Safety Netting
Because "come back if it gets worse" is the clinical equivalent of "drive carefully" — technically true, completely useless.
Safety netting is one of the most important — and most badly done — consultation micro-skills in general practice. This page gives you a structured way to do it well every time, keeps you out of medico-legal trouble, and scores you marks in the SCA without sounding like a robot.
📥 Downloads
Handouts, summaries, and teaching extras — ready when you are. Includes CRUK materials, the RCGP safety netting guide, the literature review from BJGP 2019, and a teaching session for trainers.
path: SAFETY NETTING
- decision making and safety netting in acute presentations.docx
- diagnostic safety netting.docx
- methods of safety netting.pdf
- safety netting - 10 top tips by macmillan.pdf
- safety netting against cancer.pdf
- safety netting by cruk.docx
- safety netting esp against cancer.pptx
- safety netting for primary care - evidence from a literature review 2019.pdf
- safety netting for the patient the doctor and the practice - bruce eden oxford.doc
- safety netting guide by rcgp.pptx
- safety netting in healthcare settings - what it means and for whom.pdf
- safety netting in primary care.pdf
- safety netting table by cruk.pdf
- safety netting with text messages.jpg
- safety-netting.docx
- stages of safety netting.pdf
- teaching safety netting.docx
🌐 Web Resources
A hand-picked mix of official guidance and real-world GP training resources. Because sometimes the best pearls are not hiding in the official documents.
🏛 Core official guidance
- NICE NG12 — Safety netting in suspected cancer (updated 2023)
- CRUK — Safety netting in primary care (flowchart + principles)
- RCGP SCA feedback statements — including guidance on safety netting without over-doing it
- RCGP Primary Care Cancer Toolkit — Early Diagnosis & Referral
⚖️ Medico-legal & defence bodies
📚 Evidence & further reading
⏱ One-Minute Recall
If you only read one section, read this one. The panic-before-clinic / panic-before-SCA summary.
The whole topic in 10 bullets
- What it is: contingency planning for when the diagnosis is uncertain, or the illness could deteriorate.
- Why it matters: reduces missed diagnoses, protects the patient, protects you medico-legally.
- The core framework — EDF: Expect · Deviation · Follow-up.
- Be specific. "Come back if it gets worse" scores zero marks and helps no one.
- Name the symptoms. 3–5 alarm features max — any more and the patient forgets everything.
- Name the time frame. "2 weeks" not "2–3 weeks". "By next Friday" is even better.
- Name the route. GP? 111? 999? Walk-in? A&E? Don't make the patient guess.
- Safety-net investigations and referrals too — patients assume "no news is good news". They are wrong.
- Not every consultation needs it. Hair loss does not need a 999 warning. Use clinical judgement.
- Document it. "Safety netted" alone is not enough. Name the features you discussed.
🎯 Why This Matters in GP
Management of uncertainty is a constant worry for GP trainees. We can never be 100% certain that a set of symptoms is not the start of something serious. Safety netting is how we manage that uncertainty without burning the house down.
👤 For patients
Less chance a serious illness is missed. A patient told what to look out for can return early — and that early return can mean a longer, better life. Inadequate safety netting is repeatedly associated with delays in cancer diagnosis.
🩺 For doctors
Sleep better. Reduce the fear of "did I miss something?". Reduce medico-legal risk. A clearly documented, specific safety net is one of the strongest pieces of evidence in a complaint or claim.
🏥 For the system
Earlier detection = cheaper care. Serious and chronic conditions picked up early cost less — to the NHS and to the patient. Good safety netting is good stewardship.
💡 The statistic that should haunt you
Studies show that 40–80% of medical information provided by healthcare practitioners is forgotten immediately, and almost half of what is remembered is remembered incorrectly (Kessels, J R Soc Med 2003). That is why vague, wordy safety netting fails. Short. Specific. Memorable. Written down if needed.
📘 Core Knowledge
What safety netting actually is, where it came from, and when it absolutely must be done.
So what is safety netting?
Safety-netting is the information given to a patient or carer during a consultation about what to do if their condition fails to improve, changes, or if they have further concerns. In plain English — it's contingency planning. You share with the patient the plan for "what if I'm wrong, or what if this turns out to be something else?"
It is a teachable, practisable consultation micro-skill. And it should be taught, because most trainees do it badly at first.
📜 Where it all began — Roger Neighbour's three questions
Safety netting was formally introduced nearly 30 years ago by Roger Neighbour, who defined it as a process whereby the GP answers three questions:
- If I'm right, what do I expect to happen?
- How will I know if I am wrong?
- And what would I do then?
It appears in Neighbour's own consultation model and in the Calgary–Cambridge model under "closing the session". Every EDF, SAFER, or other modern mnemonic is essentially a more specific version of Neighbour's three questions.
When is safety netting essential?
There are three high-risk clinical situations where safety netting is non-negotiable. Miss these, and both you and the patient are exposed.
1. Diagnostic uncertainty
The diagnosis is uncertain and the differential includes serious illness — particularly illness that can progress rapidly (e.g. meningitis, sepsis, testicular torsion).
2. Known risk of complications
The diagnosis is certain, but carries a known risk of serious complications (e.g. asthma, cellulitis, UTI in pregnancy).
3. Higher-risk patient
The patient — because of age, frailty, or comorbidity — has an increased risk of serious illness or deterioration (e.g. older patient with a cough, immunosuppressed patient with a fever).
👥 Particularly important when consulting with
Children · the acutely unwell · patients with multimorbidity · patients with mental health problems · older patients living alone. But honestly — do it well with everyone.
Why uncertainty is often unavoidable — and why that's OK
Diagnostic uncertainty isn't a sign that you're doing GP badly. It's a feature of the job. A lot of the time, the patient is simply presenting very early in the illness — before the classic features have appeared — which is exactly why so many early presentations look like "medically unexplained symptoms" (Jones et al, BJGP 2019). You're not missing something; the illness just hasn't declared itself yet.
💡 Uncertainty as empowerment, not weakness
If the diagnosis genuinely is uncertain, say so. Communicating that uncertainty honestly isn't admitting defeat — it empowers the patient to re-consult if things change. It also reduces the risk of false reassurance, which is one of the biggest drivers of delayed diagnosis. Most patients appreciate the honesty; the ones who struggle with it usually struggle more when the "definite" diagnosis turns out to be wrong.
Can the patient actually safety-net themselves? Silverston's 5 C's
Before you hand over monitoring to the patient, check that they can actually do it. Paul Silverston's 5 C's is a quick sanity check. If any C fails, consider bringing them back for formal review rather than relying on them to self-monitor.
| The C | Ask yourself… |
|---|---|
| Capable | Is the patient capable of checking for the features that would trigger medical reassessment? |
| Competent | Are they mentally, physically, and medico-legally competent to monitor themselves (or the person in their care)? |
| Comply | Can they practically and logistically comply with the advice — do they have a phone, transport, literacy, childcare, etc.? |
| Comprehend | Can they comprehend the advice you are giving — language barriers, learning difficulties, distress? |
| Confirmed | Have you confirmed they have actually understood and appreciated the significance? Teach-back is the gold standard here. |
Research shows patients aren't satisfied with the way GPs do safety netting — because the advice is often too vague to be useful (Cabral et al, "They just say everything's a virus", Patient Educ Couns 2014). The whole point of EDF is to force you to be specific — giving the patient only the most useful, relevant information, in a form they will actually remember.
Expect
Tell the patient what you expect to happen — the natural trajectory of the working diagnosis. When will it get better? What might crop up on the way? What indicates things are on track?
Deviation
Tell the patient the specific symptoms or signs that indicate a deviation from the expected course — red flags, worrying features, or simply the illness not settling.
Follow-up
Give a specific follow-up plan — how to make contact, who to contact, and when to do it. No vague "come back if worried".
E — Expect: what you expect to happen
Tell them the natural history of the working diagnosis in plain language. Answer three things in the patient's head:
- What might develop next? (e.g. "you might get a bit of a cough in the next day or two — that's normal")
- When should it get better and resolve? (be specific — "within 7 days", not "soon")
- What would indicate things are not going to plan? (this bridges into D)
D — Deviation: the specific symptoms that should worry them
Now spell out the worrying features. This is where most trainees fall over — they either say nothing useful ("come back if it gets worse") or drown the patient in a list of 15 red flags.
⚠️ The 3–5 rule
Limit yourself to a maximum of 3–5 alarm features, plus the catch-all: "…or even if you simply feel more unwell". Patients can't remember more. If you genuinely need to mention more than 5, write them down for the patient to take home.
F — Follow-up: the precise plan
Specific advice on how to make contact, who to see, and when. Don't leave any of these for the patient to guess.
📞 How
"Make an appointment", "call us", "ring 111", "call 999" — tell them the exact route.
👤 Who
Me? Any GP? Emergency GP? Out of hours? A&E? Say it.
⏰ When
Pick ONE time frame. "2 weeks". "By next Friday". Not "2–3 weeks".
🧠 Memory trick — anchor the date to a day
"Today is Tuesday. So if things are not better by next Friday, please come back" is easier to remember than "come back in 10 days". Anchoring the deadline to a named day makes it stickier.
Examples of "how / who / when" phrasing
- "Please make an appointment for next Friday to see me again."
- "Call 111 if you…"
- "You must call 999 and ask for an ambulance if…"
- "If you haven't heard from the fast-track cancer team within two weeks, please ring reception urgently so they can chase it up."
- "Please see me in 2 weeks so I can double-check how your headaches are getting on."
- "However, if you notice any double vision or vomiting, please come and see the emergency doctor immediately."
- "If we're closed, call 111 and speak to the emergency doctor there."
- "If you start getting any fits, faints, or funny turns, call 999 and ask for an ambulance."
🚫 Avoid ambiguity
- Don't say "if it gets worse, come back" — how long should they wait? You are the doctor, that decision is yours.
- Don't say "2–3 weeks" — pick one. Ambiguity is a clinical decision you're outsourcing to the patient.
- Don't say "if any new symptoms develop" — name the symptoms.
EDF worked example — gastroenteritis
Spot the EDF: E — settle within 10 days, gradually improving · D — 3 specific features (diarrhoea > 10 days, new vomiting, PR bleeding) · F — graded action plan (GP within a week / emergency GP same day / A&E or 999).
🔤 SAFER — Another useful mnemonic
Paul Silverston's SAFER mnemonic helps you generate symptom-based, patient-safety-focused safety netting advice. Particularly useful when you're not sure which features to warn about. It pairs beautifully with EDF — SAFER helps you decide what to include in the D (Deviation), while EDF gives you the overall structure.
| Letter | Question to ask yourself |
|---|---|
| S — Serious | What serious causes and complications do I need to make the patient aware of? |
| A — Alternative | What alternative diagnoses do I need to make the patient aware of? |
| F — Findings | What specific findings would mandate medical reassessment? |
| E — Early | What early / atypical presentations of serious illness do I need to flag? |
| R — Red flag | What red flag symptoms and signs should I name explicitly? |
🎯 The 3–5 Rule — don't overwhelm the patient
One of the most important — and most ignored — rules in safety netting. It matters in real GP work and it matters in the SCA.
❌ What overwhelming looks like
"Come back if you get fever, rash, headache, neck stiffness, drowsiness, photophobia, confusion, weakness, vomiting, seizures, cold hands, cold feet, mottled skin, non-blanching rash, bulging fontanelle, stridor…"
Result: patient forgets all of it and remembers none of it. A list longer than 5 items is effectively zero items.
✅ What sticky looks like
"Three things should bring you back straight away — a rash that doesn't fade when you press a glass on it, they become floppy or drowsy, or even if you just feel they're not right."
Result: patient remembers. Every. Word.
💡 The rule of thumb
- 3–5 specific features — pick the highest-yield, most recognisable to a layperson.
- Always add "…or even if you simply feel more unwell / you feel they're not right".
- If more than 5 features are genuinely important, write them down or send a text.
- Use everyday language — "blood in your poo" lands better than "rectal bleeding".
- Prioritise features that are recognisable to the patient, not the most academically complete list.
Why specificity matters — the lingering cough example
Here's why vague advice like "come back if the cough gets worse" fails patients. Imagine someone with a viral upper respiratory tract infection. The same symptom — a cough — means something very different depending on when they're still coughing. The patient can't make this call. You can.
A cough at week 2 is reassuring; the exact same cough at week 4 is not. Without a specific time frame in your safety netting, the patient has no way of knowing which of these they're in.
🙅 When NOT to Safety-Net (or when to keep it very light)
This is the section the original page didn't have — and the SCA feedback statements specifically warn about. Over-safety-netting is a real problem. It wastes consultation time, it worries patients unnecessarily, and in the SCA it can actually cost marks.
📣 What the RCGP itself says
"Safety-netting may cause unnecessary alarm if performed without care." (RCGP SCA feedback statements). If a candidate has been told their safety netting is inadequate, the RCGP specifically warns against over-emphasising it in subsequent consultations — because it can then appear formulaic or take up too much time.
When safety netting is unnecessary or disproportionate
- Routine, benign, self-resolving presentations where there's no realistic risk of serious deterioration — e.g. mild androgenic alopecia, a resolving verruca, dry skin.
- Chronic stable conditions where the management plan is about long-term review, not acute deterioration — e.g. well-controlled hypertension with a 6-month review already booked.
- Administrative consultations — med3 / fit note continuation, sick note, referral letter queries.
- Screening or results consultations with reassuring findings — unless there's a specific reason to re-present.
- When you've already agreed a definite follow-up plan that itself is the safety net — don't then read out a catastrophic what-if list on top.
😅 The hair-loss example
A patient with male-pattern hair loss does not need: "If you suddenly notice you're losing 10,000 hairs an hour, call 999 immediately." They need: "This is gradual and expected. Come back if anything changes — like bald patches appearing suddenly, scalp soreness, or scarring." That's it. Proportionate safety netting is a clinical skill.
How to know if safety netting is needed at all
Brief, proportionate closure.
A single line is fine — or none at all.
Full EDF safety net.
Expect · Deviation · Follow-up.
🎯 The SCA implication
Examiners want to see clinical judgement, not a memorised script. Tailored, proportionate safety netting in a low-risk case scores better than a flood of red flags in a consultation that didn't need them. Judgement > completeness.
🚨 Red Flags — things to build into your safety nets
These are the generic "if you see this, escalate now" scenarios where safety netting can be the difference between a good outcome and a coroner's court. Tailor to the specific presentation — these are a prompt, not a script.
🚨 Immediate action (999 / A&E)
- Non-blanching rash in an unwell child/adult.
- Sudden severe "thunderclap" headache.
- New focal neurology or stroke-like features (FAST).
- Severe or worsening chest pain, especially with sweating, nausea, or breathlessness.
- Cauda equina features — saddle anaesthesia, urinary retention, bilateral leg weakness.
- Testicular torsion features in a male with acute scrotal pain.
- Massive PR bleed, haematemesis, or shock physiology.
- Anaphylaxis / airway compromise.
⚠️ Urgent GP / 111 same day or next day
- Symptoms worse than expected for the presumed diagnosis.
- Failure to improve within the time frame you gave.
- Atypical features creeping in (weight loss, night sweats, PR bleeding).
- Recurrent presentations with the same unexplained symptoms.
- Concerning features in a vulnerable patient (immunosuppression, pregnancy, frailty).
- Any safeguarding concern.
🧠 The "3 strikes and you're in" principle
If a patient presents 3 times with the same unexplained symptoms, that is itself a red flag. Your safety net on visit 1 and visit 2 should make it easy for them to come back. By visit 3, the question becomes: why haven't I investigated this?
🪤 Common Pitfalls — mistakes trainees make
Every one of these has been flagged repeatedly in SCA feedback, BJGP reviews, and real complaints. None of them are new. All of them are avoidable.
1. "Symptoms getting worse" — but which symptoms?
2. "If any new symptoms develop" — which new symptoms?
3. "If it doesn't get better, come back" — in what time frame?
4. Safety netting only around symptoms — forgetting investigations and referrals
5. Drowning the patient in 10 red flags instead of the key 3–5
6. Over-safety-netting low-risk presentations
7. Telling them to "come back" without saying where
8. Documentation that reads "safety netting advice given"
9. Not legitimising repeat visits
10. Assuming the patient has understood — without checking
📱 Electronic Safety Netting
A weakness of traditional safety netting is that it relies on the patient to re-appraise their symptoms and make a follow-up appointment. Worry about wasting the doctor's time and not wanting to bother the GP are well-documented barriers to help-seeking in the UK (Hirst et al, Europe PMC).
Combine that with the fact that 40–80% of medical information is forgotten immediately, and almost half of what is remembered is remembered incorrectly (Kessels 2003), plus the fact that miscommunication is commonplace in primary care consultations (Morgan 2013) — and the case for technology-supported safety netting becomes very strong.
Where an SMS can fit into the safety netting pathway for a patient presenting with potential cancer symptoms (adapted from Hirst et al).
SMS safety netting
Many clinical systems have SMS built in. Add-ons like AccuRx and MJOG make the process painless. Create a set of generic texts, and after a face-to-face safety net, schedule an automated text to send at a suitable interval (e.g. 2 weeks).
Example SMS templates
✅ Benefits
- Improves cancer screening uptake (evidence base exists).
- Helpful in monitoring acute and chronic conditions.
- Reduces cost of care and improves adherence.
- Gives the patient concrete advice to refer back to — solves the "I forgot what you said" problem.
⚠️ Limitations
- Timing lag — they get the text at 2 weeks, but booking can take another 2 weeks.
- Risk of widening digital inequalities.
- Can increase workload if not automated.
- Medico-legal issues around undelivered / unread texts.
📝 Documentation — because if it's not written, it didn't happen
Good safety netting that isn't documented is medico-legally worthless. A specific, dated, detailed entry is your best friend if a complaint or claim ever lands.
❌ Weak documentation
"Safety netting advice given."
This tells a reviewer, coroner, or defence body absolutely nothing.
✅ Solid documentation
"Safety netted — discussed alarm features (haemoptysis, weight loss, voice change). Advised review in 2 weeks if symptoms persist, earlier if any red flag features. Patient understood. Written info given."
📋 What to include in safety-netting documentation
- Which specific features you discussed — name them.
- The time frame given.
- The route advised — GP / 111 / 999 / A&E.
- Whether written information was given.
- Whether the patient understood (teach-back).
- Who else was involved (e.g. carer, interpreter).
💎 Insider Pearls — what trainees wish they'd known earlier
Distilled wisdom from real trainee experience, SCA feedback, and the patterns that come up again and again in mock consultations.
💡 Safety net while explaining, not after
Don't leave safety netting as a separate awkward block at the end. Thread it into the management explanation: "I think this is a viral infection, so I'm expecting it to settle over the next week — but the things that would change my mind are…". It sounds more natural and saves time.
💡 The "golden 90 seconds"
If you transition to management at minute 6, aim to start safety netting by minute 10. That leaves 90 seconds for safety net + teach-back + closing — and it feels unhurried.
💡 Stop performing — start planning
The single biggest tell of a candidate who will fail safety netting: it sounds rehearsed. The single biggest tell of a candidate who will pass: it sounds like a plan they're making with the patient. Slow down by 10%. Make eye contact. Mean it.
💡 Book the follow-up yourself in high-risk cases
If the case is genuinely concerning but doesn't yet meet referral thresholds, book the follow-up appointment yourself rather than leaving it to the patient. Don't let your practice's "appointment shortage" anxiety override patient safety. In a coroner's court, "I booked them back in myself" lands far better than "I told them to make another appointment".
💡 Honesty about uncertainty scores marks
Saying "I'm not completely certain what's causing this, so I want you to be my eyes and ears over the next week" is a strength, not a weakness. Examiners love hearing doctors communicate uncertainty honestly. Patients also prefer it — and it reduces false reassurance, which is one of the biggest sources of delayed diagnosis.
💡 Think about carers as well as patients
If the patient is a child, a frail older adult, or someone with cognitive impairment — safety net the carer, not the patient. And document who you spoke to. "Safety netted with mum — advised non-blanching rash, drowsiness, or not drinking as reasons to return immediately" is stronger than "safety netted".
🎯 What actually gets you marks (composite insight)
Trainees who excel in SCA safety netting tend to do three things: (1) they integrate safety netting into the explanation rather than bolting it on, (2) they name specific features using the patient's language, and (3) they check understanding with a brief teach-back. Do those three things, and the mark takes care of itself.
🌍 Wisdom from the Wider GP Training Community
Distilled insight from UK GP training forums, trainee accounts, and UK-focused medical educators — filtered so that only advice that aligns with official RCGP, NICE, and GP educator guidance is included. Nothing here conflicts with the rest of the page.
Where this wisdom comes from — and how it's been filtered
Filter rule: if any tip from Tier 2 or 3 contradicts Tier 1 guidance, the tip is dropped. Everything below has passed that check.
1. Safety netting is partly a defence against your own cognitive bias
UK GP educators have increasingly framed safety netting not just as a patient-protection tool, but as a protection against your own thinking errors as a clinician. Most missed diagnoses in primary care aren't from a lack of knowledge — they're from cognitive blind spots (anchoring, premature closure, availability bias).
🧠 The mindset shift
Safety netting is the safety rail between System 1 thinking (fast, pattern-matching, prone to bias) and reality. When your gut says "this is just a viral thing", a specific safety net is what catches it if your gut was wrong.
"We will never be free from bias or error. Our security lies not in perfection, but in recognising and learning from our imperfections." Growth mindset, applied to clinical reasoning.
2. Why this matters — the sources of diagnostic error
When UK researchers analyse diagnostic errors in primary care, the dominant cause isn't what trainees expect. Knowledge gaps are not the biggest enemy. Cognitive shortcuts are.
Schematic — proportions drawn from UK primary care diagnostic error literature (BJGP 2015 missed opportunities study; BJGP 2019 safety netting review). Actual percentages vary by study and setting.
🎯 The implication
A specific safety net protects against all four slices. The patient returns if your clinical reasoning was flawed (slice 1), if you didn't know something (slice 2), if the system dropped the result (slice 3), or if your explanation didn't land (slice 4). It's not just a closing ritual — it's a genuine risk-management tool.
3. Patients need explicit "permission" to come back
A consistent finding from UK GP educator content, trainee write-ups, and the 2022 BMJ realist review: patients frequently don't re-present even when their symptoms persist or worsen, because they feel they'd be wasting the doctor's time. This is particularly true of:
👥 Patients who especially need "permission"
- Older patients — "I don't want to bother you".
- Parents of young children — fear of being seen as over-protective.
- Men with non-specific symptoms.
- Patients from communities with traditionally deferential interactions with healthcare.
- People who've previously been made to feel dismissed.
✅ Phrases that give permission
- "Coming back is exactly what we're here for."
- "I'd much rather see you again and find nothing than have you sitting at home worried."
- "If things aren't right, please don't battle through — we'd want to know."
- "A second visit for the same thing isn't failing — it's how we pick up the ones that evolve."
4. Generic safety netting can feel dismissive to patients
A recent UK retrospective cohort study of out-of-hours consultations (BJGP 2024) examined nearly 1,900 records. Most had some safety-netting advice — but just over half was still generic only ("any problems come back"), and qualitative research has shown patients sometimes experience this as the clinician ending the conversation rather than genuinely sharing a plan.
Data pattern from UK OOH primary care cohort, 2013–2020 (BJGP 2024).
⚠️ The lesson
"Any problems come back" is better than nothing — but most GPs stop there. That's the gap. Moving from generic to specific safety netting is the single biggest quality-improvement opportunity in UK primary care. In the SCA, it's also the single biggest differentiator between a pass and a clear pass in this domain.
5. Practical tips from trainees who passed the SCA
Recurring patterns from trainee accounts of successful SCA performance — each cross-checked against RCGP SCA toolkit guidance before inclusion.
🔁 "Practise one new phrase at a time"
Trainees who improved fastest didn't try to overhaul their entire consultation style. They picked one new phrase per week and embedded it in real clinic before moving on. "Tell me more about that" one week. "What were you hoping I could do today?" the next. Slow, deliberate, sticky.
🎬 "Every real patient is an SCA practice"
The trainees who reported feeling calmest in the actual exam were the ones who had treated every real-life consultation as practice for months beforehand. Same structure. Same phrases. Same safety-netting habit. By exam day it was reflex, not performance.
🎧 "Record, watch, wince, fix"
The single most transformative preparation activity reported by successful trainees is the one most trainees avoid: watching recordings of their own consultations. Focus on just the closing 90 seconds. Is there an E? A D? An F? How many features did you name? Is there a specific time frame?
🗣 "Use softer phrasing"
Instead of "We should…", try "We could consider…". Instead of "I think you need…", try "How do you feel about…?". Subtle — but it repositions the patient from recipient to collaborator. Hugely powerful in SCA shared-decision-making marks.
⏱ "Don't let safety netting be your 11:55 panic"
Successful trainees move to management at minute 6, so safety netting naturally lands by minute 10. If you hit minute 11 and still haven't started safety netting, your time management has already failed — not your safety netting.
🎯 "Don't be too risk-averse"
Counter-intuitive trainee insight, endorsed by the RCGP's own SCA tips: over-investigating, over-referring, and over-safety-netting all lose marks. The SCA rewards confident, proportionate GP management — not defensive medicine. Make the plan. Own the uncertainty. Provide the safety net. Don't throw everything at every case.
6. The "generic vs specific" decision — a visual rule of thumb
A useful mental model for deciding how much depth your safety net needs, drawn from UK educator discussions and matching the principles in the RCGP toolkit.
Well-defined condition, low chance of deterioration, reassuring presentation.
1–2 specific features + generic catch-all. Brief is better than absent.
Uncertainty present, or symptoms evolving, or vulnerable patient.
3–5 specific features, time frame, clear route, teach-back, consider written info.
Serious differential possible, at-risk patient, or early-in-illness presentation.
Everything above + book the follow-up yourself, text reminder, carer involvement, document exhaustively.
7. Documentation — a shared frustration across the community
A theme that comes up repeatedly in UK GP educator content and trainee discussions: documentation of safety netting is the single biggest gap between what clinicians say they do and what their records show they did. UK researchers have developed a coding tool (SaNCoT) specifically to help clinicians audit their own safety-netting documentation.
| What many records show | What the community recommends instead |
|---|---|
| "S/N given" | "Discussed: red flag features (haemoptysis, weight loss, voice change). Review in 2 weeks. 111 if concerns before then. Written info given. Patient understanding confirmed." |
| "Safety-netted" | "Safety-netted — mum advised non-blanching rash, drowsiness, poor fluid intake as features that should prompt immediate 999 call. OOH otherwise. Confirmed understanding." |
| "Advised to return if no better" | "Expected to settle within 7 days. Review in 10 days if persisting. Earlier if [named features]. Patient confident with plan." |
8. Teach-back — the community's unsung hero
One of the most consistent pieces of advice across UK GP educators and trainee blogs: ask the patient to tell you, in their own words, what they'd do if things didn't go to plan. The RCGP SCA toolkit explicitly recommends this. And it's the single fastest way to catch misunderstandings before they become missed diagnoses.
🧠 Why teach-back works
- Catches misunderstanding in real time.
- Identifies language barriers you hadn't spotted.
- Activates memory consolidation — the patient remembers what they've said, better than what you've said.
- In the SCA, it visibly demonstrates to the examiner that you've checked understanding.
✅ Three phrasings that work
9. The "three strikes" pattern — a community teaching point
A theme mentioned in UK GP educator discussions and echoed in the BJGP "missed opportunities" study: three attendances with the same unexplained symptom is itself a red flag, regardless of how benign each individual presentation looks. One experienced GP described it as "three strikes and you're out" — meaning by the third visit, investigation is almost always warranted.
🎯 How this shapes your safety netting
Make it genuinely easy for the patient to return. If they're on their second visit, make that obvious in your notes so the next clinician sees the pattern. And if you are seeing them for the third visit with the same unexplained symptom — the safety net has arguably already been breached. It's time to investigate, not to safety net again.
✅ What was deliberately NOT included
In line with the filter rule, the following common trainee/forum/YouTube recommendations were rejected because they conflict with RCGP or educator guidance, or because they're outdated:
- ❌ "Safety net aggressively on every case to be safe" — contradicts RCGP SCA feedback on proportionate safety netting.
- ❌ "Always book follow-up whether needed or not" — contradicts RCGP guidance on realistic follow-up.
- ❌ "Use standard scripts verbatim" — penalised in SCA as formulaic.
- ❌ Pre-2020 CSA-specific advice — SCA format differs, much no longer applies.
- ❌ Defensive over-referral tips — flagged as a specific trainee error in RCGP feedback.
🧑🏫 For Trainers — teaching safety netting
Safety netting is a teachable consultation skill. Don't just correct it when it's missing — train it deliberately.
Common trainee blind spots
- Vagueness — "come back if worried" is the default failure mode. Call it out every time.
- Over-inclusion — listing too many red flags. Teach the 3–5 rule explicitly.
- Under-inclusion — particularly in cases trainees think are "clearly viral" or "clearly benign".
- Missing the route — telling the patient to come back but not saying where.
- Ignoring the admin pathway — not safety-netting referrals and investigations.
- Poor documentation — "safety netting advice given" and nothing else.
Tutorial ideas
- EDF drill: give the trainee 5 quick presentations and ask them to EDF each in 30 seconds. Child with fever · chest infection · headache · minor ankle sprain · fit note renewal. Note which ones don't need safety netting.
- Video review: audio-COT a real consultation. Transcribe only the safety netting. Critique against EDF.
- Role-play over-safety-netting: deliberately ham up a 12-red-flag safety net and ask the trainee to critique.
- Documentation audit: pull 5 of the trainee's recent notes — rate the safety netting documentation out of 5 on: features named, time frame, route, understanding checked.
- Complaints / SEAs: any SEA involving delayed diagnosis often has a safety-netting angle — unpack it together.
Reflective prompts for tutorials
- "Tell me about a patient you safety-netted well this week. What made it work?"
- "Tell me about a consultation where you didn't safety net but wish you had."
- "What do you do differently for a worried parent compared to an anxious young adult?"
- "How do you decide when not to safety net at all?"
- "When did you last safety net an investigation or referral pathway?"
How to assess learning on this topic
- COT / audio-COT: assess the close of the consultation specifically. Did they EDF?
- CbD: pick a case with diagnostic uncertainty and explore how the trainee safety-netted.
- Random case analysis: pull 10 consecutive notes and look at the safety-netting documentation.
- Simulated consultations: build safety netting into mock SCA scoring.
🧠 Memory Aids & Cheat Sheets
🟢 EDF — the master framework
Expect (what will happen) · Deviation (what shouldn't) · Follow-up (what then)
🔵 SAFER — for generating the content
Serious · Alternative · Findings · Early · Red flag
🟣 5 C's — before handing over to the patient
Capable · Competent · Comply · Comprehend · Confirmed
🟠 The 3–5 rule
Max 3–5 specific features in any safety net. More than that → write it down or text it.
🎯 The "perfect safety net" checklist
- Named the expected course with a time frame.
- Named 3–5 specific features to look out for.
- Added the catch-all "…or even if you just feel more unwell".
- Specified who to contact (GP / OOH / 111 / 999).
- Specified when to contact (single time frame or day).
- Safety-netted any investigations or referrals too.
- Legitimised coming back.
- Did a quick teach-back.
- Documented features, time frame, route, understanding.
🎯 SCA High-Yield Tips
Safety netting is a core marking component under Clinical Management in the SCA. Miss it in a station where it mattered, and you can lose marks in that domain. Overdo it where it didn't belong, and you can lose marks too. This is a judgement skill, not a checklist.
🎯 What examiners specifically want to see
- Tailored safety netting — matched to this patient, this diagnosis, this context.
- Specific features named, in the patient's language.
- Time-bound — a clear, single time frame.
- Graded — the patient knows the difference between "ring the GP next week" and "call 999 now".
- Proportionate — not bolted onto every consultation regardless of risk.
- Integrated with the rest of the plan — not a rushed afterthought at 11 minutes 55.
Using EDF in the SCA
EDF works beautifully in the SCA because it forces specificity in under 30 seconds. Here's how it maps onto the closing phase of the consultation:
| EDF | In the SCA consultation | Example phrase |
|---|---|---|
| E — Expect | Set the patient's expectation for the natural course. | "I'd expect this to settle over the next 7 to 10 days." |
| D — Deviation | Name 3–5 specific features that should prompt review. | "But if you notice X, Y, Z — or even if you just feel more unwell…" |
| F — Follow-up | Give a precise time frame, route, and person. | "…book a GP appointment within a few days. If urgent, call 111. If severe, 999." |
⚠️ Common trainee mistakes in SCA safety netting
- Running out of time and squeezing safety netting into 15 rushed seconds at the end.
- Listing too many red flags — sounds like a textbook, feels robotic.
- Generic "come back if worried" with no specifics.
- Not giving a time frame at all.
- Not giving a route ("come back" — where? to whom?).
- Safety-netting a case that didn't need it — e.g. a fit note request.
- Parroting the same script for every station — examiners can see it.
💡 Quick wins for extra marks
- Use the patient's own words for what they'd notice ("if your hand stops working properly").
- Link safety netting back to their ICE — "I know you were worried it might be…".
- Offer written info or say you'll send a text — shows awareness of forgetting.
- Use a teach-back — "What would bring you back to us?".
- Make the follow-up appointment yourself in high-risk cases, rather than relying on the patient.
- Explicitly legitimise returning — "Please don't feel you'd be wasting anyone's time".
🎯 SCA Consultation Pearls
The single best piece of advice for SCA safety netting is this: stop thinking of it as a closing ritual, and start thinking of it as a safety plan. A plan has specificity, time-boxing, and named actions. A ritual is formulaic and transparent — and the examiners have seen it a thousand times. If it sounds like a plan the patient could follow tomorrow, you're doing it right.
🚩 Red flags you must mention (when relevant)
Don't assume the examiner infers what you're thinking — verbalise it. Even if the consultation hasn't required you to act on red flags, saying: "The things I want you to watch out for are…" and naming them clearly is what scores the mark. Silent clinical reasoning is unmarked clinical reasoning.
Pick the 3–5 highest-yield features for this presentation — not every red flag in the textbook. Tailoring beats completeness.
🗣 Useful Consultation Phrases — the EDF in action
Natural, usable phrases you can adapt tomorrow. Not a script — a toolkit. Read them once, try one in your next clinic, and adjust to your own voice.
💡 Sticky phrasing tips
- Use "or even if you simply feel more unwell" as your catch-all. Short, memorable, covers the atypical.
- Start the safety net with "There are three things I want you to look out for…" — numbering primes the patient's memory.
- Anchor the time frame to a named day, not a number of days.
- Use the patient's own words for features — "blood in your poo" beats "per rectum bleeding".
- Finish the consultation with legitimising, then teach-back, then a clear "thank you" — it feels complete.
❓ FAQ
Do I need to safety-net every consultation?
How many red flags should I mention?
What if I'm genuinely not sure of the diagnosis?
Is it enough to say "come back if you're worried"?
Should I safety-net referrals and investigations?
What should I actually write in the notes?
What if the patient can't self-monitor?
What about using text messages?
IMG question — what's the difference between 111, 999, OOH, and A&E?
- GP / surgery: routine and same-day appointments during surgery hours.
- Out of hours (OOH): GP service outside normal surgery hours — usually accessed via 111.
- 111: non-emergency NHS telephone triage service. Used when you need help fast but it's not life-threatening.
- 999: life-threatening emergencies only — chest pain, stroke, severe bleeding, anaphylaxis.
- A&E: accident and emergency department — for serious injuries and severe acute illness.
🏁 Final Take-Home Points
The bits to remember tomorrow — in clinic and in the SCA.
- Safety netting is a plan, not a ritual. A plan has specificity, time frames, and named actions.
- Use EDF: Expect · Deviation · Follow-up. Simple, specific, repeatable.
- Name 3–5 features max. More than that and you've lost the patient.
- Always add "…or even if you simply feel more unwell" as your catch-all.
- Pick one time frame. "2 weeks" — not "2–3 weeks". Anchor to a day if you can.
- Name the route. GP? 111? 999? Don't let the patient guess.
- Not every consultation needs it. Hair loss doesn't need a 999 warning. Judgement matters.
- Safety-net admin too. Referrals, investigations, results — "no news" is not reassurance.
- Legitimise returning — "that's what we're here for".
- Teach-back — "what would make you ring us?" is the 5-second mark-scorer.
- Document specifically. Features named, time frame, route, understanding.
- In the SCA: tailor, don't recite. Integrate, don't bolt on. Plan, don't perform.
ABSOLUTELY BRILLIANT! Planning a teaching on safety netting soon. Hope it is OK to use this resource ( with acknowledgement of course).
Of course, you can use it. Hope the session goes well. Remember to make it interactive with some practice… otherwise, the learners don’t assimilate it. xxx Ram