Crystalloids should be second choice for goal-directed fluid therapy

Editor, I would like to congratulate László et al.1 for their well performed comparison of crystalloid and colloid fluid for goal-directed volume therapy during free flap surgery that was recently published in the European Journal of Anaesthesiology. Most evaluations of goal-directed fluid therapy have used colloids but some studies, and many clinicians, have turned to crystalloids. Therefore, László's study is pertinent to the current practice of anaesthesia. However, I still have difficulties in understanding why crystalloids are used for this purpose. The reason underlying my difficulties in understanding is that the acute rise in cardiac index (CI) induced by a bolus infusion of crystalloid fluid is short-lived. When crystalloid fluid is administered rapidly, within 3 to 5 min, a redistribution phase will be very prominent. Almost half of the induced plasma volume expansion will be lost within 10 min (Fig. 1a). This is the case during general anaesthesia and surgery except when there is a sudden drop in arterial pressure, which transiently stops the redistribution.2Fig. 1: Plasma volume expansion following three boluses of Ringer's solution infused with an interval of 40 min (a). The expansion resulting from the first bolus is maintained with a gradual step-down infusion (b). Computer simulation-based kinetic data from infusions of larger amounts of fluid during thyroid surgery.3The redistribution effect becomes smaller with the infusion time, and is of negligible consequence for lengthy infusions. Then, the rate of elimination is the key factor determining plasma volume expansion, which has in fact cast doubts over the relative potency of these fluids in several intensive care studies.4 When administering crystalloid as a bolus, clinicians must continue the infusion at a lower step-down rate to maintain the plasma volume expansion that has just been achieved (Fig. 1b). If not, the bolus infusion needs to be repeated much more frequently than with a colloid, since the latter lacks a redistribution phase and better remains in the circulation. Clinicians apparently handle this issue by accepting a lower overall CI and giving bolus infusions more frequently, which was the also case in the László et al.'s study.1 However, their bolus infusions were administered over a maximum of 15 min, which somewhat reduces the distribution effect. A shorter infusion time, which is more common, aggravates the problems associated with the redistribution phase. I give those who advocate crystalloids the possibility that I am wrong in my scepticism. When urinary excretion is very prompt, as in well hydrated conscious volunteers, the infusion of a small amount of fluid does not always hydrate the interstitial fluid space because excretion may occur faster than the distribution.5 This is unlikely during ongoing anaesthesia and surgery because urinary excretion is strongly inhibited in that setting (−90%).2 Moreover, the results of László et al. indicate that their infusions were subject to distribution in the way I suspect. My kinetic considerations are not unique for infusion fluids but pertain to all drugs with exponential distribution and elimination. Hence, a crystalloid fluid should be continued at a lower rate after a bolus infusion to achieve a stable plasma volume expansion profile similar to that of a colloid. Such a step-down is quite possible to perform, but never practiced as far as I know. This is the rationale behind my opinion that crystalloid fluids should be a second-choice for goal-directed therapy during surgery.