The haemodilution enhanced onset of coagulation as measured by the thrombelastogram is transient.

BACKGROUND AND OBJECTIVE: Crystalloid haemodilution has been widely found to enhance coagulation onset, but the duration of this effect has never been documented. METHODS: Twelve healthy, consenting volunteers had a rapid infusion of 14 mL kg-1 of normal (0.9%) saline. Blood samples were taken, prior to (control), and immediately after (30 min) the rapid saline infusion was completed (30 min). They were then repeated at regular intervals up to 120 min. Haematocrit/platelet counts were taken to determine the degree of dilution and thrombelastograms, with and without platelet antagonists (ReoPro, Abciximab), were measured in all samples. Antithrombin levels were selectively measured. RESULTS: The haematocrit and platelet count showed a rapid dilutional decrease at 30 min (mean of -12.2% and -14.4%, respectively), with values returning towards baseline within 15 min after finishing the infusion. There was a significantly faster onset of coagulation (decrease in r-time) in the post-infusion sample (30 min) compared to control (P<0.05), again returning towards normal as the dilution effect was reversed. Similar thrombelastograms findings were evident in the plasma factor only group (platelets inhibited by ReoPro). Antithrombin levels changed in keeping with the haemodilution effect (P<0.0001). There was a linear relationship between antithrombin and thrombelastograms r-time (P=0.012). CONCLUSION: The faster onset of coagulation brought on by haemodilution return towards normal as the dilutional effect is reversed. This effect is mediated through plasma clotting factors. Of interest is the significant inverse correlation of the onset of coagulation increasing as the antithrombin levels decreased with dilution.

The effects of a one unit blood donation on auto-haemodilution and coagulation.

The effect of haemodilution on coagulation has been extensively investigated. We investigated auto-haemodilution following a 10% blood loss (480 ml) and its effect on coagulation. Ten healthy, unstarved volunteers were enrolled. One unit of blood was taken from each volunteer. Concurrently blood was taken from the opposite arm prior to and immediately after the blood donation, and at 1, 2, 4 and 6 hours. It was tested for thrombelastography, haematocrit and endorphins. There was a significant decrease in r-time from the control sample to the sample taken immediately post blood donation. This value returned to baseline at 1 hour post donation and did not change again. There were no other significant changes in thromboelastographic parameters. Fractional plasma noradrenaline changes were significantly raised at 1 hour post donation (P = 0.048), returning to baseline by 2 hours post donation. The haematocrit showed a rapid (approximately 4%) fall during donation followed by a slow, but progressive decrease over six hours, falling by a mean of 8.3% from pre-donation values. A state of relative hypercoagulability is found immediately after a rapid 10% loss in circulating blood volume. This may be related to the rapid immediate haemodilution. It is unlikely that the sympathetic response to blood loss plays a role. However, after the initial drop, slow restoration of circulating blood volume by autodilution takes six to eight hours, and is not associated with enhanced coagulation. Of interest is that a 10% blood loss in a healthy person does not require volume replacement.

Effects on coagulation of intravenous crystalloid or colloid in patients undergoing peripheral vascular surgery.

BACKGROUND: This study investigated whether haemodilution-enhanced coagulation can be demonstrated under regional anaesthesia, whether this occurs before surgery, and whether the fluid used influences the effect. METHODS: Patients were randomly allocated to receive either crystalloid or colloid intravenous fluid. An epidural was administered. Samples of venous blood were taken before fluid administration, after completion of the epidural and initial fluid load, during surgery before heparin, and after 24 h. Thrombelastograph analysis was performed, and full blood count, international normalised ratio, activated partial thromboplastin time, D-dimers and thrombin-antithrombin complex were measured. RESULTS: In the crystalloid group, enhanced coagulation compared with baseline was demonstrated after initial fluid load (mean (SD) r-time 10.1 (4.9) min; P < 0.033; k-time 3.5 (1.7) min; P < 0.01; alpha-angle 54.9 (13.9) degrees; P < 0.01) and before heparin administration (r-time 8.8 (3.9) min; P < 0.01; alpha-angle 54.9 (12.6) degrees; P < 0.02). There was no enhancement of coagulation in the colloid group. There were no changes from baseline after 24 h. CONCLUSIONS: This study confirms that the enhanced perioperative coagulation mechanism is related to dilution, rather than surgery, and is triggered by rapid crystalloid haemodilution. Consideration should be given to the use of colloid rather than crystalloid solutions for rapid fluid loading in vasculopathic patients undergoing surgery.

Haemodilution-induced enhancement of coagulation is attenuated in vitro by restoring antithrombin III to pre-dilution concentrations.

Moderate haemodilution enhances coagulability in vitro and in vivo as measured by thrombelastography (TEG). The mechanism has never been established. We have conducted an in vitro study to determine whether the effect can be moderated or prevented when the reduction in antithrombin III caused by dilution is prevented by supplementation. Blood from 20 volunteers was divided into four samples. One sample was not diluted and served as control (C). Another was diluted (by 20%) with saline (S). The third was diluted by 20% with saline plus two units of antithrombin (AT III) (SA). The fourth remained undiluted, with two units of added AT III (CA). Coagulation was measured in all four samples using the TEG. In a separate laboratory study, the levels of AT III were measured in control samples and compared with levels after 20% dilution, and 20% dilution with two units of AT III added to the diluent. Enhanced coagulation was demonstrated in saline-diluted samples (S) by shortening of r- and k-times, and increased alpha angle. In the SA samples, r-time shortening was prevented; k-time shortening and alpha-angle increase persisted, but to a reduced degree (difference from saline-only dilution P<0.051). There were no differences between samples C and CA. A predictable drop of AT III (24.2%) occurred with saline dilution, while AT III levels in the AT III/Saline group were similar to the undiluted control. Haemodilution-induced coagulation enhancement is attenuated, but not prevented, if AT III levels are maintained in the normal range. This is in keeping with the established concept of an antithrombin threshold preventing positive coagulation feedback into the intrinsic pathway.

Effect of 20% in vitro haemodilution with warmed buffered salt solution and cerebrospinal fluid on coagulation.

We have conducted an in vitro coagulation study consisting of two separate groups of 20 subjects using the thrombelastograph. In the first group, haemodilution was performed with a physiological balanced salt solution similar to plasma, with the exception of calcium, and buffered to a normal pH (Plasmalyte B) at 37 degrees C on blood obtained from consenting volunteers. In the second group, a protein-poor body fluid (cerebrospinal fluid (CSF)) obtained from parturient patients undergoing spinal anaesthesia for Caesarean section was used as the diluent. There were statistically significant differences between the warmed Plasmalyte B treated samples and their untreated controls for all variables measured by the thrombelastograph, except for maximum amplitude, and between the CSF treated samples and their untreated controls for all variables. We conclude that electrolyte and acid-base composition of the diluent fluid had no effect on the observation that crystalloid haemodilution produces hypercoagulability. The marked increase in coagulability produced by addition of CSF cannot be explained on a simple haemodilution basis and confirms previous suggestions of the presence of a procoagulant factor in CSF.

Pro-coagulant effect of in vitro haemodilution is not inhibited by aspirin.

We have conducted an in vitro coagulation study, using the thrombelastograph (TEG), to determine if the enhanced coagulability of whole blood after haemodilution with normal saline can still be demonstrated after administration of an antiplatelet agent. Aspirin inhibits the platelet-endothelial interaction that is part of the coagulation process. We investigated the role of aspirin in the phenomenon of haemodilution-induced coagulability to identify if the platelet-endothelial system is involved in the process. Previous work showed that the TEG is not altered by oral ingestion of aspirin. Blood from 20 volunteers was divided into two aliquots of 4 ml each. One sample was diluted by 20% by addition of 0.9% saline 1 ml while the other was not diluted and served as a control. Coagulation studies were performed using the TEG and enhanced coagulation was seen in the saline diluted samples. Subjects then received soluble aspirin 375 mg daily for 3 days, after which the tests were repeated. There was no difference in the control TEG values and saline enhancement of coagulation was preserved in all subjects after 3 days of aspirin administration. We conclude that aspirin had no effect on the observation that haemodilution with saline enhances the coagulability of whole blood.

In vivo investigation into the effects of haemodilution with hydroxyethyl starch (200/0.5) and normal saline on coagulation.

We have investigated the effects of haemodilution with either saline or hydroxyethyl starch (200/0.5) (HES) on blood coagulation in healthy volunteers in vivo. Standard haematological tests (packed cell volume (PCV), platelets, prothrombin time (PT), activated partial thromboplastin time (aPTT), fibrinogen, antithrombin III, bleeding time and platelet aggregation), and thrombelastography (TEG) were performed before and after administration of either 0.9% saline 1000 ml or HES 1000 ml i.v. over a 30-min period. Dilution of PCV and platelet concentrations as a result of volume load were 9% in the saline group and 19% in the HES group. Reductions in fibrinogen (18.6% and 28.8%) and antithrombin III (25.5% and 37.8%) were significantly greater than could be explained by haemodilution alone in both groups. Indices of platelet aggregation were significantly enhanced by saline haemodilution, but not by HES, which inhibited epinephrine-induced aggregation and prolonged bleeding time. TEG in the saline group showed significantly shortened r and k times (24% and 26%, respectively), and increased alpha angle (24%) and maximum amplitude (MA, 6%). HES haemodilution decreased MA (11%) but did not affect other TEG variables. We conclude that haemodilution of normal blood exerted a procoagulant effect, possibly by enhancement of thrombin formation. Circulating concentrations of antithrombin III were depleted more than could be explained by haemodilution alone, leading to a hypercoagulable state. This effect was offset by an antiplatelet action of HES, which was not seen with saline. The mechanism is unknown.

Haemodilution induces a hypercoagulable state.

It has been suggested that haemodilution with saline may increase whole blood coagulation. This study was conducted in two parts. First, we investigated the effect of in vitro dilution of blood with saline on whole blood coagulation as measured by the thrombelastogram (TEG). Blood (4 ml) was diluted with 0.9% saline 1 ml and coagulation compared with that of an undiluted control specimen obtained concurrently from the same subject. In the second part, the study was repeated using a modified gelatin colloidal solution (Haemaccel) as the diluent. The r time, k time and r + k time were decreased relative to control in both diluent groups. The alpha angles were increased compared with control in both groups while maximum amplitude was unchanged in the Haemaccel diluted group. We conclude that haemodilution per se increases the coagulability of whole blood in vitro, but that saline haemodilution has a more marked effect on final clot strength.