Transfusing blood to treat anaemia in normovolaemic, critically ill patients, is common practice. It has been estimated that one-third of all patients admitted to intensive care units in the developed world receive blood transfusions.1 Most intensivists prescribe blood with the belief that it enhances oxygen transport and may, therefore, relieve covert tissue hypoxia. Increasing an anaemic, critically ill patients plasma haemoglobin concentration by transfusion undoubtedly increases global oxygen delivery but there is no guarantee that an increase in regional oxygen delivery will follow. It is even less certain if increased oxygen delivery after transfusion can promote increased oxygen consumption to relieve tissue hypoxia (pathological supply dependency).
Investigators have tried to establish the existence of pathological supply dependency and define optimal oxygen transport goals in both critically ill and perioperative patients.2 3 However, they manipulated oxygen transport using fluid, inotropic and vasoactive treatments rather than blood transfusion. A transfusion threshold of 10 g dl1 was standard in these studies. Furthermore, a systematic review suggested that although some perioperative patients may benefit from augmented oxygen delivery, there was little evidence to support this approach in critically ill patients.4 There have been numerous studies evaluating the effects of blood transfusion on oxygen kinetics in a wide variety of critically ill patients.510 However, all of these studies had small patient numbers and only modest total volumes of blood were administered. Although an increase in oxygen delivery following transfusion was a consistent finding, an associated rise in oxygen consumption was observed only occasionally. A decrease in plasma lactate following transfusion was reported in only one of the studies. These results raised many questions, which include the on-going controversy over the very existence of pathological supply dependency, but it can nevertheless be concluded that there is little evidence that blood transfusion relieves tissue hypoxia in critically ill patients.
Many factors influence a clinicians decision to prescribe blood for a normovolaemic, critically ill patient including: how acutely the anaemia developed, the severity of the associated symptoms, age, and comorbidity. Most intensivists use a numeric transfusion trigger, usually a plasma haemoglobin value, as a threshold below which transfusion is indicated.11 However, this approach has been criticized as physiological variability, in terms of compensatory increases in cardiac output and oxygen extraction in response to anaemia, limits the usefulness of a single laboratory number as a general transfusion guide. This has led other workers to suggest that clinicians should use a physiological trigger, such as an oxygen extraction ratio greater than 50%, in anaemic patients as an indication for blood transfusion.12 However, this physiological trigger was calculated from studies of non-critically ill animals in which anaemia was induced by controlled euvolaemic haemodilution. Furthermore, impaired oxygen extraction is not uncommonly observed in human critically ill subjects. Even if the oxygen extraction ratio were a reliable indicator of sub-optimal oxygen delivery resulting from anaemia, it would remain difficult to define a number that could be utilized safely as a transfusion trigger in critically ill patients. Such difficulties surrounding physiological triggers have contributed to the widespread adoption of numerical transfusion triggers in clinical practice.
In 1982, Allen and colleagues suggested the 10/30 rule (10 referring to plasma haemoglobin of 10 g dl1 and 30 to the haematocrit) as a clinically useful transfusion trigger.13 Although the 10/30 rule was derived from perioperative patient studies, the rule was enthusiastically adopted throughout medical practice. It became particularly popular in critical care medicine because: the daily measurement of plasma haemoglobin was a cheap and routine investigation; a haemoglobin of 10 simplified oxygen kinetic calculations; and it was an easy number to remember! Nevertheless, the 10/30 rule as a transfusion trigger in the critically ill has never been supported by a randomized controlled trial.
Despite the 10/30 rules past popularity in critical care medicine, the increasing costs and decreasing availability of blood products have forced clinicians to reconsider the rule and ask if a transfusion trigger of less than 10 g dl1 could be used safely. Other reasons to reconsider transfusion triggers include the rare but well-recognized risks of transfusion such as infection with human immunodeficiency and infectious hepatitis viruses, and ABO mismatch. There are three other risks of transfusion that are particularly relevant to critically ill patients. First, blood transfusion can cause leucocytosis that may be of uncertain pathological significance but can certainly cause diagnostic confusion.14 Secondly, microcirculatory risks may also follow transfusion because storage of blood decreases red cell 2,3-diphosphoglycerate concentrations thus impairing the deformability of the cell. Poorly deformable red cells flow through the microcirculation inefficiently and this may partly explain the poor correlation between blood transfusion and increased oxygen consumption. Marik and colleagues demonstrated increased splanchnic ischaemia following transfusion with old blood in patients with sepsis.8 Furthermore, in a rat sepsis model, transfusion with fresh blood increased oxygen consumption whereas old blood failed to do so.15 In summary, blood transfusion (particularly with old blood) can paradoxically decrease microcirculatory oxygen delivery and contribute to tissue hypoxia.
Thirdly, in vitro studies have suggested that blood transfusion causes immunosuppression by decreasing cell-mediated immunity, reducing non-killer cell activity, suppressing macrophage antigen presentation, altering T-cell ratios, and decreasing the concentrations of cytokines (TNF, IFN-, and GM-CSF) that are vitally important in immune responses.16 Randomized controlled trials studying the effects of allogeneic transfusion in colorectal surgery, demonstrated a correlation between volume of transfusion and the incidence of postoperative infections.17 18 Although these were perioperative patient studies, they nevertheless suggest that transfusion-associated immunosuppression may occur in critically ill patients. There is evidence that the white cells present in transfused blood are the cause of this immunosuppression.16 In another study of colorectal surgery patients, there was a lower incidence of postoperative infections in patients transfused with leucocyte-depleted blood.19 Although leucocyte depletion is a costly process, it has been widely adopted throughout Europe.
In contrast to the abundance of evidence that allogeneic blood transfusion may be detrimental, there is a paucity of literature addressing how well critically ill patients tolerate anaemia. It has been documented that healthy animals and surgical patients cope well with profound anaemia.20 Other authors have suggested that perioperative bleeding is a more important factor than plasma haemoglobin concentrations in determining postoperative mortality and morbidity. In a surgical study, plasma haemoglobin values as low as 6 g dl1 were well tolerated provided operative blood loss was less than 500 ml.21 In a large study of Jehovahs Witness patients undergoing major surgery, a 1.4% mortality was attributed to anaemia, with 90% of the deaths occurring in patients undergoing cardiovascular surgery.22 These studies suggest that plasma haemoglobin measurements significantly less than 10 g dl1 are generally well tolerated by surgical patients. However, it remains unclear if this applies to critically ill patients.
At what plasma haemoglobin concentration do the risks of anaemia outweigh the potential risks of blood transfusion in critically ill patients? When is blood transfusion justified? Until 2 yr ago, there had been only five randomized controlled studies comparing different transfusion strategies in critically ill and perioperative patients.2327 Perhaps not surprisingly all of these small studies failed to answer this question. Until recently the only published studies available to help clinicians decide when to transfuse anaemic critically ill patients have been either observational or small, poorly controlled, clinical studies based on perioperative rather than critically ill patients.
Two years ago, an attempt was made to answer this question when the Transfusion Requirements in Critical Care (TRICC) Investigators, on behalf of the Canadian Critical Care Trials Group, published the results of a large multicentre, randomized, controlled clinical trial of transfusion requirements in critical care.28 After conducting an observational study to assess transfusion practice variation throughout Canada,29 the investigators enlisted 838 patients over 3 yr from 25 centres. Only normovolaemic, anaemic (plasma haemoglobin <9 g dl1) patients who were expected to stay in the intensive therapy unit for more than 24 h were included in the study. Important exclusion criteria were evidence of active bleeding (>3 units transfusion over 24 h), chronic anaemia (plasma haemoglobin <9 g dl1 in the preceding month) and patients undergoing cardiac surgery. The enlisted patients were randomized to receive either a restrictive transfusion strategy (transfusion trigger of 7 g dl1 and a maintenance range of 79 g dl1) or a liberal strategy (transfusion trigger of 10 g dl1 and a maintenance range of 1012 g dl1). Single unit transfusions were administered and plasma haemoglobin checked before further transfusion was considered.
The procedures were applied successfully as average daily haemoglobin concentrations were 8.5 and 10.7 g dl1 in the restrictive and liberal strategy groups, respectively. On average, a total of 2.6 units of blood were administered to patients randomized to the restrictive strategy group compared with 5.6 units for patients in the liberal strategy group. Patients in the restrictive strategy group received 54% less blood by volume than patients in the liberal strategy group. One-third of the patients in the restrictive strategy group did not require transfusion whereas all patients in the liberal strategy group were transfused. There was a non-significant trend towards decreased 30-day all-cause mortality (the primary outcome measure), in favour of patients in the restrictive strategy group and the same pattern was observed for the secondary outcome measures of 60-day all-cause and ICU mortalities. Also, there was a significantly lower adjusted multiple organ dysfunction score (secondary outcome measure), again in favour of patients in the restrictive strategy group. Subgroup analysis revealed that young patients (<55 yr old), and patients with APACHE II <20 were the only subgroups to have a significantly better primary outcome measure after randomization to the restrictive strategy group. There was no significant difference between the two strategies for any other subgroups including patients with septic shock and with cardiovascular disease. There were significantly fewer cardiac complications, including acute myocardial infarction and pulmonary oedema, observed in patients in the restrictive strategy group.
The TRICC investigators concluded that a restrictive strategy is at least comparable with a liberal strategy in critically ill patients and may be better because it is associated with: a trend towards improved 30-day all-cause mortality; a significant decrease in hospital mortality, adjusted multiple organ dysfunction score, and cardiovascular complications; and decreased transfusion by 54% and transfusion exposure by 33%. The TRICC investigators recommended the restrictive strategy as best practice for most patients including those with cardiovascular disease, but with the possible exception of critically ill patients with on-going coronary ischaemia. Doubt about the appropriateness of a restrictive strategy approach for such patients was supported by the publication of a later subgroup analysis that suggested that patients with severe cardiac disease randomized to the restrictive strategy group had a non-significant increase in 30-day all-cause mortality.30 Unfortunately, the TRICC study excluded patients with chronic anaemia and those undergoing cardiac surgery and so it remains difficult to recommend a transfusion strategy for either of these groups on the basis of this study.
Most studies addressing transfusion practice in critically ill patients have been conducted in North America. However, a prospective, epidemiological survey of anaemia and transfusion practice in Western Europe (Anaemia and Blood Transfusion in Critical Care Study) has been completed but the results await publication. In the long term, new therapies such as human recombinant erythropoietin and artificial blood substitutes may decrease blood transfusion requirements. However, at this moment the simplest, safest, and most cost-effective way to reduce transfusion requirements in critically ill patients is the adoption of a lower transfusion trigger as suggested by the TRICC investigators, rather than the premature introduction of clinically unproven, expensive, new treatments with uncertain safety profiles.
L. McCrossan
G. Masterson
Intensive Care Unit
Royal Liverpool University Hospital
Prescot Street
Liverpool L7 8XP
UK
References
1 Wallace EL, Churchill WH, Surgenor DM, et al. Collection and transfusion of blood and blood components in the United States. Transfusion 1995; 35: 80212[ISI][Medline]
2 Gattinoni L, Brazzi L, Pelosi P. A trial of goal-orientated hemodynamic therapy in critically ill patients. N Engl J Med 1995; 333: 102532
3 Shoemaker WC, Appel PL, Kram HB. Prospective trial of supranormal values of survivors as therapeutic goals in high-risk surgical patients. Chest 1988; 94: 117686[Abstract]
4 Heyland DK, Cook DJ, King D, Kernerman B, Brun-Buisson C. Maximizing oxygen delivery in critically ill patients: a methodological appraisal of the evidence. Crit Care Med 1996; 24: 51724[ISI][Medline]
5 Shah DM, Gottlieb ME, Rahm RL, Stratton HH, Barie PS, Paloski WH. Failure of red blood cell transfusion to increase oxygen transport or mixed venous PO2 in injured patients. J Trauma 1982; 22: 7416[ISI][Medline]
6 Dietrich KA, Conrad SA, Hebert CA, Levy GL, Romero MD. Cardiovascular and metabolic response to red blood cell transfusion in critically ill volume-resuscitated nonsurgical critically ill patients. Crit Care Med 1990; 18: 9404[ISI][Medline]
7 Steffes CP, Bender JS, Levison MA. Blood transfusion and oxygen consumption in surgical sepsis. Crit Care Med 1991; 19: 5127[ISI][Medline]
8 Marik PE, Sibbald WJ. Effect of stored-blood transfusion on oxygen delivery in patients with sepsis. J Am Med Assoc 1993; 269: 30249[Abstract]
9 Lorente JA, Landin L, De Pablo R, Renes E, Rodriguez-Diaz R, Liste D. Effects of blood transfusion on oxygen transport variables in severe sepsis. Crit Care Med 1993; 21: 13128[ISI][Medline]
10 Conrad SA, Dietrich KA, Hebert CA, Romero MD. Effect of red cell transfusion on oxygen consumption following fluid resuscitation in septic shock. Circ Shock 1990; 31: 41929[ISI][Medline]
11 Cane RD. Hemoglobin: how much is enough? Crit Care Med 1990; 18: 10467[ISI][Medline]
12 Gould SA, Rosen AL, Sehgal RC. Oxygen extraction ratio: a physiological indicator of transfusion need. Transfusion 1983; 23: 416
13 Allen JB, Allen FB. The minimum acceptable level of hemoglobin. Int Anesthiol Clin 1982; 20: 122
14 Fenwick JC, Cameron M, Naiman SC, et al. Blood transfusion as a cause of leucocytosis in critically ill patients. Lancet 1994; 344: 8556[ISI][Medline]
15 Fitzgerald RDS, Martin CM, Dietz GE, Doig GS, Potter RF, Sibbald WJ. Transfusion red blood cells stored in citrate phosphate dextrose adenine-1 for 28 days fails to improve tissue oxygenation in rats. Crit Care Med 1997; 25: 72632
16 Kaplan J, Sarnaik S, Gitlin J. Diminished helper/suppressor lymphocyte ratios and natural killer activity in recipients of repeated blood transfusions. Blood 1984; 64: 30810[Abstract]
17 Houbiers JGA, Brand A, van der Watering LMG, Hermans J, Verwey PJM, Bijnen AB. Randomised controlled trial comparing transfusion of leucocyte-depleted or buffy-coat-depleted blood in surgery for colorectal cancer. Lancet 1996; 344: 5738
18 Jenson LS, Anderson AJ, Christiansen PM, Hokland P, Juhl CO, Madsen G. Postoperative infection and natural killer cell function following transfusion in patients undergoing elective colorectal surgery. Br J Surg 1992; 79: 5136[ISI][Medline]
19 Jenson LS, Kissmeyer-Nielsen P, Wolff B, Qvist N. Randomised comparison of leucocyte-depleted versus buffy-coat-poor blood transfusion and complications after colorectal surgery. Lancet 1996; 348: 8415[ISI][Medline]
20 Weiskopf RB, Viele MK, Feiner J, et al. Human cardiovascular and metabolic responses to acute, severe, isovolemic anaemia. J Am Med Assoc 1988; 279: 21721
21 Spence RK, Carson JA, Poses R. Elective surgery without transfusion: influence of preoperative hemoglobin level and blood loss on mortality. Am J Surg 1990; 159: 3204[ISI][Medline]
22 Kitchens CS. Are transfusions overrated? Surgical outcome of Jehovahs Witnesses. Am J Med 1993; 94: 1179[ISI][Medline]
23 Weisel RD, Charlesworth DC, Mickleborough LL, Fremes SE, Ivanov J, Mickle DAG. Limitations of blood conservation. J Thorac Cardiovasc Surg 1984; 88: 2638[Abstract]
24 Johnson RG, Thurer RL, Kruskall MS, Sirois C, Gervino EV, Critchlow J. Comparison of two transfusion strategies after elective operations for myocardial revascularization. J Thorac Cardiovasc Surg 1992; 104: 30714[Abstract]
25 Fortune JB, Feustel PJ, Saifi J, Stratton HH, Newell JC, Shah DM. Influence of hematocrit on cardiopulmonary function after acute hemorrhage. J Trauma 1987; 27: 2439[ISI][Medline]
26 Hebert PC, Wells GA, Marshall JC, Martin CM, Tweedale M, Pagliarello G. Transfusion requirements in critical care. A pilot study. J Am Med Assoc 1995; 273: 143944[Abstract]
27 Blair SD, Janvrin SB, McCollum CN, Greenhalgh RM. Effect of early blood transfusion on gastrointestinal haemorrhage. Br J Surg 1986; 73: 7835[ISI][Medline]
28 Hebert PC, Wells G, Blajchman MA, et al. A multicenter, randomised, controlled clinical trial of transfusion requirements in critical care. N Engl J Med 1999; 340: 40917
29 Hebert PC, Wells G, Martin C, et al. A Canadian survey of transfusion practices in critically ill patients. Crit Care Med 1998; 26: 4827[ISI][Medline]
30 Hebert PC, Yetisir E, Martin C, et al. Is a low transfusion threshold safe in critically ill patients with cardiovascular diseases? Crit Care Med 2001; 29: 22733[ISI][Medline]