-free and focused on systems: task, team, environment, organisation and patient factors.
- Every review should end with named actions, a date and a check that the change worked.
- Screen each event for the duty of candour, CQC notifications and national recording.
What counts as a significant event
Significant event analysis is a structured way for a team to look at something that happened, understand why and change what needs changing. In primary care it covers incidents that caused harm, near misses that could have, and events where things went unusually well. CQC has a dedicated GP mythbuster on significant event analysis[1], and inspectors in dental and private settings ask similar questions: do you record events, review them properly and learn from them?
Typical events include a delayed referral, an abnormal result not acted on, a prescription issued for the wrong patient, a needlestick injury, a confidentiality breach, a patient who collapsed in the waiting room, or a complaint that revealed a process gap. Good-practice events might include a receptionist who spotted a deteriorating patient on the phone, or a nurse who caught a dosing error before it reached the patient.
How often things go wrong
Most care is safe, but errors are not rare. The NHS Patient Safety Strategy cites international reports indicating that errors occur in 2% to 3% of primary care encounters and that one in 25 of those involve serious harm to a patient[2]. The most common problems identified were difficulties making the right diagnosis, delays in referring patients to hospital and prescribing errors[2].
- At 2% error rate
- At 3% error rate
Source: NHS England and NHS Improvement, 2019[2]
Show the data
| Item | At 2% error rate | At 3% error rate |
|---|---|---|
| Errors | 200 | 300 |
| Errors involving serious harm (1 in 25) | 8 | 12 |
Example: applied to a practice with 10,000 encounters a quarter, those rates would suggest between 200 and 300 errors, of which perhaps 8 to 12 could involve serious harm. A practice that records only two or three significant events a year is almost certainly missing most of them. The aim is not to count errors but to make it normal to report them.
Pressure raises the risk. In a BMA survey of more than 1,300 GP practices, 74% saw workload rise after online access changes, 68% reported more stress and 55% reported negative effects on patient care[3]. Contributory factors like these belong in every analysis.
Keep it blame-free
The patient safety strategy is direct about why blame fails: too often individuals have not felt safe to admit errors and learn from them, and the mistaken belief persists that patient safety is about individual effort, when it depends on how normal human behaviour interacts with systems[2]. A review that ends with 'staff member reminded to be more careful' has usually stopped too early.
- Start every meeting by saying its purpose is learning, not blame.
- Ask what made the error possible or likely, not who made it.
- Look for the same weakness elsewhere: if one result was missed, how many others could be?
- Thank the person who reported, especially if they reported their own mistake.
- Keep any separate performance or conduct process entirely apart from the review.
A seven-step process
1. Record
Anyone can log an event on the day: what happened, when, who was involved, immediate actions taken. Keep the form short so people use it.
2. Screen
A named lead reviews new events within two working days, decides whether a full analysis is needed and checks for duty of candour, CQC notification or other reporting.
3. Gather facts
Build a simple timeline from records, messages and short conversations with those involved. Stick to facts.
4. Meet
Hold a 30-minute meeting with the people involved and at least one person who was not, including a non-clinical team member where relevant.
5. Analyse
Ask what happened, why it happened and what could stop it happening again, working through task, team, environment, organisation and patient factors.
6. Act
Agree two or three specific actions, each with an owner and a date. Prefer changes to systems over reminders.
7. Close and check
Share the learning with the whole team, then check after an agreed period that the change has worked and stuck.
Running the meeting in 30 minutes
Long meetings put people off reporting. Most significant events can be reviewed in half an hour if the facts are gathered beforehand and the chair keeps to a simple agenda. Book the slot within two weeks of the event, while memories are fresh, and circulate the timeline a day ahead.
- Two minutes: the chair restates the purpose, which is learning, and the confidentiality of the discussion.
- Five minutes: the timeline is read out and corrected by those involved.
- Ten minutes: the group discusses why it happened, working through each type of contributory factor.
- Eight minutes: the group agrees two or three actions, each with an owner and a date.
- Three minutes: the group decides who needs to hear the learning and how, and whether the patient has been kept informed.
- Two minutes: the chair confirms the review date and closes the meeting.
Rotate the chair between partners, managers and senior nurses so ownership of safety is shared. A trainee or a receptionist chairing occasionally, with support, is a strong signal that every role has a voice.
A one-page template
| Section | What to write |
|---|---|
| Event summary | Two or three sentences in plain language, without names of patients |
| Date, location, people involved | Roles rather than names where possible |
| Immediate actions | What was done on the day to keep the patient safe |
| Timeline | Key times and decisions, from records |
| What went well | Anything that limited harm or showed good practice |
| What could have gone better | The gaps, described as system issues |
| Contributory factors | Task, team, environment, organisation, patient |
| Actions | Each with owner, deadline and how success will be checked |
| Candour and reporting | Duty of candour applied? CQC notified? Recorded nationally? |
| Learning shared | Where and when, and with whom |
| Review date | When the actions will be checked |
Candour, notifications and national recording
Every event should be screened against three duties. The duty of candour under regulation 20 applies to notifiable safety incidents: the provider must tell the relevant person as soon as reasonably practicable, in person first by one or more representatives, followed by written notification[4]. Some incidents must also be notified to CQC. And patient safety events can be recorded nationally on the Learn from Patient Safety Events service, which CQC's mythbuster list references directly[1].
The analysis and the candour conversation are separate but linked. The patient or family should know an analysis is happening, have a chance to contribute their account and hear what was learned. Our guide to Learn from Patient Safety Events explains how national recording works.
Making the learning stick
The most common failure in significant event analysis is not the meeting but the follow-through. Actions are agreed, minuted and forgotten, and the same event happens again a year later. A few habits prevent this.
- Keep a single action log for all events, reviewed at every practice meeting.
- Prefer actions that change the system, such as a new search, a template or a checklist, over actions that rely on memory.
- Re-run any search or audit that revealed the problem after three months to check the fix worked.
- Review themes across events every six months: are the same kinds of event recurring?
- Share anonymised summaries with the wider team, including reception, so everyone sees what changed.
Example: a missed abnormal result
Example (fictional): a patient's raised blood glucose result arrives while the requesting GP is on leave. It is filed as normal by a locum working through a backlog. Three weeks later the patient phones about symptoms and the result is found. The receptionist who took the call reports the event the same day.
The review timeline shows no buddy system for results during leave and a results inbox filtered by requesting clinician. Contributory factors include a heavy backlog, an unfamiliar system for the locum and no guidance on which results need action. The practice introduces a named buddy for every absence, a daily check that the results inbox is clear, and a short guide for locums. The patient receives an apology and an explanation of the changes. Three months later a search confirms no results sat unactioned for more than two working days.
Learning from what went well
Analysing good practice is as valuable as analysing errors, and it changes how the team feels about the process. When a receptionist spots a caller with stroke symptoms and gets them to 999 within minutes, or a nurse picks up a safeguarding concern during a routine dressing, run a short review: what made it go well, and how can the practice make that more likely to happen again? Recording these events alongside incidents gives a fuller picture of safety and makes reporting feel less like confession.
Frequently asked questions
What is significant event analysis?
A structured, blame-free review of an event that caused or could have caused harm, or that showed good practice, aimed at understanding why it happened and changing systems so learning is acted on.
How many significant events should a practice record?
There is no target, but errors are estimated to occur in 2% to 3% of primary care encounters, so a practice recording only a handful a year is probably under-reporting. Aim for a culture where reporting is normal.
Who should attend an SEA meeting?
The people involved, at least one person who was not, and someone who can authorise changes. Include reception or admin staff when the event touches their work.
Does every significant event need duty of candour?
No. The duty of candour applies to notifiable safety incidents as defined in regulation 20. Screen every event for it, along with CQC notifications and national recording.
References
Lincolnshire Local Medical Committee (2026). CQC mythbuster updates December 2024 (list of GP mythbusters) (opens in a new tab) ↑
NHS England and NHS Improvement (2019). The NHS Patient Safety Strategy: safer culture, safer systems, safer patients (opens in a new tab) ↑
British Medical Association (2026). Pressures in general practice data analysis (opens in a new tab) ↑
legislation.gov.uk (The National Archives) (2014). The Health and Social Care Act 2008 (Regulated Activities) Regulations 2014, regulation 20: Duty of candour (opens in a new tab) ↑