Because the primary aim of our experiment was to investigate the effect of high COP on fluid redistribution, we used an equal fluid volume for both the saline solution-treated and the albumin-treated groups. Thus, the present data may not be extrapolated to the clinical situations involving fluid resuscitation. Furthermore, the albumin-treated groups in our study created a higher degree of reduction in Hb concentration, compared to the saline solution-treated groups (Fig 2). Such an effect, which may be related to a larger degree of hemodilution (as judged by the lack of hemolysis, which was tested in four plasma samples), can affect the blood oxygen carrying capacity.
Our observations are not consistent with a number of studies that have evaluated the effect of crystalloid and colloid infusions during increased capillary permeability on extravascular lung water content. For example, Rackow et al found no difference in lung and muscle extravascular water content after saline solution or hetastarch infusion in five septic rats with cecal ligation and perforation compared to a group of five control rats. However, Rackow et al found significantly lower plasma COP levels (9.3 ± 0.5 mm Hg) in saline solution-infused rats than in hetastarch-infused rats (21.6 ± 0.5 mm Hg) compared to control animals (16.1 ± 1.2 mm Hg). Similarly, Rutili et al found that dextran infusion in lung-injured dogs did not increase lung water content compared with saline solution infusion. Moreover, septic baboons have shown pulmonary dysfunction that is independent of the plasma COP. These contrasting reports may reflect the effect of dynamically changing capillary permeability during sepsis.
In the absence of increased capillary permeability, low oncotic pressures predispose animals to increased levels of extracellular lung water. For example, in hypoproteinemic dogs, without acute lung injury, crystalloid infusion increased the amount of lung water compared with colloid infusion. McKeen et al also found that saline solution infusion can increase extravascular lung water in sheep with hemorrhage-induced hypoproteinemia compared to colloid infusion.