Post-Cardiac Arrest Care (ROSC)

📘 1 · Washington Manual of Critical Care, 4th Ed

Washington Manual Summary — Post-Cardiac Arrest Syndrome

"Restoration of spontaneous circulation (ROSC) after cardiac arrest leads to a complex, multi-organ pathophysiological state termed post-cardiac arrest syndrome (PCAS). PCAS encompasses four interacting components, each of which independently worsens outcome: post-cardiac arrest brain injury, post-cardiac arrest myocardial dysfunction, systemic ischaemia/reperfusion injury, and the precipitating pathology that caused the arrest."

Washington Manual of Critical Care, 4th Ed. Wolters Kluwer 2023. Chapter: Post-Cardiac Arrest Care.

The Four Components of PCAS

🧠 Brain Injury (most important)

Hypoxic-ischaemic encephalopathy · cerebral oedema · seizures (including non-convulsive) · Responsible for 68% of in-hospital deaths after ROSC

❤️ Myocardial Dysfunction

Transient global myocardial stunning · reduced EF (often <30%) · haemodynamic instability · Usually reversible over 24–72h if supported

⚡ Systemic I/R Injury

Resembles septic shock phenotype · elevated lactate · coagulopathy · activation of inflammatory cascades · endothelial dysfunction

🎯 Precipitating Pathology

Must be identified and treated · ACS, PE, hypoxia, electrolyte, toxin · VF/VT → coronary angio (if STEMI) · PEA → rule out 5 Hs & 5 Ts

Immediate ICU Priorities at ROSC

Airway: Intubate if not already done, confirm placement (ETCO₂)
Haemodynamics: Target MAP ≥65–70 mmHg, SBP ≥100 mmHg; start noradrenaline if hypotensive
Oxygenation: SpO₂ 94–98% — avoid hyperoxia (PaO₂ >300 mmHg independently worsens neurological outcome)
12-lead ECG: Identify STEMI → immediate PCI; no STEMI → delay angio (see Section 3)
Fever prevention: Paracetamol ± active cooling to maintain T° ≤37.5°C
Glucose: Target 6–10 mmol/L; avoid both hypoglycaemia and marked hyperglycaemia
Seizure detection: EEG monitoring within 24h — non-convulsive SE occurs in 30–40%

📗 2 · Marino's The ICU Book, 5th Edition (2025)

Marino Physiology — Cerebral Ischaemia & Reperfusion

🔬 How post-cardiac-arrest brain injury develops

TriggerCardiac arrest → global cerebral & whole-body ischaemia

ATP depletion → Na⁺/K⁺-ATPase failure → Ca²⁺ overload (reverse Na⁺/Ca²⁺ exchanger)

ROSC → reperfusion injury — free-radical burst, inflammatory activation

Brain: no-reflowmicrovascular obstruction, vasospasm, lost autoregulation

Systemic: myocardial stunning + SIRS"sepsis-like" post-arrest shock

Pathological statePost-cardiac-arrest syndrome — anoxic brain injury + multi-organ dysfunction

"Global cerebral ischaemia during cardiac arrest sets in motion a cascade of events that continue and worsen after restoration of circulation. The paradox of reperfusion injury — where restoring blood flow to ischaemic tissue causes additional damage — is mediated by calcium overload, free radical generation, and inflammatory activation. This means that post-arrest management is not simply about maintaining perfusion; it is about actively preventing secondary injury."

Marino PL. The ICU Book, 5th Ed. Ch.45: Altered States of Consciousness. Wolters Kluwer; 2025.

📗 Marino Ch.45 — The No-Reflow Phenomenon

After global ischaemia, restoration of circulation does not guarantee adequate cerebral perfusion. Three mechanisms contribute:

Microvascular obstruction: Oedematous astrocytes, platelet aggregates, and leucocyte adhesion occlude capillaries
Vasospasm: Ca²⁺ influx into smooth muscle cells causes arteriolar constriction
Loss of autoregulation: Cerebral autoregulation is impaired for 12–24h post-arrest; MAP fluctuations directly affect cerebral blood flow

Clinical implication: Maintaining MAP ≥65–70 mmHg is critical in the first 24h to overcome no-reflow and restore adequate CBF.

📗 Marino Ch.45 — Calcium Overload & Free Radical Cascade

During ischaemia, ATP depletion causes Na⁺/K⁺-ATPase failure → intracellular Na⁺ accumulation → reverse Na⁺/Ca²⁺ exchanger → massive Ca²⁺ influx.

Ca²⁺ excess activates:

Phospholipases A₂ and C → membrane phospholipid degradation → arachidonic acid release → prostaglandins + free radicals
Proteases → cytoskeletal destruction
Nitric oxide synthase → excess NO → peroxynitrite (potent oxidant)

Reperfusion paradox: Reoxygenation provides the substrate (O₂) for the free radical burst that was "primed" during ischaemia. This is why the first minutes after ROSC carry the highest risk of secondary neuronal death.

📗 Marino Ch.45 — Why Fever is Catastrophic Post-Arrest

Each 1°C rise in brain temperature above 37°C increases cerebral metabolic rate (CMRO₂) by approximately 8%. In the post-arrest brain with impaired CBF and autoregulation, this metabolic demand cannot be met → secondary ischaemia → accelerated neuronal death.

Clinical threshold: Temperature >37.5°C ("post-arrest fever") independently worsens neurological outcome. The goal is not cooling — it is fever prevention.

📋 3 · AHA 2025 CPR Guidelines & Key Trials

AHA 2025 Guidelines — Post-ROSC Management

⚡ AHA 2025 Major Change from Prior Guidelines

Targeted Temperature Management (TTM) is no longer recommended for all ROSC patients. The TTM2 trial (2021, NEJM, n=1900) showed NO benefit of targeting 33°C vs fever avoidance at 37.5°C for shockable rhythms. AHA 2025 now recommends fever avoidance (T° ≤37.5°C) rather than active cooling to 32–36°C for most patients.

Exception: Some centres still use active cooling for selected comatose patients with witnessed, shockable cardiac arrest with short no-flow time — but this is no longer routine practice.

Temperature Management

Target: Maintain T° ≤37.5°C for at least 72h after ROSC STRONG 1B — AHA 2025
Fever (>37.5°C) must be actively prevented with paracetamol (1g QDS IV) ± cooling blankets.
Do NOT target 32–36°C routinely — TTM2 trial showed no benefit over fever avoidance TTM2 2021
Duration: At least 72h post-ROSC (secondary temperature surges can occur).

Coronary Angiography Timing (AHA 2025)

STEMI on ECG → Immediate PCI regardless of neurological status STRONG 1A
No STEMI → NO routine immediate coronary angiography STRONG 1A
Evidence: TOMAHAWK trial (2021, NEJM, n=554): immediate angio = 30-day mortality 54% vs 46% in delayed group TOMAHAWK 2021
COACT trial (2019, NEJM, n=552): no survival benefit to immediate angio in shockable arrest without STEMI COACT 2019
Conclusion: Coronary angio should still be performed in non-STEMI ROSC patients — but it can be done within 24–72h once haemodynamically stable, unless there is a compelling clinical suspicion (e.g., ECG changes, anginal prodrome, cardiogenic shock).

Oxygenation & Ventilation Targets

SpO₂: 94–98% — avoid hyperoxia STRONG 1B — AHA 2025
PaO₂: Target 75–100 mmHg (10–13 kPa). PaO₂ >300 mmHg independently increases in-hospital mortality (large observational data).
PaCO₂: Target 35–45 mmHg (normocapnia). Hypocarbia causes cerebral vasoconstriction and worsens cerebral ischaemia.
TV: 6–8 ml/kg IBW (lung-protective even in non-ARDS post-arrest lungs).

Haemodynamic Targets (AHA 2025)

MAP ≥65–70 mmHg (some centres target MAP ≥80 in first 12h based on observational data) CONDITIONAL 2C
SBP ≥100 mmHg
Noradrenaline is first-line vasopressor for post-arrest shock
Dobutamine if myocardial dysfunction (EF <30%, cardiogenic shock phenotype) — add to noradrenaline
Fluid resuscitation: Balanced crystalloids (PlasmaLyte or LR); avoid large-volume normal saline.

Glucose Management

Target blood glucose 6–10 mmol/L (108–180 mg/dL) STRONG 1B
Both hypoglycaemia (<4 mmol/L) and marked hyperglycaemia (>14 mmol/L) independently worsen neurological outcomes.
Use IV insulin infusion with hourly monitoring. NICE-SUGAR protocol: avoid tight control (4–6 mmol/L) — excessive hypoglycaemia risk.

Seizure Management

EEG monitoring within 24h of ROSC — non-convulsive seizures in 30–40% of comatose post-arrest patients STRONG 1B
Prophylaxis: Not routinely recommended by AHA 2025 (no proven benefit)
Treatment of confirmed seizures: Levetiracetam 1–3g/day IV (first-line) or sodium valproate 1200–2000 mg/day; benzodiazepines for acute termination.
Refractory status post-arrest: Poor prognostic sign; propofol or midazolam infusion for burst suppression.

🇮🇳 Indian ICU Context

ECPR (ECMO-CPR): VA-ECMO during refractory cardiac arrest is available at select tertiary centres. If <65y, witnessed arrest, <15 min no-flow, consider early ECPR referral if available.

Resource constraints: Continuous EEG not available at most centres — use intermittent 20–30 min EEG recordings every 12h in comatose post-arrest patients. Spot EEG can detect burst-suppression and overt seizure patterns.

Out-of-hospital arrest: India has poor out-of-hospital survival (bystander CPR uncommon); most cardiac arrests presenting to Indian ICUs are in-hospital. IHCA has better neurological outcome than OHCA.

💊 4 · Drug Doses & Physiological Targets

Drug Reference — Post-ROSC Care

Drug / Intervention	Indication	Dose / Target	Notes
Noradrenaline	Post-arrest haemodynamic shock	0.05–0.5 μg/kg/min IV infusion; titrate to MAP ≥65–70	First-line; no superiority of vasopressin over NE for PCAS
Dobutamine	Post-arrest myocardial stunning	2–20 μg/kg/min IV, add to NE when EF <30% or CI <2.2	Transient — stunning usually resolves in 24–72h; reassess daily
Paracetamol (IV)	Fever prevention (≤37.5°C)	1g IV QDS (every 6h) routinely	Start immediately on ICU admission post-ROSC; continue 72h
Insulin (IV infusion)	Glucose control	Titrate to BG 6–10 mmol/L (hourly checks initially)	NICE-SUGAR protocol; avoid hypoglycaemia (<4 mmol/L)
Levetiracetam	Confirmed post-arrest seizures	1g IV loading dose → 500mg–1.5g BD IV/PO	Preferred over phenytoin; minimal drug interactions; renally cleared
Midazolam	Acute seizure termination	0.1 mg/kg IV bolus (up to 10mg); repeat × 1 if needed	Follow with levetiracetam load; IM lorazepam 4mg if no IV access
Propofol	Sedation + burst suppression for refractory SE	1–4 mg/kg/h IV; titrate to burst suppression on EEG	Propofol infusion syndrome risk at >4 mg/kg/h for >48h — monitor TG, pH
Cooling blanket / Ice packs	Fever management if paracetamol insufficient	Surface cooling to maintain T° ≤37.5°C	Target fever avoidance, not hypothermia; avoid T° <35°C
PCI / Thrombolysis	Confirmed STEMI on ROSC ECG	Immediate PCI (door-to-balloon <90 min); thrombolysis if PCI unavailable	Do not withhold reperfusion for coma — neurological outcome often improves post-PCI

Physiological Targets at a Glance

Post-ROSC Bundle Targets

Temperature: ≤37.5°C for 72h minimum
MAP: ≥65–70 mmHg (consider ≥80 if cerebral perfusion concerns)
SpO₂: 94–98% | PaO₂: 10–13 kPa (75–100 mmHg)
PaCO₂: 4.5–6.0 kPa (35–45 mmHg) — normocapnia
Glucose: 6–10 mmol/L
Lactate: Target clearance >10% per hour; <2 mmol/L by 6h
Haemoglobin: ≥80–100 g/L (no anaemia in brain-injured patients)
Sodium: 135–145 mmol/L (normonatraemia)
EEG: Continuous or frequent monitoring for ≥24h in comatose patients

🗂 5 · Clinical Flowchart

Post-ROSC Management — Hour-by-Hour Algorithm

Immediately at ROSC (0–5 min)

Confirm ETT position (ETCO₂ waveform)
12-lead ECG → STEMI? → Activate cath lab immediately
FiO₂ to achieve SpO₂ 94–98% — wean O₂ if SpO₂ >98%
IV access × 2, ABG, bloods (troponin, lactate, glucose, electrolytes, FBC, coagulation)
Start noradrenaline if MAP <65 — target ≥65–70 mmHg

Within 1 Hour of ROSC

Transfer to ICU with continuous monitoring
Paracetamol 1g IV for fever prevention (start proactively)
Glucose check → insulin infusion if >10 mmol/L
Bedside echo: LV function, pericardial effusion, wall motion abnormality, RV size (PE?)
Identify and treat 5 Hs & 5 Ts (reversible causes)

6–24 Hours — Stabilisation Phase

Coronary angio decision (no-STEMI): Stabilise haemodynamics → angio within 24–72h if stable; strong clinical suspicion (ACS story, ECG changes) → earlier
CT brain: exclude haemorrhagic cause if not done pre-arrest
EEG monitoring: start within 24h in all comatose patients
Target temperature ≤37.5°C — active cooling if paracetamol insufficient
Mechanical ventilation: TV 6–8 ml/kg IBW, normocapnia
Consider nasogastric feeding if expected coma >24h

24–72 Hours — Monitoring Phase

Continue fever prevention — secondary temperature surges are common day 2–3
Reassess haemodynamics — post-arrest myocardial stunning begins to recover
Daily neurological assessment (pupils, GCS withdrawal, corneal reflex)
EEG findings: continuous monitoring if burst-suppression or seizures identified
Repeat echo at 24h to reassess LV function
Troponin peak (usually 24h post-arrest) — not reliable for diagnosing ACS alone in this context

≥72 Hours — Prognostication Phase

Neuroprognostication: Begin multimodal assessment at ≥72h from ROSC (or ≥72h after TTM if used)
Assess: GCS, pupil reactivity, corneal reflex, EEG pattern, SSEP (N20), CT/MRI brain, NSE levels
No single test is sufficient for withdrawal of care — at least 2 poor prognostic indicators required
Involve neurology, family, ethics if prognosis poor

Coronary Angio Decision Tree (AHA 2025)

ROSC + ECG

STEMI present → Immediate PCI (door-to-balloon <90 min) regardless of consciousness level

No STEMI → No routine immediate angio. Stabilise in ICU. Perform angio within 24–72h if:
• Strong clinical suspicion (anginal prodrome, dynamic ECG changes, high troponin)
• Haemodynamic instability not explained by other cause
• Shockable rhythm (VF/VT) with no obvious non-cardiac cause

PEA/Asystole (non-shockable): Lower yield for coronary disease — focus on reversible non-cardiac causes first

🔬 6 · Neuroprognostication

Neuroprognostication — When and How

AHA 2025 Timing Rule

Do NOT make withdrawal of life-sustaining treatment decisions before 72h from ROSC. If active fever management was used, wait ≥72h after temperature target achieved. Sedative effects must be excluded before assessment.

✅ Favourable Prognostic Signs

Return of pupil reactivity by 72h
Motor response (M≥3) at 72h
Normal EEG background
Absent burst-suppression at 48h
Normal CT brain at 24h
Low NSE at 48h (<17 μg/L)
Present cortical SSEP N20 waves

⚠️ Poor Prognostic Signs

Bilaterally absent pupil light reflex at 72h
Bilaterally absent SSEP N20 waves
Malignant EEG: burst-suppression or suppression-burst
NSE >60 μg/L at 48h or >33 μg/L at 72h
CT brain: generalised cortical oedema, loss of grey-white differentiation
Status epilepticus on EEG
GCS motor = 1 (no response) at 72h + 2 other poor indicators

Multimodal Approach — ERC 2021 / AHA 2025

A minimum of two independent poor prognostic indicators are required before considering withdrawal. The most robust single predictor is:

Bilaterally absent SSEP N20: False positive rate for poor outcome <1%. If available, this is the most reliable single predictor. N20 = primary cortical somatosensory evoked response at 20 ms after median nerve stimulation.
NSE (Neuron-Specific Enolase): Released from damaged neurons. Trend from 24h to 72h is more informative than single value. NSE >60 μg/L at 48h = strong predictor of poor outcome (high specificity).
CT/MRI brain: Early CT showing grey-white ratio (GWR) <1.10 = severe cortical oedema, very poor prognosis. MRI at 5–7 days: DWI restriction in basal ganglia/cortex = poor prognosis.

❌ 7 · Common Mistakes

Common Mistakes in Post-Arrest Care

❌ Mistake 1 — Hyperoxia

Giving 100% FiO₂ "to be safe" after ROSC. PaO₂ >300 mmHg is independently associated with increased in-hospital mortality. Wean FiO₂ immediately to achieve SpO₂ 94–98%. This is analogous to avoiding hyperoxia in prematurity (retinopathy).

❌ Mistake 2 — Targeting 33°C Hypothermia Routinely

Persisting with TTM 32–36°C for all post-arrest patients. TTM2 (2021) definitively showed no benefit over fever avoidance. Active cooling to 33°C without evidence is harmful (arrhythmias, electrolyte shifts, increased infection risk). Fever prevention (≤37.5°C) is the AHA 2025 standard.

❌ Mistake 3 — Early Prognostication and Premature Withdrawal

Declaring "brain death" or withdrawing care before 72h. Sedative effects, metabolic derangement, and hypothermia all suppress neurological responses — mimicking poor prognosis. Many patients with coma at 24–48h have meaningful recovery by day 5–7. Withdraw only after multimodal assessment at ≥72h.

❌ Mistake 4 — Routine Immediate Coronary Angio for All Non-STEMI ROSC

TOMAHAWK and COACT showed immediate angio in non-STEMI ROSC may INCREASE 30-day mortality (due to contrast-related complications in haemodynamically unstable patients, delayed ICU management). Stabilise first, angio when haemodynamically stable unless STEMI or strong clinical indication.

❌ Mistake 5 — Neglecting Glucose and Permitting Hypoglycaemia

In attempting tight glucose control, inducing hypoglycaemia (<4 mmol/L) is as damaging as hyperglycaemia. The post-arrest brain cannot extract adequate glucose at low serum levels. Target 6–10 mmol/L; check glucose hourly while on insulin infusion.

❌ Mistake 6 — Missing Non-Convulsive Status Epilepticus

Non-convulsive SE occurs in 30–40% of comatose post-arrest patients — it is clinically silent and can only be detected on EEG. Without EEG monitoring, it will be missed. These patients appear "comatose" but are actually seizing. EEG within 24h is essential for all comatose post-arrest patients.

📑 8 · References

References

AHA 2025 Cardiac Arrest Guidelines. American Heart Association. CPR & Emergency Cardiovascular Care Science with Treatment Recommendations. Circulation 2025.
ILCOR CoSTR 2023. International Liaison Committee on Resuscitation. Consensus on Science with Treatment Recommendations. Resuscitation 2023.
Dankiewicz J et al. (TTM2 Trial). Hypothermia versus Normothermia after Out-of-Hospital Cardiac Arrest. NEJM 2021;384:2283–2294.
Desch S et al. (TOMAHAWK Trial). Angiography after Out-of-Hospital Cardiac Arrest without ST-Segment Elevation. NEJM 2021;385:2544–2553.
Lemkes JS et al. (COACT Trial). Coronary Angiography after Cardiac Arrest without ST-Segment Elevation. NEJM 2019;380:1397–1407.
Nolan JP, Sandroni C, Böttiger BW et al. ERC and ESICM Guidelines for Post-Resuscitation Care 2021. Intensive Care Med 2021;47:369–421.
Marino PL. The ICU Book, 5th Ed. Ch.45: Altered States of Consciousness. Wolters Kluwer; 2025.
Washington Manual of Critical Care, 4th Ed. Chapter: Post-Cardiac Arrest Care. Wolters Kluwer 2023.
Sandroni C, D'Arrigo S, Nolan JP. Prognostication after Cardiac Arrest. Crit Care 2018;22:150. (systematic review on neuroprognostication accuracy)