A cohort study of coagulation parameters and the use of blood products in surgery of the thoracic and thoracoabdominal aorta☆☆☆

Article Outline

• Abstract
• Patients and methods
  • Surgical techniques
  • Transfusion practices
  • ANH technique
  • Postoperative care
  • Study procedure
  • Statistical analysis
• Results
• Discussion
• Conclusions
• Acknowledgements
• References
• Copyright

Abstract 

Objectives: We sought to study the hemostatic profile and the use of blood products in patients undergoing thoracic and thoracoabdominal aortic aneurysm repair. Methods: This is a cohort study comparing three groups of patients: 7 undergoing elective acute normovolemic hemodilution (ANH); 15 undergoing elective procedures without ANH (non-ANH); and 8 undergoing repair of ruptured thoracoabdominal aneurysms. A control group of 10 patients was used for comparison of preoperative hemostasis. The parameters studied were platelet concentration, partial thromboplastin time (PTT), and international normalized ratio (before and after surgery), packed red blood cells, fresh frozen plasma, platelets, cryoprecipitates, donor exposure, and use of desmopressin and epsilon-aminocaproic acid. Analysis of variance and multiple stepwise regression analysis were used. Results: Before operation the patients with aneurysms had PTTs prolonged compared with control subjects (P < .05). After operation the ANH group had higher platelet counts than the ruptured group (P = .001) and higher platelet counts (P = .05) and lower PTTs (P = .001) than the non-ANH group. The ANH group was transfused fewer platelets than the non-ANH group (P = .001) and less of every blood product than the ruptured group (P = .05); statistically significant differences were not observed for packed red blood cells, fresh frozen plasma, or platelets. The ANH group was exposed to an average of 65 donors fewer than the ruptured group (P < .001) and 34 fewer than the non-ANH group (P < .05). These differences could not be explained by baseline coagulation status or by the intraoperative use of desmopressin or epsilon-aminocaproic acid. Conclusions: The coagulation abnormality identified before surgery is that of higher PTT values, suggesting a disturbance of the extrinsic coagulation pathway. Blood losses, donor exposure, and platelet use were highest in the ruptured group and lowest in the ANH group. After surgery the ruptured group exhibited the worst coagulation parameters, and the ANH group exhibited the best with higher platelet count and lower PTT values than the other groups. The ANH technique appears to be an useful adjunct in the anesthetic management of these patients. (J Vasc Surg 2001;33:462-8.)

Coagulation disorders represent an important aspect of the management of patients with thoracoabdominal aortic aneurysms (TAAAs). A number of case reports of disseminated intravascular coagulopathy (DIC) associated with aortic dissection,¹ TAAA,² and abdominal aortic aneurysms (AAAs)³, ⁴, ⁵, ⁶ have appeared, together with case series suggesting that milder coagulation disturbances may be relatively common in these populations.², ⁶, ⁷ Preoperative coagulation abnormalities identifiable with routine laboratory investigations in patients with TAAAs have not previously been described.

Surgery for this condition is associated with major blood loss necessitating large volume transfusion, and complex intraoperative and postoperative coagulopathy may result.⁸, ⁹, ¹⁰ No study has quantitated total intraoperative and postoperative use of blood products in TAAAs. Problems with large transfusion requirements and coagulopathy are amplified in patients with ruptured aneurysms and possibly ameliorated by acute normovolemic hemodilution (ANH).¹¹ This technique involves intraoperative large-volume phlebotomy with reinfusion of crystalloid or colloid before cross-clamping: the phlebotomized fresh autologous whole blood is stored for reinfusion at the end of the procedure.

The purpose of this work is to identify preexisting coagulation abnormalities in patients with thoracic and thoracoabdominal aneurysm compared with matched control subjects without aneurysms. We describe the blood product use for this surgery and document that near-normal postoperative coagulation parameters may be obtained.

Back to Article Outline

Patients and methods 

This is a single-center prospective consecutive cohort study: inclusion criteria were patients with thoracoabdominal aneurysms who underwent operation either electively or for rupture by a single surgical/anesthetic team between 1990 and 1995. Patients with dissections or traumatic rupture were not included. Patients provided informed consent for surgery and ANH; data were collected prospectively. Twenty-two patients underwent elective procedures, of whom 15 (before 1993) were managed with standard anesthetic technique (non-ANH), and seven (after 1993) with acute normovolemic hemodilution (ANH). An additional group of eight patients underwent emergency surgery for ruptured TAAAs. A control group for comparison of preoperative coagulation parameters comprised 10 consecutive patients undergoing carotid endarterectomy who were not receiving anticoagulants, had no clinical evidence of aneurysmal disease, and whose abdominal aorta did not exceed 3 cm in maximum diameter by ultrasound scanning. These were thought to represent patients similar in age and nonaneurysmal vascular disease to the population with aneurysms described.

Patients with TAAAs were classified in four groups according to Svensson and Crawford.¹² Distal arch aneurysms are defined as aneurysms involving and localized to that segment of the aortic arch from which the left subclavian artery arises.

Surgical techniques 

All patients, including emergency cases, underwent standard general anesthesia with a left-sided double-lumen endotracheal tube. One or two arterial lines were inserted, followed by a pulmonary artery catheter and three large-bore central venous lines. Patients underwent operation with Crawford's “clamp and sew” technique, with mild general hypothermia, no cell-saver device, and no systemic administration of heparin. In patients with group I and group II TAAAs undergoing elective procedures, a catheter was placed in the subarachnoid space for drainage and pressure monitoring.

Transfusion practices 

Hemoglobin concentration, hematocrit, prothrombin time (PT), partial thromboplastin time (PTT), and platelet concentration were determined before induction and at frequent intervals during the case. All fluids were administered through a blood warmer, with the object of keeping the patient mildly hypothermic at a temperature above 34°C throughout the cases (monitored by nasopharyngeal probe). Ringer's lactate was the preferred solution for crystalloid infusion and was administered at a rate of 12 mL/kg/h with the objective of keeping the total infusion less than 8 L. Normal saline solution was used to flush infusion lines for blood products, resulting in an additional volume infusion of 1 to 3 L; in addition, approximately 1 L of 5% albumin was also used in each patient. Stored packed red blood cells (PRCs) were transfused when the hematocrit was less than 30% or the hemoglobin concentration was less than 100 g/L. In the context of hypotension or observed brisk bleeding, blood was transfused replacing estimated losses, volume by volume. Fresh frozen plasma (FFP) was infused as volume replacement (ie, according to visual estimates of volume loss and measured central pressures), with a ratio 1:1 with PRC and, when volume status permitted, to normalize the internal normalized ratio (INR) and PTT. The ratio of PRC to FFP was adjusted to keep the hematocrit greater than or equal to 30%. When the intravascular volume did not allow transfusion of FFP, cryoprecipitate was given to correct coagulation abnormalities. Platelets were transfused when the platelet count was less than 100 × 10⁹/L. Epsilon-aminocaproic acid and desmopressin were used only when coagulation abnormalities persisted in the face of the aggressive replacement of coagulation factors described above. This transfusion strategy was used consistently for all patient groups.

ANH technique 

A modification¹¹ of the standard ANH technique¹³, ¹⁴, ¹⁵, ¹⁶ was used. Phlebotomy was initiated soon after the patient was positioned on the operating table. During this phase, each liter of the patient's blood was replaced with an equal volume of 5% albumin. Generally about 1 L was removed, the goal being to lower the patient's hemoglobin concentration to approximately 100 g/L (hemodilution phase). Phlebotomy was then continued until another 1 to 1.5 L was obtained. During this phase, each liter of autologous blood was replaced with 600 mL stored PRC and 400 mL 5% albumin (partial exchange transfusion). Immediately before the proximal aorta was clamped, an additional 450 to 750 mL of autologous blood was removed from the patient without volume replacement. The reinfusion of autologous blood was started after the surgeon began the distal aortic anastomosis.

Postoperative care 

After operation patients from all groups were transferred to the intensive care unit, where they were cared for according to standard practice by physicians who were unaware of the research group's interest in transfusion requirements. Prospective data on all blood products transfused were collected for the entire intensive care unit and in-hospital stay.

Study procedure 

Before operation patients were questioned about and examined for evidence of spontaneous bruising or petechiae. Hemoglobin, platelet count, INR, and PTT were determined before operation and at the point of arrival in the intensive care unit after operation. Blood losses were estimated from the blood in the suction containers and the weight of saturated sponges. The intraoperative and postoperative use of blood products, desmopressin, and epsilon-aminocaproic acid was recorded.

Statistical analysis 

Data were analyzed with BMDP–7.0 (SPSS Inc, Chicago, Ill). Univariate and multivariate analysis of variance and multiple stepwise regression analysis were used. Post hoc comparisons were made by use of Tukey's studentized range method. For continuous variables all data are expressed as mean ± SD.

We hypothesized that group I, group IV, and distal arch aneurysms would represent aneurysms with lower risk of blood loss and that this risk would be higher for group II and group III aneurysms. Because the size of this data set did not permit the examination of each group as a separate variable, we investigated this hypothesis by collapsing the classification into two strata, putative high transfusion risk (group II and III) and low transfusion risk (group I, group IV, and distal arch aneurysms).

Back to Article Outline

Results 

Twenty-six of the 30 patients were men. The average age was 69 ± 10 years. Patients with ruptured aneurysms were older than patients undergoing elective procedures, but differences were not statistically significant. The demographics, aneurysm type, and comorbidity are summarized in Table I. No patients was taking antiplatelet or anticoagulant medications during the week before surgery.

Table I. Demographics

Three patients in the elective group and six in the ruptured group died before discharge. In the remaining 21 patients, the 1-year survival rate was 89%. Two patients had paraplegia, one in the elective group and one in the ruptured group, both with group II TAAAs, and both of them died of respiratory complications and kidney failure.

No patients had a history of easy bruising or physical evidence of petechiae, and none had thrombocytopenia. Descriptive statistics of preoperative hemoglobin concentration and coagulation parameters are summarized in Table II. Preoperative hemoglobin, INR, PTT, and platelet count differed among the three groups (ruptured, electively repaired aneurysms, and control subjects) by multivariate analysis of variance (P < .05). The presence of an aneurysm, and whether it was ruptured, accounted for 52% of the observed variance in these parameters (η² = 0.52). PTT was significantly prolonged in patients with aneurysms compared with controls (mean difference 4.5 seconds, 95% CI 0.5-9.0 seconds; post hoc t test, P < .05). Hemoglobin levels were lower in the ruptured aneurysm group than in patients with elective aneurysms or nonaneurysmal control subjects (mean difference 33 g/L, 95% CI 16-62 g/L; post hoc t test, P < .001). Within the elective group of patients, preoperative hemoglobin, INR, PTT, and platelet count did not differ between patients subsequently treated with or without ANH.

Table II. Coagulation parameters and estimated blood loss

Table II summarizes the estimated intraoperative blood loss, and the postoperative hemoglobin level and coagulation parameters. The estimated intraoperative blood loss was lower in the ANH group, but the difference did not achieve statistical significance (P = .40).

There were significant differences in postoperative hemoglobin, INR, PTT, and platelet count among the three groups, ANH, non-ANH, and ruptured. Group membership (ANH, non-ANH, or ruptured) accounted for 80% of the variance in these parameters (η² = 0.80). The extent of TAAAs (group I, group IV, and distal arch aneurysms vs group II and III) was not a significant predictor of postoperative parameters. Statistically significant differences between groups were confined to the platelet count and PTT. Postoperative platelet count was lower in the ruptured (P = .001) and non-ANH (P = .05) groups compared with the ANH group. PTT was more prolonged in the non-ANH group compared with the ANH (P = .001).

Visceral ischemia time was not significantly different between the ANH group and non-ANH group (P = .60) or between the ruptured and elective group (P = .53) (Table I). Four patients in the ruptured group (50%), four in the non-ANH (26%), and none in the ANH group received desmopressin or epsilon-aminocaproic acid.

The total (intraoperative and postoperative) use of PRC, FFP, platelets, and cryoprecipitate was significantly different among the groups, ruptured, non-ANH, and ANH, with group accounting for 64% of the variance in the number of units of blood products used (η² = 0.64, P = .001). This analysis was adjusted for extent of TAAA (group I, group IV, and distal arch aneurysms vs group II and III), which was also found to be predictive of patterns of blood product use. Transfusion practices for both individual and total blood products are summarized in Table III.

Table III. Total blood products transfused

The ANH group received no cryoprecipitate, fewer platelets than the non-ANH group (P < .01), and less of every blood product than the ruptured group (P < .05). The non-ANH group received fewer units of cryoprecipitate than the ruptured group, but statistically significant differences were not observed between these two groups for PRC, FFP, or platelets. Fig 1 is a graphical summary of intraoperative and postoperative blood product transfusion.

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 1. 

  Utilization of blood products. Plt, Platelets; Cryo, cryoprecipitate.

Similar results were obtained when these time periods were analyzed separately. For the total hospital stay, the mean transfusion requirements across all groups was 63 ± 39 units (median 53 units). During operation, the mean blood products use was 44 ± 23 units (median 40 units).

In a separate analysis, we used stepwise linear regression to examine the relationship between the total number of units of blood products transfused (the sum of PRC, FFP, platelets, cryoprecipitate for each patient), and estimated blood loss, TAAA group (ruptured vs elective ANH vs elective non-ANH) and the extent of the aneurysm (group I, group IV, and distal arch aneurysms vs group II and III). Each of these variables proved statistically significant (P < .05). On average, compared with the ruptured group, patients in the ANH group were transfused 65 fewer units, and in the non-ANH group 31 fewer units (Fig 2).

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 2. 

  Relationship between total blood products transfused and estimated blood losses for each of three TAAA groups. Regression lines are for each group analyzed separately. Slopes of lines are not significantly different one from another.

Among the patients in the elective group, the non-ANH group was transfused 34 units more than the ANH. Unsurprisingly, the number of units transfused increases with estimated blood losses (5 units transfused for each additional liter lost). The extent of the aneurysm was again shown to affect the total transfusion, with patients with group II and III receiving on average 27 units more than those with group I, group IV, and distal arch aneurysms. Overall these clinical factors accounted for 69% of the variance in total units of blood products transfused (R ² = 0.69).

Back to Article Outline

Discussion 

Occasional patients in whom DIC coexists with elective³, ⁴ and ruptured AAAs⁵, ⁶ and elective TAAAs² have long been recognized. These case reports and case series do not permit the incidence of the problem to be estimated. However, the prospective study of Fisher et al² describes frank DIC in three of 32 thoracoabdominal and none of 44 abdominal aortic aneurysms. Isolated elevation of fibrin degradation products (FDP) without clinical bleeding or depressed platelet counts was noted before operation in 14% of TAAA and 8% of infrarenal AAA, suggesting that a compensated process occurs in some of these patients. In a series of 84 patients with ruptured and unruptured AAA reported by Getaz et al,⁶ none had clinical evidence of DIC. Similarly, in our series of 30 ruptured and unruptured TAAAs, no overt DIC was noted before operation.

Of the patients with ruptured aneurysms reported by Getaz et al,⁶ 50% had platelet counts less than 150 × 10⁹/L and 56% coagulopathy. In the same study, 28% of patients with unruptured aneurysms had thrombocytopenia. However, only 67% of ruptured and 18% of unruptured aneurysms had the relevant laboratory determinations, and this may have led to overestimation of the incidence of abnormalities. In a series of patients with aortic disease undergoing elective procedures⁷ a 3% incidence of preoperative abnormalities in PT and PTT was reported, and no patient had abnormal levels of platelets or fibrinogen. The inclusion in this report of a number of patients with nonaneurysmal disease likely results in underestimation of the incidence of coagulation abnormalities. Another small series of electively treated AAA¹⁷ found entirely normal PT, PTT, and thrombin clotting time in these patients. In contrast, in a larger series of ruptured AAA,¹⁸ 16 of 43 patients had prolongation of INR and the same proportion thrombocytopenia. In our series no patient with electively treated or ruptured aneurysm had a platelet count less than 150 × 10⁹/L, 2 of 22 electively treated and 1 of 8 ruptured had PT prolongation (> 1.2 seconds), and 3 of 22 electively treated and no patient with a ruptured aneurysm had PTT prolongation (> 34 seconds). Our finding of a subtle preoperative prolongation in PTT compared with control subjects would be consistent with mild consumption coagulopathy, but our work is limited by the lack of more sensitive markers of coagulation and fibrinolysis. Abnormalities sufficient to cause frank thrombocytopenia or clinically important prolongation of clotting parameters are rare in elective AAAs or TAAAs, but may be more common in ruptured aneurysms.

Other studies have reported more subtle disturbances in coagulation. A report of 20 consecutive patients with TAAAs or thoracic dissection found levels of anti-thrombin III, FDP, thrombin-antithrombin III complex and plasmin-alpha 2 plasmin inhibitor to be elevated before operation, with normal levels of fibrinogen.⁹ Very similar results have been obtained in AAAs, with the additional documentation of normal clotting parameters by routine testing.¹⁰ Gertler et al¹⁹ reported that in TAAAs, clotting factors and fibrinogen levels decreased whereas D-dimer and prothrombin fragment F1.2 increased 30 minutes after supraceliac aortic cross-clamping, suggesting that mesenteric ischemia is responsible for coagulation factor consumption and increased fibrinolytic activity during TAAAs repair. Renal and limb ischemia also likely contribute: Illig et al²⁰ found normal hemostatic parameters before surgery except for elevation in D-dimer, and that both FDP and D-dimer increased after cross-clamping compared with patients with infrarenal aneurysms. The increase in FDP was significantly more pronounced in the group undergoing supraceliac clamping. In our work the visceral ischemia time was not different between groups and is unlikely to account for the observed differences in coagulation parameters and blood product use. With standard tests, there was no significant correlation between the duration of visceral ischemia and the extent of postoperative coagulation abnormalities, or with the total number of units of blood products transfused (data not shown).

The clinical implications of these findings are restricted to the most extreme cases. In ruptured AAA, thrombocytopenia and prolongation of the INR have been shown to be statistically significant multivariate predictors of death.¹⁸ With an aggressive blood product transfusion strategy, relatively normal postoperative coagulation parameters may be obtained.

Svensson et al²¹ reported the median number of units of blood and blood products used during surgery in Crawford's experience from 1960 to 1991 in 1509 patients who underwent operation for TAAA with the use of a cell saver: PRC 7 units (range, 1-46), FFP 16 units (range, 0-132), platelets 20 units (range, 0-110), cryoprecipitate 0 unit (0-100), and autotransfused cells 8 units (range, 0-68). In the latter part of this series, most patients prophylactically received aminocaproic acid. The median number of intraoperative blood products transfused was 51 units. Median intraoperative blood products transfusion in our study was 40 units, even though a cell saver device was not used.

Godet et al²² described the use of blood products in 33 patients undergoing TAAA repair. A strategy of avoiding transfusion of blood products (PRC, FFP, and platelets) unless indicated by hematocrit or abnormal coagulation parameters was used. In the absence of the results of appropriate tests, transfusions were given only after clinically manifest coagulopathy had developed. In this series the utilization of FFP was 8 ± 7 units, PRC 9 ± 6 units, autotransfused cells 12 ± 9 units. The mean blood products transfused during surgery was 29 units. Eight patients in this series had development of clinically significant intraoperative hemorrhagic complications not attributable to surgical techniques, and three of these patients died during surgery.

We have documented the typical blood product requirement to produce near-normal postoperative coagulation parameters. To facilitate clinical management and blood-bank planning, we have separately reported intraoperative and postoperative blood product use in ruptured and electively repaired aneurysms. For accurate interpretation of results and to avoid bias, we think that it is critical that studies of blood product requirements and transfusion strategies include details of the postoperative period. To our knowledge, such data for TAAAs repaired without the use of a cell-saver device are not available. Our results are comparable with the intraoperative results reported by Svensson and Crawford,¹² with an autotransfusion device: 15 units of PRCs, 16 units of FFP, and 20 units of platelets were used in this series.

Our analysis has focused on total blood products transfused, much of which is platelets, FFP, and cryoprecipitate. The total transfusion is an index of the demand for a scarce resource and is a marker of the number of donors to whom patients are exposed. This exposure is highest for patients with ruptured TAAAs, followed by patients treated without ANH, and is least in patients treated with ANH. The parallel regression lines between groups shown in Fig 2 (ie, no statistically significant difference in slope of the different regression lines) indicates that this difference is consistent for any amount of blood loss. Because the use of coagulation adjuncts was highest in the ruptured (50%), intermediate in non-ANH (26%), and lowest in the ANH group (nil), the differences observed cannot be attributed to the administration of desmopressin or epsilon-aminocaproic acid. It is possible that the technique of ANH may be responsible for these differences; however, other factors such secular improvements in intensive care practices, learning-curve effects for either the anesthetist or surgeon, or differences in case mix may be cause. We believe that a randomized controlled trial of this issue is warranted.

Consistent with our a priori hypothesis, we observed that group I and IV TAAAs and distal arch aneurysms received significantly fewer transfusions compared with group II and III. This information may also facilitate blood bank planning and has not, to our knowledge, previously been reported.

Back to Article Outline

Conclusions 

Clinically important disseminated intravascular coagulation is rare in TAAAs. A mild preoperative prolongation of PTT was identified in patients with aneurysms compared with control subjects. Transfusion of blood products was quantitated for both intraoperative and postoperative periods and was highest in patients with ruptured aneurysms, intermediate in patients undergoing elective procedures treated without ANH, and lowest in patients undergoing elective procedures treated with ANH. With the described strategy of aggressive transfusion, postoperative coagulopathy was mild, although differences were still observed between electively repaired and ruptured aneurysms. Group I, IV, and distal arch aneurysms were shown to have lower transfusion requirements than group II and III.

Back to Article Outline

Acknowledgements 

We thank the anonymous reviewers of an earlier draft for their insightful comments and suggestions.

Back to Article Outline

References 

1. Fine NL, Applebaum H, Elguezabal A, Castleman L. Multiple coagulation defects in association with dissecting aneurysms. Arch Intern Med. 1967;119:522–526
   • View In Article
   • MEDLINE
2. Fisher DF, Yawn DH, Crawford ES. Preoperative disseminated intravascular coagulation associated with aortic aneurysms. Arch Surg. 1983;118:1252–1255
   • View In Article
   • MEDLINE
3. Siebert WT, Natelson EA. Chronic consumption coagulopathy accompanying abdominal aortic aneurysm. Arch Surg. 1976;111:539–541
   • View In Article
   • MEDLINE
4. Thompson RW, Adams DH, Cohen JR, Mannick JA, Whittemore AD. Disseminated intravascular coagulation caused by abdominal aortic aneurysm. J Vasc Surg. 1986;4:184–186
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (253 KB)
   • CrossRef
5. TenCate JW, Timmers H, Becker AE. Coagulopathy in ruptured or dissecting aortic aneurysms. Am J Med. 1975;59:171–176
   • View In Article
   • Abstract
   • Full-Text PDF (1204 KB)
   • CrossRef
6. Getaz EP, Louw JH. The coagulopathy associated with aortic aneurysms. Postgrad Med J. 1977;53:668–671
   • View In Article
   • MEDLINE
   • CrossRef
7. Mulcare RJ, Royster TS, Phillips LL. Intravascular coagulation in surgical procedures on the abdominal aorta. Surg Gynecol Obstet. 1976;143:730–734
   • View In Article
   • MEDLINE
8. Fernandez LA, MacSween JM, You CK, Gorelick M. Immunologic changes after blood transfusion in patients undergoing vascular surgery. Am J Surg. 1992;163:263–269
   • View In Article
   • MEDLINE
   • CrossRef
9. Oba J, Shiiya N, Matsui Y, Goda T, Sakuma M, Yasuda K. Alterations in coagulation and fibrinolysis after surgery for aortic aneurysm. Surg Today. 1995;25:532–535
   • View In Article
   • MEDLINE
   • CrossRef
10. Aramoto H, Shigematsu H, Muto T. Perioperative changes in coagulative and fibrinolytic function during surgical treatment of abdominal aortic aneurysm and arteriosclerosis obliterans. Int J Cardiol. 1994;47(Suppl 1):S55–S63
    • View In Article
    • MEDLINE
    • CrossRef
11. Cinà CS, Bruin G. Acute normovolemic hemodilution (ANH) in surgery of the thoracoabdominal aorta: a cohort study to evaluate coagulation parameters and blood product utilization. J Cardiovasc Surg. 1999;40:37–43
    • View In Article
12. Svensson LG, Crawford ES. Aortic dissection and aortic aneurysm surgery: clinical observations, experimental investigations, and statistical analyses Part 1. Curr Probl Surg. 1992;29:817–911
    • View In Article
    • MEDLINE
13. Stehling L, Zauder HL. Acute normovolemic hemodilution. Transfusion. 1991;31:857–868
    • View In Article
14. Triulzi MD, Ness PM. Intraoperative hemodilution and autologous platelet rich plasma collection: two techniques for collecting fresh autologous blood. Transfusion Science. 1995;16:33–44
    • View In Article
    • Abstract
    • Full-Text PDF (962 KB)
    • CrossRef
15. Feldman JM, Roth JV, Bjoraker DG. Maximum blood savings by acute normovolemic hemodilution. Anesth Analg. 1995;80:108–113
    • View In Article
    • MEDLINE
    • CrossRef
16. Mortelmans YJ, Vermaut GA, Van Aken HA. Simple method for calculating component dilution during fluid resuscitation: the Leuven approach. J Clin Anesth. 1994;6:279–287
    • View In Article
    • MEDLINE
    • CrossRef
17. Gibbs NM, Crawford GP, Michalopoulos N. Postoperative changes in coagulant and anticoagulant factors following abdominal aortic surgery. J Cardiothorac Vasc Anest. 1992;6:680–685
    • View In Article
18. Davies MJ, Murphy WG, Murie JA, Elton RA, Bell K, Gillon JG, et al.  Preoperative coagulopathy in ruptured abdominal aortic aneurysm predicts poor outcome. Br J Surg. 1993;80:974–976
    • View In Article
    • MEDLINE
    • CrossRef
19. Gertler JP, Cambria RP, Brewster DC, Davison JK, Purcell P, Zannetti S, et al.  Coagulation changes during thoracoabdominal aneurysm repair. J Vasc Surg. 1996;24:936–943
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (928 KB)
    • CrossRef
20. Illig KA, Green RM, Ouriel K, Riggs P, Bartos S, Whorf R, et al.  Primary fibrinolysis during supraceliac aortic clamping. J Vasc Surg. 1997;25:244–254
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (1576 KB)
    • CrossRef
21. Svensson L, Crawford ES, Hess KR, Coselli JS, Safi HJ. Experience with 1509 patients undergoing thoracoabdominal aortic operations. J Vasc Surg. 1993;17:357–370
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (1331 KB)
    • CrossRef
22. Godet G, Samama CM, Ankri A, Barre E, Soughir S, Kieffer E, et al.  Mechanisms and prediction of hemorrhagic complications during surgery of thoraco-abdominal aortic aneurysms. [French] Annales Francaises d Anesthesie et de Reanimation. 1990;9:415–422
    • View In Article