At this year’s APEM-PERUKI conference (2026), we presented a live, interactive, simulated PEM Adventure, with a bit of a difference. The audience helped the sim team decide how to manage Sophie, a 15-year-old girl brought into the Emergency Department by paramedics at 11 pm with ligature marks to her neck.
Here’s the case, with the evidence base for her care.
Scene 1: Arrival and assessment
Sophie had recently been diagnosed with autism and ADHD, and had been struggling since the end of Year 10, socially isolated, low in mood, and no longer seeing the school counsellor who’d been supporting her.
Over the summer, she’d been groomed into county lines drug running. Today’s trip, her furthest yet, went wrong, and she was sexually assaulted and non-fatally strangled.
Too frightened to disclose any of that, she initially told the team she’d simply lost her wallet and phone while visiting friends, which meant, entirely reasonably on the information available, that the team began by treating this as a likely case of serious self-harm.
As Sophie’s distress grew and she tried to leave the department, the team reached their first crossroads.
Vote 1: Should Sophie be stopped from leaving — and under what legal framework?
- Yes, prevent her from leaving (with physical restraint if needed) under common law
- No, do nothing, as she is Gillick competent and can decide to leave when she wishes; there is no legal basis for keeping her in ED at this point
- Yes, use doctors’ holding powers under the Mental Health Act (Section 5(2))
A 15-year-old with unexplained ligature marks, months of low mood, a history of self-harm, is highly agitated and trying to leave. Normally, if Sophie was Gillick competent, her decision should be respected. But she was acutely distressed and potentially at risk to herself, and that made assessing her competence in the moment very challenging.
Being fifteen doesn’t automatically mean the team could keep her in hospital, and being mentally unwell doesn’t automatically give anyone power to detain her in the Emergency Department. Section 136 is a police power, and doesn’t apply once a patient is already with the medical team. Other Mental Health Act holding powers, including Section 5(2), the doctor’s holding power, only apply to a patient who is already an inpatient, which Sophie wasn’t.
Parental consent also cannot be used as the sole legal framework for restraining her either. At this age, a decision like this would be outside the Zone of Parental Responsibility, and to appropriately respect her dignity and autonomy it would need resolution by the course.
If clinicians genuinely believe that letting a patient leave would expose them to an immediate and serious risk of death or serious self-harm, common law gives a basis for temporary intervention, but it’s temporary intervention to prevent serious immediate harm, not a general holding power. In practice, that means minimum force, minimum time, and a clear exit plan, and being able to explain exactly why leaving right now creates an immediate, serious risk to life, and working towards the next decision rather than treating restraint as the plan itself.
It’s also worth discussing Sophie’s neurodivergence. Some of what looked like “non-compliance” may have been communication: Sophie’s autistic, has ADHD, was frightened, and was in a noisy Emergency Department at midnight, being asked difficult questions by strangers. Reasonable adjustments, such as less noise, fewer people, more space, food or a drink, a trusted face, and predictable information about what happens next, should start from the very beginning of an assessment, not only once things have escalated. Keeping someone safe doesn’t have to mean taking control away from them immediately.
The bottom line: if leaving would immediately and seriously endanger Sophie’s life, common law may allow temporary intervention, but with minimum force, minimum time, maximum dignity, and constant reassessment.
Scene 2: Restraint and rapid tranquilisation
The team decided Sophie was likely experiencing a mental health crisis and that restraint was justified to keep her safe.
Security stepped in briefly, using minimum force, while the team dimmed the lights, reduced the number of people in the room, and tried to work out what would actually help her settle.
Sophie’s distress didn’t resolve, and the team reached their next decision point.
Vote 2. Restraint & Rapid tranquilisation – what would you give?
- Lorazepam
- Promethazine
- Olanzapine
- Ketamine
- Droperidol
- Nothing – continue physical restraint
There isn’t a strong paediatric evidence base for any of these drugs, and our practice is extrapolated heavily from adult practice. A few years ago, the Project BETA group in Western America published a Delphi consensus study, with experts in child and adolescent psychiatry to provide an answer to the question: What is the best first-line drug in undifferentiated agitation? Oral lorazepam was the preferred first-line in most scenarios. It’s relatively safe with predictable side effects and has a much shorter half-life than diazepam. But there’s an important exception: benzodiazepines are not suitable for neurodivergent young people with a previous paradoxical reaction.
If there’s a known or suspected paradoxical benzodiazepine reaction, a sedating antihistamine or olanzapine might be preferable depending on the child and local protocol. Bring in learning disability or neurodevelopmental expertise wherever available.
The PEAChY-O trial, in Australasia, recently compared oral olanzapine with oral diazepam as first-line treatment for acute behavioural disturbance in 9-17 year olds. They found both were safe, with no difference in efficacy of sedation between the two drugs. However, they did find that olanzapine wafers were much easier to take than diazepam tablets. But, in 40% of cases, these medications did not achieve behavioural containment, and there was aggression towards staff in 15%.
Ketamine and droperidol are both used increasingly in EDs for patients with severe agitation. Ketamine is worth considering when the level of agitation means the restraint itself has become unsafe for the patient or staff. It’s extremely rapid, but it needs full resuscitation monitoring, carries significant potential complications, and by that point it isn’t really “rapid tranquillisation” any more; it’s emergency sedation, which can go badly wrong if it isn’t done properly. Droperidol is rapid and heavily sedating too, with theoretical QTc concerns, so not ideal either.
Whatever drug is used, it’s worth remembering what the intervention feels like from Sophie’s side: an unconsented IM injection is frightening. If restraint is needed, keep it as brief as possible, keep talking to your patient throughout, explain what’s happening, and debrief with the team afterwards. And document it properly: why restraint was needed, why medication was needed, what alternatives were tried, and what happens next.
The bottom line: the evidence base for rapid tranquilisation in children is limited. Follow your local protocol, consider alternatives for neurodivergent children and young people, and document clearly the rationale and debrief afterwards.
Scene 3: Disclosure
In the end, Sophie exhausted herself before any medication was given. She broke down sobbing, and security was stood down.
As the team built trust with Sophie, she disclosed what had actually happened to her that day: county lines exploitation, the sexual assault, and the strangulation.
Referrals for child criminal exploitation involving girls have been rising sharply, and neurodivergent children and young people, like Sophie, are recognised as being at particularly high risk. They can be less able to spot that they’re being groomed, and less able to tell someone once they have been. Sophie’s story might feel extreme, but sadly it isn’t as rare as we’d like to think.
After that disclosure, the team focused on what to do about the ligature marks on Sophie’s neck, particularly whether to organise imaging.
Vote 3. What determines the need for imaging in non-fatal strangulation?
- Suspicion of cervical spine injury
- Stridor/signs of upper airway obstruction
- Dyspnoea / surgical emphysema
- Facial or conjunctival petechiae
- Loss or near-loss of consciousness
- All of the above
The new Institute for Addressing Strangulation (IFAS) guidance for children and young people was published in July 2026. Under the adult guideline, too many children were ending up with CT angiograms of their neck for low-yield reasons.
In the new guidance, the key triggers for imaging are: clinical suspicion of a cervical spine injury, stridor or airway obstruction, and dyspnoea. Petechiae and a brief loss of consciousness are explicitly de-emphasised as automatic triggers in the paediatric guidance. Instead, other neurological signs should prompt consideration of a CT head: a GCS below 13, new neurological signs, or a seizure after the event. Airway concerns may need ENT review rather than imaging. Observation to ensure significant swelling does not develop over time is also called out.
The IFAS guidance emphasises the importance of considering safeguarding – sexual assault, domestic abuse, and exploitation.
The bottom line: use the IFAS guidance for children and young people when considering whether a patient with non-fatal strangulation needs imaging, and explore safeguarding concerns with your patient.
Scene 4: The plan
The team talked Sophie through what needed to happen next: a police referral, a SARC (Sexual Assault Referral Centre) referral for emergency contraception and further support, a safeguarding referral, CAMHS follow-up for her mood and self-harm, and a specialist youth worker.
Before any of that could happen, though, there was one more decision to make.
Vote 4. Where is the best place for Sophie to spend the night?
- Emergency Department
- Paediatric Short Stay / Assessment Unit
- General paediatric ward
- If CAMHS clear her from a mental health perspective, could she go home?
We’re seeing more and more adolescents spending a disproportionate amount of time in the Emergency Department, and it’s easy to think it’s unfair, not what we trained for, or not emergency medicine. Many of us recognise that feeling. The system has repeatedly placed young people in mental health crisis into Emergency Departments, sometimes for hours or days.
But it’s worth pushing back slightly on “this isn’t emergency medicine.” Capacity, legal authority, medication, safeguarding, and risk assessment are absolutely Emergency Medicine decisions. Practically, that means a named clinician: someone who has actually thought through what happens if Sophie deteriorates, what happens if she tries to leave again, and when she’ll next be reviewed.
The same thinking applies to the restraint earlier in the case. Sophie hit out at staff, and that’s genuinely frightening, but a frightened 15-year-old isn’t simply a smaller violent adult. Her development, trauma, and neurodivergence all matter, and her behaviour is how she’s communicating.
At the same time, there’s an opposite failure mode worth guarding against: everyone trying so hard to be kind that nobody escalates early enough, until suddenly four or five people are restraining someone all at once, which isn’t trauma-informed either.
Again, reasonable adjustments can help: fewer people, less noise, food, a trusted face, and predictability. Sometimes time is the treatment, but there needs to be a clear plan to reassess after a period of time. Security should be briefed before a crisis develops; restraint should never be the first moment the system becomes organised. And it carries extra weight in Sophie’s case specifically, because she had just disclosed a sexual assault.
Staff safety matters too. Trusts have legal duties as employers to manage foreseeable workplace violence; a serious assault, or an injury causing more than seven days off work, can become a RIDDOR-reportable incident. Trauma-informed care isn’t consequence-free care: boundaries can still be set, staff who’ve been hurt can be acknowledged, and Sophie can be helped to understand the impact of her behaviour, without shaming her for it.
Which brings us back to the vote: where should Sophie spend the night? It’s worth asking a more fundamental question – why does she need to stay at all? Is it because the team needs to observe her, because a safeguarding process needs to happen, or because everyone feels too uncomfortable to let her go?
If Sophie is being kept in overnight, the team should be able to say clearly what they’re observing for, when she’ll be reassessed, who owns that decision, and what happens if she wants to leave. The overnight “treatment” may be regulation, food, sleep, reduced stimulation, a trusted adult, predictability, then reassessment. That’s a therapeutic plan, as long as it’s actually named as one rather than left implicit.
Sophie’s own preference should be asked for directly: it doesn’t decide everything, but after a night in which so much control has already been taken away from her, handing some of it back can help. And because the facts of the case have changed, the risk assessment has changed too: what actually happens if Sophie wants to leave?
There’s a bigger issue sitting underneath all of this. Parity of esteem between mental and physical health has been talked about for well over a decade. Suicide remains the leading cause of death in adolescents. So it’s worth asking honestly whether it actually feels like parity at 3 am. If Sophie had a physical problem carrying equivalent risk, the team would expect a clinical plan, defined escalation, and someone named as responsible for her care. Parity means holding mental health presentations to the same standard of ownership and escalation as physical illness.
Scene 5: Coda
Sophie’s parents arrived just after 1am. With her permission, the team helped her explain what had happened. The CAMHS liaison team and SARC saw her the next morning.
This led to our final vote.
Vote 5. How does this case make you feel?
Cases like Sophie’s are hard. So often, we do everything right, and the outcome still isn’t great.
The gap between the care we want to give and the care we can actually give is moral injury. Sometimes we hear tough words to describe cases like Sophie’s (“heart sink,” “behavioural,” “another mental health presentation”). That language is stigma, and it changes the care a young person actually gets, from how fast we go in to see them to whether we believe them the first time.
And so we’ll leave this blog with the words we left the audience with at the APEM-PERUKI conference.
Look after your patients.
Look after each other.
And, please, look after yourselves too.
References
Croak B, Lamb D, Bhundia R, Rafferty AM, Greenberg N, Stevelink SA. Moral injury in healthcare workers: causes & interventions. British medical bulletin. 2026 Mar;157(1):ldag011.
deSouza MI, Thode Jr PH, Shrestha MP, Allen MR, Singer MA. Rapid tranquilization of the agitated patient in the emergency department: A systematic review and network meta-analysis. The American Journal of Emergency Medicine. 2022 Jan 1;51:363-73.
Gerson R, Malas N, Feuer V, Silver GH, Prasad R, Mroczkowski MM. Best Practices for Evaluation and Treatment of Agitated Children and Adolescents (BETA) in the emergency department: consensus statement of the American Association for Emergency Psychiatry. Focus. 2023 Jan;21(1):80-8.
Greenwald A, Kelly A, Mathew T, Thomas L. Trauma-informed care in the emergency department: concepts and recommendations for integrating practices into emergency medicine. Medical Education Online. 2023 Dec 31;28(1):2178366.
Hoffmann JA, Pergjika A, Konicek CE, Reynolds SL. Pharmacologic management of acute agitation in youth in the emergency department. Pediatric emergency care. 2021 Aug 1;37(8):417.
Institute for Addressing Strangulation (IFAS). Guidelines for clinical management of non-fatal strangulation in acute and emergency care services (Feb 2024, adult/adolescent) and paediatric (under-16) draft guidance (announced 01.07.2026). ifas.org.uk
Nizza IE, Bryan G, Phillips K, et al. “I don’t think they were used to mental health cases”: Young people, carers and staff experiences of mental health admissions to general paediatric wards. 2026.
Perkins J, Ho JD, Vilke GM, DeMers G. American Academy of Emergency Medicine position statement: safety of droperidol use in the emergency department. The Journal of emergency medicine. 2015 Jul 1;49(1):91-7.
Siafis S, Philipona F, Nomura N, Milosavljević F, Chalkou K, Schneider-Thoma J, Bighelli I, Lorenz C, Dib JE, Tharyan P, Chan EW. Comparative effectiveness and safety of pharmacological treatments for rapid tranquilisation in emergency settings: a systematic review and individual participant data network meta-analysis. The Lancet Psychiatry. 2026 Jun 8.
Ward JL, Vázquez-Vázquez A, Phillips K, Settle K, Pilvar H, Cornaglia F, Gibson F, Nicholls D, Roland D, Mathews G, Roberts H. Admission to acute medical wards for mental health concerns among children and young people in England from 2012 to 2022: a cohort study. The Lancet Child & Adolescent Health. 2025 Feb 1;9(2):112-20.
National Institute for Health and Care Excellence (NICE). Challenging behaviour and learning disabilities: prevention and interventions for people with learning disabilities whose behaviour challenges (NG11), 2015. [
Royal College of Emergency Medicine. Acute Behavioural Disturbance in Emergency Departments (2023) and A Brief Guide to Section 136 for Emergency Departments (2025).
Railway Children / Catch22. Neurodivergence and vulnerability to criminal exploitation in children and young people.
RCEM Best Practice Guideline – Acute Behavioural Disturbance in Emergency Departments Version 2 (2023)
