This study was designed to evaluate a new point-of-care test (HemoSTATUS) that assesses acceleration of kaolin-activated clotting time (ACT) by platelet activating factor (PAF) in patients undergoing cardiac surgery. Our specific objectives were to determine whether HemoSTATUS-derived measurements correlate with postoperative blood loss and identify patients at risk for excessive blood loss and to characterize the effect of desmopressin acetate (DDAVP) and/or platelet transfusion on these measurements.
Demographic, operative, blood loss and hematologic data were recorded in 150 patients. Two Hepcon instruments were used to analyze ACT values in the absence (channels 1 and 2: Ch1 and Ch2) and in the presence of increasing doses of PAF (1.25, 6.25, 12.5, and 150 nM) in channels 3-6 (Ch3-Ch6). Clot ratio (CR) values were calculated with the following formula for each respective PAF concentration: clot ratio = 1-(ACT/control ACT). These values also were expressed as percent of maximal (%M = clot ratio/0.51 x 100) using the mean CRCh6 (0.51) obtained in a reference population.
When compared with baseline clot ratios before anesthetic induction, a marked reduction in clot ratios was observed in both Ch5 and Ch6 after protamine administration, despite average platelet counts greater than 100 K/microliter. There was a high degree of correlation between clot ratio values and postoperative blood loss (cumulative chest tube drainage in the first 4 postoperative hours) with higher concentrations of PAF: CRCh6 (r = -0.80), %M of CRCh6 (r = -0.82), CRCh5 (r = -0.70), and %M of CRCh5 (r = -0.85). A significant (P < 0.01) improvement in clot ratios was observed with time after arrival in the intensive care unit in both Ch5 and Ch6, particularly in patients receiving DDAVP and/or platelets.
Activated clotting time-based clot ratio values correlate significantly with postoperative blood loss and detect recovery of PAF-accelerated coagulation after administration of DDAVP or platelet therapy. The HemoSTATUS assay may be useful in the identification of patients at risk for excessive blood loss and who could benefit from administration of DDAVP and/or platelet transfusion.
Key words: Coagulation: platelet count; platelet procoagulant function. Monitoring: coagulation. Surgery: cardiac.
Patients undergoing cardiac surgery with cardiopulmonary bypass (CPB) are at increased risk for excessive perioperative blood loss that requires transfusion of blood products. This risk is dependent on the type of procedure [1,2]and the duration of CPB, [3,4]in which mechanisms that mediate excessive activation of platelets, coagulation, and fibrinolysis may lead to consumption of platelets, labile coagulation factors, and excessive fibrinolysis, even in the presence of standard, high-dose heparin-induced anticoagulation. [2,3,5]Transient platelet dysfunction is considered to be the most common and important defect in hemostasis in the early postoperative period after CPB, [3,5-8]although several other abnormalities in hemostasis may contribute to abnormal bleeding, such as decreases in plasma coagulation factors. [2,9,10]Numerous platelet abnormalities have been described, including prolonged bleeding time, thrombocytopenia, [2,11]loss of platelet receptors for fibrinogen and von Willebrand factor, [12,13]and a variety of platelet or platelet activation-related abnormalities. [3,7,14,15]Because the management of excessive bleeding after CPB poses diagnostic and therapeutic challenges, use of rapidly performed screening tests has been recommended. Use of a transfusion algorithm approach, based on point-of-care measurement of whole blood prothrombin time (PT), activated partial thromboplastin time (APTT), and platelet count, has been shown to minimize blood loss, reduce blood product administration, and decrease operative time through targeted therapy for acquired hemostatic defects. However, one limitation of this approach is the lack of assessment of qualitative platelet disorders that might respond to pharmacologic interventions. Although the template bleeding time and other tests, such as thromboelastography, have been evaluated as possible point-of-care tests for the identification of platelet dysfunction and prediction of excessive blood loss, none have been shown to be sufficiently practical, rapid, easy, and useful in individual patients to gain wide acceptance.
The activated clotting time (ACT) is the most commonly used method to monitor the state of heparin-induced anticoagulation during CPB, based on its ease of use and early studies that demonstrated a reduction in postoperative bleeding when the ACT was used to monitor heparin and protamine therapy. Recently, a new whole blood point-of-care test based on ACT measurements has been developed. The HemoSTATUS (Medtronic HemoTec, Parker, CO) assay evaluates the acceleration of coagulation platelet activating factor (PAF) using the kaolin-activated ACT. Our primary objectives in this study were to determine whether Hemo-STATUS-derived measurements correlate with postoperative blood loss, identify patients at risk for excessive blood loss when compared with routine coagulation tests, and characterize the effect of DDAVP (Rhone-Poulenc Rorer Pharmaceuticals, Collegeville, PA) or platelet transfusion on HemoSTATUS measurements. Secondary objectives were to evaluate the time course of HemoSTATUS-derived measurements during CPB, determine whether these measurements are affected by preoperative antithrombotic medications, and assess cartridge/instrument measurement reproducibility.
Materials and Methods
Intraoperative Management and Data Collection
Blood was obtained from 150 consecutive adult patients undergoing cardiac surgery that required CPB at Barnes Hospital/Washington University Medical Center, St. Louis, Missouri, after approval of the study by the Institutional Human Studies Committee and obtaining of informed consent. Patients enrolled in this study were managed by one of six cardiac surgeons and were anesthetized with an opioid-based technique supplemented with inhalational anesthetic agents, muscle relaxants, and benzodiazepines. Cardiopulmonary bypass was accomplished with a Biomedicus centripetal pump (Medtronic, Minneapolis, MN) and a Cobe membrane oxygenator (Cobe Biomedical, Arvada, CO). The CPB system was routinely primed with 1.5-2 liters of Plasmalyte solution (Baxter Health Care, Irvine, CA), 50 milliequivalents of sodium bicarbonate (NaHCO3), 25 g mannitol, and 5-10,000 units porcine heparin. During cardioplegia, systemic hypothermia was maintained at 28 degrees Celsius. Systemic anticoagulation for CPB was accomplished with porcine heparin according to a previously published protocol based on measurements of ACT and whole blood heparin concentration. Transfusion of red cells was not standardized but was based on a number of patient-specific factors (e.g., age, myocardial function, other coexisting disorders, etc.). After rewarming the patient to 37 degrees Celsius, extracorporeal circulation was discontinued, heparin was neutralized with a protamine dose based on on-site measurements of residual whole blood heparin concentration, and patients were observed for excessive microvascular bleeding. Excessive microvascular bleeding was defined as diffuse bleeding from the surgical site without an identifiable surgical source, and this diagnosis was made by the surgical team members, who were unaware of results from the HemoSTATUS assay. Treatment of excessive intraoperative bleeding after CPB was standardized using a previously described algorithm based on on-site tests for whole blood PT, APTT, and platelet count.
The following demographic and perioperative data were collected for each patient: age, gender, height, weight, body surface area, preoperative antithrombotic medications (e.g., aspirin, heparin, warfarin) administered within 1 week before surgery, history, of a previous cardiac surgical procedure, operative procedure, duration of aortic cross-clamp, duration of extracorporeal circulation, doses of heparin and protamine administered, administration of DDAVP, core body temperature after arrival in the intensive care unit (ICU) initially and at postoperative hours 3, 6, and 12, hourly mediastinal chest tube drainage during the first 24 postoperative hours after arrival in the ICU, exploration for excessive bleeding and source of bleeding, blood components transfused in the perioperative interval, and hematologic/hemostasis measurements. To compare blood loss and HemoSTATUS results among patients who received intraoperative therapy (platelet transfusion or DDAVP) for microvascular bleeding and those who did not require such therapy, a post hoc analysis of the data involved segregation of patients into the following three subgroups: patients with excessive bleeding who received DDAVP, patients with excessive bleeding who received platelet transfusion, and patients who did not receive either DDAVP or platelets. To examine the potential effect of preoperative use of antithrombotic medications, clot ratio measurements were compared among patients who received either aspirin, warfarin, or heparin and a reference group that received none of these medications.
Laboratory-based hematologic tests, including complete blood count with platelet count, PT, APTT, and bleeding time, were performed preoperatively and postoperatively on arrival in the cardiac intensive care unit in all patients. The PT was performed on citrated plasma using Dade rabbit brain thromboplastin (Baxter Diagnostics, Deerfield, Ill) and an Electra 1000C (Medical Laboratory Automation, Pleasantville, NY); and the APTT was performed on citrated plasma using the Electra 1000C and Actin CaCl/Dade Ci-crol reagent (Baxter Diagnostics, Deerfield, Ill). Bleeding time was performed with the Simplate technique (Simplate IIR, Organon Teknika, Durham, NC). Platelet counts were performed electronically with the Coulter S + 4, an automated Coulter hemocytometer (Coulter Electronics, Hialeah, FL).
On-site, Whole Blood Hematologic Assays.
Single blood specimens obtained from either radial and/or femoral intraarterial catheters after removal of six dead space blood volumes or from the CPB arterial cannula were used for on-site, whole blood assays. Heparin and heparin anticoagulant effect were monitored with whole blood heparin concentration and kaolin ACT using the Hepcon instrument (Medtronic HemoTec, Englewood, CO). Whole blood PT, APTT, and platelet count were performed on specimens obtained from patients with excessive bleeding after CPB to guide specific therapy. Whole blood PT and APTT were determined using the battery-powered portable Biotrack 512 (Boehringer Mannheim, Indianapolis, IN) that uses disposable plastic reagent cartridges, and whole blood complete blood count with platelet counts were measured electronically with the Coulter T540 instrument during the intraoperative period.
Two Hepcon instruments used a six channel PAF-containing cartridge (HemoSTATUS assay) to measure PAF-accelerated coagulation in duplicate. The Hemo-STATUS test measures the shortening of the kaolin-activated ACT induced by PAF (1-O-Alkyl-2-Acetyl-sn-Glyceryl-3-Phosphorylcholine). Software adjustments were made to the Hepcon instruments by Medtronic HemoTec so that plunger lift and drop rates were preset, to minimize mechanical activation of platelets. To increase the sensitivity of the method, ACTs were prolonged, by adding either 3 or 4 U/ml heparin to each of 6 channels, with the exception of the cartridges used during the CPB interval, which did not contain heparin. Channels 1 and 2 (Ch1, Ch2) were devoid of PAF (control ACT), whereas channels 3-6 (Ch3-Ch6) contained increasing doses of PAF (1.25, 6.25, 12.5, and 150 nM). Platelet activating factor-accelerated coagulation was expressed as a clot ratio value and was calculated with the following formula for each PAF concentration: clot ratio = 1 - (ACT/control ACT). Clot ratio values also were expressed as percent of maximal (%M) and were calculated using the mean clot ratio obtained with channel 6 (0.51), obtained from a normal reference population (22 volunteers, Medtronic Hemo-Tec) using the following formula: %M = CR/0.51 x 100. Clot ratio values (mean +/- SD) in this series of normal volunteers were as follows: CRCh3, 0.20 +/- 0.10; CRCh4, 0.42 +/- 0.07; CRCh5, 0.48 +/- 0.06; CRCh6, 0.51 +/- 0.06. Measurements of PAF-accelerated coagulation were obtained in duplicate during the following time periods: period 1 (baseline), before induction of general anesthesia; period 2 (CPB), shortly before discontinuation of CPB in a subset of patients (n = 125); period 3 (post-CPB), after the neutralization of heparin with protamine; period 4 (ICU), after arrival in the ICU; and after administration of either DDAVP or platelet transfusion in patients with excessive postoperative bleeding. All test results were expressed as the mean of duplicate measurements.
Student's paired or unpaired t test and one-way analysis of variance with Bonferroni correction were used to compare demographic, hematologic, operative interval, transfusion, and postoperative blood loss variables that were expressed as mean values for each of these groups. Chi-squared and Fisher's exact analyses were used to compare variables that were expressed as percentages. Ordinary (nonweighted) least squares linear regression was used to estimate the relation between cumulative blood loss and multiple demographic, operative, and hemostatic measurements, and P < 0.05 was considered statistically significant. Variables significantly associated with cumulative blood loss with univariate analysis were subjected to stepwise, multivariate linear regression analysis using a backward elimination procedure. At each stage of the elimination procedure, the variable with the largest current P value was eliminated until all remaining variables were or became statistically significant (P < 0.05). Bias analysis was used to test agreement between duplicate HemoSTATUS-derived measurements obtained with different cartridges, using two Hepcon instruments.
Bayes' theorem was used to evaluate the diagnostic performance of assay systems and requires the definition of two variables: 1) a disease state and 2) an abnormal test result. The disease state in our analysis was postoperative blood loss, and this was defined as cumulative mediastinal chest tube drainage greater than: 200 ml in 4 h, 300 ml in 4 h, or 400 ml in 4 h. Test results were positive when HemoSTATUS-derived clot ratio values were equal to or greater than a defined %M value. For each blood loss state and positive test result, Bayes' theorem was used to calculate predictive indices. Sensitivity was defined as the percentage of patients with excessive blood loss that have a positive test result. Specificity was defined as the percentage of patients without excessive blood loss that have a negative test result. Positive predictive value was defined as the percentage of patients with a positive test result that exceeded the threshold of excessive blood loss. Negative predictive value was defined as the percentage of patients with a negative test result that did not exceed the threshold. Accuracy was defined as the proportion of correctly predicted cases, positive or negative. Logistic regression analysis was used to generate receiver operating characteristic curves to characterize the predictive values of measurements throughout the range of %M values.
Relation Between HemoSTATUS Clot Ratio Values and Postoperative Blood Loss in Patients Not Receiving Hemostatic Therapy Perioperatively: Use of Univariate and Multivariate Models.
Of 150 patients enrolled, 12 patients who required reexploration and who had a surgical source of bleeding identified were excluded from further analysis. The relation between clot ratio values and postoperative blood loss was examined in all patients and in the series of patients who did not receive either DDAVP or platelets during the perioperative interval. The demographic, operative, and blood loss data for these patients are summarized in Table 1. Correlations obtained between clot ratio values and cumulative chest tube drainage in all patients (n = 138) and in patients not receiving hemostatic therapy perioperatively (n = 83) are summarized in Table 2. Using univariate analysis, the highest correlations (r = -0.85, -0.82) were obtained between mediastinal chest tube drainage in the first 4 postoperative hours and postoperative clot ratio values (%M) measured after arrival to the ICU in channels 5 (CTD4 = -7.2 CRCh5 + 777) and 6 (CTD4 = -7.3 CRCh6 + 947), respectively. The data from channels 5 and 6 are illustrated in Figure 1and Figure 2, respectively. To determine the best predictive model for chest tube drainage in the first 4 h after surgery, all demographic, operative, and hemostatic measurements were included in a multivariate linear regression model; this analysis was repeated using either clot ratio measurements or %M values for each of the channels. Mediastinal chest tube drainage in the first 4 postoperative hours was related linearly to both %M values for CRCh5 (partial correlation coefficient = -0.73) and postoperative bleeding time measurements (partial correlation coefficient = 0.32); this linear relation (CTD4 = -6.0 CRCh5 + 15.8 bleeding time + 602, r2= 0.75) is also illustrated in Figure 3.
Prediction of Excessive Blood Loss in Patients Not Receiving Hemostatic Therapy Perioperatively.
The predictive indices for estimation of excessive postoperative chest tube drainage at each of three defined criteria (> 200, 300, or 400 ml in the first 4 postoperative hours) by %M values for CRCh5 and CRCh6 are listed in Table 3. The percentage of patients with mediastinal chest tube drainage greater than 200 ml, 300 ml, and 400 ml in 4 h was 57%, 40%, and 21%, respectively. Excessive blood loss using all three blood loss criteria (> 200 ml, 300 ml, or 400 ml in the first 4 postoperative hours) was predicted by both HemoSTATUS-derived clot ratio values and bleeding time, whereas PT, APTT, and platelet count had no predictive value. The area under the receiver operating characteristic curve for > 200 ml in the first 4 postoperative hours for each of the assays are as follows: %M clot ratio values in channel 5 (0.95), bleeding time (0.87), PT (0.58), APTT (0.48), and platelet count (0.50). Similarly, the greatest areas under the receiver operating characteristic curve for excessive blood loss were obtained with HemoSTATUS-derived clot ratio values for each respective criterion (> 300 ml: %M CRCh5; > 400: %M CRCh6), as shown in Figure 4and Figure 5.
Changes in HemoSTATUS-derived Clot Ratio Values in Response to DDAVP or Platelet Transfusion.
(Figure 6) illustrates the changes in clot ratio values in response to platelet transfusion; clot ratios were obtained before and after administration of either 6 units of random donor or one unit of single donor pheresis platelets (n = 40). Platelet counts increased significantly (P = 0.0002) after administration of platelets (preadministration, 86 +/- 38; postadministration, 120 +/- 42). Figure 7illustrates the changes in clot ratio values in response to administration of 0.3-0.4 micro gram/kg DDAVP (n = 25). Platelet count measurements before administration of DDAVP were similar (P = 0.65) to those obtained after administration of DDAVP (preadministration, 126 +/- 35; postadministration, 119 +/- 33).
Blood Loss and HemoSTATUS-derived Clot Ratio Values in Patients with and without Excessive Intraoperative Microvascular Bleeding.
Post hoc analysis of the relation between blood loss and hemostatic measurements was performed in three groups of patients: patients with excessive intraoperative bleeding who received either DDAVP (n = 19) or platelet transfusion (n = 25) intraoperatively and patients without excessive bleeding who did not receive such therapy during the intraoperative period (n = 83). Table 1summarizes demographic, operative, and blood loss data, and Table 4details the hematologic data obtained for each of these groups. Impaired PAF-accelerated coagulation was evident in patients with intraoperative evidence of excessive microvascular bleeding who subsequently were treated with either platelet transfusion or DDAVP when compared with patients who did not have excessive bleeding and who did not require hemostatic therapy (DDAVP or platelets) intraoperatively (Table 4). Figure 8illustrates %M values of CRCh6 values in patients who received DDAVP and/or platelet transfusion intraoperatively (n = 49), as compared with patients who did not require perioperative treatment (n = 83) at four peri-CPB time points. Mean clot ratio (%M) values for channel 6 were significantly lower during and after CPB in both groups of patients when compared with baseline values, and decreased to a greater degree, after protamine, in patients who required treatment intraoperatively (DDAVP or platelets). Clot ratio values were normal in patients who required treatment intraoperatively on arrival in the ICU when compared with lower values obtained from patients who did not require therapy.
Effect of Preoperative Antithrombotic Medications on Clot Ratio Values and Between-Cartridge/Instrument Reproducibility.
Preoperative clot ratio values for patients who did not receive aspirin, warfarin, or heparin (n = 28) for each respective PAF concentration (channels 3-6) were as follows: 1.25 nM: 0.12 +/- 0.08; 6.25 nM: 0.34 +/- 0.11; 12.5 nM: 0.45 +/- 0.11; 150 nM: 0.56 +/- 0.09. Warfarin use (n = 7) was detected in channel 3 (1.25 nM), as demonstrated by significantly (P = 0.03) lower clot ratios (0.04 +/- 0.04) when compared with patients who did not receive antithrombotic medications (Table 5). Aspirin use (n = 47) was detected in channel 4 (6.25 nM), as demonstrated by significantly lower (P = 0.04) clot ratios (0.28 +/- 0.09) when compared with patients who did not receive antithrombotic medications (Table 5).
Paired clot ratio measurements (n = 649) were similar in all channels, as demonstrated by paired t test; Ch3 (P = 0.28), Ch4 (P = 0.26), Ch5 (P = 0.84), and Ch6 (P = 0.88). Cartridge and instrument reproducibility was evaluated using bias analysis (bias +/- SD), and the following results were obtained for each respective channel: CRCh3 (-0.005 +/- 0.11), CRCh4 (-0.004 +/- 0.09), CRCh5 (0.0005 +/- 0.07), and CRCh6 (-0.0004 +/- 0.06).
Excessive microvascular bleeding with CPB appears to be strongly associated with complex, CPB-related hemostatic abnormalities, [2-5]and, in particular, platelet dysfunction, which may be induced by adherence of platelets to the CPB surfaces and platelet activation/degranulation/desensitization, [3,7,8,12,14,15]as well as the potential effects of heparin and/or hypothermia. The findings presented in the current study demonstrate that a simple, rapid test of PAF-accelerated coagulation (HemoSTATUS) may be useful in predicting postoperative mediastinal chest tube drainage. This is in agreement with the results of previous studies that demonstrated a relation between excessive post-CPB bleeding and impaired platelet aggregation, [11,18]as assessed by laboratory-based turbidometric tests in platelet-rich plasma that are complex and time-consuming. Of numerous hemostatic measurements included in this study, clot ratio values and bleeding time were the only variables that correlated with postoperative chest tube drainage. Predictive indices (sensitivity, specificity, etc.) were derived using Bayes' theorem to assess the diagnostic performance of the HemoSTATUS assay in identifying patients with excessive blood loss. Our data indicate that optimal diagnostic performance (greatest accuracy values) for clot channel 5 occurs at values 75, 60, and 50, and for channel 6, occurs at values 95, 90, and 75 for each respective blood loss criterion (200, 300, 400 ml). The values summarized in Table 3allows the clinician to evaluate the predictive indices at a specific measurement value. Using positive predictive values as an example, if a clot ratio measurement (%M) of 60 or less is obtained in channel 5, then 100%, 92%, and 58% of patients will have greater than 200, 300, and 400 ml blood loss in the first 4 postoperative hours, respectively. The diagnostic performance of measurements of PAF-accelerated coagulation in channels 5 and 6 with respect to identification of patients at risk for excessive blood loss for each criterion is graphically presented throughout the range of values in Figure 4and Figure 5.
Our findings, which demonstrate a direct relation between bleeding time measurements and mediastinal chest tube drainage, differ from previous evaluations that examined this issue using preoperative measurements of bleeding time. [19,20]Although it was suggested that routine preoperative coagulation tests can predict blood loss after cardiac surgery, others have not confirmed this with coagulation tests obtained either pre- or postoperatively. In this study, we found no relation between routine pre- or postoperative coagulation measurements and blood loss. However, these findings must be interpreted with caution, because the patients with intraoperative bleeding and abnormal coagulation results received hemostatic blood components that may have reduced their subsequent blood loss. Thromboelastography was shown, in some studies, to predict the risk of postoperative bleeding. [23,24]However, others failed to confirm the usefulness of the thromboelastography in predicting either intraoperative or postoperative bleeding. A point-of-care assay that measures platelet-dependent clot retraction was shown to correlate with blood loss after cardiac surgery in a small series. 
Another major objective of assays that evaluate hemostasis is to specifically assess platelet-related abnormalities in patients that display clinical evidence of excessive bleeding, because platelet dysfunction probably represents the most important hemostatic derangement related to use of extracorporeal circulation. Although useful methods for rapid, on-site evaluation of platelet concentration have been developed, rapid, user-friendly tests for evaluation of platelet function are not available. Our findings that PAF-accelerated coagulation, as assessed with the HemoSTATUS assay, is significantly enhanced by platelet transfusion (Figure 6) or DDAVP (Figure 7) and is decreased in patients who receive aspirin preoperatively using low in vitro concentrations of PAF in channel 4, seem to support the concept that the shortening of the ACT by PAF is largely due to direct or indirect effects of PAF on the availability of procoagulant activity on platelets. The effect of aspirin on low concentrations of PAF in the HemoSTATUS test is consistent with what has been demonstrated using laboratory-based, turbometric measurements of platelet aggregation-higher agonist (collagen) concentrations overcome the mild inhibitory effects of aspirin. In patients who did not receive intraoperative hemostatic therapy, the partial recovery of PAF-accelerated coagulation on arrival to the ICU is consistent with previous observations that platelet abnormalities may disappear within a few hours after CPB. In contrast, the complete recovery of clot ratio values on arrival to the ICU in patients who received either DDAVP or platelets suggests that these interventions were efficacious in restoring PAF-accelerated coagulation (Figure 8). Previous suggestions that ACT measurements can potentially be affected by activation or depression of platelet function were confirmed in a recent trial that demonstrated that ACT measurements are prolonged with platelet glycoprotein IIb/IIIa integrin blockade. Although any strong platelet activator could potentially be used with an ACT-based system, PAF, a biologically active phospholipid that was shown to induce platelet aggregation in an early rabbit model, was chosen for this test mainly because of its stability. Platelet activating factor is a potent, lipid mediator that also stimulates aggregation and degranulation of neutrophils, chemotaxis, chemokinesis, superoxide formation, protein phosphorylation, glycogenolysis, production of tumor necrosis factor, and numerous metabolites. [31,32]Therefore, abnormalities in PAF-accelerated whole blood coagulation may reflect not only platelet contributions but also leukocyte contributions. Whether and to what extent the shortening of ACT measurements by PAF is related to its interaction with cells other than platelets is uncertain, and requires further investigation. In addition, another shortcoming of the current study involves our lack of comparison of HemoSTATUS test results to established laboratory tests of platelet function, which should be the subject of future investigations.
Desmopressin acetate was shown to be efficacious in patients with von Willebrand disease and mild hemophilia A, and with uremia, cirrhosis, and certain cardiac surgical patients such as those who required prolonged use of CPB and patients receiving platelet-inhibiting drugs. Although prophylactic administration of DDAVP was not shown to be clinically beneficial, certain cardiac surgical patients identifiable by thromboelastography may benefit. The findings, in our study, of a significant increase in clot ratio values that reflect enhanced recovery of PAF-accelerated coagulation after administration of DDAVP (Figure 7), together with the findings of significantly shortened bleeding times after arrival in the ICU and markedly reduced blood loss in the series of patients that received DDAVP (Table 1), are consistent with the concept that DDAVP can improve platelet function in cardiac surgical patients and that the new, on-site ACT-based test can detect the beneficial effects of DDAVP in this setting. However, because our findings were derived from a post hoc analysis of the data and a number of confounding variables were not controlled, the findings must be confirmed by further, prospective, controlled studies. Patients in our study received DDAVP when point-of-care coagulation results were not grossly abnormal and a qualitative platelet defect was suspected (Table 4) because physicians managing these patients were blinded to results from the HemoSTATUS assay. The mechanisms by which DDAVP might shorten the ACT in the presence of PAF are uncertain. Administration of DDAVP may result in an increase in plasma levels of von Willebrand factor related to receptor-mediated endothelial cell release. Desmopressin acetate was shown to shorten bleeding time results and increase ristocetin-induced aggregation, but other studies could not demonstrate an effect of DDAVP on platelet adhesion. Although increases in von Willebrand factor can enhance platelet subendothelial and platelet-platelet interactions, other mechanisms may be operative, such as DDAVP-mediated expression of glycoprotein lb receptors, generation of platelet microparticles/enhanced procoagulant activity, increased platelet von Willebrand factor release, and expression of P-selectin by endothelial cells. 
Activated clotting time-based clot ratio values correlate strongly with postoperative blood loss and detect amelioration of depressed PAF-accelerated coagulation subsequent to DDAVP or platelet therapy. Therefore, the HemoSTATUS assay may be useful in the identification of patients at risk for excessive blood loss and who may benefit from administration of DDAVP or platelets. Further studies are needed to compare clot ratios with platelet aggregation measurements, to assess the differential contribution of platelets versus white cells on PAF-accelerated coagulation, and to more clearly define the role of the HemoSTATUS test in the management of patients with CPB-related platelet dysfunction and excessive bleeding,