ACLS Updates: Critical Changes Every Emergency Provider Must Know

Chester "Chet" Shermer, MD, FACEP · May 18, 2026

About the author: Dr. Chester Shermer, MD, FACEP

ACLS Updates: Critical Changes Every Emergency Provider Must Know

Updated for the 2025 AHA adult CPR/ECC guidelines: prompt defibrillation without a fixed two-minute delay, uncertain benefit from double sequential defibrillation, IV-first access, and current post-arrest temperature care.

The 2025 American Heart Association Guidelines for CPR and Emergency Cardiovascular Care are out. This post was originally written against the 2020 guidelines and has been updated to reflect the 2025 adult recommendations. Follow your current local protocol and use the linked guidelines for the full recommendations.

If arrest is unwitnessed or unmonitored, start CPR immediately while a defibrillator or AED is obtained and applied. Analyze the rhythm and deliver a shock promptly when indicated; do not impose a universal two-minute CPR delay before using an available defibrillator. Minimize interruptions before and after the shock.

If you only read one paragraph: compressions and shocks are still the two things that save the patient, and everything else in the algorithm exists to avoid getting in their way. What changed in 2025 is mostly about the things we had started doing around those two — a defibrillation technique many departments adopted early, a vascular-access reflex, a temperature target, and a handful of drugs that keep getting pushed out of habit.

Here is what changed, what did not, and what to do about it in your department.

The short version

Defibrillation: the change most likely to catch you out

This is the one to read twice.

After the DOSE-VF trial in 2022, a lot of departments and EMS systems wrote double sequential external defibrillation (DSED) and vector change (VC) into their refractory-VF protocols. Some bought a second monitor for it.

The 2025 guidelines do not establish the usefulness of either approach. Part 9 gives a Class 2b recommendation for changing the pad vector or using double sequential defibrillation after three or more shocks: benefit remains uncertain. Do not treat either as a proven routine intervention.

The reasoning matters more than the verdict, because it changes how you should think at the bedside. When VF is still on the monitor at the next rhythm check, the intuitive read is "the shock failed, so shock harder." The evidence says that in roughly four out of five of those moments, the shock did not fail — it terminated the VF, and the VF came back. Recurrence and refractoriness look identical on a rhythm check three seconds after a shock, and they are not the same problem. You do not fix recurrence with more energy.

What to do instead. Go back to the things that reduce recurrence and improve the odds of the next shock: compression quality between shocks, pad placement you can actually defend (good skin contact, correct anterolateral or anteroposterior position, hair and patches dealt with), full recoil, and short pre-shock pauses. Pre-charge the defibrillator during compressions so the pause is a shock and nothing else. Then look hard for the reversible cause driving the recurrence.

If your protocol already has DSED in it, this is a conversation to have with your medical director now rather than at your next protocol review. The guideline does not say DSED is harmful. It says the case for it is not made. Whether you keep it as an option after three failed shocks is a local decision — but it should be a decision, made on the current evidence, not a leftover.

High-quality CPR: unchanged, and still where the misses are

Nothing here moved, which is worth saying out loud, because "nothing changed" is exactly the section people skip.

Rate: 100–120 per minute. Faster is not better — speed eats depth.
Depth: at least 2 inches (5 cm) for adults, and no more than 2.4 inches (6 cm).
Full recoil: leaning on the chest kills venous return and coronary perfusion. It is the commonest silent defect in a tired team.
Pauses under 10 seconds: rhythm checks, shocks, intubation, everything.
Ventilation: do not over-ventilate. With an advanced airway in place, one breath every 6 seconds, without stopping compressions.

Use the number in context. End-tidal CO2 can help assess CPR physiology; an abrupt increase may indicate ROSC. Changes can also reflect ventilation, drugs, airway position, and arrest cause, so do not use one reading alone to judge CPR quality or decide when to stop. Rotate compressors about every two minutes, or sooner if quality falls.

Vascular access: IV first, and the honest reason why

Part 9 recommends intravenous access first during adult cardiac arrest. Intraosseous access is reasonable when IV access is not feasible or would delay treatment.

That will read as a step backwards to anyone who has watched an IO go in during a code in twelve seconds. So here is the evidence behind it, because the size of the difference is the part that should shape your behaviour.

Two large randomised trials published together in October 2024 compared an IO-first strategy with an IV-first strategy in out-of-hospital cardiac arrest. PARAMEDIC-3 randomised 6,082 patients and found 30-day survival of 4.5% with IO-first against 5.1% with IV-first — no advantage to going IO first. IVIO randomised 1,506 patients and found no significant difference in sustained ROSC.

Read that as what it is. Neither trial showed IO to be dangerous. They showed that reaching for the drill first buys you nothing, and IV-first is therefore the sensible default. What they absolutely do not support is delaying drugs to chase a difficult vein. If the IV is not going to happen quickly, drill. The point is the order of preference, not a prohibition.

One route did get removed. The endotracheal tube is no longer recommended as a route for drug administration. If that is still in a protocol or a code-cart card in your department, take it out.

Drugs: what to give, when, and what to stop giving

Epinephrine, 1 mg. It raises coronary perfusion pressure and improves the odds of ROSC. The timing depends on the rhythm:

Non-shockable rhythm (PEA, asystole): give it as early as possible.
Shockable rhythm (VF, pulseless VT): prioritize defibrillation. Give epinephrine after initial defibrillation attempts have failed.
Repeat every 3–5 minutes either way.

The three that keep getting given anyway. Routine calcium, sodium bicarbonate, and magnesium are not recommended in cardiac arrest. Each has real indications, and outside those indications they are noise:

Calcium: hyperkalemia, calcium-channel-blocker toxicity, hypocalcemia.
Sodium bicarbonate: sodium-channel-blocker toxicity (tricyclics, cocaine), salicylate toxicity, hyperkalemia.
Magnesium: torsades de pointes, or known hypomagnesemia.

If you cannot name the indication out loud when you order it, that is your answer. In a code with an unclear cause, the honest move is to keep looking for the cause rather than to empty the drawer into the patient.

Airway: the guideline stopped choosing for you

The 2025 position on airway is a deliberate refusal to name a winner, and it is more useful than a mandate would have been. Either bag-mask ventilation or an advanced airway may be reasonable during CPR — the right answer depends on the setting and on the skill actually present in the room.

Out of hospital, where intubation success is low or training is limited: supraglottic airway.
Out of hospital, where intubation success is high and training is good: supraglottic airway or endotracheal tube.
In hospital, with expert clinicians: supraglottic airway or endotracheal tube.
Once an advanced airway is in: one breath every 6 seconds, continuous compressions, no pausing to ventilate.

Continuous waveform capnography is not optional. It confirms the tube, it monitors the tube, it grades your compressions, and it is usually the first thing that tells you ROSC has happened.

The practical read. Do not let an airway procedure delay CPR or a needed shock. Start with effective ventilation, and choose an advanced airway when the patient and team's skills warrant it. Plan the procedure to minimize interruptions.

Knowing when to stop, and when to escalate

Point-of-care ultrasound may help an experienced operator investigate reversible causes, but its usefulness during arrest is not well established and it can prolong CPR interruptions. Cardiac standstill on ultrasound should not be used alone to decide when to stop resuscitation. If ultrasound is used, plan image acquisition around an existing brief rhythm check and minimize interruptions.

ECPR now has a full system-of-care chapter rather than a passing mention. The new recommendations are about the machinery around it: which patients are selected, how cannulation is organised, and how a region concentrates the volume needed to do it well. That framing is the point. ECPR is not a piece of equipment you buy — it is a pathway with inclusion criteria, a team, and a destination hospital, decided before the patient arrests.

And a new ethics chapter. The 2025 edition carries a dedicated chapter on the ethics of CPR and emergency cardiovascular care, from neonatal through geriatric. If you have ever run a code you knew should not have started, that chapter is written for you.

After ROSC: the part that decides the outcome

Getting a pulse back is the midpoint, not the end. Part 11 was substantially rewritten.

Temperature control: 32–37.5 °C, for at least 36 hours. Two changes from what you learned in 2020. The range is wider at the top, and the duration is longer — 36 hours rather than 24. The wider range reflects the trials, TTM2 among them, that failed to show an advantage for deep cooling over controlled normothermia. The practical core has not changed and never really did: pick a target and hold it, and do not let the patient get hot. Fever after arrest is the thing that reliably hurts.

Worth noting alongside it: prehospital cooling has not been shown to help. Cooling in the back of the truck is not the win it once looked like.

Blood pressure: MAP at or above 65 mmHg. Post-arrest hypotension is strongly associated with worse outcomes. Treat it early and treat it deliberately.

Oxygen: use 100% FiO2 until oxygenation can be measured reliably, then titrate. Target an oxygen saturation of 90–98%, or a PaO2 of 60–105 mmHg. Avoid both hypoxemia and hyperoxemia; do not leave FiO2 at 100% after reliable measurement is available without a clinical reason.

The cath lab: urgency depends on the ECG and the patient, not on the arrest alone. Emergency coronary angiography is for STEMI, shock, or ongoing instability. It is not routine for the comatose patient without ST elevation. But angiography before discharge is recommended for survivors with a suspected cardiac cause — particularly with an initial shockable rhythm, unexplained left ventricular dysfunction, or evidence of severe ischemia. Do not let "no ST elevation" become "no workup."

Neuroprognostication: slow down, and use more than one modality. Delay a definitive multimodal prognostic impression until at least 72 hours after ROSC in patients who have not received hypothermic temperature control, or at least 72 hours after return to normothermia in those who have. Sedation, paralytics, and temperature can confound the exam. The 2025 guidelines also distinguish prediction of favorable from unfavorable outcomes.

Team dynamics: still the multiplier

None of the above survives contact with a disorganised room.

Named roles. Compressor, airway, drugs, recorder, leader. Say them out loud at the start.
Closed-loop communication. "Give 1 mg epi." — "1 mg epi." — "1 mg epi given." Three beats, every time.
A leader with their hands off the patient. The moment the team leader is doing a procedure, nobody is watching the whole code.
Debrief every arrest. Short, structured, non-punitive: what went well, what did not, what is a system problem rather than a person problem.

In twenty-five years I have seen far more codes go wrong from coordination than from knowledge. The compressor who never got relieved. The shock that waited on a conversation. The epinephrine given twice because two people heard the same order. None of those are gaps in the guideline.

What to change on Monday

Pull the endotracheal drug route out of your protocols and code-cart cards.
Re-open the DSED question with your medical director. Decide it deliberately.
Make IV the default first attempt for access, with the drill immediately available and no delay tolerated.
Move your post-arrest temperature order set to 32–37.5 °C for at least 36 hours, and make fever prevention explicit.
Add a post-ROSC oxygen wean to the order set. 100% is a starting point, not a plan.
Check neuroprognostication timing in your ICU pathway: at least 72 hours after ROSC, or after return to normothermia when hypothermic temperature control was used, with confounders excluded.
Debrief every arrest for a month and count how many of the problems were coordination rather than knowledge.

Dr. Chet's Take

The headline in this edition is a walk-back, and I think that is healthy. We adopted double sequential defibrillation fast — faster than the evidence ever justified — because it felt like doing something in the exact moment when doing something is what everyone in the room wants. Then the science came back and said the thing we were treating was mostly not the thing we thought we were treating. Four out of five times, that VF on the monitor is a rhythm that came back, not a shock that bounced. More joules do not fix a recurrence.

That is the pattern worth taking from the whole 2025 edition. Almost every change is the guideline telling us to stop adding and go back to doing the two boring things extremely well. Compressions without leaning. Shocks without long pauses. Then find the cause. The IO finding is the same story wearing different clothes — we all believed the drill was better because it is faster and more satisfying, and six thousand patients say it makes no difference to whether anybody walks out of the hospital.

Where I would not soften anything is the post-ROSC section. We are still, in 2026, walking away from patients with a pulse and 100% oxygen running, and we are still letting them get warm. Those are not close calls and they are not resource problems. They are attention problems, and they happen in the twenty minutes after everyone in the room has already decided the hard part is over.

Frequently asked questions

Should I stop doing double sequential defibrillation?

The 2025 guidelines say its usefulness after at least three shocks has not been established; they do not establish that it causes harm. If it is in your protocol, ask your medical director to review the current evidence. Persistent VF on a later rhythm check does not automatically mean the previous shock failed.

Does IV-first mean I should stop using IO access in cardiac arrest?

No. IV is the preferred first attempt, and IO is reasonable if IV access is not feasible or would delay treatment. Both large 2024 trials found IO-first no better, not IO worse. Do not delay indicated treatment to chase a difficult vein.

What temperature should I target after ROSC?

Anywhere in 32–37.5 °C, held for at least 36 hours in a patient who remains unresponsive. The wide range reflects trials that showed no advantage for deep cooling over controlled normothermia. The non-negotiable part is preventing fever.

Did compression rate or depth change in 2025?

No. Rate 100–120 per minute, depth at least 2 inches (5 cm) and no more than 2.4 inches (6 cm), full recoil, pauses under 10 seconds.

When can I prognosticate neurologically?

Generally at least 72 hours after ROSC, or at least 72 hours after return to normothermia if hypothermic temperature control was used, using multiple modalities and accounting for sedation, paralysis, and temperature.

Sources

About the Author: Chester "Chet" Shermer, MD, FACEP, is the Founder & Medical Director of Global MedOps Command. He is a board-certified emergency physician with over two decades of clinical experience in emergency medicine, simulation education, and medical leadership. Dr. Shermer is passionate about advancing clinical readiness and optimizing team performance through innovative training methodologies, including the EM-Sim program.

Apply the 2025 ACLS Guidelines on EM-Sim

Knowing the guidelines is the first step — executing them under pressure is the skill that saves lives. The Cardiac Arrest — VF/Pulseless VT simulation on EM-Sim puts you through a complete resuscitation scenario with real-time feedback on rhythm recognition, drug timing, shock delivery, and post-ROSC management.

→ Start the Cardiac Arrest Simulation