Anesth Analg 2004;98:578-584
© 2004 International Anesthesia Research Society
doi: 10.1213/01.ANE.0000100682.84799.E8
CARDIOVASCULAR ANESTHESIA
Prophylactic Treatment with Desmopressin Does Not Reduce Postoperative Bleeding After Coronary Surgery in Patients Treated with Aspirin Before Surgery
Hilde Pleym, MD*,
Roar Stenseth, MD PhD*,
Alexander Wahba, MD PhD
,
Lise Bjella, MD*,
Arve Tromsdal, MD,
Asbjørn Karevold, MD, and
Ola Dale, MD PhD*,
Departments of *Anesthesiology and
Cardiothoracic Surgery, St. Olav University Hospital; and
Department of Circulation and Medical Imaging, Faculty of Medicine, Norwegian University of Science and Technology, Trondheim, Norway
Address correspondence and reprint requests to Hilde Pleym, MD, St. Elisabeth Department of Cardiothoracic Surgery, Hans Nissens gate 3, N-7018 Trondheim, Norway. Address e-mail to hilde.pleym{at}stolav.no
 |
Abstract
|
|---|
The synthetic vasopressin analog desmopressin has hemostatic properties and may reduce postoperative bleeding after coronary artery bypass grafting (CABG). A study on the effects of recent aspirin ingestion on platelet function in cardiac surgery showed a greater impairment of platelet function in patients treated with aspirin <2 days before the operation. We evaluated the effects of desmopressin on postoperative bleeding in CABG patients who were treated with aspirin 75 or 160 mg until the day before surgery. The study was a prospective, randomized, double-blinded, placebo-controlled, parallel group trial. One-hundred patients were included and divided into two groups. One group received desmopressin 0.3 µg/kg and the other received placebo (0.9% NaCl) after the neutralization of heparin with protamine sulfate. Postoperative blood loss was recorded for 16 h. The mean (SD) bleeding was 606 (237) mL in the desmopressin group and 601 (301) mL in the placebo group (P = 0.93), representing no significant difference (95% confidence interval, -107 to 117 mL). We conclude that desmopressin does not reduce postoperative bleeding in CABG patients treated with aspirin until the day before surgery.
IMPLICATIONS: Continuation of aspirin until the day before coronary artery bypass grafting may increase postoperative bleeding. The administration of desmopressin to these patients after the neutralization of heparin with protamine sulfate does not reduce postoperative bleeding.
|
Introduction
|
|---|
Patients with coronary artery disease are routinely treated with antiplatelet medication, most often aspirin. Unstable coronary patients are often taking aspirin until the day before surgery, and this is also the case for an increasing number of patients with stable angina. Investigations on the effect of aspirin treatment on postoperative bleeding after cardiac surgery have shown that patients treated with aspirin until surgery have increased postoperative bleeding (1), and also an increased need for transfusions of blood products (1,2).
The synthetic vasopressin analog desmopressin has hemostatic properties (3,4). Several studies have documented the ability of desmopressin to reduce hemorrhage in a variety of congenital and acquired platelet disorders, as well as after aspirin ingestion (5–7). In these situations, desmopressin is considered to act by increasing the concentration of von Willebrand factor, an important mediator of platelet adhesion. A number of investigators have studied whether the prophylactic use of desmopressin reduces postoperative bleeding after cardiac surgery (8–14). Salzman et al. (8) found, in a double-blinded, placebo-controlled study, that desmopressin significantly reduced blood loss in patients undergoing various cardiac operations. Patients with uncomplicated primary coronary artery bypass grafting (CABG) were not included in this study. Other investigators have studied the effect of desmopressin on bleeding and transfusion requirements after routine primary CABG, and after valvular heart operations (9–15). No effect of desmopressin on bleeding or transfusion requirements was found in investigations in which few or no patients had been pretreated with aspirin (9–12). However, in studies in which all patients had been treated with aspirin within the 7 or 5 days before surgery (13–15), desmopressin reduced the postoperative blood loss (13,14) and the transfusion requirements (14,15) compared with placebo.
Gibbs et al. (16) found a greater impairment of platelet function in patients treated with aspirin <2 days before the operation compared with patients who had a longer preoperative aspirin-free interval. In this study, we therefore investigated the effect of prophylactic treatment with desmopressin on postoperative bleeding in a group of patients treated with aspirin until the day before surgery.
|
Methods
|
|---|
The study was a prospective, randomized, double-blinded, placebo-controlled, parallel group trial. After approval of the Regional Committee for Medical Research Ethics, Health Region IV, Norway, and the Norwegian Medicines Control Authority, 100 patients with stable angina pectoris, scheduled for first-time CABG, were included in the study after giving written informed consent. All patients were asked to continue their aspirin medication until the day before surgery. Patients were excluded if they were receiving treatment with heparin or low-molecular-weight heparin, oral anticoagulants, nonsteroidal antiinflammatory drugs, or other platelet inhibitors.
The Unit for Applied Clinical Research at the Norwegian University of Science and Technology randomized the patients into two groups by means of a computer program, one group receiving desmopressin and the other group receiving placebo. Desmopressin 15 µg/mL or placebo was prepared at the hospital pharmacy and delivered in identical 20-mL syringes. The desmopressin group received desmopressin 0.3 µg/kg, and the placebo group received a corresponding volume of a 0.9% sodium chloride solution. The injections were given over 10 min at the end of cardiopulmonary bypass (CPB), immediately after the administration of protamine sulfate to neutralize heparin.
A standard anesthetic technique with morphine 0.12 mg/kg and scopolamine 0.0048 mg/kg for premedication, and fentanyl 4 µg/kg, thiopental 2–3 mg/kg, diazepam 2.5–10 mg, and pancuronium 0.08 mg/kg for induction was given. Anesthesia was maintained with nitrous oxide 30%–50% (before CPB), isoflurane, and small additional doses of fentanyl as required (total dose approximately 10 µg/kg). During CPB, anesthesia was maintained with midazolam 1 mg every 15 min. Hypertension was treated with glyceryltrinitrate 0.5–5 µg · kg-1 · min-1 before and after CPB and in the intensive care unit (ICU), and with sodium nitroprusside 0.5–5 µg · kg-1 · min-1 during CPB. Hypotension was treated with incremental doses of ephedrine 2.5–5 mg or phenylephrine 0.1–0.2 mg. Before CPB, heparin 300 U/kg (Leo, Copenhagen, Denmark) was given through a central venous catheter to achieve a kaolin activated clotting time (ACT) (Medtronic Blood Management, Parker, CO) of >480 s. The ACT was measured in duplicate, and the mean value was used. Additional heparin was given when needed to keep the ACT above the target. During CPB the ACT was monitored every 20 min. The perfusion circuit was primed with 1800 mL of Ringer’s acetate solution to which 7500 U of heparin was added. A membrane oxygenator without heparin coating was used. Cold antegrade hyperkalemic crystalloid or blood cardioplegia and moderate hypothermia to 34°C were used during CPB. Pump flow was 2.4 L/min/m2 during normothermia and 2.0–2.4 L/min/m2 at 34°C. The patient was warmed to a rectal temperature of at least 36°C before termination of CPB. After CPB, protamine sulfate 1 mg for every 100 U of previously administered heparin (prime heparin not included) was given to achieve an ACT within 10% of the baseline value. Additional doses of protamine were given when necessary.
At the end of surgery the patients were transferred to the ICU. All patients were sedated with incremental doses of midazolam 1–2 mg until they were considered ready for tracheal extubation. Patients were considered ready for tracheal extubation when they had a PaO2 of >9.0 kPa with a fraction of inspired oxygen <0.5, a PaCO2 <6.0 kPa, a pH >7.25, a rectal temperature >36.5°C, a urine output of >0.5 mL · kg-1 · h-1, a stable circulation, and postoperative bleeding of <200 mL/h. When these criteria were fulfilled the sedation was stopped, and when the patients were awake and responded to verbal commands, they were tracheally extubated.
No attempts were made to measure blood loss during the operation. Cardiotomy suction was used while the patient was fully anticoagulated, and the blood was returned to the patient without centrifugation. The time from termination of CPB to skin closure was recorded as a measure for the time spent on surgical hemostasis during the final part of the operation.
The primary study outcome was postoperative blood loss. From the time of arrival in the ICU, postoperative blood loss from the mediastinal and pleural drains was recorded for 16 h. The blood was collected in a sterile cardiotomy reservoir (Cardiotomy Resevoir, filtered; Sorin Biomedica, UK) and autotransfused to the patients for up to 8 h. Incidents of reexploration for bleeding were recorded. The criteria for transfusion of packed red cells were a hematocrit value <20% during CPB and <25% in the postoperative phase. Postoperative treatment with fresh frozen plasma, platelets, tranexamic acid, and desmopressin was given at the discretion of the attending physician when he or she considered this to be appropriate for the treatment of increased postoperative bleeding. Tranexamic acid was given when the postoperative bleeding exceeded 200 mL during the first postoperative hour or if bleeding of >150 mL/h persisted for >3 h. Desmopressin and transfusions of fresh frozen plasma or platelets were given to patients with persistent postoperative bleeding of >200 mL/h. The attending physician was at this point still blinded as to whether the patient had received desmopressin or placebo, as the randomization code for all patients was first disclosed 1 mo after the inclusion of the last patient. Transfusions and doses of tranexamic acid and desmopressin were recorded for the first 16 h and for the whole hospital stay. Mean postoperative hospital stay after CABG at our institution is 6.7 days.
Preoperative hemoglobin and serum creatinine levels were recorded. Hematocrit, platelet count, international normalized ratio, activated partial thromboplastin time, fibrinogen, D-dimer, MB isoenzymes of creatine kinase (CK-MB), troponin T, and aspartate aminotransferase (ASAT) values were recorded preoperatively and in the morning the first postoperative day. Electrocardiogram (ECG) recordings were made the day before the operation and in the morning of the first postoperative day. After the inclusion was finalized, all ECGs were interpreted by a cardiologist who was blinded to treatment group. The diagnosis of a perioperative myocardial infarction was based on the detection of new development of Q waves in the postoperative ECG combined with an increase of serum levels of the cardiac enzymes CK-MB, troponin T, and ASAT. A CK-MB level >50 µg/mL and/or a troponin T level >1.00 µg/mL and/or an ASAT level >100 U/L were considered to indicate myocardial necrosis.
The sample size calculation was based on the following assumptions. We considered that a difference in blood loss of <200 mL between groups would not be clinically meaningful to detect. In a previous study on the effects of tranexamic acid on postoperative bleeding after coronary surgery (17), we found a standard deviation (SD) for blood loss of approximately 300 mL, a value that we considered could also be applied to this study. Consequently, with a power of 90% and 
= 0.05, 48 patients had to be included in each group.
Data are given as means (SD) and medians (ranges), respectively. Statistical analyses were performed using the program package SPSS for Windows®, version 10.0. For comparative statistics, Fisher’s exact tests, 
2 tests, Mann-Whitney U-tests, and Student’s t-tests were used as appropriate. P values < 0.05 were considered statistically significant.
|
Results
|
|---|
One-hundred patients were included in the study and randomized. As shown in Figure 1, 8 patients were excluded from the study after randomization. Baseline patient data and data on surgical procedures from the remaining 92 patients are presented in Table 1. There were no differences between the two groups with respect to previous diseases such as myocardial infarction, heart failure, diabetes, chronic obstructive pulmonary disease, peripheral vascular disease, cerebrovascular disease, or smoking habits. Preoperative medication, heart rate, and blood pressure did not differ between groups. The aspirin dose was either 75 mg (12 patients in the desmopressin group and 11 patients in the placebo group) or 160 mg (34 patients in the desmopressin group and 35 patients in the placebo group). There were no differences in the ACT values before heparin, after heparin, or after protamine, and no differences in the total dose of heparin or the protamine dose given between the 2 groups (Table 2).
View larger version (22K):
[in this window]
[in a new window]
|
Figure 1. Flow of participating patients through the investigation. Reasons for patients not receiving treatment as allocated: one patient (desmopressin group) was lost after randomization, two patients (one in each group) went through off-pump coronary surgery, and in two patients (one in each group), there were intraoperative complications. Reasons for patients being withdrawn: one patient (placebo group) received tranexamic acid before cardiopulmonary bypass, two patients (one in each group) were reexplored because of surgical bleeding.
|
|
The preoperative and postoperative laboratory test results are presented in Table 3. There were no differences in these test results between the two groups. The mean (SD) difference in postoperative compared with preoperative cardiac enzyme values in the desmopressin and placebo groups, respectively, was 15.7 (18.5) versus 15.8 (18.1) µg/mL (P = 0.99) for CK-MB, 0.37 (0.43) versus 0.30 (0.26) µg/mL (P = 0.38) for troponin T, and 30.6 (30.0) versus 30.4 (22.5) U/L (P = 0.96) for ASAT. Two patients in the desmopressin group and four patients in the placebo group had increased cardiac enzyme levels to the extent that myocardial necrosis could be suspected. There was, however, no development of new Q waves in any ECG, and therefore no patients were diagnosed as having developed a perioperative myocardial infarction.
The time from termination of CPB to skin closure, a measure of the time spent on surgical hemostasis, did not differ between the two groups (32 [10] min in the desmopressin group versus 33 [9] min in the placebo group, P = 0.76).
The total postoperative bleeding volumes are presented in Figure 2. There was no difference in total postoperative bleeding between the 2 groups (606 [237] mL in the desmopressin group versus 601 [301] mL in the placebo group [P = 0.93], giving a difference in bleeding of 5 mL, 95% confidence interval of the difference -107 to 117 mL). All patients had mediastinal shed blood retransfused, and the volume retransfused was the same in the 2 groups (516 [257] mL in the desmopressin group versus 497 [250] mL in the placebo group, P = 0.72). Four patients were reexplored because of postoperative bleeding. Two of these patients were excluded from the analysis because they had a surgical cause of bleeding (Fig. 1). The remaining two patients were both in the placebo group. They bled 650 and 710 mL, respectively, during the first 4 h after surgery. No surgical source of bleeding was found, and it was concluded that the patients had diffuse microvascular bleeding.
View larger version (20K):
[in this window]
[in a new window]
|
Figure 2. Box-plot and whisker diagram showing total postoperative bleeding in the two groups. Center box shows median and interquartile range with whiskers indicating range of data, except for outliers, which are marked individually. The upper range is defined as the 75 percentile + 1.5 x (75 percentile - 25 percentile); the lower range is defined as the lowest value because there were no outliers in this direction.
|
|
There was no significant difference in the number of transfusions of blood products between the groups. Twenty-four percent (11 of 46) of the patients in each group received at least 1 transfusion. The number of blood product donor exposures per patient was 0.57 in the desmopressin group and 1.07 in the placebo group (P = 0.28). The number of patients receiving transfusions of packed red cells, fresh frozen plasma, and platelets, and the number of patients treated with tranexamic acid or desmopressin postoperatively are given in Table 4. No statistically significant differences were found between groups in these variables.
One patient in the desmopressin group had an episode of hypotension immediately after injection of the study drug that required treatment with ephedrine and phenylephrine. Another patient in this group had sternal dehiscence and renal failure and had a prolonged stay in the ICU. Finally, 1 patient in the desmopressin group had a pulmonary embolism 1 mo after the operation. All patients made a full recovery. No other major complications were recorded in the study patients.
|
Discussion
|
|---|
Previous studies on the effect of desmopressin in aspirin-treated patients (13,14) showed that desmopressin reduced postoperative bleeding in patients treated with aspirin until less than five or seven days before surgery. Despite the findings by Gibbs et al. (16) that patients who ingested aspirin less than two days before surgery had a greater impairment of platelet function than did patients who had stopped taking aspirin two days before the operation, we were not able to reproduce the results from the previous investigations. One reason for this discrepancy may be the difference in total amount of postoperative blood loss in the different study populations. A meta-analysis on the effects of desmopressin on reducing blood loss in cardiac surgery suggests that desmopressin significantly reduces blood loss only in cardiac operations in which the mean blood loss in placebo-treated patients is >1100 mL (18). The patients in the placebo group in the study by Gratz et al. (13) bled 1176 mL postoperatively, and the patients in the placebo group in the study by Sheridan et al. (14) bled 2376 mL (reported as total blood loss). In contrast, the postoperative blood loss in our study was 600 mL.
According to our routine, tranexamic acid and desmopressin and transfusions of fresh frozen plasma and platelets were given as treatment for increased postoperative bleeding. The choice of treatment was at the discretion of the attending physician, as described in the Methods section. This treatment may have reduced the postoperative bleeding, and because more patients in the placebo group received such treatment, the difference in postoperative bleeding between the two groups could have been reduced. However, only four patients received transfusions of fresh frozen plasma and/or platelets, and these were the patients who had the largest postoperative bleedings, the outliers in Figure 2. A possible reduction in bleeding caused by the treatment given to these patients would not have influenced the main results of the study to any major extent. Two previous investigations on the effect of tranexamic acid in cardiac surgery have shown that tranexamic acid must be given before CPB to exert maximal reduction in postoperative bleeding (19,20). Also, De Prost et al. (21) have investigated the effects of desmopressin in patients with established postoperative bleeding and an increased bleeding time after cardiac surgery. The results of this study show that desmopressin is not useful in reducing bleeding in these patients. We therefore conclude that the "rescue" treatment of some patients with tranexamic acid, desmopressin, fresh frozen plasma, or platelets because of increased postoperative bleeding did not significantly influence the results.
The study was designed to detect a difference in blood loss of 
200 mL, which we considered to be clinically relevant. Thus, a smaller difference in blood loss between the two groups could a priori not be expected to be revealed. However, given the present results with a 95% confidence interval from -107 to +117 mL, it is highly unlikely that desmopressin may reduce the bleeding with >107 mL. Such a small effect would hardly justify the prophylactic use of desmopressin in these patients.
One patient in the desmopressin group experienced a pulmonary embolism one month after the operation. No other patients had severe thromboembolic complications. The possibility for thromboembolic complications, and in particular myocardial ischemia and infarction must, however, be considered when giving desmopressin to CABG patients. In a meta-analysis on pharmacological strategies to decrease excessive blood loss in cardiac surgery, Levi et al. (22) found that desmopressin was associated with a 2.4-fold increase in the risk of perioperative myocardial infarction. No patients in our study were diagnosed as having a perioperative myocardial infarction, but our study was not powered to detect a difference in the incidence of myocardial infarctions between the two groups. It is important to be aware that controlled clinical studies are not suitable for detecting rare adverse drug reactions (23).
In the desmopressin group, one patient had an episode of hypotension and needed treatment with vasoactive drugs immediately after the infusion. Previous investigations on the hemodynamic responses to desmopressin administration after cardiac surgery have shown that desmopressin produces a decrease in arterial blood pressure that is secondary to a decrease in systemic vascular resistance (24,25). The authors of these studies concluded that desmopressin should be given with caution after cardiac surgery.
In summary, we have shown that prophylactic treatment with desmopressin has no effect on postoperative bleeding after CABG in patients treated with aspirin until surgery. The prophylactic administration of desmopressin to reduce postoperative bleeding in primary CABG patients can therefore not be recommended.
|
Acknowledgments
|
|---|
HP is supported by The Norwegian Health Association Grant 6432.
The authors thank the Unit for Applied Clinical Research at the Norwegian University of Science and Technology for help with the randomization procedure.
|
References
|
|---|
- Taggart DP, Siddiqui A, Wheatley DJ. Low-dose preoperative aspirin therapy, postoperative blood loss, and transfusion requirements. Ann Thorac Surg 1990; 50: 425–8.
- Weightman WM, Gibbs MN, Weidman CR, et al. The effect of preoperative aspirin-free interval on red blood cell transfusion requirements in cardiac surgical patients. J Cardiothorac Vasc Anesth 2002; 16: 54–8.[ISI][Medline]
- Kohler M, Harris A. Pharmacokinetics and haematological effects of desmopressin. Eur J Clin Pharmacol 1988; 35: 281–5.[ISI][Medline]
- Mannucci PM. Desmopressin, a non-transfusional form of treatment for congenital and acquired bleeding disorders. Blood 1998; 72: 1449–55.
- Mannucci PM, Vincente V, Viannello L, et al. Controlled trial of desmopressin in liver cirrhosis and other conditions associated with a prolonged bleeding time. Blood 1986; 67: 1148–53.[Abstract/Free Full Text]
- Mannucci PM. Desmopressin (DDAVP) for treatment of disorders of hemostasis. Prog Hemost Thromb 1986; 8: 19–45.[ISI][Medline]
- Kobrinsky NL, Israels ED, Gerrard JM, et al. Shortening of bleeding time by 1-deamino-8-D-arginine vasopressin in various bleeding disorders. Lancet 1984; 1: 1145–8.[ISI][Medline]
- Salzman EW, Weinstein MJ, Weintraub RM, et al. Treatment with desmopressin acetate to reduce blood loss after cardiac surgery. N Engl J Med 1986; 314: 1402–6.[Abstract]
- Hedderich GC, Petsikas DJ, Cooper BA, et al. Desmopressin acetate in uncomplicated coronary artery bypass surgery: a prospective randomized clinical trial. Can J Surg 1990; 33: 33–6.[ISI][Medline]
- Spyt TJ, Weerasena NA, Bain WH. The effects of desmopressin acetate (DDAVP) on haemostasis and blood loss in routine coronary artery bypass surgery: a randomized, double-blind trial. Perfusion 1990; 5 (Suppl): 57–61.
- Marquez J, Koehler S, Strelec SR, et al. Repeated dose administration of desmopressin acetate in uncomplicated cardiac surgery: a prospective, blinded, randomized study. J Cardiothorac Vasc Anesth 1992; 6: 674–6.[Medline]
- Ansell J, Klassen V, Lew R, et al. Does desmopressin acetate prophylaxis reduce blood loss after valvular heart operations? J Thorac Cardiovasc Surg 1992; 104: 117–23.[Abstract]
- Gratz I, Koehler J, Olsen D, et al. The effect of desmopressin acetate on postoperative hemorrhage in patients receiving aspirin therapy before coronary artery bypass operations. J Thorac Cardiovasc Surg 1992; 104: 1417–22.[Abstract]
- Sheridan DP, Card RT, Pinilla JC, et al. Use of desmopressin acetate to reduce blood transfusion requirements during cardiac surgery in patients with acetylsalicylic-acid-induced platelet dysfunction. Can J Surg 1994; 37: 33–6.[ISI][Medline]
- Dilthey G, Dietrich W, Spannagl M, Richter JA. Influence of desmopressin acetate on homologous blood requirements in cardiac surgical patients pretreated with aspirin. J Cardiothorac Vasc Anesth 1993; 7: 425–30.[Medline]
- Gibbs NM, Weightman WM, Thackray NM, et al. The effects of recent aspirin ingestion on platelet function in cardiac surgical patients. J Cardiothorac Vasc Anesth 2001; 15: 55–9.[ISI][Medline]
- Pleym H, Stenseth R, Wahba A, et al. Single-dose tranexamic acid reduces postoperative bleeding after coronary surgery in patients treated with aspirin until surgery. Anesth Analg 2003; 96: 923–8.[Abstract/Free Full Text]
- Cattaneo M, Harris AS, Stromberg U, Mannucci PM. The effect of desmopressin on reducing blood loss in cardiac surgery: a meta-analysis of double-blind, placebo-controlled trials. Thromb Haemost 1995; 74: 1064–70.[ISI][Medline]
- Brown RS, Thwaites BK, Mongan PD. Tranexamic acid is effective in decreasing postoperative bleeding and transfusions in primary coronary artery bypass operations: a double-blind, randomized, placebo controlled trial. Anesth Analg 1997; 85: 963–70.[Abstract]
- Casati V, Bellotti F, Gerli C, et al. Tranexamic acid administration after cardiac surgery: a prospective, randomized, double-blind, placebo-controlled study. Anesthesiology 2001; 94: 8–14.[ISI][Medline]
- De Prost D, Barbier-Boehm G, Hazebroucq J, et al. Desmopressin has no beneficial effect on excessive postoperative bleeding or blood product requirements associated with cardiopulmonary bypass. Thromb Haemost 1992; 68: 106–10.[ISI][Medline]
- Levi M, Cromheecke ME, de Jonge E, et al. Pharmacological strategies to decrease excessive blood loss in cardiac surgery: a meta-analysis of clinically relevant endpoints. Lancet 1999; 354: 1940–7.[ISI][Medline]
- Kaufmann DW, Shapiro S. Epidemiological assessment of drug-induced disease. Lancet 2000; 356: 1339–43.[ISI][Medline]
- Salmenpera M, Kuitunen A, Hynynen M, Heinonen J. Hemodynamic responses to desmopressin acetate after CABG: a double-blind trial. J Cardiothorac Anesth 1991; 5: 146–49.
- Frankville DD, Britton Harper G, Lake CL, Johns RA. Hemodynamic consequences of desmopressin administration after cardiopulmonary bypass. Anesthesiology 1991; 74: 988–96.[ISI][Medline]
Accepted for publication September 23, 2003.
Top
Abstract
Introduction
