Inhaled nitric oxide and the longitudinal distribution of PVR in ARDS.

Editors,—I read with interest the article by Ruttmann, James and Aronson1 concerning haemodilution with hydroxethyl starch and normal saline. However, I disagree with their conclusion that they have shown that haemodilution per se exerted a procoagulant effect in both groups. Instead, I propose that their conclusion should be that they demonstrated that haemodilution with normal saline resulted in these changes. They did not state that the hydroxyethyl starch was dissolved in 0.9% sodium chloride, although that is the usual formulation. If this were the case, one would expect the changes that were found in the normal saline group to be present also in the hydroxyethyl starch group, just as they described. Normal saline has a pH of 6.1, osmolality of 308 mosmol kg 1, sodium ion concentration of 154 mEq litre 1 and chloride ion concentration of 154 mEq litre 1. As would be expected, post-infusion serum values reflected those of i.v. fluids, especially when the latter were administered rapidly.2 Therefore, after normal saline has been infused, there is an increase in serum sodium, chloride and osmolality, and a decrease in serum pH. Administration of another crystalloid, such as Hartmann’s solution, results in different serum values3 because Hartmann’s solution has a higher pH (6.5), lower osmolality (273 mosmol kg 1 calculated and 254 mosmol kg 1 measured), and different electrolyte concentrations compared with normal saline. The procoagulant changes reported in their article could have been caused by the changes in serum electrolyte concentration, osmolality or pH associated with infusion of normal saline, rather than by haemodilution. It would be interesting to know if the changes in coagulation reported in this article could have been avoided by the use of a different crystalloid such as Hartmann’s solution.