After more than a decade of running ACLS simulations, one pattern shows up in almost every code: nurses who know the algorithms cold still freeze at the bedside. Not because they don't know what to do — but because knowing and doing under stress are two completely different skills.
Here are the three most common failure points we see, and exactly how to fix them.
"Knowing the algorithm is necessary but not sufficient. The bedside is not a written exam."
1. Rhythm interpretation delay
The most common mistake isn't misidentifying a rhythm — it's taking too long to call it. In a real code, every minute of delay before defibrillation meaningfully reduces the chance of survival. The fix isn't memorizing more strips. It's drilling the binary decision: shockable or not shockable? Everything else is secondary.
Notice what the binary does. It removes the hardest cognitive step — precise naming — from the most time-critical moment. You do not need to distinguish coarse VF from polymorphic VT to act correctly. You need to know that both get electricity. Refine the diagnosis afterwards, during compressions, when the cost of thinking is zero.
Providers hesitate because they are trying to be right when they should be trying to be fast and safe. A shock delivered to organised VT that turned out to be something else is a recoverable error. A three-minute pause to be certain is not.
2. No one is actually running the code
The second failure is structural rather than clinical. Everyone in the room knows the algorithm. Nobody has said out loud who is leading, so five competent people make six overlapping decisions and no one owns the timeline.
The symptoms are easy to spot from outside and almost invisible from inside: two people reaching for the same syringe, nobody watching the clock, drugs given without anyone announcing them, and a rhythm check that happens whenever someone happens to notice it has been a while.
- Name yourself "I'm running this code." It sounds theatrical the first time and it ends the ambiguity instantly. If someone more senior arrives, hand over explicitly — "you have it" — never by drifting.
- Assign by name, not by gesture "Sarah, compressions. Mike, airway. Priya, drugs. James, timer and documentation." Pointing produces two people doing one job and one job undone.
- Put someone on the clock explicitly Two-minute cycles, epinephrine intervals, total downtime. Without an owner, time in a code compresses subjectively — everyone underestimates how long they have been going.
- Demand closed-loop communication "Epi one milligram IV" — "Epi one milligram IV, in." An order without a confirmation is a hope, and a drug given without an announcement gets given twice.
- Say the plan out loud before the next cycle "Two minutes of compressions, then rhythm check, charge to 200 during the cycle." The room now moves together instead of reacting to you.
The leader should have empty hands. If you are compressing, drawing up drugs or bagging, you are not leading — you are a task-doer who is also being asked to hold the whole picture. That is where important things get dropped.
3. Fixating on the drugs while the basics decay
The third pattern is the most consequential. Attention flows to the interesting problems — which vasopressor, which antiarrhythmic, whether to give bicarbonate — while the things that actually determine survival quietly degrade.
Compressions get shallow. The compressor works past the two-minute swap because they feel fine, and does not notice their depth falling. Interruptions stretch to twenty and thirty seconds for an IV attempt or an intubation. Ventilation rate climbs. Nobody is watching, because everybody is thinking about pharmacology.
No drug in the algorithm compensates for poor perfusion. Epinephrine given during compressions that are not generating output does not circulate. The unglamorous work is the work that changes outcomes.
Assign someone to watch compression quality and interruption length, and give them the authority to interrupt you. A team where the most junior person can say "that pause is too long" and be thanked for it will outperform a more experienced team where they cannot.
Why simulation fixes what reading cannot
Every failure above is a performance problem, not a knowledge problem. You cannot read your way out of freezing, and you cannot fix diffuse leadership with a better flowchart. The only reliable remedy is having done it before — enough times that the sequence runs without consuming your full attention.
That is the entire argument for practising codes rather than studying them. The first time you say "I'm running this code" should not be the first time a patient needs you to.
Scenarios that put you in the leader's position with the clock running and vitals that respond to what you actually do. You can hesitate on a rhythm and watch the cost, or fixate on drugs while the pressure drops. Failing here is free — which is exactly what makes it useful.
What every clinician should remember
- Reduce the first rhythm decision to a binary — shockable or not — and refine the diagnosis later
- Say out loud that you are running the code, and assign every role by name
- The leader should have empty hands; a leader doing tasks is not leading
- Use closed-loop communication for every drug and every shock, without exception
- Put someone on compression quality and interruption length, and let them overrule you