Blood Products (Anesthesia Text)

Red Blood Cells

Introduction to RBC Use

Blood Compatibility for Red Cell Transfusions

Major cross matching occurs when the donor erythrocytes are mixed with the recipient’s plasma (minor cross matching mixes the donor plasma with recipient erythrocytes). In an urgent situation, one can do a partial cross match (in which macroscopic agglutination is looked for) in < 5 minutes. In truly urgent situations, give type-specific non-cross-matched blood. O-negative should be the last option because it can contain high levels of anti-A and anti-B antibodies.
Screening simply analyzes donor blood for common antibodies. The risk of a significant hemolytic reaction following screening in type-specific blood is 1:10,000 units. [Stoelting RK. Basics of Anesthesia, 5th ed. Elsevier (China) p. 356, 2007]

Stored at 4° C in an anticoagulated preservative that contains citrate, phosphate, and dextrose (CPD). Citrate binds calcium and anticoagulates. Phosphate retards breakdown of 2,3-DPG, and dextrose serves as fuel. In these conditions, RBCs are viable for at least 21 days [Am Assoc Blood Banks Tech. Manual p. 37. p. 635, 1990]. That said, 2,3-DPG declines with age, which may have adverse effects – A study of 63 trauma patients who received 6-20 units of blood identified mean age of blood, number of units older than 14 days, and number of units older than 21 days as independent risk factors for MOF. [Zallen G et. al. Am J Surg 178: 570, 1999]

PRBCs can be combined with saline through a Y-connector to give approximately equal amounts as in whole blood, but be sure not to use LR as the solvent because it contains calcium which will clot the PRBCs [Am Assoc Blood Banks Tech. Manual p. 341, 1990]. RBCs are usually passed through filters which can impede the flow of blood – they should be changed after ~ 4 units of blood. Smaller filters (< 170 microns) may reduce pulmonary complications of RBC transfusions but this is unproven [Ann Emerg Med 17: 327, 1988]. Warming blood has two effects – the first and most important is prevention of hypothermia [Transfusion 31: 558, 1991], but also reduction in viscosity with increases in flow rates of up to 50% [J Trauma 21: 480, 1981]. The ideal temperature is 33 – 35° C, as temperatures approaching 37° C can lead to hemolysis. [Transfusion 31: 558, 1991]


Changes in hemoglobin and hematocrit are misleading because they depend on plasma volumes – clinical studies have shown that there is a poor correlation between changes in hematocrit/hemoglobin and changes in RBC volume. [Br J Hemtaol 6: 228, 1990; Sur Gynecol Obstet 175: 243, 1992]

Anemia decreases the viscosity of blood, and while RBC mass decreases, the loss in oxygen transportation capacity may be partially attenuated by the resultant increases in blood flow. Early studies suggested that hemoglobin < 7 g/dL would raise cardiac output, but more recent studies suggest that the figure is closer to 4.5 g/dL [Circulation 39: 503, 1969; Crit Care Med 7: 380, 1979]. Similarly, the addition of RBCs can decrease cardiac output [J Trauma 22: 741, 1982]. A recent study of 11 patients at UCSF showed that a Hgb of 5 mg/dL produces no changes in VO2, lactate concentration, or EKG appearance provided that volume is maintained (albumin/autologous plasma were used) [JAMA 279: 217, 1998] – in fact, VO2 increased from 3.07 mL O2/kg/min to 3.42 mL O2/kg/min (p < 0.001). Lower hematocrits have even been shown to increase blood flow preferentially to the cardiac and cerebral circulations. [Crit Care Clin 20: 225, 2004]

Also, tissue oxygen extraction increases in response to anemia, and in unconscious primates on 100% oxygen the hematocrit can drop to 10% before lactate levels rise [J Surg Res 42: 629, 1987], although the exact threshold in human beings on suboptimal oxygen is not known.

Some have suggested that oxygen extraction ratio is a more reliable marker of transfusion requirements than hematocrit or hemoglobin [J Surg Res 42: 659, 1987; J Trauma 32: 769, 1992] – this can be measured continuously using a pulse oximeter and mixed venous oximetry. O2ER values > 0.5 may be an ideal transfusion trigger.

The efficacy of RBC transfusions does not live up to its theoretical limit, as RBCs can increase blood viscosity and lower cardiac output [J Trauma 22: 741, 1982]. Marino’s data on RBC transfusions into euvolemic anemic post-op patients suggests that there is tremendous variability in the effects of blood on VO2 – it should not be taken for granted that RBCs will increase tissue oxygenation.

Indications for Transfusion[edit]

Indications for Transfusion
Evidence of impaired tissue oxygenation: VO2 < 100 mL/min/m2
O2ER > 0.5 in patients with adequate cardiac output
Correction of Hb < 7.0 in critically ill patients [NEJM 340: 409, 1999]

The decision about when to transfuse the operative patient is a difficult one. Clearly, at some point DO2 can be adversely affected by low hemoglobin, but this risk must be weighed against the rare, but real risks associated with transfusion. Unfortunately, there is very little quality data available to guide the practitioner.

First, one must determine whether the patient simply needs volume (which can be provided in the form of crystalloids or colloids), or increased oxygen carrying capacity (i.e., red blood cells). The traditional indications for volume resuscitation are unreliable at best – tachycardia (insensitive and non-specific in the anesthetized patient), hypotension, decreased CVP, and low urine output are all traditionally thought of as indicators for volume infusion, and while most practitioners would agree that a hypotensive, tachycardic patient with a CVP of 6 mm Hg and no urine output would benefit from an infusion of volume, the data on individual parameters and fluid resuscitation in general is mixed. [Chappel D et. al. Anesthesiology 109, 723: 2008]

Very few controlled studies examine blood transfusion in the surgical patient [Stoelting RK. Basics of Anesthesia, 5th ed. Elsevier (China) p. 357, 2007]. The majority of studies are in non-surgical patients, thus requiring both extrapolation and skepticism. Based on Ely’s study, and assuming that critically ill patients have similar transfusion requirements as surgical patients, it seems reasonable to transfuse at 7.0 mg/dL in patients with no cardiac history (data from a subset of patients with a cardiac history in Ely’s study was indeterminate).

Important Studies[edit]

Transfusion Criteria [Hebert et al. NEJM 340: 409, 1999][edit]

838 critically ill patients with euvolemia after initial treatment who had hemoglobin concentrations of < 9 g/dL within 72 hours after admission to the ICU and randomly assigned 418 patients to a restrictive strategy of transfusion, in which red cells were transfused if the hemoglobin concentration dropped below 7.0 g/dL and hemoglobin concentrations were maintained at 7.0 to 9.0 g/dL, and 420 patients to a liberal strategy, in which transfusions were given when the hemoglobin concentration fell below 10.0 g/dL and hemoglobin concentrations were maintained at 10.0 to 12.0 g/dL.

Results: The mortality rate during hospitalization was significantly lower in the restrictive-strategy group (22.2 percent vs. 28.1 percent, p = 0.05). Overall, 30-day mortality was similar in the two groups (18.7 percent vs. 23.3 percent, p= 0.11). Overall mortality rates were significantly lower with the restrictive transfusion strategy among patients who were less acutely ill — those with an APACHE II score of «20 (8.7 percent in the restrictive-strategy group and 16.1 percent in the liberal-strategy group, P=0.03) — and among patients who were less than 55 years of age (5.7 percent and 13.0 percent, respectively; P=0.02), but not among patients with clinically significant cardiac disease (20.5 percent and 22.9 percent, respectively; P=0.69) [Hebert et al].

EPO in the ICU [JAMA 288: 2827, 2002][edit]

Prospective, randomized, double-blind, placebo-controlled, multicenter trial in medical, surgical, or a medical/surgical ICU in 65 participating institutions. 40,000 units of rHuEPO or placebo on ICU day 3 and continued weekly for patients who remained in the hospital, for a total of 3 doses. Patients in the ICU on study day 21 received a fourth dose. Patients receiving rHuEPO were less likely to undergo transfusion (60.4% placebo vs 50.5% rHuEPO; P<.001). There was a 19% reduction in the total units of RBCs transfused (1963 vs 1590). Increase in hemoglobin from baseline to study end was greater in the rHuEPO group (mean [SD], 1.32 [2] g/dL vs 0.94 [1.9] g/dL; P<.001). Mortality (14% for rHuEPO and 15% for placebo) and adverse clinical events were not significantly different.

Transfusion and Infection [Crit Care Med 34: 2302, 2006][edit]

Prospective, observational, cohort study. Single-center, medical/surgical, closed ICU. Of the 2,085 patients enrolled, 21.5% received red blood cell transfusions. The posttransfusion nosocomial infection rate was 14.3%, significantly higher than that observed in nontransfused patients (5.8%; p < 0.0001, chi-square). In a multivariate analysis controlling for patient age, maximum storage age of RBC, and number of PRBC transfusions, only the number of transfusions was independently associated with nosocomial infection (odds ratio 1.097; 95% confidence interval 1.028-1.171; p = .005). When corrected for survival probability, the risk of nosocomial infection associated with red blood cell transfusions remained statistically significant (p < .0001). Leukoreduction tended to reduce the nosocomial infection rate but not significantly.

Hemoglobin and SAH [Neurosurgery 59: 775, 2006][edit]

Of 103 patients, patients who died had lower hgb than survivors on Days 0, 1, 2, 4, 6, 10, 11, and 12 (p <= 0.05). Higher mean hgb was associated with reduced odds of poor outcome (p = 0.008) after correcting for Hunt and Hess grade, age, and vasospasm; results for hgb on Days 0 and 1 were similar. Higher Day 0 (p = 0.05) and mean hgb (p = 0.009) predicted a lower risk of cerebral infarction independent of vasospasm. There were no associations between hgb and other prognostic variables.

Transfusion in Cardiac Surgery (RCT) [Hajjar LA et al. JAMA 304: 1559, 2010][edit]

Single center RCT in Brazil involving over 500 patients randomized to liberal (transfuse to >30% hematocrit) vs conservative (>24%) transfusion strategy in patients undergoing routine CABG and/or valve surgery requiring bypass. The intervention significantly reduced PRBCs transfusion (78% and 47% receiving, respectively). Morbidity and mortality was the same in both groups.

Transfusion in Cardiac Surgery (Retrospective) [Bennett-Guerrero E, JAMA 304: 1568, 2010][edit]

Researches using the STS database found transfusion practices among 700 US hospitals for 100,000 patients undergoing CBP showed wide variance in the rates of product transfusion (RBC (7.8%-92.8%), plasma (0%-97.5%), and platelet (0.4%-90.4%)). These differences persisted after adjusting for hospital and patients factors. Like the TRACS trial, there was no apparent difference in mortality or morbidity.

Post-Operative Transfusion in Hip Surgery (Prospective, RCT) [Carson JL et al. NEJM 365: 2453, 2011][edit]

Carson JL et al. randomized 2016 patients considered at “high risk” for cardiac morbidity (either a history of or risk factors for cardiovascular disease) to transfusion thresholds of 8 versus 10 post-operatively. 40% of patients had known CAD in both groups. The primary outcome (death or an inability to walk 10 ft [or across a room] without human assistance at the 60-day follow-up) was no different between groups (35.2% vs. 34.7% in liberal and restricted, respectively). Importantly, only 3% of the patients in this study required the provision of critical care in an intensive care unit. Furthermore, the median units of blood transferred were 2 and 0 units in the liberal and restricted groups, respectively [Carson et al.].

Complications of Blood Product Transfusion[edit]

Blood Bank Concepts[edit]

Some authors (Miller’s Anesthesia, 6th edition, Chapter 55) have questioned whether or not a cross match is really needed. They cite the following statistics – only 1% of previously transfused or pregnant patients (i.e. some prior exposure to non-native red blood cell antigens) will have any non-A or non-B antibodies. Many of these are not reactive at temperatures above 30C. If anti-Rh(D) is accounted for, only 0.1% of these patients will have reactive (A, B, D) antibodies. Thus, a simple ABO-Rh type reduces the risk of a transfusion reaction to 99.8%. Screening lowers this risk to 99.94%, and crossmatching lowers it to 99.95% [Polesky HF, Walker RH, ed. Safety and Transfusion Practices, Skokie, IL: College of American Pathologists p. 79, 1982]. Is the crossmatch worth this extra 0.01% of risk (i.e. 1:10,000)?

Type and Cross[edit]

What does a “type and cross” actually mean? As always, both the recipient and donor cells are ABO-Rh typed – this alone reduces the risk of a transfusion reaction to 0.2%. In order to further decrease this risk, a small sample of donor blood can be mixed with the recipient’s serum. In the first phase (immediate phase, 1-5 mins, room temp), ABO errors antibodies and MN, P, and Lewis system antibodies are sought. In the second phase (incubation phase, 10-20 minutes), the first phase reactants are heated to 37C in salt solution (or for 30-45 minutes in albumin), additional antibodies (mostly to Rh, but also partial or incomplete Ab) are detected, as the salt solution and/or albumin can facilitate agglutination. In the last phase, (antiglobulin phase), antiglobulin sera are added, further increasing the ability of the crossmatch to detect incomplete antibodies (ex. Rh, Kell, Kidd, Duffy). This third step is not essential.

Type and Screen[edit]

In a type and screen, as in a type and cross, the recipient and donor cells are ABO-Rh typed (risk of transfusion reaction 0.2% after simply doing an ABO-Rh type). Subsequently, “standard” blood cells (with known, significant non-ABO-Rh-antibodies) are added to the patient’s serum. The major advantage of the type and screen is that it can be performed prior to the operation (ex. a patient’s blood and serum can be typed and screened well in advance of the day of surgery), and, if negative, a 99.94% transfusion risk is assured with ABO-Rh-matched blood. If an antibody is found, the blood bank can then give donor blood negative for the identified antibody, although the blood bank may choose to crossmatch the donor’s blood with the recipient’s serum. An important point is that while typing and screening cannot eliminate all potential transfusion reactions (clinically insignificant reactions may still approach 1%), the vast majority of transfusion reactions following a type and screen are benign – in fact, in a study of 13,950 patients, Oberman et al. found only eight “clinically significant” antibodies that were detected by complete crossmatch but not during antibody screening[Oberman et al.] (thus, the risk of hemodynamically significant transfusion reactions after a negative type and screen is approximately 0.057%).

Adverse Reactions[edit]

Risks of Homologous Blood Transfusion
Complication Frequency
Fever, chills, urticaria 1:100
Hemolysis 1:6000
Fatal hemolysis 1:100,000
Anaphylaxis 1:500,000
Bacterial contamination 1:25,000
Hepatitis B 1:250,000
Hepatitis C 1:1,000,000
HIV 1:2,000,000

(Acute lung injury 1:5000 – < 10% fatality, leading cause of death)

Transfusion Reactions[edit]

Common symptoms of Acute hemolytic reaction

Transfusion reactions can be difficult to detect, as general anesthesia tends to ameliorate the most common signs and symptoms [Kopke PM, Holland PV. Transfus Clin Biol 8: 278, 2001]. One should be suspicious of increased peak airway pressures, hyperthermia, or changes in urine output or color in the context of a blood transfusion [Stoelting RK. Basics of Anesthesia, 5th ed. Elsevier (China) p. 360, 2007]. One should also be suspicious of urticaria, hypotension, tachycardia, and microvascular bleeding. [Stoelting RK. Basics of Anesthesia, 5th ed. Elsevier (China) p. 361, 2007].

Hemolytic Reactions[edit]

Caused by wrong blood type transfusion, acute hemolytic reactions are an antibody reaction to ABO surface antigens on donor erythrocytes and are rarely life threatening (~ 1:100,000) – one unit of RBC can by lysed in an hour, leading to a severe immunologic response that can be fatal. Hemolysis, spontaneous hemorrhage, complement activation, and renal failure are possible [Stoelting RK. Basics of Anesthesia, 5th ed. Elsevier (China) p. 361, 2007]. This error is usually clerical [Heart Lung 20: 506, 1991]. Severe reactions can occur with as little as 10 mL of blood [World J Surg 11: 25, 1987] – common initial signs include fever, dyspnea, chest pain, and low back pain. Hypotension can develop very suddenly. Hypotension is the only obvious sign under general anesthesia, thus always maintain a high level of suspicion. Severe reactions are accompanied by a consumptive coagulopathy and MOD. Acute renal failure occurs in 5 – 10% of cases [Anesth Intens Care 21: 15, 1993]. Treat by immediately discontinuing the transfusion, and consider fluid resuscitation with the addition of mannitol or furosemide. Bicarbonate has not been proven to be helpful [Stoelting RK. Basics of Anesthesia, 5th ed. Elsevier (China) p. 361, 2007]

Treatment of Acute Hemolytic Reactions
1. Stop the transfusion
2. Check blood pressure
  • If BP dropping, give volume
  • If BP dropping, start vasopressors (no data as to which are superior)
Once the patient is stabilized…
3. Obtain a blood sample and inspect plasma for pink/red hue of hemoglobin
4. Obtain a fresh urine specimen and dipstick for blood
5. Send blood for direct Coomb’s

Febrile Reactions[edit]

While not as potentially devastating as acute hemolytic reactions, febrile non-hemolytic reactions are much more common (0.5% of transfusions). This is a reaction of antibodies to leukocytes in the donor’s blood (Abs produced in prior transfusions or pregnancies). These fevers will begin within 6 hours (after 6 hours look to a different etiology, ex. hemolysis). The goal is to exclude the possibility of hemolysis, so go through the same algorithm (see above). Some recommend routine culture of donor and recipient blood (if there is any sign of systemic illness, such as rigors, dyspnea) for the remote possibility of infection. > 50% of these patients will never have another transfusion reaction, so leukocyte-poor blood is not needed unless a second reaction occurs.

Allergic Reactions[edit]

Allergic reactions (rash, anaphylaxis) are a result of sensitivity to donor plasma proteins, usually beginning with urticaria and possibly with fever. Mild urticaria does not require intervention in the absence of fever, however it is common practice to stop the transfusion and give diphenhydramine 25-50 mg PO or IM q6h – the only benefit is relief of pruritus. If the patient has true anaphylaxis, treat it as such (also test these patients for IgA deficiency and avoid future transfusions if at all possible).

Acute Lung Injury[edit]

Acute lung injury is possible, but only occurs in 1:5000 transfusions [Intensive Care Med 14: 654, 1988]. The theory is that donor antileukocyte antibodies bind host granulocytes, sequestering them in the pulmonary microcirculation and leading to ARDS. Unlike most cases of ARDS, this variety is fatal in < 10% [Crit Care Med 34S: S114, 2006]. Dyspnea and/or hypoxemia may arise within a few hours, and CXR may show diffuse infiltrates. You CANNOT get pulmonary edema from PRBC because the osmotic pressure is too high – if you see what you think is edema, it’s TRALI/ARDS. The process generally resolves within a week. Stop the transfusion and manage as you would ARDS.

Infectious Disease[edit]

Highest risk of viral transmission is hepatitis B (1:220,000). < 10 fatalities per year in United States from bacterial infections


TRALI is an acute syndrome of dyspnea, hypoxemia, and non-cardiogenic pulmonary edema usually occurring within 6 hours of transfusion and is now the leading cause of mortality following blood transfusion (as of 2005). Stop all transfusions if ongoing, and if possible consider suctioning fluid from the endotracheal tube to send for protein count.


Allogenic transfusions suppress cell-mediated immunity, and may place patients at risk for post-operative infection.

Data Refuting a RBC/Infection Association[edit]

Vamvakas et. al. studied the records of 492 patients who underwent colorectal cancer resection and calculated the probability of infection in association with transfusion with and without adjustment for the effects of chronic systemic illness, number of days with urinary catheter, endotracheal intubation, impaired consciousness, and specific risk factors for wound infection. After adjustment for the effects of the aforementioned variables, allogeneic transfusion was not associated with postoperative infection at any site (p = 0.407). However, in a secondary analysis (data mining?) that adjusted for the effects of only the 18 confounders considered by previous authors, transfusion was the most significant predictor of infection. In that analysis, the risk of postoperative infection increased by 14 percent per unit of red cells transfused (p < 0.001). [Vamkakas EC et. al. Transfusion 36: 1000, 1996]

In Hébert et al.’s prospective, randomized study of critically ill patients, there was no statistically significant difference in pneumonia, bacteremia, catheter sepsis, or septic shock. [Hébert et al. NEJM 340: 409, 1999]

Vamkakas repeated this study on 416 CABG patients – on univariate analysis, patients who did (n = 64) or did not develop infection received 956.6 +/- 180.6 and 321.3 +/- 39.6 mL of plasma, respectively (p<0.0001). In multivariate analyses, the volume of transfused allogeneic plasma was not associated with postoperative pneumonia and/or wound infection (p = 0.24), pneumonia (p = 0.21), or wound infection (p = 0.74). [Vamvakas et. al. Transfusion 42: 107, 2002]

Ali et. al. prospectively studied 232 patients undergoing cardiac surgery, 50% of whom received blood product transfusion. There were no differences in the frequency of chest infection (20% versus 15%, p = 0.38), urinary infection (3.5% versus 5.3%, p = 0 0.75), wound infection (3.5% versus 8.0%, p = 0.16), or overall infection (28% versus 30%, p = 0.89) comparing the transfused versus untransfused groups. There was no evidence to suggest that administration of blood products was associated with infection (odds ratio 0.92, p = 0.77). [Ali ZA et. al. Ann Thorac Surg 78: 1542, 2004]

Data Supporting a RBC/Infection Association[edit]

Leal-Noval et. al. studied 738 patients in a post-operative ICU, and examined the influence of 36 variables on the development of severe postoperative infections (SPIs) in general and individually for pneumonia, mediastinitis, and/or septicemia. After multivariate analysis, the variables associated with SPI (incidence, 9.4%) were reintubation, sternal dehiscence, mechanical ventilation (MV) for > or = 48 h, reintervention, neurologic dysfunction, transfusion of >= 4 U RBCs, and systemic arterial hypotension. The mortality rate (patients with SPI, 52.8%; non-SPI patients, 8.2%; p < 0.001) was greater for the infected patients. The transfused patients also had a greater mortality rate (13.3% vs 8.9%, respectively; p < 0.001) and a longer mean stay in the ICU (6.1 +/- 7.2 days vs 3.7 +/- 2.8 days, respectively; p < 0.01) than those not transfused. Keep in mind that this was not a prospective study, thus the sicker patients may have been transfused. [Leal-Noval SR et. al. Chest 119: 1461, 2001]

Bochicchio GV et. al. conducted a prospective observational cohort study of 766 trauma patients admitted to the intensive care unit (ICU), who received mechanical ventilation (MV) for >= 48h, and who did not have pneumonia on admission. Logistic regression analyses controlled for all variables related significantly to VAP by univariate analysis (sex, Injury Severity Score, and ventilator days and ICU length of stay prior to VAP) and found that transfusion of blood products was an independent risk factor for VAP. All blood products were associated with a higher risk of VAP (RBC: odds ratio [OR] 4.41; 95% confidence interval [CI] 1.00, 19.54; p = 0.05; FFP: OR 3.34; 95% CI 1.18, 9.43; p = 0.023; platelets: OR 4.19; 95% CI 1.37, 12.83; p = 0.012). [Bochicchio GV et. al. Surg Infect (Larchmt) 9: 415, 2008]

Summary of RBC/Infection Risk[edit]

No prospective, randomized studies show a difference in any infection rates following transfusion of blood products. Retrospective or prospective non-randomized studies have produced mixed results.


Stored blood usually contains an excess of hydrogen ions (due to both the preservative as well as the continued metabolic activity) as well as potassium, neither of which is are clinically significant, even following massive transfusions [Stoelting 359-60]. Stored blood does contain progressively less 2,3 DPG, thus the older blood is, the less oxygen releasing capacity available – a study of 63 trauma patients who received 6-20 units of blood identified mean age of blood, number of units older than 14 days, and number of units older than 21 days as independent risk factors for MOF [Zallen G et. al. Am J Surg 178: 570, 1999]. Citrate (used to prevent coagulation) can cause two problems – a metabolic alkalosis, as well as hypocalcemia (very rare). Supplemental calcium is only indicated in three situations – infusion > 50 cc/min, liver disease (i.e., inability to metabolize citrate to bicarbonate), or for neonates. For these reasons, hypocalcemia is mostly an issue in liver transplant patients. Hypothermia can also be an issue if blood is not warmed through a Ranger.

Autologous Blood[edit]

Predeposited autologous blood (PAD) is usually not very cost-effective – most patients donate 10.5 cc/kg every 5-7 days, with the last unit donated 72 hours before surgery. Most are on iron supplementation, and some go on EPO (which increases available blood by 25%). A randomized, controlled trial of 48 CABG patients showed no significant decrease in total units of blood used (39.1% of non-PAD patients received blood products, versus 47.8% of PAD patients, p=0.73), but a significant decrease in exposure to allogenic blood products (39.1% versus 16% received allogenic blood, p=0.036). [Bouchard D et. al. Can J Surg 51: 422, 2008]

Intraoperative blood is contraindicated if a site infection or malignancy is present. Complications of the Cell Saver include dilutional coagulopathy, excessive heparinization, hemolysis, air embolism, and even DIC. A recent, randomized trial of 213 first-time CABG patients, however, showed no difference in the percentage of patients ultimately exposed to allogenic blood products (32% in both groups, p = 0.89) or the total number of units transfused (24 vs 25 in controls, p = 0.88) [Klein et al.].

Complications of Autologous Blood[edit]

The most common complication is dilutional coagulopathy, more serious (and rare) include consumptive coagulopathy and lung injury.


Introduction to Platelets[edit]

While thrombocytopenia is defined as platelets < 150,000/mm3, the ability to form a hemostatic plug is retained down to 100,000/mm3, thus 100,000 should be the functional definition of thrombocytopenia. Further, the bleeding tendency in thrombocytopenia is more a function of structural lesions than platelet count. Traditionally it has been taught that spontaneous bleeds occur at 20,000/mm3 [JAMA 271: 777, 1994], but this is no longer true [Blood 81: 1411, 1993] – in the absence of other risk factors for bleeding, platelets below 5000/mm3 can be tolerated without bleeding. [Abstr Clin Care Guidelines 6: 4, 1994]

Pseudothrombocytopenia occurs when EDTA causes platelet clumping in vitro. It tends to occur in patients who have antiplatelet antibodies, and is found in as many as 2% of hospitalized patients [May Clin Proc 59: 123, 1984] – diagnose this by either collecting blood with citrate (not EDTA) or by looking at the smear.

On the other hand, in cases of dysfunctional adherent ability (ex. uremia), risk of bleeding can be increased despite platelet counts > 100,000/mm3. While “bleeding time” attempts to measure the ability of platelets to adhere, there has never been a correlation between “bleeding time” and risk of bleeding [Semin Thromb Hemost 16: 1, 1990] – the only way to assess bleeding risk is to have knowledge of the conditions that alter platelet adhesion (renal failure, bypass, ASA, Dextran).


Causes of Platelet Disorders in the ICU
Thrombocytopenia Abnormal Platelet Function
Heparin (1-3%) Renal insufficiency
Sepsis (>50%) Cardiopulmonary bypass
AIDS (40-60%) Aspirin or Ketorolac
DIC Acetaminophen
TTP Dextran
Large volume blood transfusions Bactrim

() = incidence of platelet disorders in given state. The only data we have on thrombocytopenia is an MICU study, which showed a prevalence of 23% [Chest 104: 1243, 1993]


HIT occurs in 1-3% of patients taking the drug, appearing within 14 days of starting heparin (usually 5-10). The likelihood is independent of the dose, as even heparin flushes and heparin-coated catheters can induce thrombocytopenia [J Vasc Surg 7: 667, 1988]. Thrombocytopenia is less common with LMWH [NEJM 332: 1330, 1995]. The major complication of HIT is thrombosis, not bleeding – VTE are found in 70% of patients with HIT and arterial thromboses in 15% [Am J Med 101: 502, 1996] – in 50% of patients, HIT is first diagnosed after a thrombotic complication, and in the other 50%, half will develop a thrombotic complication within 30 days despite the discontinuation of therapy [Am J Med 101: 502, 1996]. Diagnosis requires detection of heparin-induced IgG antibodies AS WELL AS THE CLINICAL SCENARIO (because the heparin antibody assay will also detect on-pathogenic antibodies to heparin-platelet-factor-4 complex [Chest 127S: 35S, 2005]), i.e., thrombocytopenia at least 5 days after initiation of heparin, and usually thrombosis.

The available evidence indicates that the risk for thrombosis in the days to weeks after stopping heparin therapy is at least 20% and possibly as high as 50% in HIT patients who present with isolated thrombocytopenia. This evidence supports the view that alternate therapy with a rapidly acting anticoagulant should be initiated when the heparin therapy is discontinued. [AIM 164: 361, 2004]

Treat by 1) d/c heparin 2) d/c heparin flushes 3) d/c heparin-coated vascular cath