Lead Through the Storm — A 48-Hour Crisis at Severnvale NHS Trust

A four-round, advanced crisis-leadership simulation set inside Severnvale University Hospitals NHS Foundation Trust, the designated major-trauma and emergency-surge centre for a Midlands region of 620,000 people (two sites, 840 beds, GBP 620M budget, 8,200 staff). At 18:40 on a Friday night a chemical-release incident triggers mass self-presentation while — within the same hour — a ransomware attack disables the electronic patient record and bed-management systems, forcing the whole trust onto paper. ED attendances surge toward 560 in 24 hours against a normal 350; the trust starts at 94% occupancy with only ~50 free beds. Regional command reviews the trust at hour 48; if flow is not stabilised and command is judged incoherent, external incident management is imposed. Playing the Trust Chief Executive, you steer the response across a scarce pool — 6 surge theatres/resus bays, 40 redeployable staff, 12 critical-care beds and a GBP 1.2M contingency — over four rounds drawn from Colley & Spyridonidis' work on collective leadership in turbulent times: (1) make sense under information loss, separating what you know from what you assume and naming which decisions must be pushed to those who can see the ground; (2) allocate the finite resource pool by explicit prioritisation rule while protecting a deliberate reserve for the night-two collapse; (3) design the decision-rights model — choosing centralized command, distributed/mission-command leadership, or a calibrated mix, anchored by a one-line commander's intent — against the reality that the blind centre cannot micro-control; and (4) frame the message to a frightened, exhausted frontline so it switches the distributed model on, then defend containment and stabilisation to regional command at hour 48. The math rewards the concept's correct plays — distributed authority with clear intent, rule-based triage with a protected reserve, an empowering honest message, and pacing the response — and measurably punishes the five classic errors: the central-control reflex during the blackout, distribution without intent, reactive loudest-voice allocation, treating the message as PR, and front-loading the entire contingency in the first six hours. Final KPIs track Flow Stability %, Frontline Trust, Command Coordination, plus contingency and redeployable-staff spent against hard caps.