The code blue pager goes off. Within minutes, a team assembles. Everyone roughly knows how this works. The ICU registrar leads. Anaesthetics manages the airway. The algorithm runs.
Roughly.
No one has confirmed these roles for this arrest, with this team, on this shift. You don't know whether the anaesthetic registrar is a senior trainee who has intubated hundreds of times, or a first-year who has never managed an airway on the floor. You don't know whether the medical registrar is running drugs or chasing a history. You don't know whether the ICU consultant walking in expects to take over, or to stand back and let the registrar lead.
None of this has been discussed. Two people reach for the defibrillator pads. No one is timing the adrenaline.
There are conventions. There is not a plan.
And then we ask this group of strangers to perform one of the most time-critical, coordination-dependent tasks in medicine.
The only high-stakes industry that does this
A flight crew doesn't meet for the first time at the threshold of the runway. A fire crew doesn't arrive at the scene and start deciding who drives the truck.
Every other high-stakes industry that relies on team coordination under pressure has a basic expectation: the team knows who it is before the crisis begins.
In resuscitation, we've normalised the opposite. The standard model in most hospitals is that the cardiac arrest team is assembled ad hoc from whoever is on shift, converges at the bedside having had no prior contact, and begins advanced life support while simultaneously figuring out team structure, leadership, and communication norms.
The clinical knowledge is there. The algorithm is established. The equipment is standardised. But the team, the people who have to coordinate all of it under pressure, is improvised from scratch, every single time.
The cost is measurable
In a prospective randomised simulation trial, Hunziker and colleagues compared pre-formed resuscitation teams against ad hoc teams.1 The results weren't subtle.
Pre-formed teams delivered 31 more seconds of hands-on CPR in the first three minutes. They defibrillated 40 seconds faster. They made 40% more leadership statements.
Not because they knew more. Not because they were more experienced. Because they had briefly met beforehand and agreed on who would do what.
When Nallamothu and colleagues studied hospitals with the best in-hospital cardiac arrest outcomes, they found consistent themes across 158 interviews at nine sites.2 Top performers emphasised knowing each other's names, roles, and experience levels before the emergency. The teams that performed best weren't the ones with the most expertise. They were the ones that had structured the basics before the crisis began.
The fix is embarrassingly simple
A brief huddle at the start of each shift. The on-call resuscitation team meets, confirms who is on the team, agrees on roles, and establishes a shared mental model for how they'll work together if called.
The specifics matter less than the fact that it happens. Who leads. Who manages the airway. Who runs drugs and keeps time. Whether the consultant expects to take over or mentor from the side. A few minutes of role clarity that would otherwise consume the first chaotic minutes at the bedside, minutes during which a patient is in cardiac arrest.
The principle is well-established: teams that share a mental model before the crisis outperform teams that have to build one during it.3 Pre-shift briefings are standard in aviation, in operating theatres, in military operations. What's remarkable is that resuscitation, arguably the most time-critical team task in hospital medicine, is one of the last domains to adopt it.
A system problem, not a people problem
This isn't a criticism of the clinicians running arrests. Experienced doctors and nurses are remarkably good at adapting to unfamiliar teams on the fly. But "good at adapting to a bad system" is not the same thing as "the system is well-designed."
The system is designed as though coordination is a natural byproduct of co-location — that putting the right people in the same room is enough. It isn't.
The registrar on their first night covering MET calls doesn't have ten years of adaptive expertise to fall back on. They need the system to be designed so that coordination is built in, not improvised.
References
- Hunziker S, Tschan F, Semmer NK, et al. Hands-on time during cardiopulmonary resuscitation is affected by the process of team-building: a prospective randomised simulator-based trial. BMC Emergency Medicine 2009; 9: 3.
- Nallamothu BK, Guetterman TC, Harrod M, et al. How do resuscitation teams at top-performing hospitals for in-hospital cardiac arrest succeed? A qualitative study. Circulation 2018; 138(2): 154–163.
- Pimentel CB, Snow AL, Carnes SL, et al. Huddles and their effectiveness at the frontlines of clinical care: a scoping review. Journal of General Internal Medicine 2021; advance online publication.