De-escalation
Practice urgent de-escalation scenarios with real-time tension tracking: regulate yourself, read risk, communicate clearly, coordinate help, and keep everyone safe.
What you practise
- Self-Regulation
- Risk Recognition and Safety Positioning
- Reading Agitation Type
- Empathic Reassurance Without Collusion
- Clear and Brief Communication
- Validate and Limit-Set
Who you can meet
- Marcus, 34 — 34-year-old man, agitated and physically restrained by two ED staff. Brought by police after a disturbance at a homeless shelter.
- Nurse Andersen, 52 — Experienced ED charge nurse. Sharp, competent, slightly dry. Has seen it all. Can provide background on Marcus and coordinate team actions.
- Jonas — Orderly. Mid-40s, strong, calm. Holding Marcus's left arm and shoulder.
- Erik — Orderly. Early-30s, anxious, newer. Holding Marcus's right arm.
What the feedback looks at
- Self-regulation under pressure
- Risk recognition and safety positioning
- Reading what drives the agitation
- Empathic reassurance without collusion
- Clear and brief communication
- Validation paired with limits
- Team and environmental safety
- Escalation threshold recognition
Rated against: These dimensions are house-authored for PsychBase and adapted from published verbal de-escalation practice, chiefly the ten-domain consensus in Richmond et al. 2012, Verbal De-escalation of the Agitated Patient, Project BETA, West J Emerg Med, together with the Safewards model of conflict and containment. The wording is adapted and paraphrased, never transcribed from those sources, and the numbered anchors are ours. A published rating instrument for de-escalation performance does exist, the De-escalating Aggressive Behaviour Scale, DABS, and its English version EMDABS; adopting or mapping onto it is a candidate for a future revision of this rubric, and it is not used here. Scores are produced by a language model and are not equivalent to ratings by a trained human rater. No endorsement by any author, publisher, or instrument owner is implied.
This lab is for clinical training. The patient is generated by an AI language model from a written case; no real patient and no patient data are involved. Browsing is open; starting a session needs a free account.