|
|
||||||||
Ann Thorac Surg 2001;72:527-533
© 2001 The Society of Thoracic Surgeons
a Hygeia Associates, Grass Valley, California, USA
b Department of Medicine, Sunnybrook and Womens College Health Sciences Centre, University of Toronto, Toronto, Ontario, Canada
Accepted for publication April 13, 2001.
Address reprint requests to Dr Wilkes, Hygeia Associates, 17988 Brewer Rd, Grass Valley, CA 95949
e-mail: mwilkes{at}hygeiaassociates.com
| Abstract |
|---|
|
|
|---|
Methods. Randomized controlled trials comparing albumin and HES in cardiopulmonary bypass patients were identified by bibliographic database searches and other methods.
Results. Sixteen trials involving 653 randomized patients were included. In 88% of randomized comparisons, postoperative bleeding was lower in the albumin group, and the standardized mean difference in bleeding favoring albumin across all trials (-0.24; 95% confidence interval, -0.40 to -0.08) was statistically significant. Bleeding differences between albumin and either high or medium molecular weight HES were similar. In trials of adults, the pooled mean blood loss in the albumin group was 693 ± 350 mL compared with 789 ± 487 mL in the HES group. The estimated proportion of adult albumin group patients with blood loss of more than 1,000 mL was 19% compared with 33% of adult HES group patients.
Conclusions. Postoperative blood loss is significantly lower in cardiopulmonary bypass patients exposed to albumin than HES.
| Introduction |
|---|
|
|
|---|
These observational studies are difficult to reconcile with randomized controlled trials supporting the equivalence of albumin and HES in cardiopulmonary bypass patients [622]. However, the randomized trials involved small numbers of patients and may have lacked statistical power. This meta-analysis tested the hypothesis that postoperative bleeding is lower in cardiopulmonary bypass patients exposed to albumin than HES.
| Material and methods |
|---|
|
|
|---|
Trial inclusion criteria
Candidate trials must have involved randomized comparison of exogenous purified albumin with HES in patients undergoing cardiopulmonary bypass. No restrictions were placed on the type of colloid usage such as preoperative, intraoperative, or postoperative administration or incorporation in the priming fluid. Postoperative bleeding data had to have been reported. Language restrictions were not applied. All inclusion criteria were established a priori.
Trial identification
Trials conforming to the above inclusion criteria were sought by an array of methods, including computer searches of MEDLINE; EMBASE; the Cochrane Controlled Trials Register; the Cochrane Medical Editors Trial Amnesty of unpublished clinical trials; and additional resources such as conference reports, abstracts, compilations of references, and full-text journal articles available on the Internet as identified using the Altavista, Northernlight, Hotbot, and Excite search engines. Authors of published randomized trial reports on colloid administration were consulted. The New England Journal of Medicine, JAMA, the BMJ, and The Lancet were searched by hand for the period from January 1990 to December 2000. Also examined were the reference citations in colloid-related prior published meta-analyses; completed reviews and protocols in the Cochrane Database of Systematic Reviews; review articles; randomized trials; and nonrandomized clinical studies.
Data extraction
Two investigators (M.M.W. and R.J.N.) independently assessed the conformity of trials to the inclusion criteria of the meta-analysis and extracted data from the included trials. Disparities in trial inclusion decisions and extracted data were resolved by discussion.
Statistical analysis
For continuous data such as volume of postoperative blood loss, standardized mean difference between the randomized groups was calculated [23]. Standardization transforms study results to a common scale (standard deviation units) that facilitates pooling and is appropriate for comparisons across studies that may, for example, report blood loss data in different units (ie, mL versus mL/kg), differ in blood loss measurement methods, or involve varying surgical techniques associated with substantial differences in magnitude of blood loss. To provide a clinical frame of reference, patient-weighted pooled within-group blood loss in milliliters was calculated for trials of adults, and corresponding z scores were used to estimate the percentages of patients in the albumin and HES groups with postoperative blood loss exceeding 1,000 mL.
In the case of reoperation because of bleeding, relative risk and risk difference were calculated. In the absence of significant heterogeneity, pooled standardized mean difference, relative risk, risk difference, and 95% confidence intervals (95% CI) were derived under fixed-effects models [24]. Unless substantial heterogeneity is present, effect sizes do not differ notably between fixed- and random-effects models. Publication bias was evaluated by Eggers test. Mean values of continuous data are cited ± standard deviation.
Evaluation of methodologic quality among the trials was based on randomization method, exclusions after trial entry, and blinding. Randomization method was classified as adequate, inadequate, or unclear.
| Results |
|---|
|
|
|---|
|
|
Patient exclusion criteria were specified in 14 included trials (Fig 1). The most frequently applied exclusion criteria were coagulopathy, renal dysfunction, and anemia in 64%, 57%, and 50% of the 14 trials, respectively.
|
|
|
Various forms of autotransfusion were reported in some trials. Thus, in one trial autologous blood collected preoperatively was reinfused after bypass [20]. In nine trials, reinfusion of blood in the extracorporeal circuit was reported either with [12, 14, 16, 17] or without [9, 10, 13, 15, 18] prior hemoconcentration. Intraoperative fluid replacement in one trial [6] included autotransfusion of recovered blood. In one trial [15] 20 of 60 (33%) patients received postoperative autotransfusion of shed mediastinal blood. In another trial patients underwent postoperative autotransfusion of salvaged pump blood [18]. In four trials no autotransfusion of shed blood was performed intraoperatively [15] or postoperatively [13, 14, 17]. Otherwise autotransfusion was not reported.
The respective median total doses administered for albumin (adjusted as appropriate to the equivalent of a 5% concentration), and high and medium molecular weight HES were 15.9 mL/kg (range 7.2 to 34.8 mL/kg), 15.0 mL/kg (range 7.5 to 18.9 mL/kg), and 8.9 mL/kg (range 6.5 to 18.0 mL/kg), respectively. In no trial did HES dosage exceed the recommended maximum of 20 mL/kg.
Randomization method was inadequate in two included trials [6, 7] and unclear in all others. Exclusion of patients after study entry was reported in four of the included trials [9, 18, 19, 21, 22]. Three trials involved some form of blinding [14, 16, 19]. One trial was unblinded [20], and blinding was unspecified in the remaining trials.
Bleeding
In adult trials cumulative mediastinal bleeding over the first 24 hours postoperatively was reported in milliliters; whereas, this value was cited in mL/kg for the two pediatric trials. In 15 of 17 randomized comparisons (88%) postoperative bleeding was lower among albumin than HES recipients, as indicated in Figure 2. The pooled standardized mean difference in bleeding favoring albumin was statistically significant (Fig 2). The magnitude of the between-group difference was similar in comparisons of albumin with both high and medium molecular weight HES.
There was no statistically significant heterogeneity with respect to postoperative bleeding among all trials (p = 0.40) or the trial subsets involving high (p = 0.08) and medium (p = 0.93) molecular weight HES administration. Because of lack of heterogeneity, the calculated standardized mean difference in bleeding and 95% CI were the same under both fixed-effects and random-effects models. There was no evidence of significant publication bias (p = 0.31).
The effects of colloid usage, age, concomitant albumin administration, and time period reported on bleeding differences are shown in Table 3. The differences in bleeding for trials of priming fluids and volume expansion were statistically significant, as were those for trials in adults and with or without concomitant albumin administration. In trials reported after 1990, the bleeding difference favoring albumin was greater than that among trials reported before 1990.
Among the 14 trials of adults, the pooled mean blood loss in the albumin group was 693 ± 350 mL compared with 789 ± 487 mL in the HES group. For these trials the estimated proportion o albumin group patients with blood loss more than 1,000 mL was 19% compared with 33% of patients exposed to HES.
Reoperation
For the nine trials reporting reoperation data, the rate of reoperation for both albumin and HES group patients combined was 5.8% (26 of 448 patients). The pooled risk of reoperation was lower in albumin than HES group patients (risk difference, -3.7%; 95% CI -8.4% to 1.0%), although the difference was not statistically significant. There was no evidence of reoperation risk difference heterogeneity (p = 0.42). The reoperation risk differences versus high and medium molecular weight HES were -2.0% (95% CI -8.7% to 4.7%) and -5.2% (95% CI -11.8% to 1.4%), respectively. Among the five trials reporting one or more reoperations, the relative risk of reoperation (0.54; 95% CI 0.25 to 1.13) favored albumin; however, this result was not statistically significant. No significant heterogeneity was noted with respect to this relative risk (p = 0.60).
Ventilatory support
In each of the four trials with ventilatory support data, the duration of intubation was shorter in albumin recipients. The pooled standardized mean difference was -0.17 (95% CI -0.43 to 0.08). Thus, the length of ventilatory support was consistently shorter with albumin, although in this relatively small subset of trials the difference could not be demonstrated to be statistically significant.
Intensive care unit stay
In three of four trials with pertinent data, length of intensive care unit stay was briefer in the albumin group. The pooled standardized mean difference was -0.21 (95% CI -0.45 to 0.03), indicating nevertheless that the difference was not statistically significant.
Blood products
For 8 of 15 (53%) randomized comparisons, blood product usage was lower in the albumin than the HES group and lower in the HES group than the albumin group in 6 of 15 (40%) comparisons. In one comparison (7%), blood product usage was equal. It was not feasible to quantify the magnitude of the differences across trials because of inconsistencies in the types of blood products administered (whole blood, plasma, red blood cells, platelets, or combinations thereof) and data reported (means or counts), as well as the absence of standard deviations or other indicators of variability in some cases.
| Comment |
|---|
|
|
|---|
The observation of less bleeding in albumin recipients was striking for its consistency, being evident in 88% of randomized comparisons. Albumin resulted in significantly less bleeding than HES in comparisons involving both volume expansion and addition of colloid to the priming fluid. Furthermore, the bleeding difference was demonstrable in a generally highly selected patient population, since for example coagulopathy and renal dysfunction were exclusion criterion in most of the included trials. Among higher risk patients undergoing cardiopulmonary bypass it is plausible that the bleeding difference might be even more pronounced.
The estimated percentage of adult patients with more than a 1,000 mL blood loss in the albumin group (19%) was lower than that in the HES group (33%). These percentages are consistent with the proportion of patients (29%) experiencing this magnitude of blood loss in a previously reported study of 5,426 cardiopulmonary bypass patients [28]. At least in some clinical centers, this volume of blood loss might trigger decisions to transfuse blood products, prolong mechanical ventilation, or reexplore for bleeding [1, 2, 4, 5, 28]. Even a modest reduction in percentages of patients with excessive bleeding might favorably affect costs of care, because in the case of cardiopulmonary bypass 20% of patients account for 80% of blood products transfused [29].
Consistent reductions in frequency of reoperation, duration of ventilatory support, and length of intensive care unit stay were apparent with albumin. However, with the more limited data available for these end points, the differences could not be shown to be statistically significant. Blood product usage was lower for the albumin group in most of the included trials.
The present findings are perhaps unsurprising in view of the well-known dose-dependent disturbances in coagulation function induced by HES. It has been suggested that such disturbances might be avoided by the use of medium versus high molecular weight HES. Indeed, in one included trial [17] mediastinal drainage was lower in recipients of medium molecular weight HES than high molecular weight HES. In another trial comparing low (120 kDa) with high (400 kDa) molecular weight HES in 45 cardiopulmonary bypass patients, the reoperation rate was greater in the high molecular weight HES group [30]. However, both low and high molecular weight HES reduced von Willebrand factor and factor VIII levels to a similar extent. Furthermore, in a recent observational study of 200 patients, medium molecular weight HES was associated with significantly greater postoperative mediastinal drainage than albumin [5]. The present meta-analysis does not provide support for reduced bleeding with lower molecular weight preparations of HES. Furthermore, the risk difference for reoperation favored albumin more strongly versus medium than high molecular weight HES.
The observed difference in bleeding might be attributable to a protective effect of albumin, a deleterious effect of HES, or a combination of the two. Albumin possesses special properties that might serve to avert bleeding after cardiopulmonary bypass, an operation associated with profound alterations in hemostasis as well as the inflammatory and complement systems. These properties of albumin include antioxidant and free radical-scavenging activity; binding affinity for lipids, drugs, toxic substances, and other ligands; and the ability to act as a specific inhibitor of apoptosis in microvascular endothelial cells.
Recent data have documented increases in free radical generation and circulating lipid peroxide levels after cardiopulmonary bypass. Quenching of free radicals and binding of lipids by albumin might favorably modify the risk of hemostatic complications. Free fatty acids, for instance, may contribute to erythrocyte crenation and loss of deformability, impairing capillary flow and tissue oxygenation. Erythrocyte alterations may affect platelet function. Albumin has been shown to block erythrocyte crenation induced by cardiopulmonary bypass. In vitro simulation of extracorporeal circulation has provided evidence that albumin can coat the fluid pathway surface, diminish fibrinogen binding sites, reduce platelet granule release, and preserve the functional and morphologic integrity of platelets.
Hydroxyethyl starch reduces von Willebrand factor and factor VIII more than can be explained because of hemodilution alone. Cardiopulmonary bypass has been shown to induce a 72% decline in platelet von Willebrand factor receptor glycoprotein Ib. Platelet refractoriness in bypass patients might at least partly reflect this fall in receptor and might be exacerbated by HES-mediated reduction in the von Willebrand factor ligand for this receptor.
Investigation of HES in cardiopulmonary bypass patients has received its primary impetus from the desire for cost savings. With greater potential for bleeding taken into account, however, albumin rather than HES might offer the lower overall total costs of care in these patients. Thus, in a retrospective case-control study involving a population of 511 patients, HES was substituted for albumin in the priming solution as a cost-saving measure [4]. Hydroxyethyl starch usage led to an increase in the rate of excessive postoperative bleeding. The median unadjusted hospital cost was $3,458 higher in patients with bleeding, and the added costs of managing bleeding complications in this study population exceeded the savings afforded by the lower acquisition cost of HES. In another retrospective study of 444 patients undergoing cardiothoracic operations [3] the percentages of patients receiving at least 1 U of allogeneic red blood cells, fresh frozen plasma, platelets, or cryoprecipitate during the first 24 hours postoperatively were significantly higher with usage of HES for intraoperative volume expansion than with albumin or crystalloid. Clearly, such differences in blood product usage would affect total costs of care.
| Acknowledgments |
|---|
|
|
|---|
| Footnotes |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
Related Article
This article has been cited by other articles:
![]() |
N. A. Nussmeier and B. E. Searles The Next Generation of Colloids: Ready for "Prime Time"? Anesth. Analg., December 1, 2009; 109(6): 1715 - 1717. [Full Text] [PDF] |
||||
![]() |
J. Boldt, S. Suttner, C. Brosch, A. Lehmann, and A. Mengistu Influence on coagulation of a potato-derived hydroxethylstarch (HES 130/0.42) and a maize-derived hydroxethylstarch (HES 130/0.4) in patients undergoing cardiac surgery Br. J. Anaesth., February 1, 2009; 102(2): 191 - 197. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Hartog, F. Brunkhorst, and K. Reinhart Hydroxyethyl Starch 130/0.4 and Surgical Blood Loss Anesth. Analg., February 1, 2009; 108(2): 672 - 672. [Full Text] [PDF] |
||||
![]() |
S. A. Kozek-Langenecker, C. Jungheinrich, W. Sauermann, and P. van der Linden Hydroxyethyl Starch 130/0.4 and Surgical Blood Loss Anesth. Analg., February 1, 2009; 108(2): 672 - 673. [Full Text] [PDF] |
||||
![]() |
A. A. Schramko, R. T. Suojaranta-Ylinen, A. H. Kuitunen, S. I. Kukkonen, and T. T. Niemi Rapidly Degradable Hydroxyethyl Starch Solutions Impair Blood Coagulation After Cardiac Surgery: A Prospective Randomized Trial Anesth. Analg., January 1, 2009; 108(1): 30 - 36. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. A. Kozek-Langenecker, C. Jungheinrich, W. Sauermann, and P. Van der Linden The Effects of Hydroxyethyl Starch 130/0.4 (6%) on Blood Loss and Use of Blood Products in Major Surgery: A Pooled Analysis of Randomized Clinical Trials Anesth. Analg., August 1, 2008; 107(2): 382 - 390. [Abstract] [Full Text] [PDF] |
||||
![]() |
F. M. Brunkhorst, C. Engel, F. Bloos, A. Meier-Hellmann, M. Ragaller, N. Weiler, O. Moerer, M. Gruendling, M. Oppert, S. Grond, et al. Intensive Insulin Therapy and Pentastarch Resuscitation in Severe Sepsis N. Engl. J. Med., January 10, 2008; 358(2): 125 - 139. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Boldt, M. Wolf, and A. Mengistu A New Plasma-Adapted Hydroxyethylstarch Preparation: In Vitro Coagulation Studies Using Thrombelastography and Whole Blood Aggregometry Anesth. Analg., February 1, 2007; 104(2): 425 - 430. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Boldt, T. Scholhorn, J. Mayer, S. Piper, and S. Suttner The value of an albumin-based intravascular volume replacement strategy in elderly patients undergoing major abdominal surgery. Anesth. Analg., July 1, 2006; 103(1): 191 - 199. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. T. Niemi, R. T. Suojaranta-Ylinen, S. I. Kukkonen, and A. H. Kuitunen Gelatin and hydroxyethyl starch, but not albumin, impair hemostasis after cardiac surgery. Anesth. Analg., April 1, 2006; 102(4): 998 - 1006. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. J. Gu and P. W. Boonstra Selection of priming solutions for cardiopulmonary bypass in adults MMCTS, January 9, 2006; 2006(0109): 1198. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Dial, E. Delabays, M. Albert, A. Gonzalez, J. Camarda, A. Law, and D. Menzies Hemodilution and surgical hemostasis contribute significantly to transfusion requirements in patients undergoing coronary artery bypass J. Thorac. Cardiovasc. Surg., September 1, 2005; 130(3): 654 - 654. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Farstad, V. L. Kvalheim, and P. Husby Cold-induced fluid extravasation during cardiopulmonary bypass in piglets can be counteracted by use of iso-oncotic prime J. Thorac. Cardiovasc. Surg., August 1, 2005; 130(2): 287 - 294. [Abstract] [Full Text] [PDF] |
||||
![]() |
Evidence-based Colloid Use in the Critically Ill: American Thoracic Society Consensus Statement Am. J. Respir. Crit. Care Med., December 1, 2004; 170(11): 1247 - 1259. [Full Text] [PDF] |
||||
![]() |
M. M. Wilkes, R. J. Navickis, J. Avorn, and W. Winkelmayer Hydroxyethyl Starch-Induced Postoperative Bleeding in Cardiac Surgery Patients: More Trials Needed? Chest, July 1, 2004; 126(1): 311 - 312. [Full Text] [PDF] |
||||
![]() |
I. Celik, J. Boldt, and H.-J. Priebe Renal dysfunction after hydroxyethyl starch. Anesth. Analg., July 1, 2004; 99(1): 309 - 310. [Full Text] [PDF] |
||||
![]() |
A. Shander, T. Rijhwani, D. Moskowitz, and R. Spence Blood and Starch in Cardiac Surgery Chest, June 1, 2004; 125(6): 2369 - 2370. [Full Text] [PDF] |
||||
![]() |
M. E. Barron, M. M. Wilkes, and R. J. Navickis A Systematic Review of the Comparative Safety of Colloids Arch Surg, May 1, 2004; 139(5): 552 - 563. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. M. Moskowitz, J. J. Klein, A. Shander, K. M. Cousineau, R. S. Goldweit, C. Bodian, S. I. Perelman, H. Kang, D. A. Fink, H. C. Rothman, et al. Predictors of transfusion requirements for cardiac surgical procedures at a blood conservation center Ann. Thorac. Surg., February 1, 2004; 77(2): 626 - 634. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Boldt New Light on Intravascular Volume Replacement Regimens: What Did We Learn from the Past Three Years? Anesth. Analg., December 1, 2003; 97(6): 1595 - 1604. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Avorn, M. Patel, R. Levin, and W. C. Winkelmayer Hetastarch and Bleeding Complications After Coronary Artery Surgery Chest, October 1, 2003; 124(4): 1437 - 1442. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Sedrakyan, K. Gondek, D. Paltiel, and J. A. Elefteriades Volume Expansion With Albumin Decreases Mortality After Coronary Artery Bypass Graft Surgery Chest, June 1, 2003; 123(6): 1853 - 1857. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. M. Wilkes Reply Ann. Thorac. Surg., May 1, 2003; 75(5): 1683 - 1683. [Full Text] [PDF] |
||||
![]() |
M. Durand and O. Chavanon Does hydroxyethyl starch increase blood loss in cardiac operations Ann. Thorac. Surg., May 1, 2003; 75(5): 1682 - 1683. [Full Text] [PDF] |
||||
![]() |
E. A. Hessel II and L. H. Edmunds Jr. Extracorporeal Circulation: Perfusion Systems Card. Surg. Adult, January 1, 2003; 2(2003): 317 - 338. [Full Text] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |