Perfusion Leadership & Management
Pediatric PerfusionIssue No. 5November 16, 20263 min read

The Dose That Was Borrowed

Four hundred units of heparin per kilogram is the neonatal dose most programs use, and the paper says plainly that it came from adult protocols. This trial asked what six hundred does instead.

The Research

Wang H, Liu Q, Gao P, et al. Optimization of Anticoagulation Strategy During Cardiopulmonary Bypass in Neonates with Congenital Heart Disease: A Randomized Controlled Trial. Drug Design, Development and Therapy, 2026;20:634756. Registration ChiCTR2300069157. · PubMed 42899169

Four hundred units of heparin per kilogram is the neonatal dose most programs use, and the paper says plainly that it came from adult protocols. This trial asked what six hundred does instead.

What They Did

Randomized 76 neonates having cardiac surgery on bypass at a single center, single blind, to an initial heparin dose of 600 or 400 units per kilogram, 38 in each group. The primary endpoint was the change in prothrombin fragment 1+2 across bypass, which is a direct measure of how much thrombin is still being generated while you think the child is anticoagulated.

What They Found

Fragment 1+2 rose 0.70 nmol/L at 600 and 4.08 nmol/L at 400, a difference of 3.38 nmol/L, p below 0.001. The first-dose ACT target was reached 84.2 percent of the time at 600 against 44.7 percent at 400. Surgical preparation was about five minutes shorter. Supplemental heparin was lower, though total heparin dose did not differ significantly. Seventy-two hour chest drainage and albumin use were lower, mechanical ventilation was shorter, and red cell, plasma, and platelet transfusion requirements did not differ. One neonate died in the 600 group and none in the 400 group.

The Hallway

On the smaller dose the blood made nearly six times more of the clotting signal. That is not a close call.

Palm TreeMean ΔF1+2 was 0.70 vs 4.08 nmol/L in the 600 and 400 U/kg groups, respectively (treatment-minus-control mean difference, -3.38 nmol/L; 95% confidence interval, -4.58 to -2.18; P < 0.001).

Where the Evidence Is Strong

Randomized, registered and prospective, with a mechanism-based primary endpoint rather than a soft clinical one. Thrombin generation is the thing heparin is supposed to suppress, so a roughly six-fold lower fragment rise is direct evidence the dose did what it was meant to do, not a proxy for it. The ACT difference is big enough to change an evening, since two out of three neonates on 400 did not reach target on the first bolus. The complete-case sensitivity analysis supported the primary finding.

Where the Evidence Is Limited

Seventy-six neonates at one center. The trial was not powered for clinical outcomes, and the secondary signals, drainage, albumin, ventilation, are unadjusted for multiple comparisons, so they are hints and not results. One death in the higher-dose group is exactly the kind of finding that at this size could be chance and equally cannot be waved away. Transfusion requirements did not move, so no bleeding benefit is established. And neonatal practice at one center may not transfer to another.

The InFlow-mation

Critical Conclusion

Four hundred units per kilogram left most of these neonates short of the ACT target and allowed substantially more thrombin generation. Six hundred did both better, and the trial was too small to prove or exclude a clinical difference.

Take It Back to the Pump

Watch the ACT response, not the dose on the label. A neonate missing target on 400 is the drug underdosing, not the child misbehaving. I will take 600 to our next pediatric protocol review, and I will say out loud that the safety case rests on 76 patients.

Pump off.

On the source

Peer reviewed, single-center randomized single-blind trial, registered ChiCTR2300069157. Neonates only, 76 patients, secondary endpoints not powered for clinical outcomes. Do not carry this to older children or adults, and do not read transfusion safety from it.

Previous issueOne Number Survived