Written By Dr. J Shriki

// May 13, 2026 · ECMO · Anticoagulation

Anticoagulation for Emergent ECMO Cannulation

// 2 AM Quick Look-Up
Mnemonic — Remember "5" (the anchor digit)
Agent Bolus Infusion Target
Heparin 50–100 U/kg weight-based drip ACT 180–220 / PTT 50–75 / anti-Xa 0.3–0.7
Bivalirudin 0.5 mg/kg 0.15–0.3 mg/kg/h aPTT 50–80 / ACT 160–220
Argatroban none 0.3–0.5 mcg/kg/min aPTT 50–80 / ACT 180–220
// HIT or heparin contraindicated
  1. Bivalirudin first — only agent that reduces device thrombosis
  2. Argatroban fallback — does not reduce circuit thrombosis
  3. Neither available, cannot defer? One-time heparin ~2,500 U for circuit init only → transition to DTI immediately

The Decision Tree

Standard case (no HIT, no contraindication)

UFH remains the most widely used agent for both VV and VA ECMO. Bolus 50–100 U/kg at cannulation, then weight-based drip to ACT 180–220 s, aPTT 50–75 s, or anti-Xa 0.3–0.7 IU/mL. Caveats: AT-III dependent (acquired deficiency is common on ECMO and masquerades as heparin resistance — check the level before escalating dose), HIT risk, and unpredictable PK. ACT correlates poorly with actual heparin concentration (r² ≈ 0.03 between bolus dose and post-bolus ACT) — use anti-Xa as an adjunct when available.

HIT confirmed or strongly suspected

Bivalirudin is first line. Bolus 0.5 mg/kg IV → infusion 0.15–0.3 mg/kg/h. Target aPTT 50–80 s or ACT 160–220 s. Half-life ~25 min with predominantly enzymatic proteolysis (~80% non-renal), so it offsets fast even without a reversal agent. Reduce infusion ~50% with concurrent CRRT.

Argatroban is the fallback if bivalirudin is not immediately available — no bolus, infusion 0.3–0.5 mcg/kg/min, aPTT 50–80 s. Important: argatroban is not pharmacologically equivalent to bivalirudin in this context (see below).

Emergent cannulation, both DTIs unavailable

If cannulation cannot be deferred — imminent hypoxemic arrest, no DTI on the shelf — a single reduced-dose heparin bolus (~2,500 U IV) for circuit initiation only is defensible. Heparin must be scrupulously avoided before and after the procedural window. Transition to argatroban or bivalirudin immediately post-cannulation. The ACCP guidelines support short procedural heparin re-exposure in HIT when the alternative is death. The institutional fix is upstream: keep both DTIs stocked and bedside-accessible.

Why Bivalirudin Beats Argatroban

The 2023 Chen et al. network meta-analysis (23 trials, 2,522 patients) is the cleanest head-to-head comparison available:

Outcome Bival vs UFH Argatroban vs UFH
Device thrombosis ↓ OR 0.51 ✓ No difference
Patient thrombosis ↓ OR 0.44 ✓ No difference
Major bleeding No difference No difference
Mortality No difference No difference

Bivalirudin vs argatroban head-to-head for device thrombosis: OR 0.14 (95% CI 0.03–0.51) favoring bivalirudin. The 2025 Hu et al. meta-analysis confirms the signal — lower mortality (OR 0.74), lower thrombotic events (OR 0.52), higher decannulation success (OR 1.87) vs UFH. Bivalirudin is no longer just an "alternative" — it is increasingly the preferred agent in centers with experience.

Argatroban: Dose Depends on the Indication

One of the most common errors is applying the FDA-label HIT dose to ECMO patients. They are not the same regimen.

Indication Bolus Infusion Target
ECMO (off-label) none 0.3–0.5 mcg/kg/min aPTT 50–80 / ACT 180–220
HIT, FDA label none 2 mcg/kg/min (max 10) aPTT 1.5–3× baseline (≤100)
HIT + hepatic impairment none 0.5 mcg/kg/min aPTT 1.5–3× baseline
PCI in HIT 350 mcg/kg over 3–5 min 25 mcg/kg/min (max 40) ACT 300–450
Post-PCI continued none 2 mcg/kg/min aPTT-guided
Pearl The ECMO dose (0.3–0.5 mcg/kg/min) is roughly 50–80× lower than the PCI infusion dose. Critical illness, hepatic congestion from RV failure or post-arrest shock liver, and reduced clearance are the reasons. Starting at the FDA HIT dose in an ECMO patient will produce supratherapeutic levels and bleeding.

Half-life ~40–50 min. Avoid in PCI patients with AST/ALT ≥3× ULN. Mayerhöfer et al. (2024) found argatroban guided by hemoclot™ targets (0.4–0.6 µg/mL) may prolong thromboembolism-free time vs anti-Xa guided UFH in VV ECMO — but this is a monitoring strategy, not a thrombosis-reduction signal vs UFH.

Weight Basis and Dose Caps

All three agents are dosed by actual body weight. There is no published ECMO guideline specifying a maximum absolute heparin bolus. A 200 kg patient receives 10,000–20,000 U per protocol; there is no evidence basis for capping at 10,000 U.

The CPB obesity literature does not transfer cleanly: those patients receive ~300 U/kg targeting ACT >400, where overdosing concerns scale up. ECMO uses ~50–100 U/kg targeting ACT 180–220, so the obesity-related overdose risk is proportionally smaller. Practical move in morbid obesity: start at the low end (50 U/kg ABW), check ACT within minutes, titrate. The bolus is a starting point — individual heparin responsiveness is the dominant variable.

For bivalirudin and argatroban specifically, the published cohorts (Tsu Dager n=135, Elagizi Davis n=121) show no clinically significant outcome differences in obese vs nonobese patients dosed by ABW.

BMI Is Not a Contraindication

EOLIA excluded BMI >45, but that was a trial-specific criterion, not a clinical guideline. The actual outcome data go the other way:

  • ELSO registry (Peetermans, n=18,529): BMI ≥35 associated with reduced in-hospital mortality (OR 0.878, p=0.008).
  • ECMObesity (Rudym, n=790): obesity (BMI ≥30) associated with lower ICU mortality (OR 0.63, p=0.018).
  • Hatton 2025 (ELSO, n=24,796): nonlinear relationship — BMI 40 had 18% lower risk of death vs BMI 25.

Scoping review consensus (Ripoll 2024): obesity by BMI alone should not be an exclusion criterion. Technical considerations remain — larger cannulae (25–31 Fr drainage), ultrasound-guided access mandatory, higher flow requirements, increased renal complications — but these are operational, not absolute contraindications.

Pearls

1 Bivalirudin is the only agent proven to reduce both device and patient thrombosis vs UFH without increasing bleeding.
2 Argatroban is a fallback, not an equivalent. It does not reduce thrombosis vs UFH.
3 ACT is unreliable in isolation. Add anti-Xa when available.
4 "Heparin resistance" on ECMO is often AT-III deficiency. Check the level before chasing the dose up.
5 Argatroban dose is indication-specific. ECMO ≠ HIT label ≠ PCI. Confirm context before titrating.
6 Stock both DTIs at the ECMO bedside. The emergent-cannulation scenario without a DTI on hand is an avoidable institutional failure.
References
  1. Bernhardt AM, et al. ISHLT/HFSA Guideline on Acute Mechanical Circulatory Support. J Card Fail. 2023.
  2. Guglin M, et al. Venoarterial ECMO for Adults: JACC Scientific Expert Panel. JACC. 2019.
  3. Chen J, et al. Anticoagulation Strategies in ECMO: Network Meta-Analysis. Pharmacotherapy. 2023.
  4. Hu Y, et al. Bivalirudin Anticoagulation in Adult ECMO: SR/Meta-Analysis. Medicine. 2025.
  5. Lofy T, et al. Bivalirudin vs UFH in ECMO — Single-Center Cohort. Ann Pharmacother. 2026.
  6. Lanoiselée J, et al. Heparin Dosing in VA ECMO: PK Analysis. Pharmaceutics. 2024.
  7. Tsu LV, Dager WE. Bivalirudin Dosing in Obese HIT Patients. Pharmacotherapy. 2012.
  8. Elagizi S, Davis K. Argatroban Dosing in Obesity. Thromb Res. 2018.
  9. Haas E, et al. Heparin Management During CPB in Obese Patients. Eur J Anaesthesiol. 2016.
  10. Vienne M, et al. Adjusted Heparin Calculation Model During CPB: RCT. Eur J Anaesthesiol. 2018.
  11. Peetermans M, et al. BMI and Respiratory ECMO Outcomes: ELSO Registry. Intensive Care Med. 2023.
  12. Rudym D, et al. ECMObesity Study. AJRCCM. 2023.
  13. Hatton KW, et al. BMI and Mortality in VV ECMO. Crit Care Med. 2025.
  14. Ripoll JG, et al. Obesity as Exclusion Criterion for ECMO: Scoping Review. Anesth Analg. 2024.
  15. Argatroban (Acova/argatroban) prescribing information, FDA.
Dr. J Shriki
Next

The 1980s Should Be Revisited But Not Relived!