The objective of this study was to determine if coagulation is different between 6%
hetastarch in
normal saline (NS) and 6%
hetastarch in
lactated Ringer's solution (LR), with use of an ex vivo thromboelastography (TEG) model with healthy donated volunteer blood. We simulated
hemodilution that occurs during clinical
resuscitation of hemorrhagic or
hypovolemic shock, using healthy human donor whole blood (WB) ex vivo. Coagulopathy related to low, medium, high, or very high dilution of WB with NS or a high-molecular-weight
hetastarch-based plasma expander, 6%
hetastarch in NS (HSNS) or 6%
hetastarch in
lactated Ringer's [Hextend (HSLR)], was analyzed by thromboelastography (TEG). No changes were noted in the TEG profile of undiluted WB controls during the 6-hour period of use (P > 0.95). Dilution with HSNS and HSLR significantly impaired coagulation compared to both WB control and NS. Progressive dilution with NS impaired coagulation but to a lesser extent than
colloids (P < 0.01). Low dilution of blood with NS increased clot strength by 12% (not significant; P = 0.097). We conclude that WB containing
citrate obtained from healthy donors for TEG analysis yields reproducible data over a minimum of 6 hours. Either
hetastarch, when present at concentrations comparable to the manufacturer's maximum recommended dose of 20 mL/kg (equivalent to the high dilution used in these experiments), decreases clot tensile strength to levels associated with an increased risk of
bleeding. Substitution of
lactated Ringer's for NS in 6%
hetastarch appears to offer no advantage in avoiding
hemostatic compromise in an in vitro model.