Risks of transfusion: outcome focus

Transfusion risks have been reported in texts with a focus on pathogen transmission. Today we have made strides in making our blood supply safer. Most blood bankers will herald their accomplishments by saying that the blood supply is the safest it has ever been and they are correct.1-4 Tremendous amounts of research dollars and technologic advances have occurred in making hepatitis C and human immunodeficiency virus (HIV) transmission extremely rare.2 This article will discuss those risks as well as a number of other routinely mentioned risks (i.e., ABO-Rh mismatch, transfusion related acute lung injury [TRALI], graft-versus-host disease [GVHD], etc.) In the past few years, however, new and groundbreaking research is being performed examining transfusion and outcome. Much of the focus of this article will be on that work and asking the all important question: Does transfusion do what it is intended to do—improve outcome or prevent adverse outcomes? Red cell (RBC) transfusions are administered to increase oxygen-carrying capacity. Unfortunately there is no single best clinical monitor that can tell us when to transfuse. There is disagreement as to when a transfusion is necessary or beneficial. This disagreement is because there is little research outlining when transfusions improve outcome. Transfusion is one of the most widely utilized medical interventions and has always been held in high esteem by practioners. It has not undergone prospective randomized testing to examine efficacy. We have a great deal of information about risks but very little data regarding efficacy. The trigger to transfuse has evolved over time. Evolved is the key word here because there have not been extensive (different disease entities and specific patient groups) bench or human studies delineating when it is either necessary or wise to transfuse. In the early 1900s, immediately after modern blood banking was born, the trigger to transfuse was a hemoglobin (Hb) level of approximately 3.5 to 4 g per dL.5 Interestingly, that number is the number at which human critical oxygen demand is found with anemia. The early transfusionists described that as the level they would encounter shock and congestive heart failure if patients did not receive a blood transfusion. In the 1920s the number rose to about 5 to 7 m per dL with the idea that transfusion should be a prophylactic treatment, not waiting until the adverse events of critical DO2 were encountered.5 That transfusion trigger remained until after the Second World War. In 1947, a single influential physician published his opinion that the best Hb level to transfuse at for surgery was 10 g per dL and that if a patient had lost approximately 15 percent of his or her circulating volume a unit (or two) of blood should be administered. This was not based on bench research or human outcomes research, but was one man's opinion. It was reprinted in every anesthesia and surgical text for the next 40 years. With the evolution of practice, it took on a near religious observance and can still be found in some institutions today. Only after the HIV crisis of the late 1980s have we realized that a more critical look at a transfusion trigger and outcome needs to be performed.1,6 To date only a very few small studies have been performed with proper prospective data gathering to help us understand the risk-to-benefit ratio of transfusion. For a number of years, from the 1970s through the 1980s, the risks of hepatitis B and C as transmitted by transfusion were widely recognized.5 It was estimated that that risk for patients to seroconvert from transfusion was approximately 10 percent. It was widely accepted that transfusion was “good,” and there was no research conducted comparing risks and benefits of transfusion. The following statistics for resulting adverse outcomes are merely speculative but are staggering.6 If one assumes that between the early 1970s and the late 1980s some 30 to 50 million people were transfused with a viral seroconversion rate of 10 percent that means that some 3 to 5 million cases of new hepatitis were transmitted. Of these approximately 50 percent proceeded to chronic active hepatitis (1.5-2.5 million) and of those approximately 20 percent (300,000-500,000) proceeded to cirrhosis. That means that over that 20-year period by iatrogenic means alone we have created a major epidemic of cirrhosis and hepatomas, some of which we are still caring for today. In my career as a hepatic transplant anesthesiologist I have cared for at least 50 patients with cirrhosis undergoing liver transplantation secondary to transfusion hepatitis. Were those transfusions life-saving or necessary? When the HIV crisis became a major epidemic in the developed world, transfusion medicine came under very direct scrutiny. Questions were raised as to the vigilance of blood bankers and how information had been disseminated with regards to the risks of transfusion. The 1990s saw tremendous strides in reduction of the pathogen transmission risks of transfusion and outcome. Variant Creutzfeldt-Jacob disease (vCJD) or mad cow disease has captured the political events of Europe, and recently a single cow infected that crossed the border between Canada and the United States caused widespread financial events. For many years the blood banking industry prided itself on the fact that no single case of vCJD had been transmitted by transfusion. Just within the past 6 months the UK has announced the first case of transfusion-transmitted vCJD. Others certainly a wait us in the future. Advances in surrogate testing, enzymes, and voluntary disqualifications and the eventual establishment of nucleic acid testing (NAT; 1996-1997) radically reduced the risk of hepatitis C and HIV via transfusion. Today the risks of these infectious complications being transmitted through transfusion are vanishingly small (Table 1).2 Hepatitis B remains a significant risk but other adverse outcomes are more prevalent. NAT, an outgrowth of the genetic revolution, has made our blood supply safer not just from HIV and hepatitis C. In the past 3 years the specter of West Nile virus has arisen. Within 2 years of appreciation of it as a transfusion pathogen risk, a new NAT was devised and that has been applied to all blood harvested, the end result being that the risk of West Nile virus transmission has fallen from 1 in 400 to 4000 units transfused in 2001 (summer) to about 1 in 2.4 million units today.7 New viruses will be constantly cropping up in our blood supply, and we cannot test for them all. Transfusion-transmitted virus (TTV) is present in 52 percent (8-82 percent reported) of all units infused and it is simply not tested or detected at this time.8 Cytomegalovirus and Epstein-Barr virus are present in a large number of units. If a patient is negative for these viruses, he or she may well receive the virus through transfusion. If a patient is immunosuppressed and negative for these viruses, however, blood tested for these viruses is available. There is the threat of severe acute respiratory syndrome and avian flu virus looming on the horizon as well as any number of other unknown viruses that could be finding their way into our blood supply. Universal pathogen destruction is under testing for FDA approval at this time. If it is successfully brought forward, however, this pathogen reduction technology will be costly and not available for every unit of blood transfused. A wide number of bothersome and less widely publicized adverse events with blood transfusion are often discussed in the literature (Table 2).2 These are thought to be less important often in such discussions simply because they do not gather the big headlines of HIV and hepatitis. If you were a patient and they happened to you, however, they would be an adverse outcome. Febrile reactions are very common and cause transfusions to be halted, work-ups to be begun, and perhaps even longer stays in the hospital. Allergic reactions can be minor or all the way to catastrophic with anaphylactic events. Minor ABO-Rh mismatch is far too common and delayed RBC destruction is also far too common. Thankfully, major ABO-Rh RBC lysis with disseminated coagulation and death occurs only in about 1 in 100,000 units transfused. That is still more common than hepatitis B today and is fully preventable. Bacterial contamination of banked units is a rare but catastrophic event. RBC bacterial contamination occasionally happens and is discussed by the FDA and the CDC through announcements and warnings. The RBC units, although held for up to 42 days, are maintained at 4°C and are therefore less likely to grow bacteria than products maintained at normothermia or room temperature. Occasionally Yersinnia enterocolitica and Serratia marcescens are found in a transfusion disaster. Platelet (PLT) units, either random single-donor or plateletpheresis units are maintained for up to 5 days in the blood bank at room temperature. Bacterial contamination has been estimated to occur in a significant number of these units, perhaps as often as 1 in 400 to 1 in 2000 units. The risk to patients is hard to discern because voluntary reporting to the FDA is rather poor. It is generally accepted, however, that the risk of a fatal septic PLT transfusion could be as high as 1 in 25,000 to 50,000 units. In the past several months the AABB has asked that all hospitals and blood centers begin to screen their PLTs for bacterial contamination. With that being implemented a significant percentage of PLT units are being discarded. One cannot jump to the conclusion that every unit discarded is bacterially contaminated but some must be infected. Most physicians do not recognize this as a potential risk of transfusion and would look for other causes of hypotension if confronted with vital sign changes upon transfusion. There are pressures to increase the storage time for PLTs from 5 to 7 days but no decision has been made in that regard. Adverse events with transfusion can therefore take one of the more classically discussed outcomes above. These are the classic viral risks and some of the other, “bothersome” risks of transfusion. Remember, if it happens to you, even if it is only a bothersome risk, it still has adverse consequences. When surgeons and anesthesiologists write about adverse outcomes in their frame of reference they refer to the following: death, myocardial infarction (MI), perioperative infection (pneumonia, wound infection, peritonitis, urinary tract infection, prosthetic infection, and subacute bacterial endocarditis), stroke, congestive heart failure (CHF), cancer recurrence (metastasis), bleeding, and others. Transfusion of blood products has traditionally been given to prevent some of these adverse events. Especially when one considers the use of RBC products in relation to the potential for ischemic adverse outcomes and with regard to bleeding and the use of PLTs, fresh frozen plasma (FFP), and cryoprecipitate. RBC products are administered with the idea of either preventing an oxygen delivery debt or enhancing oxygen-carrying capacity. PLTs, FFP, and cryoprecipitate are administered either to treat active bleeding or to prevent bleeding in patients with real or perceived coagulopathies. It has always been assumed that these products work and make patients better. RBC transfusions today are largely dependent on the use of RBCs stored at 4°C in the blood bank. Fresh blood is rarely available but it should be noted that it can be obtained in some centers if desired. As RBCs are stored and aged they undergo a number of physiologic changes.9-13 The RBCs are largely using anaerobic glyocolisis to maintain cellular integrity. As such the blood rapidly becomes acidotic and by Day 28 of storage it can have a very low pH value. Potassium leaks out of the cells and it can rise in the plasma to levels of 78 meq per L or higher.9-13 The intracellular adenosine diphosphate is relatively well maintained but the intracellular 2,3-diphosphoglyceratae (2,3-DPG) levels plummet within 24 hours of harvest.9-13 The shift in available 2,3-DPG causes the Hb in stored cells to profoundly bind and hold oxygen. Therefore, when initially transfused, stored RBCs take up oxygen with tremendous affinity. They, however, do not release the oxygen to tissues. It has been estimated that stored RBCs take oxygen from circulating plasma, other normal RBCs, and even from tissue myoglobin. A recent prospective study of oxygen delivery to tissues in patients after cardiac surgery found that the transfusion of 1 or 2 units of RBCs did nothing to improve oxygen delivery to striated muscle. The shift to 100 percent O2 breathing, however, radically increased tissue oxygen delivery.14 So the idea that one is transfusing RBCs to increase oxygen-carrying capacity may not be realistic. That oxygen-carrying capacity, however, may be doing nothing or worsening the release of oxygen to the tissues that it is targeted to help. Blood utilized in the US has an average life of 19 to 21 days at transfusion. In Europe the average unit transfused is younger by about 5 days. Blood is considered “fresh” if its shelf life is younger than 5 days old. Once transfused the cellular dysfunctions of RBCs begin to be repaired as the cells undergo metabolism. The potassium is reabsorbed and the 2,3-DPG is replenished. But the return to normal physiology can take between 5 and 24 hours to restore. Of note is the fact that at 42-days-old only about 70 percent of cells are even capable of surviving for 24 hours once transfused. Therefore, 30 percent are ghosts or merely cellular debris. Those cells that are still alive are swollen and unable to behave normally in the microcirculation. RBCs that are stored lose approximately 20 to 30 percent of their cell membrane lipid content and their cytoskeletons are inflexible. Cell-to-cell interactions through fibrin cross-linking lead to microaggregate formations that increase both in number and in number of cells per microaggreagte over the length of storage.15 These factors combined with decreased oxygen binding and release lead to some fascinating and counterintuitive events seen with transfusion of RBCs.15-22 In articles just beginning to be appreciated, it has been demonstrated that patients transfused either during heart surgery or in the critical care units have worse P50 values after transfusion of RBC units.16-20 The gastric wall pH as measured by gastric tonometry has shown that patients transfused with RBCs acutely have worsened gut mucosal oxygen supply than before transfusion. In animal models of the microcirculation it has been shown that if hemorrhagic shock is induced the mesenteric blood flow is dramatically decreased.21 When the animal's circulating volume is restored with fresh blood, the mesenteric blood flow returns close to normal. When banked blood is transfused, however, only about 10 to 15 percent of mesenteric blood flow is restored, and it is unclear how long after volume restoration mesenteric flow remains depressed. In studies of the microcirculation and oxygen delivery to the tissues in need, it is also clear that these tissues are either not restored to normal oxygen supply demand ratios or made worse. In studies of rat gut oxygen critical DO2 levels were reached at a higher Hb level if transfused blood was utilized.21-23 That means that shock, as defined by critical oxygen deficit, is more likely or encountered at a higher Hb level in the face of transfused RBCs as opposed to simple anemia. Considerable work from databases has been done. Most of the work examines the effect of hematocrit (Hct) and Hb on outcome. Three of these articles from the cardiac surgical experience note that low Hct is associated with adverse outcomes.24-26 Fang and colleagues24 examined patients during and after coronary artery bypass graft (CABG) surgery and noted that no increase in mortality was seen until the Hct reached about 15 percent during cardiopulmonary bypass. It was only at that level that mortality, overall, actually doubled. Hardy and colleagues26 examined the lowest Hct after heart surgery and did note that there was a correlation between lowest Hct after surgery and an increase in mortality. DeFoe and associates,25 in more than 7000 patients, examined lowest Hct on bypass. These authors found a graded increase in mortality, stroke, return to surgery, and other adverse outcomes with lowest Hct on bypass. What all three of these studies have in common is that they also did not look at transfusion. In each instance the conclusion was that low Hct might be bad and therefore should be avoided with transfusion. It is most likely, however, that transfusion was a surrogate of low Hct, and without independently investigating transfusion one cannot be sure whether it is low oxygen-carrying capacity attributed to low Hct or effects of transfusion itself that drove the adverse outcomes. A large data-based study of blindness for cardiac surgery was carried out at the Mayo Clinic.27 In that study 17 cases of blindness (0.06%) were found in 27,915 cases of cardiac surgery between 1976 and 1994. Postoperative Hct was found to have an association with blindness, and the article was written to put emphasis upon that association. When the article is read closely, however, one finds that the greatest association is between the presence of atherosclerotic vascular disease and blindness. Doing an angiogram within 48 hours of heart surgery (potentially showering the vasculature with atheroemboli) was highly associated with blindness as well. The use of a non-RBC transfusion product such as PLT transfusion, length of bypass, and transfusion (RBC) itself was associated with blindness. When bivariate analysis was carried out comparing low Hb and outcome, other factors were as or more important than low Hb alone. This study points out the problem with much of the research that exists today. There does seem to be a bias toward low Hb and Hct causing adverse events. If one examines the data that in databases from patients, it is clear that they can well. It is only at a Hb level of approximately 3 to 5 g per dL that mortality data dramatically This level of Hb to what is seen as the critical Hb level from animal studies of critical DO2 is the at which from to and the longer an critical DO2 the more likely is to In work from as well as other the Hb level to critical DO2 is approximately 3 to 3.5 g to It is therefore not that in clinical databases that mortality the level of critical animal work has shown that when aged animal blood is transfused the way as human stored banked that the Hb of critical DO2 is that shock is more likely with the use of transfused of the on more than patients undergoing surgery associated adverse events and for transfusion was in this data-based data were available on the lowest Hct on bypass but the care unit Hct did have a to outcome. Hct the of the patients Hct, volume in the transfusion trigger of the transfusion The Hct was related to outcome, and those patients with the lowest Hct had the lowest and high risk the death Transfusion was highly related to medical (i.e., transfusion not to patient That means that we as physicians do not transfuse based on patient needs but on of transfusion trigger that are These data did with when 28 potential risk and were into Hct out with an to adverse in that low Hct had the adverse outcomes. research from the literature data from the cardiac Those patients transfused the most have worse outcomes. In the surgery literature those patients with the most transfusions to have the risk for are the in the patients transfused the most as are infection These data are also when of risk are It however, seem that those patients with the most bleeding, and therefore most transfusion, might well be the most Just as in the however, transfusion out as an of adverse outcome the effects of of injury alone. It as transfusion trigger was of of injury and actually drove some of the adverse outcomes. the study examined was from the This study examined patients the of Of all patients, those a transfusion for should be to from oxygen-carrying capacity with transfusion. The had patients but after of of patients the data reported about of the It was to examine only patients over One should did transfusion in all the but with not have an effect on In that it was found that were by transfusion if patients had a low Hct on to the and also had an The that this actually outcome only about patients, approximately percent of the There was no analysis to for a of data no with or without in the and therefore it was that did not the When the and those not are however, they are far from Those patients were in the low Hct were from those in the normal The low Hct had many more more than the number of patients were as do not and had less than the interventions cardiac and surgical Most the use of blood products by transfusion was very low in this and it simply to that any today would not transfusion in a patient with an there were other such as a do not that might be associated with a of transfusion When the outcome was mortality, it is very hard to understand how with do not and can be as near Therefore, the conclusion of this widely article was that in patients with a 10 g per dL or 30 to percent Hct transfusion trigger is The data is far from and a The mortality for those patients transfused and for those not transfused. Those patients not transfused had approximately than those that one or more units of RBCs during their These data were after analysis events and with from from cardiac surgery for those transfused and not In a study of transfusion during 2000 it was found that transfusion had important effects on outcome. occurred at the rate in patients transfused to those not transfused. This finding was not only for the near but for up to the months of These authors did statistics and analysis each significant by statistics patient is a based on the number and of his or her from one are based on to other patients with and without the is thought to be the best way to from large The outcome, mortality over up and was still after with those patients transfused a death rate of those not transfused. These are but do the of and the effects of during early from interventions have demonstrated that early with and can the of early and as well as longer The that transfusion has been with both to cardiac surgery and also to effects on A large number of studies from databases have an association between transfusion and infection or has been associated with transfusion the that period the for transplantation was to make sure that patients during the time of transplantation several units of transfusion. This was that the would have less It was that a unit of blood would the rate and of In cardiac surgery a number of have shown a association between the number of units transfused and the risk of perioperative In one article it was estimated that the risk of increased 5 percent per unit on blood or PLTs That use of transfusion, however, could not be out as the or independently the most important for length of in the hospital. In surgery, a large examined the number of patients It was noted that most patients a but anemia. With the use of only 1 unit of blood transfused, however, there was a increase in perioperative infection and mortality. That ratio was by as much as when transfusions were but more adverse events were seen when transfusion outcome was related to death In surgery, the infectious complications have been and between and In a number of prospective randomized it has been shown that the risk of infectious complications from to if blood is utilized for transfusion to it is noted that blood stored can still have and therefore not all is avoided by the use of The literature on and is less with and cancer has shown some effect of transfusion at the time of and the is that of blood transfusion may increase the rate of events or either have not been as widely or have not to have as a The to be caused by a number of The blood cells which when transfused can and has been in the for up to 1 after transfusion and the effect for some period of time to the A wide of are in blood, and These the cells from and infection even these would to a This is a very literature with a large number of have that the to blood products will or the effect of by transfusion. That to be at this studies have noted effects with and large however, have shown little effect or no One study a of length of in the by days if blood was for transfusion. The authors noted this to be It however, be of little or no clinical Of one should look at the between those transfused and those not transfused. If these patients were not transfused, their length of was approximately the length of for

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.