ISSN NUMBER: 1938-7172
Issue 4.8

Michael A. Fiedler, PhD, CRNA

Associate Editor:
Mary A. Golinski, PhD, CRNA

Contributing Editors:
Penelope S. Benedik, PhD, CRNA, RRT
Joseph F. Burkard, DNSc, CRNA
Gerard T. Hogan, Jr., DNSc., CRNA
Alfred E. Lupien, PhD, CRNA
Lisa Osborne, PhD, CRNA
Dennis Spence, PhD, CRNA
Steven R. Wooden, MS, CRNA

Assistant Editor
Jessica Floyd, BS

A Publication of Lifelong Learning, LLC © Copyright 2010

New health information becomes available constantly. While we strive to provide accurate information, factual and typographical errors may occur. The authors, editors, publisher, and Lifelong Learning, LLC is/are not responsible for any errors or omissions in the information presented. We endeavor to provide accurate information helpful in your clinical practice. Remember, though, that there is a lot of information out there and we are only presenting some of it here. Also, the comments of contributors represent their personal views, colored by their knowledge, understanding, experience, and judgment which may differ from yours. Their comments are written without knowing details of the clinical situation in which you may apply the information. In the end, your clinical decisions should be based upon your best judgment for each specific patient situation. We do not accept responsibility for clinical decisions or outcomes.

Table of Contents










Attention subscribers licensed in Alabama, Alaska, Idaho, Kentucky, Nevada, and New Mexico:

This issue contains 1 PHARMACOLOGY specific CE credit.



Is that Transfusion worth the Risk?

As an anesthetist at the head of the bed, I, like all of you, am responsible for the actual administration of blood products as part of my anesthesia duties. Certainly the decision to transfuse is not a decision made unilaterally by any one individual, or in isolation. In non-trauma surgical scenarios, the decision to transfuse is made by the surgeon and the anesthesia personnel, hopefully in a collaborative way; and the decision is not always agreed upon.

I continue to be perplexed at the variation in transfusion practices amongst various members of the OR team. I have been asked to administer blood products when the hemodynamic parameters were perfectly stable and the hemoglobin is far from what I would consider too low. I have been told to withhold blood products until the hemoglobin reached a “trigger value” but I am not sure of the evidence supporting the “trigger” value. I have been asked to administer autologous pre-donated blood just so it isn’t “wasted.” The question I continue to ask myself, my colleagues in the operating room, and the blood bank is:  What does the evidence show regarding transfusion practices in non-traumatic non-hemorrhagic surgical scenarios?

To try to answer these questions I’ve reviewed the research and had excellent dialogue with my anesthesia colleagues and physicians in pathology and the laboratory. I have discovered the following:

  • There is a large variation in transfusion practice within the United States and it varies based on geographic locale, the specific institution, and the individual’s opinion in the decision making process
  • Internationally there is an even larger variation in transfusion practice; as high as twelve-fold
  • Physician bias rather than physiologic status of the patient is a constant related to differing practices

 I have also learned that:

  • Blood use increased 2-3% per year between the years 1999-2004 within the United States
  • Blood use decreased 8% in the same time period in the UK
  • Blood use in the United States is 15% higher than Europe
  • Blood use in the United States was 44% higher than Canada between the years of 1999-2004

Over 30 million blood products are used per year in the United States. The evidence regarding transfusion hemoglobin trigger values remains controversial, education is lacking, and trigger values are not agreed upon for many specific patient care situations.


What we do know from the research conducted is that there is an increase in mortality, increase in the ICU and hospital length of stay, increase in the total cost of individual care, and a decrease in long term survival all associated with blood transfusions.


We now understand that a transfusion is analogous to a “liquid transplant,” and all the sequelae associated with any other transplant. Within banked blood there exists a build- up of cytokines (pro inflammatory mediators), an increase in potassium and bio-reactive responders, and a decrease in ADP, and 2,3-DPG. Blood stored for 3 weeks contains up to 30% nonviable RBCs. Transfusion of stored blood is associated with a higher rate of infections in the recipient, inadvertent bacterial contamination, febrile and allergic reactions, and the possibility of errors leading to acute hemolytic transfusion reactions. Adverse reactions following blood transfusions are dose related with an 8-10% incidence of adverse effects per unit transfused. There is an increase in bacterial infections of 50% in transfused patients and a strong association with ventilator acquired pneumonia.

There have been many efforts to minimize the number of blood transfusions in the absence of trauma or massive hemorrhage, but alternatives to transfusion are not widely known, used, and perhaps even less well accepted. Preoperative autologous donations are one such example. For the non-anemic patient, preoperative autologous donations reduces allogeneic blood exposure by two thirds and for anemic patient preoperative autologous donations reduces allogeneic blood exposure by one third. Often times the discard rate of pre-collected units is as high as 50%. However, predonation of autologous blood is not cost effective, can worsen or cause anemia, and is still subject to errors in storage and administration. Autologous blood is not risk free! One in 16,783 autologous donations are associated with adverse reactions such as contamination and volume overload. Using a trigger value of hemoglobin less than 8 gm/dL in “appropriate patients,” ensuring adequate volume, and adequate tissue oxygenation also termed normovolemic anemia, optimizing red blood cell mass by using iron supplements and erythropoietin, being highly selective of type of anesthesia, maintaining normothermia, performing surgery using minimal blood loss techniques, and cell salvage processes may all be useful.

This issue of Anesthesia Abstracts is devoted to increasing your understanding of blood and transfusion practices. My hope is that the following articles will educate you regarding current evidence in transfusion practices, inform you of the risks with transfusions, and offer you alternatives to transfusion. I urge you to examine your “transfusion triggers” much more critically and to become more conservative about when and how much you transfuse.

Mary A. Golinski, PhD, CRNA

© Copyright 2010 Anesthesia Abstracts · Volume 4 Number 8, August 31, 2010


Vretzakis G, Kleitsaki A, Stamoulis K, Bareka M, Georgopoulou S, Karanikolas M, Giannoukas A


Intra-operative intravenous fluid restriction reduces peioperative red blood cell transfusion in elective cardiac surgery, especially in transfusion-prone patients:  a prospective, randomized clinical trial

J Cardiothorac Surg 2010;5:7

Vretzakis G, Kleitsaki A, Stamoulis K, Bareka M, Georgopoulou S, Karanikolas M, Giannoukas A


Purpose            The purpose of this study was to determine what impact, if any, peri-operative intravenous fluid restriction during open heart surgery had on reducing the need for allogeneic red blood cell transfusions. The sample of the population studied was both general cardiac surgery patients and also the subset of patients who were transfusion-prone; those with pre-operative anemia, female gender, and small BSA.

Background            Individuals who undergo open heart surgery frequently receive allogeneic blood transfusions. Two concurrent processes are responsible for a precipitous hematocrit (Hct) reduction during cardiac surgery:  blood loss and red blood cell dilution due to positive fluid balances. Hemodilution itself has been identified as a major factor in determining whether or not to transfuse. There exists a plethora of data that suggests allogeneic transfusions are associated with increased morbidity and mortality in cardiac surgery. However, one study conducted recently could not demonstrate an association between moderate blood transfusion (defined as < 6 units) and long term survival. It appears that the risk of transfusion-associated adverse outcomes may depend on the volume or numbers of packed red cell transfusions and therefore it is postulated that a reduction in the number of transfusions may be an important goal during cardiac surgery.

Methods            This was conducted as a prospective, partially blinded (see notes) study over a 20 month period within one institution. Patients were included who were undergoing elective cardiac surgery utilizing cardiopulmonary bypass (CPB) and aged between 18-85 years. Computer generated randomization allowed the researchers to place the patients in to one of two groups.

Group A            the restrictive protocol whereby intravenous fluids before CPB were limited to 500 mL. Peripheral lines infused hetastarch instead of crystalloid and were clamped off following central line placement. Based on each patient’s hemodynamic profile in real-time, fluid boluses were given quickly in 50 mL increments. Anesthetic, inotropic, and vasoactive medications were double concentrated and administered without a carrier fluid infusion. If blood was drawn for labs, it was re-infused; excessive line flushing was avoided. Before the patient went on bypass, the hemodynamic profile of each patient was managed according to this algorithm:

A.   Titration of anesthetic drugs (regardless of filling pressures) for:

a.     MAP > 55 mmHg with SvO2 > 75%, cerebral oxygen saturation > 60% and BIS values (depth of consciousness monitoring) < 35

B.    Administer vasoconstrictor (regardless of filling pressure) for: 

a.     MAP < 55 mmHg with SvO2 > 75%, cerebral oxygen saturation > 60% and BIS > 35

C.    Infuse an inotropic agent (dobutamine) for:

a.      SvO2 < 75%, PCWP > 16 mmHg, and heart rate < 90 bpm 

D.   Pace via epicardial electrodes for:

a.     SvO2 < 75% and heart rate < 40 bpm

Group B            patients were managed with Ringer’s Lactate solution through their peripheral IV line. Other drugs were ‘usual’ and ‘standard’ concentration and administered with a carrier infusion at 40 mL/h. Anesthesia personnel did not follow any specific fluid administration protocol except for intra-operative transfusion of packed cells. BIS and cerebral oxygenation data was always available for both groups; hemodynamics were managed according to the judgment of the provider for Group B.

Group A perioperative RBC transfusion was based upon the following:

  • cerebral oxygenation data was taken into consideration and:
    • If mean values from both hemispheres was < 60 or decreased by > 20% compared to the mean value during pulmonary artery cannula insertion, the patient was transfused.

Group B perioperative RBC transfusion was based upon the following: 

  • During aortic cross clamping allogeneic blood was not given if Hct was >21%
  • When Hct was < 17%, one unit of PRBCs was given
  • If Hct fell between 17-21%, anesthesia was free to act according to their own judgment.

In both groups, after the cross clamp was removed and before weaning from bypass, PRBCs were transfused for a Hct < 21%.

·      After weaning from CPB, patients were re-transfused with salvaged blood. If Hct was < 24%, packed cells were administered.

·      In the ICU, packed cells were transfused for Hct < 24%

The primary outcome variable was the total number of PRBC units transfused during the hospital stay. All pertinent fluid data was recorded, which included total IV fluid volume, urine output, priming and cardioplegic solution volumes, additional fluid administered during bypass, hemofiltration volumes, and pump residual volumes. Hematocrit values were documented at all appropriate time frames during the procedure and in the ICU.

Results             Demographic data and clinical characteristics did not differ between the two groups; Group A n = 100 and Group B (control) n = 92. A total of 137 patients were transfused with allogeneic blood during the hospitalization period, receiving a total of 289 units of PRBCs.

The total number of PRBC units transfused was significantly lower in the fluid restricted group (113) versus the control group (176) (P < 0.0001). During surgery, a total of 81 patients were transfused. The number of PRBC units administered was significantly lower in the fluid restricted group (P < 0.0001). Additionally, calculated net erythrocyte volume loss during the entire procedure was significantly lower in the fluid restricted group (P < 0.005). During the ICU stay, patients in the fluid restricted group received more transfusions but this was not statistically significant. The following were also revealed:

  • Transfused patients were significantly older in age, shorter in height, lower in weight and BSA, and had lower pre-operative Hct values compared to those not transfused.
  • BMI and discharge Hct values did not differ between the groups.
  • Male gender and those in the fluid restricted group were strongly associated with a lower probability of transfusion.

After subdividing both groups by whether they received a transfusion or not the following was discovered:

  • Among those transfused during surgery, there were significant differences in gender (more females transfused in control group), age (older in control group)  and BSA (lower in control group).

Three variables were noted to be significant predictors of transfusion:

  • Group assignment
  • Preoperative hematocrit
  • BSA

The likelihood of allogeneic transfusion was 3.12 times greater in the control group versus the fluid restricted group.

Each 1% increase of the pre-operative hematocrit value was associated with a 15% lower probability of transfusion.

Those in the fluid restricted group received 0 or 1 PRBC unit while those in the control group received 3 or more PRBC units (P < 0.0007)

There were no reported deaths during surgery for any subjects. The ICU length of stay and the time of mechanical ventilation did not differ significantly between groups.

Conclusion            This study showed that during cardiac surgery, a protocol which included fluid restriction and ended with a comparable hematocrit, resulted in less total blood transfusions than a protocol without fluid restriction. This is important because it is believed a greater number of allogeneic blood transfusions in the perioperative cardiac surgery period predisposes to a greater risk of untoward outcomes from the transfusions.



This study clearly had several strengths which included the design, statistical rigor, and a comprehensive and delineated transfusion protocol. Additionally it was performed at one institution with the same team and had minimal missing data points. However, it also possessed a few limitations, as all studies do. The design itself did not allow for full blinding (virtually impossible) and different anesthetists did participate. Group B or the non-fluid restricted patients were volume treated according to usual and customary practices of the anesthesia providers which varied. While the mortality rate was very low, it also may speak to the lower acuity of the patients who were enrolled. Generalizability to a higher acuity subset of the population may prove to be very weak. It is critical, however, that we continue to gain knowledge and build the body of evidence related to identifying methods to reduce the risks of untoward outcomes associated with blood transfusions. Research is needed that is adequately powered to determine if fluid restriction in this specific patient population not only minimizes the need for transfusion, but also achieves a reduction in morbidity and mortality.

Mary A. Golinski, PhD, CRNA 

‘Partially- blinded’ operationally defined for this study:  Surgeons, surgical assistants, perfusionists and ICU staff were not informed about the study. The anesthetists were not informed about the specific scope and aims of the study. All patients received standardized anesthesia and intra-operative care. Acute normovolemic hemodilution and retrograde autologous priming of the pump circuit were not used for any patient.

Anti-platelet medications (except ASA) were discontinued at least 72 hours before surgery. Aprotinin, aminocaproic acid, or tranexamic acid were not used for any patient.

© Copyright 2010 Anesthesia Abstracts · Volume 4 Number 8, August 31, 2010


Hebert P, Wells G, Blajchman M, Marshall J, Martin C, Pagliarello G, Tweeddale M, et al


A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care

N Engl J Med 1999;340:409-417

Hebert P, Wells G, Blajchman M, Marshall J, Martin C, Pagliarello G, Tweeddale M, et al


Purpose            The purpose of this study was to clarify the potential risks of anemia, as well as the possible benefits of transfusions, in critically ill patients. Does a restrictive approach to red cell transfusion (defined as hemoglobin concentrations maintained at 7-9 g/dl) provide equivalent outcomes compared to a more liberal transfusion strategy (maintaining hemoglobin concentrations 10-12 g/dl), in the critically ill?

Background            There continues to be a plethora of differing views amongst those caring for critically ill patients, regarding the risks of anemia and the potential benefits of treating the hemoglobin values, via blood transfusion. On the pro side:  anemia may not be well tolerated by critically ill patients. Past research has demonstrated that anemia increased the risk of death after surgery in those with coronary artery disease (CAD) and in the critically ill individual. The theory behind the benefits suggests that red blood cell transfusions augment the delivery of oxygen and therefore the deleterious effects of an oxygen debt will be minimized. This theory promoted protocols which used the Hgb value of 10 g/dl as the transfusion trigger value. On the con side:  the critically ill patient is at increased risk for the immunosuppressive and microcirculatory complications of red cell transfusions. There are serious concerns regarding the supply of product as well as the safety of administering blood products. These two concerning variables have led to treatment taking a more conservative approach. The optimal transfusion practice for critically ill patients with a variety of differing diagnoses has not been agreed upon nor has conclusive evidence been established.

Methodology            This study was carried out as a randomized, comparitive, multi-site clinical trial. Patients who were admitted to 1 of 25 pre-determined intensive care units in Canada for greater than 24 hours, who had hemoglobin concentrations of 9 g/dl or less within 72 hours after admission, and who were considered euvolemic after their initial treatment, were included. Specific appropriate exclusion criteria were upheld. Those with normovolemia were assigned to one of two treatment arms, stratified according to disease severity (APACHE II scoring):

  • Restrictive strategy treatment arm. Hemoglobin concentrations were maintained in the range of 7-9 g/dl, and a transfusion was given when the Hgb was  <7 g/dl
  • Liberal strategy treatment arm. Hemoglobin concentrations were maintained in the range of 10-12 g/dl, and a transfusion was given when the Hgb was <10 g/dl.

Blood was transfused one unit at a time and Hgb concentrations measured after each unit. Additional clinical treatment suggestions were provided which were thought to promote or enhance oxygen delivery. These included:  the use of fluids and vasoactive drugs, when necessary, and advice when a transfusion was not indicated by the study protocol. All other management decisions were left to the discretion of the attending physicians. Adherence to the transfusion protocol was required only during the ICU stay. At the time of randomization, all pertinent demographic, diagnostic, and clinical information was obtained. Additionally the multiple-organ dysfunction score was obtained for each patient. All laboratory values were recorded during the ICU stay as well as the use of red blood cell transfusion metrics, medications administered, all mechanical ventilation (or lack thereof) data, the need for dialysis, and surgical interventions.

The primary outcome measure was death from all causes within 30 days of randomization. Secondary outcomes included:

  • 60-day death rates from all causes
  • Mortality rates during the ICU stay and during the entire hospitalization
  • Survival times in the 1st 30 days
  • Measures of organ failure and dysfunction (number and rate)
  • Multiple organ dysfunction score
  • Length of ICU stay

Final statistical computations were based on the intention to treat. A P value of <0.05 was considered statistically significant.

Results            Over 2,000 patients were screened for consent;  however due to specific exclusion criteria; i.e. previous transfusions, time limitations, no next of kin, language barriers, and reasons labeled as ‘other’; 838 total patients consented. The total number of patients included those in the liberal strategy group, n = 420, and the restrictive strategy group, n = 418. Demographic data analysis demonstrated that no significant differences existed in any baseline characteristics between the 2 groups. The two most common reasons for ICU admission were respiratory and cardiac disease related. The average APACHE II score was 21. More than 80% of study patients were intubated and mechanically ventilated. Approximately 25% had an infection as either a primary or a secondary diagnosis.

Following is a summary of descriptive data for each group:

  • average daily Hgb in restrictive group = 8.5 g/dl vs. average daily Hgb in liberal strategy group = 10.7 g/dl (P < 0.01)
  • in the restrictive group an average of 2.6 units of PRBCs were administered verses 5.6 units in liberal group (P <0.01)
  • 54% fewer transfusions were administered when the restrictive transfusion criteria were used
  • 33% of restrictive group patients received no transfusions
  • 100% of liberal group patients received transfusions
  • there was no difference between groups in terms of interventions which may have influenced outcomes (vasopressor use etc)
  • Dialysis, mechanical ventilation, and surgical procedures occurred equally in both groups

The following outcomes were statistically significantly different between groups:

  • The multiple organ dysfunction scores (adjusted) were lower in the restrictive group (P = 0.03)
  • 30-day morality was significantly lower in the restrictive group for those with an APACHE II score < 20 and among those less than 55 years of age
  • Mortality rates during hospitalization were lower in the restrictive group (P = 0.05)
  • Cardiac events were more frequent in the liberal strategy group while in the ICU (P < 0.01)

The following outcomes had statistically insignificant differences between groups:

There was no difference between groups in the rate of death from all causes within 30 days (P = 0.11)
  • Mortality rates during ICU stay
  • 60 day mortality rate
  • The number of organs failing as well as length of stay in the ICU and hospital

All outcomes in the two groups were similar for those who were > 55 years of age and for those who had an APACHE II  score > 20

Conclusion            Using a restrictive transfusion strategy in the critically ill, as low as 7 g/dl, combined with the maintenance of hemoglobin concentration in the range of 7-9 g/dl, was at least as effective as a liberal transfusion strategy where hemoglobin concentrations were maintained between 10-12 g/dl. Using a restrictive approach decreased the number of transfusions by 54% and decreased exposure to receiving any transfusion by 33%. This resulted in a significant decrease in cost of care as well as decreasing a person’s exposure to complications that can result due to transfusion therapy.



This is the original TRICC (Transfusion Requirements in Critical Care) trial. This renowned research study was published in the New England Journal of Medicine in 1999, already greater than a decade ago. While it may be considered ‘aged,’ it is widely regarded by many experts in the field to be one of the most significant studies drawing attention to changing transfusion guidelines. It has prompted scientists to look more closely at the evidence and to generate evidence-based care models, related to transfusion therapy. Additionally, it has promoted numerous clinical trials related to the potential benefits of transfusion therapy as well as risks and concerns related to transfusion therapy. It has challenged all of us to consider the evidence, and to make transfusion decisions with a great degree of scrutiny, when administering an anesthetic to a patient losing blood. For example, we may assess ~ 10% of the patients’ total blood volume in the suction canister, yet for many patients who have begun the surgical procedure with normal hemoglobin values, even with a 10% or greater total blood volume decrease; blood pressure, heart rate, urine output, and other clinical parameters remain normal. There may not be a significant cardiac history. Should we be exposing these individuals to the risks of transfusion? Are we treating what we see in the suction canister, or are we treating the patient? Are we contributing to positive outcomes or are we challenging patients with a previously intact immune system in a manner that will not help healing during the post-operative period? Is our transfusion exposing the already immuno-compromised individual to immunologic changes which may hurtle them over the curve to a state of further illness? Will they be able to combat exposure to the factors that are released from damaged stored red blood cells during the time the blood has been sitting on the shelf – and respond appropriately? And will anesthesia providers be able to educate each other and change our mind-set as the evidence continues to show us that transfusion may not always be the most appropriate answer?


Mary Golinski PhD, CRNA


The Apache II Score provides an estimate of ICU mortality based on a number of laboratory values and patient signs, taking both acute and chronic disease into account. The data used should be from the initial 24 hours in the ICU, and the worst value (furthest from baseline/normal) should be used. It is a severity of disease classification system, one of several ICU scoring systems deemed reliable. After admission of a patient to an ICU an integer score from 0 to 71 is computed based on several measurements; higher scores imply a more severe disease and a higher risk of death.

This scoring system is used in many ways:

  • Specific procedures and specific medical care is only given to patients with certain APACHE II scores
  • APACHE II scores can be used to describe the morbidity of a patient when comparing outcomes with other patients
  • Predicted mortalities are averaged for groups of patients in order to specify the group's morbidity
© Copyright 2010 Anesthesia Abstracts · Volume 4 Number 8, August 31, 2010

Eder A, Herron R, Strupp A, Dy B, White J, Notari E, Dodd R, Benjamin R


Effective reduction of transfusion-related acute lung injury risk with male-predominant plasma strategy in the american red cross (2006-2008) 

Transfusion 2010;50:1732-1742

Eder A, Herron R, Strupp A, Dy B, White J, Notari E, Dodd R, Benjamin R


Purpose            The purpose of this study was to evaluate the American Red Cross risk reduction strategy which involved reducing the number of cases of severe Transfusion Related Acute Lung Injury. Risk was reduced by minimizing the preparation and distribution of plasma blood components from donors known to be WBC-alloimmunized or at increased risk of WBC alloimmunization.

Background            Transfusion Related Acute Lung Injury (TRALI) has become known as a predominant cause of illness and mortality directly associated with blood transfusion therapy. It has accounted for most of the transfusion related deaths reported to the FDA since 2004. It is specifically diagnosed by the presence of:

1)    Pulmonary edema and hypoxemia within 6 hours of transfusion therapy, and

2)    An absence of other causes of acute lung injury or circulatory overload

It is theorized that various antigens in blood components, most often white blood cell alloantibodies, may stimulate TRALI and may be the final common pathway precipitating acute lung injury. Alloantibodies that act against human leukocyte antigens (HLA) occur as the result of a “sensitization process.” Sensitization occurs during pregnancy, transfusion, or transplantation procedures. The prevalence of those with HLA positive antibodies is the same in women who have never been pregnant, non-transfused men, and transfused men. An estimated 17.3% of female blood donors and 24.4% of those with a history of previous pregnancy demonstrate HLA antibodies. The highest prevalence is among women with 4 or more pregnancies. It appears obvious therefore, that most antibody-positive blood donors are multiparous women.

Several international programs have corroborated the association of HLA-sensitized female donors with TRALI. The Red Cross found a significant number of female WBC antibody positive donors implicated in 75% of probable TRALI deaths associated with plasma component transfusions between 2003 and 2005. The recognition that WBC alloantibodies were implicated in several TRALI cases prompted blood centers and hospitals to take specific action to prevent TRALI from occurring. The blood centers were instructed by the American Association of Blood Banks to evaluate donors who were implicated or associated with TRALI and evaluate their eligibility to continue to donate. The United Kingdom National Blood Service implemented measures in 2003 to reduce TRALI by the preferential use of plasma from male donors. As evidence continued to emerge in late 2006, the Red Cross began preferentially distributing plasma collected from male donors for transfusion. Plasma collected from female donors was diverted to be manufactured into plasma derivatives.

This study described the results of implementing a male only plasma donor policy on the incidence of TRALI.

Methodology            The implementation of a male predominant plasma strategy was launched as a pilot in September 2006. Thirteen of the Red Cross’ 35 regional centers assessed the feasibility of preferentially distributing plasma only from male donors. The pilot achieved 95% male predominance for transfusable plasma within three months. Other regional centers implemented the strategy in early 2007 and expanded the number of blood components included. Apheresis plasma collection operations were converted to male only donors and/or HLA tested donors. Regional blood centers investigated all reported:

·      infectious and noninfectious complications in recipients of blood components

·      the actions taken to retrieve prior donations from the involved donors

·      actions taken to prevent recurrent events

Each Red Cross regional medical director assigned a probability code, P1-P6, when the investigation of a suspected transfusion reaction was complete.

·      P1            case rescinded

·      P2            TRALI not supported

·      P3            TRALI unlikely or not supported, other etiologies more likely cause of reaction

·      P4            TRALI cannot be excluded, but cases do not meet consensus definition

·      P5            consistent with TRALI but transfusion source not identified

·      P6            consistent with TRALI and transfusion source identified

All fatalities during the investigation were identified. These cases were independently reviewed by three separate physicians. Each case was scored for the presence or absence of the following clinical criteria:

1)    onset with 6 hours

2)    hypoxemia

3)    CXR findings

4)    other clinical signs consistent with pulmonary edema

5)    absence of circulatory overload

6)    presence or absence of risk factors for acute lung injury

The cases were classified as “probably TRALI” if the information met a consensus conference definition of TRALI. Cases were also classified if TRALI was suspected but information about intravascular volume status was equivocal.

Results            By 2007, the regions that participated reached the goal of having > 95% of transfusable plasma components from male donors. The average proportion of plasma from male donors varied according to ABO type; >99% for groups O, A, and B; 57% for group AB.

The analysis separated cases when only a single component of blood was administered as opposed to cases that involved multiple blood component types. When plasma was the only component transfused, there was a significant decrease in TRALI during the study period; 26 cases in 2006 vs. 7 cases in 2008. This was a level that was no different from the rate of TRALI following RBC transfusion. During 2006 there were 8 fatalities following plasma transfused from antibody positive female donors. During 2008 there were no transfusion-related TRALI fatalities that involved plasma from female donors.

Conclusion            Alloantibodies against human leukocyte antigens (HLA) occur as a result of pregnancy, blood transfusion, and transplantation procedures. The incidence of TRALI decreased markedly when plasma from male and other low risk donors (those negative for HLA antibodies) was distributed for transfusion.



 TRALI has now become one of the leading causes of morbidity and mortality associated with blood component therapy. It has accounted for most of the deaths reported to the FDA since 2004. It is critical that we as providers of anesthesia, which often times includes blood transfusion as part of our duties, be aware of all risks of blood transfusion therapy. While we may not always be able to avoid blood component transfusions, we must understand the physiology behind TRALI and be diligent when performing our follow up care.

There are several things we can do related to transfusion therapy:

1)    Have sound reasons for transfusion therapy.

2)    Know the laboratory values; being acutely aware of the patient’s hemodynamic and coagulation state.

3)    Perform vigilant post operative assessments and report abnormal findings.

4)    Report transfusion related morbidity and mortality to the appropriate bodies (this is a strong contribution to continuing research and the building of evidence based practice).

4)      Continue to educate our colleagues.

Mary Golinski, PhD, CRNA


·      The liquid portion of blood
·      Protein-salt solution in which red and white blood cells and platelets are suspended
·      92 percent water, constitutes 55 percent of blood volume
·      Contains albumin (the chief protein constituent) fibrinogen (responsible, in part, for the clotting of blood) and globulins (including antibodies)
·      Serves a variety of functions: maintains a satisfactory blood pressure and volume to supplying critical proteins for blood clotting and immunity. Serves as the medium for exchange of vital minerals such as sodium and potassium and helps to maintain a proper pH (acid-base) balance in the body, which is critical to cell function.
·      Frozen quickly after donation (up to 24 hours) to preserve clotting factors, stored up to one year, and thawed shortly before use. It is commonly transfused to trauma patients, those with liver disease or multiple clotting factor deficiencies.
© Copyright 2010 Anesthesia Abstracts · Volume 4 Number 8, August 31, 2010

Morton J, Anastassopoulos K, Patel S, Lerner J, Ryan K, Goss T, Dodd S


Frequency and outcomes of blood products transfusion across procedures and clinical conditions warranting inpatient care:  an analysis of the 2004 healthcare cost and utilization project nationwide inpatient sample database 

Am J Med Qual 2010;25:289-296

Morton J, Anastassopoulos K, Patel S, Lerner J, Ryan K, Goss T, Dodd S


Purpose            The purpose of this study was twofold:  1) to calculate the frequency of blood transfusion for hospitalized individuals in the United States, and 2) to assess clinical and health outcomes across the full spectrum of surgical procedures and physiologic conditions for those who received blood transfusions.

Background            Common reasons for receiving a blood transfusion include acute blood loss during surgical procedures, the inability to achieve hemostasis, and trauma. Often times in a traumatic situation, blood transfusion is necessary to save a life. We are now learning that in a non-emergent situation, transfusion is an imperfect way to manage blood loss. There are negative clinical outcomes associated with blood transfusion such as increased postoperative infection rates, transfusion related acute lung injury, and reduced long-term survival. Significant advances in both surgical and anesthetic technique have reduced the need for blood transfusion. Nevertheless, in some situations intraoperative and postoperative hemorrhage or acute blood loss require transfusion.

Increased provider and patient awareness of transfusion related problems will encourage innovative ways to minimize bleeding related to surgery. Ultimately, improving outcomes and health-related quality of life is the goal.      

Methodology            This was a retrospective cohort study using data obtained from the Agency for Healthcare Research and Quality 2004 Healthcare Cost and Utilization Project Nationwide Inpatient Sample (NIS). A sample of 8,004,571 was obtained, including patients from 37 states within the USA. Using this sample, the 10 most common primary procedures performed during the course of a hospitalization and involving blood product transfusion were identified. Discharges from these top 10 primary procedure categories were classified into 2 groups:  those who were transfused and those who were not transfused. Demographic data was obtained for all patients including age, gender, primary payer source, the Charlson comorbidity index, and admission type.

Data from the NIS included specific information regarding:

  • The frequency of blood product transfusion
  • Morbidity, including postoperative infections and noninfectious transfusion related complications such as:
    • Transfusion related acute lung injury (TRALI)
    • Anaphylactic shock caused by serum
    • Other serum reaction
    • ABO incompatibility
    • Rh incompatibility reaction
  • In-hospital mortality
  • Length of stay
  • Hospital charges

Mortality rates and the postoperative infection rates were calculated overall and for each study cohort. Logistic regression was used to calculate odds ratios for death and postoperative infection for the non-transfused cohort compared with the transfused cohort. Linear regression was used to calculate incremental differences between cohorts. Rates of transfusion related complications were calculated for the transfused cohort.

Results            Approximately 2.23 million discharges in the 2004 NIS database included a procedure code for blood product transfusion. The top 10 primary procedure categories with greatest frequency of transfusion were:

  • Hip replacement; total and partial (most common procedure associated with transfusion)
  • Upper gastrointestinal endoscopy; biopsy
  • Knee arthroplasty
  • Respiratory intubation and mechanical ventilation
  • Treatment; fracture or dislocation of hip and femur
  • CABG
  • Other vascular catheterization; not heart
  • Colonoscopy
  • Colorectal resection
  • Other operating room procedures on vessels other than head and neck

Demographic analysis showed a mean age of 61.1 years. Greater than 60% were female. Almost half of patient discharges were associated with at least one co-morbidity. Overall, 72.6% were designated as emergent or urgent. The following outcomes occurred significantly more frequently in the transfused cohort than in the non-transfused cohort after adjusting for confounding variables (P < 0.001):

§  Average Length of Stay

§  Mortality

§  Postoperative infection rates (the most frequent was pneumonia, followed by sepsis, UTI, cellulitis and abscess, bacteremia, and osteomyelitis)

§  Total charges

The odds of death were 1.7 times greater in the transfused cohort after controlling for age, gender, Charlson co-morbidity index, admission type, and Diagnostic Related Group.

Conclusion            The cohort who received transfusions in this study demonstrated increased odds of mortality, increased hospital Length of Stay, and higher infection rates. The total charges related to the in-hospital stay were higher in the transfusion cohort approximating a difference of $17,194. This is consistent with previously published data from randomized clinical trials and prospective observational studies.



The evidence is becoming more and more conclusive: the administration of blood products in a liberal manner, for example hemoglobin trigger values of 10 gm/dL in a non-crisis situation or when there is hemodynamic stability, can create more harm than good. What has been demonstrated over the past decade suggests that the host or recipient undergoes significant damaging physiologic responses to what some now call the “liquid transplant.” This liquid transplant has been positively correlated with significant untoward outcomes and in the severest of situations, death. The body reacts at the cellular level in a similar manner to any antigen and physiologically in ways far beyond the basic incompatibility cascade. The theories we have based our practice on for years are no longer supported. The comprehensive cost of a transfusion coupled with the unfavorable outcomes is causing a cry for innovative ways to reduce the incidence of transfusions. Drugs that stimulate the production of red blood cells and techniques that minimize blood loss during surgery are all being researched. We must stay abreast of the research and be involved in the research. We need to educate each other and continually practice according to what the evidence supports.

Mary Golinski, PhD, CRNA

Co-morbidity describes the effect of all other diseases an individual patient might have other than the primary disease of interest. The Charlson Co-morbidity Index is the most widely accepted, validated method, currently used to quantify co-morbidity. It contains 19 categories of co-morbidity which are primarily defined using ICD-9-CM diagnosis and procedures codes. Each category has an associated weight; this is based on the adjusted risk of 1-year mortality. The higher the score, the more critical effects caused by the co-morbidity.


Transfusion related acute lung injury (TRALI) is a life-threatening adverse effect of transfusion that is occurring at an increasing rate in the United States. In the past 2 reporting years, TRALI has been the leading cause of transfusion-related death. TRALI and acute lung injury (ALI) share a common clinical definition except that TRALI is temporally and mechanistically related to the transfusion of blood or blood components. Two different etiologies have been proposed:

1)    A single antibody-mediated event involving the transfusion of anti-HLA class I and class II or anti-granulocyte antibodies into patients whose own leukocytes express the similar antigens.

2)    A 2-event model: the first event is the clinical condition of the patient resulting in pulmonary endothelial activation and neutrophil sequestration, and the second event is the transfusion of a biologic response modifier (including lipids or antibodies) that activates these adherent polymorphonuclear leukocytes (PMNs), resulting in endothelial damage, capillary leak, and TRALI.

© Copyright 2010 Anesthesia Abstracts · Volume 4 Number 8, August 31, 2010


Feagan B, Wong C, Kirkley A, Johnston D, Smith F, Whitsitt P, Wheeler S, Lau C


Erythropoietin with iron supplementation to prevent allogeneic blood transfusion in total hip joint arthroplasty. a randomized, controlled trial

Ann Intern Med 2000;133:845-854

Feagan B, Wong C, Kirkley A, Johnston D, Smith F, Whitsitt P, Wheeler S, Lau C


Purpose            The purpose of this study was to ascertain whether a pre-operative modified dosing regimen of epoetin alfa (erythropoietin injectable) was could reduce the need for allogenic blood transfusion in those having total hip arthroplasty.

Background            It is not uncommon for patients to receive an allogenic blood transfusion during total hip arthroplasty. Advances in technology have allowed laboratory personnel to be quite successful when screening donated blood products for the detection of viruses and other contaminants. Nevertheless, there remains public concern regarding the transmission of disease from transfusions. Healthcare providers aim at avoiding transfusion altogether if possible. Autologous blood donation is one such strategy but is not without risk. First, donation of one’s own blood (and subsequent banking) may be extremely inconvenient to the patient. Second, the donation itself may cause or worsen anemia; and lead to an increased incidence of transfusion. Third, pre-donated autologous blood is at risk for contamination and human error in processing, which ultimatly can lead to life threatening transfusion reactions. And lastly, some individuals are simply not candidates for pre-donation due to co-existing medical conditions.

The administration of recombinant human erythropoietin (epoetin alfa) has been studied in the past; the evidence demonstrated it reduced the need for allogenic blood transfusions during total hip replacement. The FDA approved regimen consists of 4 subcutaneous injections, 600 u/kg, administered before surgery once a week for 3 weeks and the 4th dose on the day of surgery. The human body’s response to epoetin alpha is influenced by the co-administration of iron; a second response is dependent on the dose and timing of the injection. The researchers hypothesized that a ‘high dose’ of oral iron in conjunction with a prolonged epoetin alfa dosing schedule (> 3 weeks) may bring an optimal hematologic response compared to the standard regimen.

Methodology            This study was conducted as a double-blinded, randomized, parallel-group, multicenter clinical trial. Inclusion criteria included patients scheduled for primary total hip arthroplasty, with “adequate” hemoglobin levels, and who did not pre-donate blood. Exclusion criteria were those with rheumatoid arthritis, recent GI or intracranial bleed, iron deficiency, seizures, blood dyscrasias, or uncontrolled hypertension. After obtaining informed consent, all patients began daily oral iron therapy at least 42 days before surgery; three capsules per day if tolerated at a dose of 150 mg per capsule. Each person was also randomly assigned to one of three groups:

Group 1            received four weekly subcutaneous injections of placebo 4 weeks before surgery (28, 21, 14, and 7 days before surgery)

Group 2            received four weekly subcutaneous injections of high-dose epoetin alfa (40,000 U) 4 weeks before surgery (28, 21, 14, and 7 days before surgery)

Group 3            received four weekly subcutaneous injections of low-dose epoetin alfa (20,000 U)  4 weeks before surgery (28, 21, 14, and 7 days before surgery)

Injections were withheld if the hemoglobin was >15 g/dL, or if the systolic BP was > 200 mm Hg or diastolic BP was > 105 mm Hg. Any adverse events were noted and documented as patients were evaluated prior to injections. During surgery, blood loss was quantified according to each site’s usual standard. Transfusion of allogenic blood was performed according to the usual practice at the site and/or per provider. Patients were assessed, and lab work drawn, on postoperative days 1, 3, and 5. Duplex ultrasonography was conducted on day 5 to assess for DVT.

The primary outcome measure was the incidence of allogenic blood transfusion. Secondary outcome measures included changes in the reticulocyte counts and hemoglobin concentration. Additionally, the incidence of proximal or distal DVT, pulmonary emboli, and other serious adverse events were noted and compared between groups.

An asymmetric randomization schedule allocated a greater number (n) to the low- dose and placebo groups versus the high-dose group.

Results            A total of 201 patients were included in the intention-to-treat analysis. All four doses of the study drug were administered to 98% of subjects:

Group 1 (placebo)            n = 78

Group 2 (high dose)          n = 44

Group 3 (low dose)           n = 79

There were no statistically significant differences in the demographics between groups. The iron therapy was generally well tolerated and the treatment groups did not differ in adherence to oral iron treatment. The transfusion trigger used for all three groups was similar and did not differ significantly. The primary outcome variable, rate of transfusion, was statistically significantly different for treatment groups vs. placebo:

Group 1            44.9% received transfusion (35 of 78 patients)

Group 2            11.4% received transfusion (5 of 44 patients)

Group 3            22.8% received transfusion (18 of 79 patients)

Regarding the secondary outcome variables, patients who received epoetin alfa had a rapid increase in the reticulocyte count and hemoglobin concentration. The greatest increase in the reticulocyte count occurred in the high dose group (group 2); this was statistically significant. Increased hemoglobin concentration were observed in the high dose and low dose groups, while little changed was noted in the placebo group (P < 0.001). Following surgery, hemoglobin values decreased in similar amounts for all groups. Using multivariate analytic techniques; age, predicted blood volume, preoperative hemoglobin concentration, and preoperative reticulocyte counts were significant predictors of transfusion. The incidence of adverse events, DVT or PE, did not differ statistically between groups. One pulmonary embolus was reported in the placebo group.

Conclusion            the epoetin alfa protocols reduced the need for perioperative blood transfusion. Additionally, while lacking sufficient power to definitively exclude the possibility of an increased incidence of DVT as a result of epoetin alfa, the data did not show an increased incidence of DVT in patients who took epoetin alfa according to the protocol.



One cannot however underestimate the value of this study. While it was not the first study regarding epoetin alfa as an alternative to transfusion therapy, it influenced both clinical practice and subsequent research. We are continually gaining knowledge regarding the risks associated with the processing and transfusion of both autologous and allogenic blood. These risks involve more than the transmission of viruses. They relate to what is occurring in the blood while it is “banked.” We should continue to seek out effective treatment modalities as alternatives to transfusion. We should add to the growing body of evidence. I felt it was important to introduce the subject of alternatives to transfusion with this article. It is a relatively “clean” study, conducted with an acknowledgement of limitations, yet powerful enough to provide evidence upon which to base our practice.


Mary A. Golinski, PhD, CRNA

Reticulocyte            an immature erythrocyte. Reticulocytes normally account for less than 2% of the circulating erythrocytes. A greater proportion reflects an increased rate of erythropoiesis.

Reticulocyte Count            The reticulocyte count is used to help determine if bone marrow is responding adequately to the body’s need for RBCs and to help determine the cause of and classify different types of anemia. The number of reticulocytes must be compared to the number of RBCs to calculate a percentage of reticulocytes. The test is ordered along with a RBC Count. A hemoglobin and/or hematocrit are usually ordered in order to evaluate the severity of anemia.

The RBC, hemoglobin, and hematocrit are frequently ordered as part of a complete blood count . The CBC usually includes an evaluation of RBC characteristics, such as cell size, volume, and shape. Based on these results, a reticulocyte count may be ordered to further examine the RBCs. Reticulocytes can be distinguished from mature RBCs because they still contain remnant genetic material (RNA), a characteristic not found in mature RBCs. Circulating reticulocytes generally lose their RNA within one to two days, thus becoming mature RBCs.

Erythropoietin and Its Receptor (adopted from Erythropoiesis and Red Blood Cell Physiology by Richard Sullivan, M.D.)

Human erythropoietin is a 193-amino acid glycoprotein. Approximately 90% is produced in the kidney, and the rest is produced in a variety of extra-renal sites. Hypoxia causes erythropoietin-secreting cells to synthesize and release a cytokine. When the oxygen level within the cytoplasm of erythropoietin-producing cells falls below a critical level, erythropoietin is synthesized and secreted into the bloodstream. Considerable evidence supports the concept that the biochemical "oxygen sensor" within erythropoietin-producing cells is a type of specialized heme protein, which detects the intracellular oxygen level and translates it into a "message" that results in stimulation or suppression of erythropoietin synthesis. Once synthesized and released from the cell, erythropoietin travels in the bloodstream to the bone marrow, where it binds to receptors on erythroid cells, thereby initiating their proliferation and differentiation.

Recombinant erythropoietin- Epoetin alfa is a man-made, injectable drug typically used for treating anemia. When the natural mechanism of erythropoesis is not working, it may become necessary to stimulate the bone marrow to produce red blood cells. The erythropoietin that is used for therapy is called epoetin alfa. It does not cure the underlying cause of the anemia, and unless the underlying cause can be reversed, treatment with epoetin alfa must be continued indefinitely. Epoetin alfa belongs to a class of drugs called colony-stimulating factors because of their ability to stimulate cells in the bone marrow to multiply and form colonies of identical cells. Epogen and Procrit are both epoetin alfa, but they are marketed by two different pharmaceutical companies.

© Copyright 2010 Anesthesia Abstracts · Volume 4 Number 8, August 31, 2010

Moonen A, Thomassen B, Knoors N, Os J, Verburg A, Pilot P


Pre-operative injections of epoetin-αlpha versus post-operative retransfusion of autologous shed blood in total hip and knee replacement. a prospective randomized clinical trial

J Bone Joint Surg B 2008;90-B:1079-1083

Moonen A, Thomassen B, Knoors N, Os J, Verburg A, Pilot P


Purpose            The purpose of this study was to evaluate the use of a cost effective post-operative retransfusion system following joint replacement surgery in patients with baseline hemoglobin levels between 10 gm/dL and 13 gm/dL, compared with using less cost effective preoperative injections of epoetin alfa. “Evaluating” was operationally defined as comparing the need for allogeneic blood transfusion following surgery and comparing costs of retransfusion and epoetin alfa therapy.

Background            Patients often require blood transfusions during joint replacement surgery. As the risks and complications of blood transfusion have become more evident, the search for alternatives to allogenic transfusion is progressing quickly. Pre-operative injections of epoetin alfa and post-operative retransfusion techniques have shown great potential as alternatives to allogenic transfusions. It remains to be seen, however, what interventions are most successful at reducing the need for blood transfusion during joint replacement surgery. Previously there were no randomized clinical trials which compared pre-operative injections of epoetin alfa and the post-operative transfusion of blood shed during surgery. The specific aim of this research was to compare the differences in the need for allogeneic blood transfusions in these two groups.

Methodology            This study was conducted as a randomized, prospective, clinical trial. A total of 100 patients were consented. Those who were scheduled for elective total joint replacement surgery and devoid of known hematologic diseases, coagulation disorders, malignancies, or infections were enrolled. Each patient’s pre-operative hemoglobin value had to be between 10 gm/dL and 13 gm/dL to be eligible for randomization in to one of two groups.

Group 1 received 40,000 IU of epoetin alfa via subcutaneous injections given weekly beginning three weeks before surgery, with the final injection immediately following the operation. In addition, supplemental oral iron (200 mg 3x/day) was taken beginning three days before the first injection and completed the day of surgery. Group 2 received autologous retransfusions. The autologous retransfusions were carried out via a wound drain connected to the system which provided a suction force so the blood would accumulate in a transfusion bag. The blood was auto-transfused when the bag was full at 500 mL or six hours post-operatively.

All patients received low molecular weight heparin for thromboembolic prophylaxis starting after surgery and continuing for 2 weeks. Allogeneic blood transfusions were administered according to policy. The anesthesia provider determined the hemoglobin transfusion trigger which depended upon the risk factors of the individual and the specifics of the procedure itself (e.g. how much blood was lost during surgery). All allogeneic blood transfusions and complications were recorded. The length of follow up varied from 2 to 18 months.

Results            Statistical analysis of demographic data did not show any significant differences between the two groups. Primary total hip arthroplasty was performed in 60 patients and primary total knee arthroplasty was performed in 40 patients. Spinal anesthesia was used in 84% of the cases. The remainder received general anesthesia. Estimated surgical blood loss was similar between both groups (P = 0.75). The mean transfusion triggers in the epoetin alfa group and the retransfusion group were 8.5 gm/dL and 8.7 gm/dL respectively. Patients undergoing total hip procedures were autotransfused with a mean of 216 mL (range 0-700 mL) and those undergoing total knee procedures were autotransfused with a mean of 131 mL (range 0-500 mL).

Mean pre-operative hemoglobin values for those in the epoetin alfa group and the autotransfusion group were identical. For those in the epoetin alfa group, the pre-surgical hemoglobin levels increased by an average of 2.5 gm/dL compared to their baseline. On post-operative day one, mean hemoglobin values decreased to 11.4 gm/dL in the epoetin alfa group and to 9.7 gm/dL in the autotransfusion group. By post-operative day one, the epoetin alfa group had a mean hemoglobin level of 11.2 gm/dL and the autotransfusion group had mean hemoglobin value of 9.5gm/dL. These reductions of hemoglobin levels on post-operative days one and three were significantly different between groups (P = 0.011 and P = 0.012 respectively). Post-operative clinical complications were similar between groups.

In the epoetin alfa group, two patients (4%) received at least one allogeneic blood transfusion. In the autotransfusion group 14 patients (28%) received allogeneic blood transfusions (P = 0.002).

Assessing only the total hip patients, 7% of those in the epoetin alfa group received allogeneic transfusions compared to 30% in the autotransfusion group (P = 0.042). For total knee patients, none in the epoetin group received an allogenic transfusion while 25% in the autotransfusion group received an allogeneic transfusion (P = 0.042).

Conclusion            This study demonstrated that during hip and knee replacement surgery preoperative injections of epoetin alfa coupled with iron supplements were more effective in reducing the need for allogeneic blood transfusions in mildly anemic patients compared with postoperative autotransfusion of shed blood. It also demonstrated that epoetin alfa therapy was more expensive than using the autotransfusion drain system.


This study was conducted in a rigorous manner and carried out with sound scientific processes. Interesting to note was that the average cost of epoetin alfa injections was estimated at $1,832 compared with zero cost of drug in the autotransfusion group. The Bellovac retransfusion system was estimated at $84.00 per unit. The authors estimated the cost of one allogeneic blood transfusion as similar to the cost of the autotransfusion system. I have seen significantly higher costs quoted for a unit of transfused blood in the United States. (This study was conducted in the Netherlands). It is overwhelming, however, to think of the cost of one patient having complications of an allogeneic blood transfusion. Complications such as the development of transfusion related acute lung injury, transfusion associated circulatory overload, an increased length of stay, possible ICU admission, or equally as devastating, the development of a transfusion reaction and all the care involved in treating the transfusion reaction. Well it would far, far exceed the cost of using the epoetin alfa subcutaneous injections.

Mary A. Golinski, PhD, CRNA

Iron Supplements            The goals of providing oral iron supplements are to supply sufficient iron to restore normal iron levels and replenish hemoglobin deficits. There are a large number of iron preparations available. Iron supplements are available in regular tablets and capsules, liquid and drops, coated and extended release tablets and capsules. Oral iron preparations are available in both ferrous and ferric states. Ferrous salts are preferred because they are absorbed much more readily. The most commonly available oral preparations include ferrous sulfate, ferrous gluconate, and ferrous fumarate. All three forms are well absorbed but differ in elemental iron content. Ferrous sulfate is the least expensive and most commonly used oral iron supplement.

The effectiveness of iron supplementation is determined by the reticulocyte count, hemoglobin, and ferritin levels. Hemoglobin usually increases within 2-3 weeks of starting iron supplementation. Therapeutic doses of iron should increase hemoglobin levels by 0.7-1.0 gm/dL per week. Reticulocytosis (increase in the numbers of reticulocytes) occurs within 7-10 days after initiation of iron therapy. Serum ferritin level is a more accurate measure of total body iron stores. Adequate iron replacement has typically occurred when the serum ferritin level reaches 50 µg/L.

© Copyright 2010 Anesthesia Abstracts · Volume 4 Number 8, August 31, 2010

Stowell C, Jones S, Enny C, Langholff W, Leitz G


An open-label, randomized, parallel-group study of perioperative epoetin alfa versus standard of care for blood conservation in major elective spinal surgery

Spine 2009;34:2479-2485

Stowell C, Jones S, Enny C, Langholff W, Leitz G


Purpose            The purpose of this study was to determine whether or not patients who undergo spine surgery and are treated pre-operatively with epoetin alfa to minimize the need for allogeneic blood transfusion, are at additional risk of deep vein thrombosis (DVT) compared to those who do not receive epoetin alfa but do receive standard care for blood conservation.

Background            Current evidence strongly supports the use of recombinant human erythropoietin (epoetin alfa) in anemic patients to reduce allogeneic blood transfusions when undergoing elective, non-cardiac, nonvascular surgery with significant blood loss. Most of this evidence was from orthopedic surgery studies published in the mid 1990s. A safety analysis of these trials showed that rates of DVT were similar in those who received epoetin alfa and those who received placebo. However, these trials included the use of perioperative anticoagulation. Spine surgery patients represent an appropriate group to study for the development and risk of DVT, as prophylactic anticoagulation was not routinely used in spine patients at the study institutions. It was hypothesized that spine patients who receive epoetin alfa perioperatively, but are not anti-coagulated, are at no greater risk of developing DVT than those who were not anti-coagulated.

Methodology            This was an open-label, parallel-group, multicenter, randomized controlled clinical trial. Those greater than 18 years of age scheduled for elective spine surgery with an anticipated blood loss of 2 to 4 units and with starting hemoglobin values >10 to ≤ 13 gm/dL were enrolled. Exclusion criteria included those scheduled to receive perioperative anticoagulation. Patients were randomized to one of two groups:

Group 1            received epoetin alfa (Procrit) 600 U/kg administered subcutaneously once weekly times four doses plus standard treatment.

Group 2            received standard treatment, which included the institutions’ policy for blood conservation, and did not receive an erythropoiesis-stimulating agent.

All patients received oral iron therapy beginning 21 days prior to surgery. No perioperative anticoagulation was administered. Mechanical deep vein thrombosis prophylaxis was allowed.

Patients were screened for DVT via color-flow Doppler on post-operative day 4 or within 24 hours of discharge. Doppler results were classified as normal, acute thrombosis, not evaluable, or not done (missing data). Doppler images were reanalyzed by an independent reviewer who was blinded to which group the patients belonged to. If interpretations differed, a third party, also blinded, rendered the definitive interpretation.

The following operational definitions were applied for what was considered “DVT positive” or “Other Thrombotic Events”:

·      All DVTs diagnosed by Doppler, both acute and chronic, regardless of symptoms experienced

·      All reported adverse events of DVT, regardless of Doppler result

·      PE was not counted as a DVT if Doppler findings were normal

·      A positive DVT: one or more veins showed a DVT but was considered normal if at least 4 of 6 proximal readings were obtained and no proximal or distal vein showed DVT

·      Doppler assessment was labeled unknown if no proximal or distal vein was positive for DVT but 3 or more proximal readings were not done or were not evaluable

Other clinically relevant adverse events were included in outcome analysis. Life threatening adverse events and/or re-hospitalizations were reported through 30 days after the last dose of the study drug or 30 days following surgery. These patients were followed until their symptoms resolved, returned to baseline, or were stabilized.

The primary outcome variable was the incidence of DVT between subjects undergoing elective spine surgery who received epoetin alfa or standard care for blood conservation. The criterion used for what was considered no additional risk of DVT was a 1-sided 97.5% upper confidence limit of < 4% for the difference between the treatment group with epoetin alfa and the standard care group. Statistical calculations were completed; 572 subjects provided 80% power to demonstrate non-inferiority for the primary study end point (see notes).

Results            A total of 587 subjects completed the study. Group 1:  n = 280 and Group 2:  n = 307. The two groups did not differ significantly in terms of demographic data.

Overall, 4.7% (n = 16) of the subjects in the epoetin alfa group were diagnosed with DVT vs. 2.1% (n = 7) of subjects in the standard care group. The between-group difference for the incidence of DVT was 2.6%. This demonstrated non-inferiority (see notes).

There were 5 subjects (1.5%) in the epoetin alfa group and 3 subjects (0.9%) in the standard care group with clinically relevant thrombovascular events other than DVT. The clinically relevant events in the epoetin alfa group included: 2 CVA, 1 TIA, 1 myocardial ischemia patient, and 1 myocardial infarction patient. The clinically relevant events in the standard care group included: 3 pulmonary emboli (3 patients). It must be noted though, that these cases included 2 subjects in the epoetin alfa group who did not receive the treatment and 1 subject in the standard care group who received the study drug in error.

Seven subjects in the epoetin alfa group and 4 subjects in the standard care group had thrombovascular events that were deemed not clinically relevant. Analysis of the combined incidence of pulmonary embolism or acute DVT confirmed by Doppler performed post hoc demonstrated these events occurred in 9 of the epoetin alfa group and 8 patients in the standard care group. The overall incidence of adverse events was similar between groups as was the incidence of serious adverse events.

Conclusion            This study found that a higher incidence of DVT occurred with the use of epoetin alfa versus standard care for blood conservation during spine surgery, however, the rates of other clinically relevant thrombovascular events were similar. It is suggested to consider antithrombotic prophylaxis if using epoetin alfa in this population.



It must be noted that this study used a very conservative definition for the diagnosis of DVT. For example, any evidence of acute or chronic DVT on Doppler, irrespective of symptoms, and any reported adverse events of DVT regardless of Doppler findings, were considered “positive” for DVT. We cannot lose sight, though, that using these definitions the rate of DVT was higher in the epoetin alfa group, but the rate of symptomatic acute DVT was the same in each of the groups. I believe the operational definitions used in this study were its greatest limitation. The authors did own up to this and were quite thorough in discussing the lack of baseline ultrasound measurements which would have excluded those with preexisting evidence of DVT that ultimately may have been attributed incorrectly to the experimental arm. Additionally, it was noted that asymptomatic DVT identified by Doppler studies alone are more often than not, insignificant clinically. Another very important comment made by the authors, that added to the scientific “appropriateness,” was that they revealed the findings of a post hoc analysis. The post-hoc analysis included subjects with pulmonary emboli (excluded in the original operational definition), excluded those with chronic DVT identified before the study, and excluded those with adverse events of DVT but with normal Doppler findings. This approach did not reach statistical significance! A very notable suggestion for future research with an evidence based approach.

Mary A. Golinski, PhD, CRNA

Notes- Non-inferiority Hypothesis

Most statistical tests in the healthcare arena are performed to show whether two treatments are significantly different from each other and whether we can reject the null hypothesis. The null hypothesis is rejected if the test statistic is sufficiently large compared to the critical value or if the P-value of the test statistic is sufficiently small compared to a pre-specified level of α.

In clinical trials in the biopharmaceutical industry, testing for the non-inferiority of a new modality, such as epoetin alfa, compared with an existing approved/accepted modality, such as other standard care blood conservation processes, has become an acceptable practice. The goal of these trials - the non-inferiority trial - is to show that the new treatment is statistically (and clinically) NOT inferior to the standard approved/accepted process.

The use of non-inferiority experimental designs is well established in evaluating new clinical entities and devices in the biotech and pharmaceutical industries. Adapting this technique for identifying processes and products as non-inferior vs. equivalent (equivalent meaning there is NO difference) within a pre-specified non-inferiority margin can provide more cost effective sampling and analysis when applied to statistical quality control.

© Copyright 2010 Anesthesia Abstracts · Volume 4 Number 8, August 31, 2010

Pediatric Anesthesia

Oliver K, Tucci M, Bateman S, Ducruet T, Spinella P, Randolph A, Lacroix J


Association between length of storage of red blood cell units and outcome of critically ill children:  a prospective observational study

Critical Care 2010;14:R57

Oliver K, Tucci M, Bateman S, Ducruet T, Spinella P, Randolph A, Lacroix J  


Purpose            The purpose of this study was to determine whether or not a relationship existed between red blood cell (RBC) storage time and the development and severity of multiple organ dysfunction syndrome (MODS) in critically ill children.

Background            It is a well established fact that almost half of all critically ill children admitted to the ICU will receive a blood transfusion. While RBC transfusions are aimed at restoring hemoglobin levels and therefore oxygen delivery to vital organs, there is data that suggests transfusions can worsen the outcome of the individual. Complications from RBC transfusion have been reported and include:  increased rates of mortality, increased MODS, acute respiratory distress syndrome, deep vein thrombosis, and hospital acquired infections.

The maximum time banked RBCs can be stored is 42 days. Blood banks issue blood products such as RBCs dispersing the ‘oldest’ units first in order to avoid wasting such valuable product. The most recent studies that evaluated the effect of storage time on outcomes have been conducted on the adult population; and findings are inconsistent. Only one small study to date has assessed the effect of RBC length of storage time before transfusion on the clinical outcomes of the pediatric patient.

Methodology            This was conducted as a prospective, multi-center observational study by the Pediatric Acute Lung Injury and Sepsis Investigators Network in the USA and Canada. Inclusion criteria were:

  • < 18 years of age
  • Admitted to the PICU
  • Length of stay > 48 hours

The data that was collected on admission included:

  • Specific demographics
  • Severity of illness score (Pediatric Risk of Mortality III)
  • Organ dysfunction (Pediatric Logistic Organ Dysfunction score)
  • MODS score

Daily documentation included:

  • The number of RBC transfusions
  • Length of storage of the RBC units before infusion
  • MODS variables
  • Pertinent clinical information
  • Complications

The total number of blood transfusions were recorded as was the volume transfused per unit. The metrics of RBC storage time were operationally defined as:

  • Fresh            units stored for a period shorter the median length of all stored blood
  • Old            units stored for more than the median length of all stored blood.

If a patient received multiple transfusions, old blood was differentiated from fresh blood based upon the oldest unit received. Two outcome measures were assessed:  the primary outcome measure - which was the proportion of patients who developed concurrent dysfunction of two or more organ systems, or had progression of MODS, and the secondary outcome measure - PICU length of stay and 28 day mortality. All outcome measures were monitored prospectively and were checked after the first transfusion.

Results            A total of 30 sites participated in the study. One site was excluded later due to lack of documentation. A total of 977 consecutively admitted critically ill children were enrolled. Of those, 447 were transfused for a total of 1,991 total transfusions. Multiple transfusions occurred in 61% of patients and 86% of total transfusions were pre-storage leukoreduced. Of the transfused patients, data on the length of time the blood was in storage was available for 66% of the 447. For the patients who were assessed and data analyzed, 296 of 447, or 67%, received multiple transfusions.

The median length of storage of the RBCs was 14.0 days and the mean length of RBC storage was 17.8 + 11.6 days. Forty-nine percent (49%) of the transfused patients received their first transfusion on the first day spent in the PICU.

In terms of primary outcome variables:  the number of RBC transfusions was significantly higher in those who developed new or progressive MODS compared with those who did not (P < 0.001). Also, the total volume of RBC transfusions was higher in these patients (P < 0.001) and the proportion of patients who received at least one RBC unit stored for 14 days or longer was greater (P = 0.01). The unadjusted odds ratio for development of MODS or worsening of MODS in patients receiving at least one RBC unit stored for >14 days was significant (P = 0.01). After correcting for confounding variables (demographics, MODS at admission, etc), the adjusted odds ratio for developing new or worsening MODS in those who received older blood remained significant (P = 0.03). Those who received a single transfusion of older blood did not demonstrate new or progressive MODS of significance.

Assessment of the secondary outcome variables demonstrated a significant difference in ICU length of stay for those who received older blood (P < 0.001), however statistical significance was not observed for a difference in mortality.

Conclusion            In this study of critically ill children, transfusing RBCs stored for ≥14 days was independently associated with an increased incidence of MODS and a longer length of stay in the ICU.


While this was just one observational research project, it was an important one, adding a key piece to a much needed body of evidence. Our role as anesthetists is a bit different from those treating these patients in the ICU setting; never should we think we are not a very important piece of the entire process of attempting to make these children better. Our roles may be different, but our objectives are the same. We want to promote positive outcomes. We receive these patients and administer anesthesia; we administer blood products as part of the peri- operative care we are responsible for. Our part in the continuum of care cannot be underestimated nor undervalued. We must be effective communicators with the intensivists and surgeons. Our understanding of the benefits and the risks of blood transfusion as well as the importance of our documentation validating why we transfused, is extremely important.   


Mary A. Golinski PhD, CRNA

MODS-(adopted from:  The multiple organ dysfunction syndrome, John C Marshall, M.D.,

The Multiple Organ Dysfunction Syndrome (MODS) can be defined as the development of potentially reversible physiologic derangement involving two or more organ systems not involved in the disorder that resulted in ICU admission. Organ dysfunction in a critically ill patient can be described in one of two ways — as the clinical intervention that was employed to support the failing organ system (mechanical ventilation, hemodialysis, inotropic or vasopressor agents, parenteral nutrition etc), or as the acute physiologic derangement that made such support necessary.

The first descriptions of the syndrome generally counted the number of failing systems, and used as descriptors, the need for clinical intervention. Recently several similar descriptive scales have been developed, based on the quantification of organ dysfunction as a numeric scale. Each uses the same six organ systems to characterize MODS - respiratory, cardiovascular, renal, hepatic, neurologic, and hematologic systems. They differ in minor ways with respect to the selected parameters to describe cardiovascular dysfunction, and in the timing and weighting of the variables selected. The Multiple Organ Dysfunction (MOD) score is a scale that uses physiologic variables exclusively.

The multiple organ dysfunction (MOD) score

Organ system







(PO2/FIO2 Ratio)

> 300




≤ 75


(Serum Creatinine)

≤ 100




> 500


(Serum Bilirubin)

≤ 20




> 240


(R/P Ratio)

≤ 10.0




> 30.0


(Platelet count)

> 120




≤ 20


(Glasgow Coma Score)





≤ 6

The PO2/FIO2ratio is calculated without reference to the use or mode of mechanical ventilation, and without reference to the use or level of PEEP.
The serum creatinine level is measured in μmol/L, without reference to the use of dialysis.
The serum bilirubin level is measured in μmol/L.
The R/P ratio is calculated as the product of the heart rate and right atrial (central venous) pressure, divided by the mean arterial pressure
The platelet count is measured in platelets/mL 10-3
The Glasgow Coma Score is preferably calculated by the patient's nurse, and is scored conservatively (for the patient receiving sedation or muscle relaxants, normal function is assumed unless there is evidence of intrinsically altered mentation

Pediatric risk of Mortality III Score-A validated 3rd generation pediatric physiology-based score for mortality risk. It was developed and remodeled from the original scale as relationships between physiologic status and mortality risk needed to be reassessed. This was owed to new treatment protocols, therapeutic interventions, the development of more sophisticated monitoring strategies, and changes in patient population.

Pediatric Logistic Organ Dysfunction Score-A descriptive score which is validated to estimate the severity of cases of multiple organ dysfunction syndrome in PICUs. The data required to calculate this score are collected from baseline to discharge from the PICU or up to 2 hrs before death in the PICU. The Pediatric Logistic Organ Dysfunction score can be used to describe the clinical outcome of patients during their stay in a PICU.

Facts regarding storage of Packed Red Blood Cells-(see also: American Journal of Critical Care. 2007;16:39-48. Packed Red Blood Cell Transfusion in the Intensive Care Unit: Limitations and Consequences).

RED BLOOD CELLS | GENERAL INFORMATION (adopted from the American Red Cross):

Preparation variations include Red Blood Cells (Adenine-Saline Added); Red Blood Cells Leukocytes Reduced (LR-RBC); Red Blood Cells Apheresis; Red Blood Cells Deglycerolized; Red Blood Cells Irradiated; Red Blood Cells, Low Volume; and Red Blood Cells Washed.

Description of Components:

Red Blood Cells consist of erythrocytes concentrated from whole blood donations by centrifugation or collected by apheresis method. The component is anticoagulated with citrate and may have had one or more preservative solutions added. Depending on the preservative-anticoagulant used, the hematocrit of Red Blood Cells ranges from about 50-65% (e.g., AS-1, AS-3, AS-5) to about 65-80% (e.g., CPDA-1, CPD, CP2D).

Red Blood cells contain an average of about 50 mL of donor plasma (range 20 mL to 150 mL), in addition to the added preservative and anticoagulant solutions.

Each unit contains approximately 42.5-80 gm of hemoglobin or 128-240 mL of pure red cells, depending on the hemoglobin level of the donor, the starting whole blood collection volume, and the collection methodology. When leukoreduced, RBC units must retain at least 85% of the red cells in the original component. Each unit of Red Blood Cells contains approximately 147-278 mg of iron, most in the form of hemoglobin.

A dose of one unit of compatible Red Blood Cells will increase the hemoglobin level in an average sized adult who is not bleeding or hemolyzing by approximately 1 gm/dL or Hct by 3%. In neonates, a dose of 10-15 mL/kg is generally given, and AS-1 or AS-3 packed red cells with a hematocrit of approximately 60% will increase the hemoglobin by about 3 gm/dL.

Leukoreduced Red Blood Cells- The removal of leukocytes from the blood or blood components supplied for blood transfusion. After the removal of the leukocytes, the blood product is said to be leukoreduced.

© Copyright 2010 Anesthesia Abstracts · Volume 4 Number 8, August 31, 2010