Appropriate use of red cell transfusion

What is clinically appropriate blood transfusion? Is it appropriate to give a red cell transfusion to a man with a haemoglobin of 9 g/dl? Perhaps yes if he is dying with leukaemia and would be better able to attend and enjoy his daughter’s wedding with a Hb of 11 g/dl. Perhaps not if he recovering from an elective hip operation. But perhaps yes if he has bad coronary artery disease. Or perhaps not if the choice is between blood for him or for a child with severe malaria and a Hb of 4 g/dl. Clinical decisions are rarely simple and are nearly always a balance of risks, and also of cost and benefit. Transfusion decisions are especially so because of the real and perceived risks of transfusion, the limited supply of red cells and the very uncertain evidence base with which to guide our decision. Red cell transfusion has become an accepted part of medical practice. There is no doubt that red cell transfusion saves lives. Patients with multi trauma or post-partum haemorrhage survive rather than die, because of transfusion. We can perform complex surgery that would otherwise not be possible. In patients with bone marrow failure it improves the quality and length of life. Unfortunately any successful medical intervention has the propensity to become overused and transfusion is no exception. If it were a free good without risk this might not matter but it does of course carry risk and the measures to reduce this, especially in a risk averse western world increase the price. In a typical hospital, transfusion costs are between 1% and 2% of the annual hospital budget. Blood donation is given as an altruistic gift. It is also our responsibility to ensure that a gift is used well and thus as suppliers or prescribers of transfusion we are under obligation to maximize the benefit from this gift. Clinically appropriate transfusion encompasses not only the benefit or otherwise to the patient who may receive the transfusion, but also to the donor, to the society and to the other patients who may be deprived of other treatments if all our money is spent on unnecessary transfusion. The epidemiology of red cell transfusion has changed significantly over the last decade. In our region we have seen a 40% reduction in total units used for supporting surgery from 1999/2000 to 2009. At the same time medical use has remained stable. As a result, the proportion used for surgery has fallen from 41% to 29% [1, 2]. With increased longevity in the western world it is not surprising that the average age of the recipients is over 60 years and that patients over 80 have a rate of transfusion 10 times that of patients between 40 and 50 years. Because the elderly are less likely to be eligible to donate there is serious concern about maintenance of supplies of red cells as the population demographic changes [3]. We are therefore obliged as clinicians or as advisors to clinicians to make best use of this expensive and increasingly scarce resource. Every transfusion should be questioned and one of the key questions is what is the aim of the proposed transfusion? What do I hope to achieve by giving a red cell transfusion to this patient? The aims of red cell transfusion may be broadly divided up as follows: To prevent or correct severe acute anaemia that might otherwise cause organ damage or delayed recovery from an acute event. For e.g., before, during or after surgery/trauma/childbirth. To improve quality of life, adequate activity or physical growth in a patient with anaemia that is not otherwise correctable either due to bone marrow failure, or due to continuous bleeding or haemolysis that exceeds the bone marrow capacity for production of red cells. For e.g., Myelodysplasia or aplastic anaemia, thalassaemia or angiodysplasia of the intestine. To prepare a patient for surgery, relieve symptoms or to speed up recovery or discharge from hospital in a patient with chronic anaemia that is otherwise correctable. For e.g., severe iron deficiency. To reverse or prevent an outcome related to damage caused by the patient’s own red cells. For e.g., exchange transfusion for haemolytic disease of the newborn; transfusion or exchange transfusion for treatment of sickle cell crisis/complications. To increase haemoglobin to maximize response to radiotherapy. In each case, and sometimes more than one of these indications may be present for the same patient, the risks and benefits are very different and are also different for each product used. In acute anaemia/haemorrhage, the requirement is for a rapid increase in oxygen carriage capacity. Oxygen extraction varies between different tissues. Skin and kidneys extract only a small proportion (<10%) of the oxygen carried by the perfusing blood. This is because their blood supply is controlled for different reasons, the skin being primarily a temperature regulator and the kidneys, a plasma filter. The brain extracts about 30% but less in patients with reduced brain activity. As such these organs have a degree of reserve in that by small reduction in venous partial pressure of oxygen considerably more O2 can be extracted thanks to the shape of the Hb pO2 association curve. The heart lacks this cushion in that the myocardium extracts over 60% of the oxygen from the perfusing blood, more than any other organ other than vigorously exercising skeletal muscle. It has the added disadvantage of being required, in the anaemic patient, to pump more blood, i.e. to do more work, to compensate for the lack of oxygen carriage. Oxygen supply to the myocardium is modulated by changes in coronary blood flow in a roughly one to one ratio to the increase in cardiac work through vasodilatation of the coronary arteries. In other words there is little opportunity for increased oxygen extraction. To counter this the thinner blood flows more easily through the coronary circulation and is pumped more easily but if the demands on the heart exceed the oxygen supply, myocardial ischaemia and regional dysfunction occur. Our evidence for the value of transfusion comes from both animal and human studies. In healthy resting dogs no change in myocardial function is seen down to a Hb of around 5 g/dl. When a 75% stenosis is introduced in the anterior descending coronary artery, mimicking the situation in a human with coronary artery disease, myocardial dysfunction is seen at a Hb level of 7–8 g/dl [4]. Weiskopf et al. have shown in humans that at rest, acute normovolaemic dilution to a Hb of 5 g/dl is well tolerated with some changes in mental acuity [5]. It is possible, however, that these changes may be as much due to alterations in viscosity of blood and the effects this has on cerebral microvascular regulation as to lack of oxygen supply [6]. Carson et al. and others have shown that patients refusing blood for religious reasons mortality is not greatly increased until the Hb falls below 5 g/dl. However, if there is known prior coronary artery disease morbidity and mortality increase far earlier [7]. This evidence suggests that like dogs, we are tolerant of surprising degrees of anaemia if we have healthy hearts but less so if we do not. This conclusion should be tempered by several qualifications. It does not take account of other factors related to Hb/haematocrit. Acute anaemia usually occurs in a bleeding patient. Haemostasis has been shown to worsen with a haematocrit below 30% whether by effects on platelet flow or perhaps nitric oxide metabolism. It does not show that red cell transfusion can reverse any adverse effects of anaemia and it does not tell us much about the patient who is not resting. The situation in a patient with chronic anaemia is usually very different with the chief limitation caused by anaemia being in exercise tolerance. It is well known that professional cyclists have used erythropoietin and blood doping to increase their haemoglobin and thus performance, a proof if any of the benefit of a higher haemoglobin level when exercising to the extreme. Patients with bone marrow disorders will also testify to the beneficial effect of red cell transfusion though there is very little trial evidence to support the practice. There are few good randomised trials of red cell transfusion. Most reviews mention a handful and of these by far the largest and the most influential, both in the real world and in any meta-analysis is the Canadian intensive care study, the TRICC trial [8]. This was a well-constructed and conducted study looking at two thresholds for transfusion in patients in intensive care beds. The overall result that has altered transfusion worldwide was that patients did as well or better with a lower threshold for transfusion (7 g/dl compared to 9 g/dl). Given the influence this trial has had it deserves close scrutiny. Firstly it should be noted that the trial looked at seriously ill but haemodynamically stable patients rather than those with acute blood loss or those attempting to carry out everyday duties of daily living. Secondly that the restriction on transfusion only remained whilst the patient was on the intensive care unit. And thirdly that in a subsequent analysis a subgroup with previous cardiac history showed the opposite outcome, albeit non-significant, with slightly better survival in the higher transfusion threshold group [9]. The results may be interpreted as showing that severe anaemia is good for you if you have healthy coronary arteries, a not impossible conclusion, or that transfusion is bad for you but that the adverse effect is outweighed by the benefit of an increased Hb if you have diseased coronary arteries. The product used in this trial, plasma reduced non-leucodepleted red cells, was at the time standard in North America. Since then we have become more aware of possible adverse effects of moderate volumes of plasma, and also of detrimental effects of leucocytes both on the patient directly and on the quality of red cells during storage. A typical product today is of leuco-depleted red cells in optimal additive. The trial results may then rest heavily on the quality of the red cells used. Possibly in support of this, a similar trial was then conducted in children who it may be assumed have no coronary artery disease [10]. This study used leuco-depleted and plasma poor red cells and showed no difference between two similar transfusion thresholds, the outcome in this case being new organ dysfunction. However, a further subanalysis has shown a possible difference depending on the storage age of the blood transfused [11]. It becomes clear that we must not only look at the patient being transfused but also at the product being used. We should not assume all red cells are equal. Any clinical decision about transfusion should take not only the patient into account, but also the quality of the product available. Until we have the perfect cultured red cell for transfusion, trials of transfusion are more about the quality of the red cell than about the physiology of anaemia. So what is clinically appropriate transfusion? It depends on what you are trying to achieve. It depends on the product and the efficacy and risks it brings with it, on availability and cost, and on the best use of a scarce resource to maximize outcome. Some of the product factors to be considered are: 1. What is the infective risk of infection? In countries with highly developed health systems the infective risks of transfusion are very low. In other areas the risk of virus transmission may be relatively much higher. 2. What are the non-infectious risks? For instance, do the red cells contain large amounts of plasma that might cause TRALI? 3. What is the storage age of the product? We still do not know whether or how important storage age is for function of transfused cells or whether it only matters for some indications such as massive transfusion. Some of the patient factors to be considered are Is the anaemia acute or chronic? Is the patient symptomatic of his anaemia? Does the patient have cardiovascular or respiratory disease? Is the patient septic or actively bleeding? Is the anaemia correctable or not by means other than transfusion? What activity is expected of the patient? Is anaemia going to delay the patient’s recovery? Some of the societal factors to be considered are Is the cost of the transfusion warranted or would it be better spent on other aspects of the patient’s care? If the supply of red cells is limited would another patient benefit more? Is there any risk to the donor that outweighs the benefit to the patient? In routine clinical practice it may not be possible to assess all these in detail and more pragmatic guidance is necessary. In our hospital we use a simple mnemonic using Hb in g/dl and based on a traditional rhyme for children to impress on junior doctors that not all patients are the same. In the Netherlands the 6-7-8 rule based in mmol/l of Hb gives similar results. Guidelines such as these are valuable but should not be a substitute for careful and individual assessment of a patient’s needs and careful use of a valuable and sometimes scarce gift. The author declares that there are no potential conflicts of interest.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.