Ongoing Research
Published Abstracts
Published Manuscripts
Past Meetings
Contact Us

Prospective, Randomized Trial Of Restrictive Versus Liberal Blood Transfusion Policy In Burns Greater Than 20% TBSA
Principal Investigator: Tina L. Palmieri, MD
Research Coordinator:
Start Date: 01/01/2004
End Date: 01/01/2006


This is a prospective, randomized trial to compare clinical outcomes as they relate to blood transfusion strategies. Traditional transfusion practices, where hemoglobin levels are maintained at approximately 10-11 g/dL will be compared a more restrictive transfusion strategy where hemoglobin levels are maintained at 7-8 g/dL.


Significance.  The use of blood transfusions to augment hemodynamic status has become a common practice, with over 11 million units of red blood cells transfused in more than 3 million patients every year in the United States.1 In the intensive care unit approximately 25% of critically ill patients receive blood transfusions to ameliorate the effects of anemia.2 These physiologic responses to anemia include increased cardiac output, decreased blood viscosity, increased oxygen extraction, and capillary/tissue alterations.3. Although critically ill patients may be predisposed to the adverse effects of anemia, they are also subject to the adverse consequences of blood transfusion. Commonly cited adverse effects of red blood cell (RBC) transfusions include infection, pulmonary edema, immune suppression, and microcirculatory alterations.4 Traditionally blood transfusions have been administered when the patient’s hemoglobin is less than 10 g/dL or hematocrit less than 30%. However, a recent multicenter, prospective, randomized study of transfusion in ICU patients, the TRICC study (Transfusion Requirements in Critical Care), challenges this standard.5 A total of 838 patients were randomized to receive blood transfusion based on a liberal (maintain hemoglobin 10-12 g/dL) versus a restrictive (maintain hemoglobin 7-8 g/dL) strategy. The restrictive strategy was at least as effective as the liberal strategy in critically ill patients. Significant differences favoring the restrictive strategy included in-hospital mortality, cardiac complication rate, and organ dysfunction.

Historical Review.  The applicability of the TRICC study to the burn population is somewhat limited. First, patients were excluded if they had a drop in hemoglobin of 3 g/dL or required 3 units of blood within 12 hours of enrollment. It is common for burn patients in the acute resuscitative phase to have a significant drop in hemoglobin due to ongoing blood loss (other trauma, escharotomies), hemodilution, or hemolysis. Second, the study did not specifically indicate whether or not burn patients were enrolled. Of the 838 patients, only 165 were included in the trauma category, of which burn injury would have been a minor component. This does not provide an adequate sample size to determine the effects of a restrictive strategy on this population. Third, there was no evaluation of the effect of a restrictive strategy on wound healing, a major consideration in burn patients. The effects of blood transfusion thus remain undefined in burn patients.

The impact of the TRICC on transfusion practices in the United States has also been limited. The CRIT study, a prospective, multiple center, observational study of ICU patients, analyzed the transfusion practices of 284 ICUs in 213 hospitals in the United States.6 The mean pre-transfusion hemoglobin was 8.6±1.7 g/dL, indicating that the majority of patients were still being transfused at a hemoglobin higher than what was recommended in the TRICC study. Once again, the number of units of RBC transfusions the patients received was independently associated with longer ICU length of stay and increased mortality. However, age of blood transfusion did not appear to impact survival. The CRIT study excluded burn patients; thus, it provides no data on burn center transfusion practices and the outcomes related to those practices.

Limited data exists regarding the effects of a restrictive blood transfusion policy in adult burn patients.7  In on study by Sittig and Deitch a total of 14 patients admitted to a burn center during a 6 month interval were transfused if their hemoglobin level was <6.0 g/dL.The outcomes of patients with >20% total body surface area (TBSA) burns or patients requiring excision and grafting of >10% TBSA were retrospectively compared to a matched group of 38 patients treated the previous year using a non-restrictive policy (hemoglobin maintained above 9.5-10 g/dL). No differences existed in hospital length of stay. The patients treated with the liberal strategy received 3.5 times as much blood as their restrictive counterparts. Although this study is an important first step in the evaluation of blood transfusion in burn patients, itis limited by its retrospective nature, review bias, and inadequate number of patients.

To address the issue of blood transfusion practices in burn patients, we conducted a survey of burn unit directors regarding their blood transfusion practices. The survey evaluated transfusion triggers for adults, elderly, and children with severe (>20% TBSA) burn injury.8 The overall mean hemoglobin transfusion threshold in 51% of all respondents was 8.1 g/dL, while 26% reported a hematacrit transfusion threshold of 26%. Blood transfusion thresholds differed significantly based on both burn size and age, with respondents being more likely to transfuse patients over 60 years of age at any given burn size than their younger counterparts. The transfusion threshold for patients over 60 years with a 40-60% TBSA burn was 9.25 g/dL as opposed to 8.4 g/dL for the 40-60 year old group, and 8.0 g/dL for the 16-39 group. The hemoglobin threshold decreased (i.e. patients were likely to receive blood transfusion at a lower hemoglobin level) as TBSA increased within any given age group. Other factors which increased the likelihood of blood transfusion were the presence of cardiac dysfunction, sepsis, and the acute respiratory distress syndrome (ARDS). This study documented self-reported transfusion thresholds, but did not document the actual transfusion practice for burn patients.

To evaluate actual burn center transfusion practices, the Burn Multicenter Trials Group (consisting of 21 burn centers nation-wide), reviewed the actual use of blood transfusion in patients with burn injury >20% TBSA over a one year period.9 Data was collected from 21 different burn centers on a total of 666 patients. The mean patient age was 33.2±0.9 years, mean TBSA burn was 38.9±0.8%, and the overall mortality was 20.4%. The overall hemoglobin at which the first transfusion was administered was 9.35±0.8 g/dL for all patients. Mortality was related to age, TBSA burn, presence of inhalation injury, and the number of units of PRBC transfused. Although this study suggests that blood transfusions influence outcome, it does not provide direct evidence that maintenance of a lower transfusion trigger improves survival. A prospective, randomized, multicenter study is needed to determine the appropriate blood transfusion threshold for burn patients.


The hypothesis of the study is that a restrictive approach to red blood cell transfusion (hemoglobin maintained at 7-8 g/dL) is associated with equivalent survival compared to a liberal approach (maintaining hemoglobin 10-11 g/dL) in patients with burns ≥20% TBSA. 


Study Design and Treatment Protocols . This is a prospective, randomized trial to compare patient outcomes as they relate to transfusion strategies. Randomization will be based on “intent to treat,” meaning that each potential study patient will be assigned a study number upon admission and approached for consent. Consecutively admitted burn patients admitted with the above criteria will be assigned to one of two treatment groups (restrictive versus liberal transfusion strategy). Sealed, opaque envelopes arranged by a computer-generated random order will be prepared by the coordinating center and distributed to each participating institution, where they will be sequentially opened to determine treatment assignments. Patients assigned to the restrictive strategy of transfusion will receive blood transfusions to maintain hemoglobin levels in the range of 7-8 g/dL. In the liberal transfusion group, hemoglobin concentrations will be maintained 10-11 g/dL. The patients will receive blood transfusions one unit at a time with a hemoglobin measurement taken after each unit of blood transfused. Compliance will be assessed with daily monitoring of hemoglobin concentrations for each patient for the first two weeks. After the first two weeks hemoglobin levels will be obtained daily until the patient meets two of the following criteria: wound 90% healed (7 days after the last operation), the patient is no longer in the burn intensive care unit, the patient has three stable hemoglobins (defined as three hemoglobins within 0.2 mg/dL of each other), or intravenous narcotics are no longer needed. If the patient meets these criteria, hemoglobin levels will be obtained twice a week until discharge.

Blood transfused in the operating room will follow this protocol unless transfusion is needed to maintain hemodynamic stability during times of acute blood loss. The anesthesiologist will be notified of patient participation in the study and blood transfusion policy (restrictive vs. liberal transfusion). Hemoglobin levels will be obtained preoperatively within 8 hours of onset of the operation. Routine monitoring of patient vital signs including blood pressure, central venous pressure, arterial blood gas, mixed venous oxygen saturation, urine output, temperature, and ventilatory status will be continued throughout the operation. Prior to and immediately after each unit of blood transfused a hemoglobin level will be obtained. If the patient is hypotensive due to massive blood loss, blood will be transfused as needed to maintain hemodynamic stability without waiting for the results of the hemoglobin level. The reason for transfusion will be recorded, as will the hemoglobin at the time the transfusion was begun and finished. The operative procedure, estimated blood loss, number of units of blood transfused in the operating room, and other fluid administered (crystalloid, colloid, other blood products) during the operation will be recorded. A hemoglobin level will be obtained postoperatively within 30 minutes of completion of the surgical procedure. The operative period begins at the time the patient enters the operating room and ends when the patient leaves the operating room.

Duration.  Subjects will be followed throughout admission for acute burn injury.

Sample Size.  A power analysis has been performed by an independent statistician. The sample size required at p = 0.05, power = .8, using a one-tailed test is 295 subjects. We intend to enroll 400 subjects to compensate for subject mortality/withdrawal. Enrolled subjects that survive less than 24 hours or whose treatment includes a Do Not Resuscitate (DNR) status will be withdrawn from the study.

Safety.   After enrollment of 200 patients an interim analysis of mortality and complications will be performed by an independent safety monitoring board consisting of a statistician, a blood bank physician, an anesthesiologist, and a burn surgeon other than the principle investigator.

Subject Selection

Study Population. Patients will be enrolled in the study if they are admitted to one of the participating Burn Multicenter Research Group Centers within 96 hours of injury with a burn injury ≥20% TBSA with an anticipated need for burn excision and grafting.

Subject Recruitment .  Informed consent will be obtained by the one of the Investigators or research personnel. Eligible subjects will be approached for informed consent within 72 hours of admission. If the investigator determines that the subject lacks decision-making capacity, the investigator will seek surrogate consent.

Inclusion Criteria.

  • >20% TBSA burn with anticipated operation need on admission
  • age >18 years
  • Admission within 96 hours of injury

Exclusion Criteria.

  • <18 years of age
  • pregnancy
  • inability or unwillingness to receive blood products
  • history of chronic anemia (hemoglobin <9.0 g/dL one month prior to enrollment)
  • preexisting need for hemodialysis
  • brain death or imminent brain death
  • non-survivable burn as determined by the attending burn surgeon
  • angina or acute myocardial infarction
  • preexisting hematologic disease
  • closed head injury with Glasgow coma scale <9

Baseline Data Collection. Data will be collected on the following parameters within 72 hours of patient enrollment: age, gender, TBSA burn (second and third degree), inhalation injury, APACHE II score, multiple organ dysfunction score (MODS), tobacco use, recreational drug use, and associated illnesses. Blood alcohol and urinary toxicology screen will be obtained on all patients at admission.

Ongoing Data Collection. Organ dysfunction will be assessed with lab values obtained within the first 24 hours and daily. If routine clinical care presents multiple values on any given day, those indicating the highest level of dysfunction will be recorded.Hemoglobin concentration will be recorded daily and prior to each blood transfusion. The number of red-cell transfusions, age of red blood cells (number of days until unit expires and cannot be used), use of leukoreduced blood, use of other blood products (albumin, fresh frozen plasma, platelets, cryoprecipitate), medications given (includes vasoactive agents), need for mechanical ventilation, infection (catheter, blood, urine, pneumonia, wound as defined by the CDC) and dialysis will also be recorded. Each surgical intervention will be recorded, including blood loss, number of red blood cell units transfused during the procedure, and number of blood products transfused. APACHE and MODS score will be obtained weekly.

Patient Withdrawal from the Study . Patients may be withdrawn from the study if, in the judgment of the patient’s burn surgeon, the patient has any of the following: acute myocardial event (myocardial infarction with elevated troponin level, new onset angina); sepsis with one of the following: mixed venous saturation of <70%, elevated lactate, oxygen debt; or extensive operations by another service (i.e. trauma, gastrointestinal bleed). Patients will continue to be followed for the primary and secondary outcome measures (as defined below). The study principle investigator will be notified of all patient withdrawals.

Outcome Measures. The primary outcome measure will be death from all causes. Secondary outcomes include: length of stay in the burn intensive care unit, hospital length of stay, length of time on mechanical ventilation, organ dysfunction, time to 90% burn wound healing (defined as 7 days after the last excision and grafting procedure), and number of infectious episodes. Patients who die will be assigned a MODS score of 24. Demographics of patients in the two treatment groups will be compared using univariate descriptive statistics. Continuous outcome measures including MODS scores and total RBC units transfused will be analyzed using an independent t test. Number of infectious episodes, time to healing, and burn unit length of stay will be compared using a t test. Multiple regression analysis will be used to delineate the influence of multiple factors on mortality.


The risks associated with this study include those relating to blood transfusion (both groups are subject to these risks, as patients with burn injury of this magnitude, based on our previous study, receive an average of 13 units of PRBC during their hospital stay). The risks of blood transfusion include: transfusion-related Acute Lung Injury (ALI) (1/8000), delayed hemolytic transfusion reaction (1/1000), acute hemolytic transfusion reaction (1/250,000-1,000,000), infection with hepatitis C (1/1,600,000), hepatitis A (1/1,000,000), hepatitis B (1/220,000), HIV (1/1,900,000), bacterial contamination of blood (1/500,000), and blood type mismatch (1/14,000).10 Other potential complications associated with blood transfusion include pulmonary edema and congestive heart failure. The restrictive transfusion policy could potentially result in decreased oxygen delivery and delayed wound healing. Patients developing an acute cardiac event during the study will be withdrawn from the protocol, but will continue to be monitored for complications and mortality.


The patients in the restrictive transfusion group will likely receive fewer units of blood during the course of their hospitalization.

Risk/Benefit Ratio

Both the liberal and restrictive transfusion policies are used in current burn center practice, since no standard of care for blood transfusion in burns currently exists. Patients will receive routine monitoring of their status as well as additional hemodynamic monitoring during the study. No additional foreseeable risks are present for patients participating in this study; therefore, the benefits outweigh the risks.


Study records and participation in this study will be kept confidential in accordance with applicable state and federal laws. Records will be available only to the investigators and agencies with the authority to regulate research.


No additional costs, originating from the study, are anticipated for patients or their associated third party payers.

Principal Investigator’s Disclosure of Personal and Financial Interests

The principal investigator has no financial interest in this research study.


1. Wallace EL, Churchill WH, Surgenor DM, Cho GS, McGurk S. Collection and transfusion of blood and blood components in the United States, 1994. Transfusion 1998;38:625-636.

2. Hebert PC, Wells G, Martin C, Tweeddale M, Marshall J, Blajchman M, Pagliarello G, Schweitzer I, Calder L: A Canadian survey of transfusion practices in critically ill patients. Crit Care Med 1998; 26:482-487.

3. Marini JJ. Transfusion triggers and Occam’s rusty razor. Crit Care Med 1998;26:1775-1776.

4.Alvarez G, Hebert PC, Szick S. Debate: transfusing to normal haemoglobin levels will not improve oucome. Crit Care 2001;5(2):56-63.

5. Hebert PC, Wells G, Blajchman MA, Marshall J, Martin C, Pagiarello G, Tweddale M, Schweitzer I, Yetsir E, et al, A multicenter, randomized controlled clinical trial of transfusion requirements in critical carel N Engl J Med 1999;340:409-17.

6. Corwin HL, Gettinger A, Pearl RG, Fink MP, et al. The CRIT study: anemia and blood transfusion in the critically ill-current clinical practice in the United States . Crit Care Med 2004; 32:39-52.

7. Sittig KM, Deitch EA. Blood transfusions: for the thermally injured or for the doctor? J Trauma 1994;36(3):369-372.

8. Palmieri TL, Greenhalgh DG. Blood Transfusion in Burns: What Do We Do? J Burn Care and Rehab 2004;    .

9. Palmieri TL, Caruso DM, Foster KN, Cairns B, Gamelli RL et al. Blood transfusion practices in major burn injury: A multicenter study. Presented at the 36th Annual American Burn Association Meeting, Vancouver, CA. March 2004.

10. Goodnough LT, Brecher MD, Kanter MH, et al. Transfusion medicine: Part I: Blood transfusion. N Engl J Med, 1999; 340:438-447.