The Effects Of Crystalloid Solutions On The Human Blood Coagulation System

ABSTRACT

Crystalloid solutions are used in clinical practice for resuscitation and correction of electrolyte imbalances. However, up to 25% of individuals may develop dysfunctions of haemostasis following fluid infusions, complicating resuscitation and outcome. Studies on the effects of crystalloids solutions on human blood coagulation have produced conflicting results: either suggesting procoagulant effects or impaired coagulation. The mechanisms for these discrepant results remain unclear. However, the role of solute composition remains largely unexplored. The main objective of the study was to determine the effects of crystalloid solutions on human coagulation system under conditions of varying solute type and concentrations. Specifically, the study investigated the effects of ionic crystalloids solutions containing Na+ (sodium gluconate and sodium chloride), Cl- (sodium chloride and choline chloride), and non ionic solutions such as mannitol and dextrose, in vitro. The blood samples were obtained from consenting healthy adult human blood donors. The laboratory methods were routine coagulation tests, plasma thrombin generation, and Thromboelastography and Light transmission platelet aggregation. Data analysis was done using STATA version 11.0, Texas, USA. Analysis of variance (ANOVA) and Kruskal Wallis equality of population tests were applied to examine for differences between groups. In routine tests, undiluted control had median prothrombin time 16.1 secs (IQR 15.3-17.1; N=17); median INR 1.2 (IQR 0.99-1.32; N=14); mean activated prothrombin time 33.98secs (SE 0.94; 95% CI 31.94-36.02; N=14) and median fibrinogen level 283.9 mg/dl (IQR 249.9-317.8; N=12). These tests showed significant positive linear increase with NaCl concentration except fibrinogen concentration which decreased with NaCl concentration. Comparison with other crystalloid solutions revealed crystalloid ionic strength had the most influence, and presence and concentration of chloride ions was most significant. In thrombin generation, mannitol had no influence on the thrombin generation parameters either in the extrinsic or intrinsic activation. However, NaCl had most significant effect in intrinsic activation, where there was progressive lengthening of lag time with each concentration though curves remaining qualitatively similar. The median thrombelastography values for undiluted control were: R 12.9(IQR 11.0-17.4); K 5.1 (IQR 3.5-7.0); alpha angle 32.4 (IQR 27.7-46.9); Maximum amplitude 55.8 (IQR 54.3-57). Thrombelastography median parameters had curvilinear relationship with NaCl solutions such that R&K trough inflection point around 0.3 M, as well as alpha angle and Maximum amplitude plateau. Comparison with other ionic crystalloid solutions were qualitatively similar except that from 1200 mOsm, the order of divergence was choline chloride>NaCl> sodium gluconate. Mannitol and dextrose dose dependently decreased thromboelastography alpha and maximum amplitude without much change in R&K. Citrate anticoagulated whole blood samples showed patterns of enhanced coagulation in comparison to neat non-citrate anticoagulated samples. Platelet aggregation was inhibited dose dependently by all the crystalloid solutions irrespective of solute type indicating that platelet activity was sensitive to osmolality rather than ionic strength. This study concludes that the effects of crystalloid solutions on human blood coagulation are dependent on solute content and concentration. Further, increase in chloride concentrations beyond the normal physiological range impairs blood coagulation. It is therefore recommended that the concentrations of solute ions in resuscitation fluids should be within normal plasma levels.