Fluid administration and fluid balance are both important clinical considerations in the management of patients receiving extracorporeal membrane oxygenation (ECMO). Insufficient intravascular volume may limit the output of preload-dependent centrifugal ECMO pumps and lead to inadequate extracorporeal blood flow, whilst positive fluid accumulation has been consistently associated with increased mortality.^1–3 Some studies have even demonstrated that once patients achieve significant positive fluid accumulation and overload, they remain at greater risk of worse outcomes even if the fluid balance is corrected with mechanical fluid removal using renal replacement therapy.⁴ However, precisely quantifying fluid balance in critically ill patients receiving ECMO can be challenging because of the need to factor in insensible fluid loss from skin evaporation, for example, versus water absorption from inhaling humidified oxygen via mechanical ventilation. These challenges may be easier to overcome in small children because clinical algorithms exist to estimate insensible fluid losses and because the patients can be more readily weighed.

 

In this edition of the Journal, Nakamura and colleagues⁵ present a study of an important potential source of water loss during ECMO, namely that from the membrane lung. These authors conducted an ex-vivo study mimicking an adult patient receiving ECMO with a polymethylpentene (PMP) membrane lung and an experimental circuit flow rate of 3 L/min. They made changes to the priming solution temperature, circuit pressure, and sweep gas flow rate. Circuit pressure was not associated with water loss, but temperature and gas flow were. Water loss was approximately 8 mls/hr at 35 ^OC and 13 mls/hour with 5 L/min sweep gas. While these figures may not seem particularly high, cumulative water loss over of a course of ECMO lasting several days, weeks, or months, could be considerable. These findings are consistent with other studies of this phenomenon