http://www.medscape.com/viewarticle/533726
Blood Transfusions Impair Recuperation From Severe Burns and Open Heart Surgery
NEW YORK (Reuters Health) Jun 02 - Patients undergoing coronary artery bypass grafting (CABG) or being treated for major burns risk adverse outcomes if they are given blood product transfusions, according to two articles in the June issue of Critical Care Medicine.
Researchers at the Cleveland Clinic Foundation in Ohio, led by Dr. Colleen Gorman Koch, point out that multiple studies have revealed increased morbidity associated with administration of packed red blood cells (PRBCs). However, as they note in their paper, these studies suffered from small sample size, failure to include all transfusions received during hospitalizations, or from limiting their evaluation to a single morbid outcome.
The Ohio-based team examined outcomes for nearly 12,000 patients who underwent CABG at their institution between 1995 and 2002. Nearly half received at least one unit of PRBCs. In their analysis, the investigators adjusted for multiple risk factors, including demographics, cardiac condition, lab values, comorbidities and operative conditions.
The adjusted odds ratio for in-hospital mortality was 1.77 for one unit compared with no transfusions (p < 0.0001). There was a dose-dependent relationship, with risk escalating rapidly after about 5 units.
The investigators also observed significantly increased risk for renal failure, need for ventilator support beyond 3 days, serious infection, cardiac morbidity, and neurologic complications (p < 0.001 for each morbid event).
Predictors for transfusion included reoperation, older age, lower preoperative hematocrit, lower BMI, and renal dysfunction.
Dr. Koch's team attributes the increased risk associated with PRBCs on an intense inflammatory response and adverse immunomodulatory effects, such as fewer circulating lymphocytes, activation of immune cells, and down-regulation of antigen-presenting cells.
They advise that "better identification of groups at high risk for transfusion and potentially modulating this risk through interventions such as optimizing the preoperative hematocrit, meticulous attention to crystalloid administration, and reexamining the hematocrit threshold for perioperative transfusion ... may circumvent the need for transfusion."
In the second paper, Dr. Tina L. Palmieri from the University of California-Davis Medial Center in Sacramento and her associates note that there is only limited information regarding risk associated with transfusions for burn patients.
Their study investigated outcomes among patients treated during the year 2002 at 21 regional burn centers throughout the US and Canada. Included were 620 patients with burns covering at least 20% of the total body surface area, among whom 463 (74.7%) received blood transfusions (mean 13.7 units/patient). Ninety patients (14.5%) who survived for more than 24 hours died in-hospital.
The investigators found that, overall, transfusion was not associated with risk of death. However, those who died were more likely than survivors to receive transfusions outside of the operating room.
In multiple logistic regression analysis, independent predictors for mortality included total number of blood transfusions, as well as age, extent of burn, cardiac disease, and blood stream infection.
Furthermore, logistic regression modeling revealed that the risk of developing an infection increased 13% for each unit of blood transfused (p < 0.001).
The authors note that immunosuppression always follows severe burn injury. "Further compromise of the immune system by blood transfusion may increase the patient's susceptibility to infection and may affect mortality," they suggest.
Dr. Palmieri's group calls for "a prospective, randomized, multicenter study...to determine the appropriate indications for blood transfusion in burn patients. Until then, they recommend that blood transfusion for burn patients should be administered only to those with a demonstrated physiologic need.
Crit Care Med 2006;34:1602-1616.