Comparison of methods for intravenous infusion of fat emulsion during extracorporeal membrane oxygenation

Abstract

Study objectives: To characterize the effects of infusing fat emulsion during neonatal extracorporeal membrane oxygenation (ECMO) by comparing results from patients receiving fat emulsion through the ECMO circuit with those receiving fat emulsion through separate intravenous access. A second goal was to identify the optimal route for administration.

Design: Prospective, randomized, open-label trial.

Setting: Neonatal intensive care unit in a 106-bed quaternary care pediatric hospital.

Subjects: Nine neonates receiving ECMO who required intravenous nutrition. Intervention. Patients received 1-3 g/kg/day of fat emulsion into either the ecmo circuit or separate intravenous access.

Measurements and main results: The ECMO circuit and samples of blood were evaluated hourly for phase separation, layering out of the emulsion from blood, agglutination, and blood clots. After completion, the oxygenators were dissected and examined. Data were compared with an unpaired t test. The characteristics of the groups were similar, except for a higher mean weight in the ECMO circuit group (3.6 +/- 0.3 kg vs 2.8 +/- 0.4 kg, p=0.03). The mean +/- SD triglyceride level during the study was 87 +/- 79 mg/dl, with no significant difference between the two groups. Two patients in each group had elevated triglyceride levels. No cases of phase separation occurred. In the five patients who received fat emulsion into the ECMO circuit, three had layering out of the emulsion and agglutination, and all developed clots in the circuit despite adequate anticoagulation. Of the four patients in the intravenous-access group, one had layering and agglutination, and two had blood clots.

Conclusions: Although both methods were associated with layering out, agglutination, and clot formation, these effects occurred more frequently with administration into the ECMO circuit, particularly in areas of stasis. This may result in disruption of normal ECMO blood flow and impaired delivery of calories. Fat emulsion should therefore be administered through separate intravenous access during ECMO whenever possible.