There is recent safety data in pregnancy from an RCT of oral iron vs oral iron plus IV iron polymaltose performed by the Launceston group. Infusion related incidents were minimal.

Pregnancy outcome was assessed as a secondary / post-hoc outcome in terms of birth weight etc over 200 women which isn't huge numbers but is a start and may not have been powered to detect a significant difference if number of adverse events small.

Interestingly they targeted mild-moderate anaemia and intervened early at around 26 weeks ie. this was more of a safety study and study on treetment of iron deficiency rather than iron deficiency anaemia. Patients with Hb <85 due to IDA were excluded and treated with IV iron but data on these was not reported - would be interesting to see what the increment was and how quickly this occurred.

It's likely that the effect of IV iron was diluted by the concomitant use of oral iron following the infusion and the early intervention at ~26 weeks which may have allowed for oral iron to exert it's full effect.

I tend to get referred these patients late, with persistent Hb <90 or less after failing oral iron or being intolerant of oral iron. Iron polymaltose appears to be a safe and effective option in this select group.

A prospective randomized, controlled trial of intravenous versus oral iron for moderate iron deficiency anaemia of pregnancy

I have noted some of the comments regarding the age of Blood and outcomes.

Firstly, one issue raised was the problem of supply of blood at remote locations & particularly where there is no pathology laboratory service on site. This issue has been dealt with in our local Hunter New England Area by: 

• The use of our e-Blood application at sites with no on-site pathology laboratories, which allows the computer release of Blood & Blood Products by trained non-pathology staff (eg. Nursing, Medical staff, etc )
• This has meant that blood is no longer tagged in case it is needed, instead it is easily & quickly computer released as required for actual transfusion from a stock of un-allocated blood.
• This has then meant that, at these sites, we only stock them with 'fresh' blood (ie. once blood reaches < 14 days to expiry it is rotated back to our main sites & replaced with the freshest stock available)
• So at these remote sites patients are actually transfused 'fresher' blood.
• NB. this practice has NOT resulted in increased expiry rates for blood, which for us is < 2%

Back to the issue of Blood Age & Outcome. I think most of the data seems to indicate that the Age of Blood is an issue for Trauma patients especially when given in the 1st 24 hrs. The mechanisms for this are not fully understood.

I also note that the FDA in the US has been asked to review this issue with the suggestion that the expiry of blood be reduced to ~ 30 days instead of the current 42 days post collection. One comment was that blood > 28 days is a 'volume expander' with no real benefit to the patient & in fact many negative consequences.

I feel that if you put in place appropriate policies & have systems that effectively manage you blood supplies, then in fact you can ensure that blood doesn't become 'old' when it is used. This also relates to a lot of other 'issues' relating to blood management, ie. with appropriate policies & systems many 'issues' aren't issues!

(NB. While writing this I had a quick look at 2008 for our Hunter Region & found that for the ~ 15,000 units Packed cells transfused the average age of blood transfused was 21 days with a median age of 19 days)

Several issues have been brought up in the last few days to which I would like to respond.

Regarding critical haemorrhage in the obstetric setting

This is a situation in which it is universally agreed that allogeneic blood transfusion saves lives the main ongoing issues warranting further discussion include:

• Reasons for the apparent/real increase in perinatal massive transfusion and its potential complications
  • Kuklina EV, Meikle SF, Jamieson DJ, et al. Severe obstetric morbidity in the United States: 1998-2005. Obstet Gynecol. Feb 2009;113(2 Pt 1):293-299.
• Logistics (near patient availability of blood products, local technical & haematological expertise)
• Early control of bleeding and minimizing allogeneic blood transfusion
• Diagnosis and management of coagulopathy
• Role of haemostatic agents such as rVIIa
• Massive life saving blood transfusion as an independent risk factor impacting on clinical outcomes (TRALI, ARDS, MOF, sepsis etc), role of age and volume of transfused blood
• The management of anaemia following perinatal haemorrhage when the patient has been stabilised. In this setting the tolerance of anaemia and the role of transfusion in improving short and long-term outcomes  requires further debate and better evidence. Several women in such circumstance died of transfusion transmitted HIV in the 1980’s in whom tolerance of anaemia and appropriate haematological management would have them alive today to care for and enjoy seeing their children grow up.

In response to some of Barry Walter’s comments I make the following comments:

I agree with Barry’s comments amount the problems of iron depletion and deficiency in pregnancy and its current suboptimal management. However, I have to take issue with several of his opinions and statements on other important aspects of transfusion medicine.

Allogeneic blood transfusion and blood component therapy undoubtedly has an established role in modern medical practice. However, regularly questioning current transfusion medicine practices is good medicine and good science. Many of Barry’s statements are not supported by an extensive literature questioning the role of red blood cell (RBC) transfusions in improving clinical outcomes in many clinical settings, especially in perioperati ve haemodynamically stable patients. Awareness of knowledge of this literature amongst the medical community has been slow and most clinicians regard RBC transfusion as safer than they have ever been.  Most doctors are not aware of the real risks and costs of blood transfusion as it is only now that we are getting a better understating of these issues. Nobody currently knows the total cost of transfusion medicine to the community and as already mentioned the focus on transfusion outcomes has expanded the spectrum of transfusion risks and the paucity of evidence for benefit in many clinical settings. Efficacy of most blood component replacement therapy with the labile blood components (RBC, FFP and Platelets) has been assumed and the risks accepted and minimised as far as possible. In recent years a paradigm is evolving that blood transfusion is safer than it has ever been and benefit in improving clinical outcomes is accepted. However, the ever expanding literature I have mentioned is questioning the efficacy of red blood cell (RBC) transfusion and implicating allogeneic RBC transfusion as an independent risk factor for adverse clinical outcomes in many clinical circumstances.

In my opinion it is no longer acceptable or ethical to continue adopting a laissez faire approach to the assumed benefits and accepted risks of allogeneic blood transfusion. RBC Transfusion can no longer be regarded as the default clinical decision when there is clinical uncertainty. Opinion-based medicine should no longer determine the clinical practice of transfusion medicine. Managing a patient’s own blood appropriately is now the focus of attention. If allogeneic blood transfusion is regarded as an appropriate intervention evidence is needed that the patient’s ultimate clinical outcome will be improved. Surrogate endpoints are necessary for many clinical interventions, but we need to be sure that these immediately measurable surrogate endpoints correlate with better long-term clinical outcomes. Barry mentions transfusion for treating hyperdynamic heart failure in a young postpartum women.  A hyperdynamic circulation is nature’s adaptive response to anaemia and is not heart failure. Sure, if the anaemic patient has co-morbidities such a reduced cardiac or respiratory reserve short-term correction of the anaemia becomes a more important priority. To use an analogy of blood sugar levels, potassium levels and leukocyte counts is not appropriate. Humans have a large total red cell mass to allow for reserve and adaptation to stress. A normal person can lose up to 80% of their red cell mass and survive if normovolaemia is maintained. Using the analogy suggested , I am concerned that Barry would justify transplanting a kidney into a patient who only had one well functioning kidney.

It is unreasonable and judgmental to call the hundreds of national and international researchers into the efficacy and safety of allogeneic blood transfusion as being zealots, unscientific and not concerned about patient welfare. Most of these researchers are busy clinicians in a wide range of clinical specialties, who are increasingly concerned that RBC transfusions are exposing many patients to risk with minimal or no evidence for benefit. They may be zealous skeptics that are challenging the deniers to come up with better evidence and data supporting the assumed benefits and safety of labile blood products. These blood components are tissue transfusions/transplants and have the greatest variety of potential adverse clinical outcomes of any medical intervention. It should also be emphasised that there are ethical considerations for both patients and blood donors that transfusion is used appropriately and for the greatest number of patients in need.

There are compelling reasons why the precautionary principle should be used when considering the transfusion of labile blood components, primum non nocere. Support for adopting a precautionary approach to the transfusion of labile blood components. Benchmarking studies identify a wide variation in RBC transfusion practices between, countries, hospitals and individual clinicians. There are more appropriate and alternative therapies and interventions available to avoid or minimize the use of allogeneic blood components. As there have been no comprehensive activity-based costings of RBC transfusion, no valid cost-effectiveness comparison studies have been possible.   

Evidence-based transfusion medicine should view a patient’s own blood as a valuable and unique natural resource that should be conserved and managed appropriately. Altruistically donated allogeneic blood transfusion is also a valuable, unique and costly resource that should only be used as therapy when there is evidence for potential benefit, there are no alternatives, a quality product is available and the risks are appropriately considered and balanced against the benefits and with appropriate explaination for the patient.

BY WAY OF CLARIFICATION - MY COMMENT BELOW.

The expert from OS ran a Unit where patients were referred with very low Hb for whom "blood was not an option" hence they had a complex package, including long term ventilation with 100% oxygen, etc, and did not use blood at all. Fair enough - no argument - but that does not apply to 99.999% of patients and has no role in our decisions with our available options.

I am unaware of any negligent use of blood transfusion in the areas I practice, as an Obstetric Physician ( I see pregnant women with medical problems) or in General Medicine, my other area. I started practice in 1975 and attend both public and private hospitals. Doctors are acutely aware of risks with transfusion and costs thereof, and the scarcity of blood and I remain sceptical that there is any overuse of blood, at least in my state.

I attended a travelling road show of "experts" in the area and was horrified when a question was asked of a junior Doctor whether she would transfuse a patient with Hb of 40. The interrogator went on to detail his experience with ICU type treatments of people with Hb of less than 30, for weeks on end, in another country, and virtually brow beat the audience. I fervently defended the junior doctor and reminded the speaker of the requirements of education.

The blood transfusion debate has been hijacked by interest groups, and only a small component of them represents those interested in science or indeed patient welfare. There is too much prestidigitation and playing with mirrors as all sorts of irrelevant pseudoscience is introduced in the debate. In diabetes we seek normal blood sugar levels, if potassium is low we correct it, if there is leucopenia we worry, why should we not seek something closer to normal in a severely anaemic patient. We don't aim for 120, but we should try to keep Hb above 80, in postpartum women and particularly in old or inform people. Doctors who see people in the weeks and months after they have left hospital have a greater appreciation of the situation than those who see patients for a few hours or days only in hospital.

Dr Isbister in this column mentions "religious" commitment to transfusing at certain levels.  I suspect a "religious" commitment to avoiding transfusion at any level, and to using very expensive products to avoid blood.

I remain unconvinced by their specious sophistry and will continue to teach students and young Doctors that human physiology seeks to maintain Hb at the levels with which we are all familiar for the same reasons that it maintains all other components of the "milieu interieur".

Sure, we must be very careful with blood and observe indications and contraindications. Nevertheless, please let us stop the political campaigning and zealotry.

Obstetricians, Anaesthetists and Gastroenterologists have been invited to join the transfusion question debate dated September 09.

Dr Boyle, thank you for posting. We haven't directed this current debate at obstretricians so I'm not sure we will get answer- however, if there are any obstretricians on-line or who stumble upon this debate, please feel free to jump in with your comments.

Carolyn, Blood Watch, Clinical Excellence Commission.

Can we hear some coments form Obstetrician please

Interesting debate about the value of quoting a particular reference (Basran et al Anesth Analg 2006;103(1):15-20).  Steve Shafer's letter/editorial openly discusses the problems with the paper and defends leaving the results in the public domain.

As has already been mentioned, a larger study published this year also found a strong association between age of stored red cells and poor outcomes.  Although both these studies are retrospective, I find it hard to see how they could have been biased.  What hidden systematic factors would cause sicker patients to get older stored blood?

What to do about this? Imagine the supply problems if we stopped using cells that have been stored for more than 15 days??  The main thing is to be as sure as possible that every unit transfused is clinically justifiable.

In 2008 I have beocme involved in the Blood Watch Program initiated by the Clinical Excellence Commission (CEC).  The aim is to reduce unnecessary blood transfusions in NSW.  As an Orthopaedic Surgeon I am in a group of Surgeons who account for more elective blood transfusions than most others. 

On looking at the facts about blood transfusion I have become aware that allogenic blood produces significant immune depression which can lead to increased infection after surgery or any large transfusion.  Also the age of the stored blood can have an impact on morbidity and mortality.  We all know that human error in cross matching and delivery can occur and blood is costly and supplies are limited. 

We now realise that patients do worse the more we transfuse.  There is enough outcome data to support transfusing less. 

As a Surgeon I was brought up on the idea of "liberal transfusion practices".  It is now realised that this approach results in about 30% of blood transfusions being inappropriate.  The idea of a haemoglobin (Hb) below 10 grams triggering a blood transfusion of two units is now discredited.  Instead we should be undertaking "restrictive transfusion practices".  This involves doing a risks benefit analysis realising that an Hb below 10 grams does not need transfusion.  Instead patients with an Hb between 7 and 10 grams may need transfusion depending on laboratory and clinical indications while those with an Hb of less than 7 grams are likely to need transfusion.  If transfusion is deemed necessary then consider giving only one unit of blood followed by a Hb check and just as importantly a clinical assessment before more transfusion.  Patients with co-morbidity such as cardiac disease are more likely to need transfusion. 

As an Orthopaedic Surgeon doing a lot of elective joint replacements I have endeavoured to set in place a restrictive transfusion practice.  This does require close assessment of patients after their surgery and being aware of the co-morbidities such as cardiac disease.  It also means putting patients on iron tablets and having them followed up by their GPs more often.  On a practical note often the best indication for transfuions is the decrease in hypostatic blood pressure when the patient first gets out of bed and "feels faint".

In looking at restrictive transfusion practices I am impressed by Fairfield Hospital and Sydney South West AHS where they are achieving one third reduction in blood transfusions following joint replacements by a co-ordinated effort from medical and nursing staff.  This is something Hunter New England could also consider doing by including a similar "blood transfusion stamp".

Hi, I am responding to the second posting below this, which raises some legitimate concerns about the robustness of some of the cited journal. The author is also critical of the CEC for posting them. I make the following comment: I agree that the journal articles should be robustly assessed by clinicians, however, I don't see it as the role of the CEC to vet or censor published refereed journal articles. The articles are there for all to see and review and should be used to illuminate the debate rather than as a support or justification for current practice. I would welcome other comments on this. The important message is, surely, that if the transfusion risk outweighs the benefit to the patient, it should not be given.

Bill can you send me the reference re Bilat knees?  We have a very significant complication rate in public hospital even with "young" and "fit" patients.  Most surgeons have given up doing them but there are a few die hards who tell me they don't see complications in private and there is no evidence to suggest bilats do worse.  Thanks

For clinical practice change, high quality evidence needs to be the driver.
Unfortunately, one quoted reference certainly does not meet these criteria, and in fact, may distract the point that you are endeavoring to make.

I have been asked what would I do on the warad if I did not have all the fancy "do-dads" to assess oxygen supply demand and delivery.  Well, I believe you can look at heart rate, blood pressure, pulse oximetry and get an idea if the patient is stable. If the patient is stable then do nothing- he/she does not need a transfusion. Just because he/she hit your arbitrary trigger does not mean giving a transfusion will prevent some badness.

However, if the patient has some other major medical problem, and of course 80% of our patients do have something like hypertension, diabetes, coronary disease etc, etc, then you do have to make a judgement call. It simply comes down to your clinical accumen. What I don't do  is labor under the misconception that I know a unit of blood is either good for the patient or "needed". Hgb/Hct is a surrogate for oxygen carrying capacity and a very poor one for delivery. Once again, the lower the Hgb goes the more likely you are to find some manifestation of ischemia and potential organ dysfunction. What we don't have is solid data to say that by transfusing in advance of such events we prevent aht in a large population or in individuals. Clearly we must transfuse, and I do, but what I am trying to force is physicians to think.  Every study I have seen in which patients have been randomized to lower transfuion versus higher, or like the Spence article at AABB wherein he looked at over 100,000 patients and saw those transfused up to 10gm/gl before surgery did worse. The Habib article with renal failure and Surgenor's most recent article out of the Northern New England Data base for Cardiac Surgery wherein they found that low Hgb was a risk but transfusion made the risk worse all make me question.  They all make me question the standard teaching and a belief system.

I don't have all the answers! I will say that people are always asking me hypotheticals ending with-"well wouldn't you transfuse in this situation". And yes, most often I would, because all of us can think aobut situations wherein we would transfuse. But, today I think it is healthy to say, we simply need a whole lot more data and better monitors as well as perhaps really pressurizing the blood banking industry and pharmaceuticals to create a more effective product- normal funcitoning red cells!

Thanks for the great comments Bruce. My question for you and the panel is as follows: you talk alot about some of the more fancy measures of perfusion/oxygen delivery, techniques that are not readily accesible outside of a critical care environment. If you took away all these fancy techniques and were just left with your patient,  your clinical skills, simple lab tests (Hb/Hct) and other measures available on a normal post op ward, how would you decide whether a patient with a borderline Hb/Hct needed a transfusion (in a relatively quick and simple way)? 

I have read with great interest everyone's inputs. I do agree that there is considerable risk that we will be over zealous with our desire to not transfuse. So often I get misquoted as saying: "all blood transfusions are bad and one should never transfuse". Clearly that is not the case- I have saved many, many lives with transfusion! Unfortunately we are in such a huge need for more good science. We should transfuse for physiologic reasons!

This is perhaps my biggest take home point. If you transfuse based upon physiology you will usually do the right thing. If you transfuse because you hit a trigger you will almost always be wrong. Why?

Tirggers are a Hgb or Hct level which is meant to be a surrogate marker for oxygen carrying capacity. But as a surrogate they are lousy!  In your patient at 60 versus 100gm/L of HGB  we certainly know nothing of cardiac output, microvascular perfusion, volume status or oxygen demand. So those two numbers as triggers are radically different but if some one writes a guideline and says for example: 60gm/L is the tirgger, then they ar picking an arbitrary number.  We need to be physicians and medical care professionals.

Oxygen delivery is the key to survival on this planet! Every physician should be intimately familiar with the critical oxygen delivery curves and the effects of anemia upon oxygen extraction, lactate production etc.  Hint- 100gm/L is not the "best oxygen delivery".  That is old data from work done in the 1960's looking at calculated data in larger arteries (aorta, to third generation arteries). The microcirculation very carefully auto regulates itself in terms of Hct and oxygen delivery. What changes in the microcirc is red cell speed, oxygen extraction and capillary density (more open up if the tissues are in need of oxygen).

So here is another tough thing. As soon as you transfuse most of our key measurements of oxygen delivery and usage (mixed venous oxygen saturation) goes artifially up. The Hgb you just gave from the transfusion does not unload its oxygen to tissues so if you use that to guide your therpay as soon as you give the blood you will think you are doing good. Indeed data shows from Amy Tsai's lab and Marcos Intaglietta that the tissue oxygen levels fall another 400% when you transfuse. But we cannot let people bleed to death.

So, yes you have to decide when to transfuse. The Hgb and Hct levels are bad surrogates and many of our critical care medicine measurements are artifactually affected as soon as you transfuse. But, be a physician and a physiologist. Don't make your transfusion decisions just based upon a trigger guideline. And, yes there is a risk of under transfusing. What is that risk (under transfusing)compared to the risk of say -TRALI which is between 1-73 patients transfused to 1-5000 patients trransfused?  We simply have no idea!! We need so much basic research in transfusion to debunk the myths. The lay media publishes a lot of hype saying we have the safest blood ever. True, but that is for HIV and hepatitis, not TRALI, not een all the other viruses we do not test for. Why is it that those patients transfused have 3-5 times more perioeprative infections, twice the length of hospital stay and twice the mortality out five years after discharge?  It is because those associations have nothing to do with HIV and Hep C but they are real reproducible and meaningful. We simply do not know when to transfuse.

I am scared of under transfusing, but I do not over transfuse to compensate for that fear!!  And, I might add, I live in a country filled with as many lawyers as you have on your Great Barrier Reef!

Thanks Bill and Roger for you comments.

The issues you bring up are a regular subject of debate and the literature and data is rather conflicting. As I have mentioned elsewhere in the debate the problem you refer to is typically in the elderly fragile female. There is no clear answer, but my personal view is blood volume (red cell mass) and macrocirculation is the predominant issue and not the presumption that the transfusion is delivering more oxygen to the tissues. Having said that, if a patient becomes for mobile, aerobic performance is more related to red cell mass rather than actual haemoglobin level. In the ideal world comprehensive patient blood management will minimise or prevent the need for transfusion. Maximising Hb preoperatively, minimising blood loss intraoperatively and iron +/- epo postoperatively has been shown to be effective. Transfusion has tended to be the default decision as the alternatives are not always available or reimbursed. If a transfusion is considered to be clinically indicated with a haemoblobin >80gm/L the reasons should be clearly documented in the patient's notes and authorised by the consultant.

Conlon, N. P., E. P. Bale, et al. (2008). "Postoperative anemia and quality of life after primary hip arthroplasty in patients over 65 years old." Anesth Analg 106(4): 1056-61, table of contents.
    BACKGROUND: It is uncertain whether anemia in elderly patients after primary hip arthroplasty has an effect on their quality of life. METHODS: We conducted a prospective observational study over 3 mo to investigate the association between discharge hemoglobin levels and subjective experience of quality of life at 2 mo postoperatively in patients aged over 65 yr who were scheduled for primary hip arthroplasty. Quality of life was measured preoperatively and at 2 mo postoperatively using the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36) and the Functional Assessment of Cancer Therapy Anemia (FACT-Anemia) subscale. Pearson correlation coefficients between change in SF-36 and FACT-Anemia subscale scores (from preoperatively to 2 mo postoperatively) and hemoglobin on Day 8 were calculated. RESULTS: Eighty-seven patients were included in the study. Data were available at 2 mo postoperatively from 79 of these patients. The correlation between Day 8 postoperative hemoglobin and change in SF-36 was 0.49 (P < 0.0005) and change in FACT-Anemia subscale score was 0.46 (P = < 0.0005). The correlation was not significantly changed after adjusting for advancing age, presence of significant cardiovascular disease, or whether the patient was transfused. CONCLUSIONS: We found a positive correlation between hemoglobin levels on discharge and change in quality of life scores from preoperatively to 2 mo postoperatively in patients over 65 yr old after primary hip arthroplasty.
 

Guerin, S., C. Collins, et al. (2007). "Blood transfusion requirement prediction in patients undergoing primary total hip and knee arthroplasty." Transfus Med 17(1): 37-43.
    The aim of this study was to identify the clinical factors associated with the need for peri-operative blood transfusion in non-anaemic patients undergoing hip or knee arthroplasty. We prospectively evaluated 162 consecutive patients who underwent total hip or knee arthroplasty. Analysis was performed to establish the relationship between all independent variables and the need for postoperative transfusion. Univariate analysis revealed a significant relationship between the need for postoperative blood transfusion and the pre-operative haemoglobin levels (P= 0.001), weight (P= 0.019) and age (P= 0.018). Multivariate analysis identified a significant relationship only between the need for transfusion and the pre-operative haemoglobin level (P= 0.0001). The pre-operative haemoglobin level of the patient was the only variable to independently predict the need for blood transfusion after primary hip or knee arthroplasty.

 

Wong, C. J., M. K. Vandervoort, et al. (2007). "A cluster-randomized controlled trial of a blood conservation algorithm in patients undergoing total hip joint arthroplasty." Transfusion 47(5): 832-41.
    BACKGROUND: The optimum strategy for reducing allogeneic blood transfusion in patients undergoing total hip joint arthroplasty (THJA) is unknown. STUDY DESIGN AND METHODS: The effectiveness of a comprehensive blood conservation algorithm (BCA) was evaluated by means of a cluster randomization trial. Thirty hospitals performing primary THJA were randomly assigned to implement the algorithm or to continue with usual care (UC). Subsequently, the institutional rate of allogeneic transfusion was determined for 60 consecutive patients who underwent surgery at each site. The BCA consisted of patient and provider education, hemoglobin-based recommendations for specific blood conservation strategies (recombinant human erythropoietin [rHuEPO] or autologous blood donation [ABD]) and transfusion guidelines. The main outcome measure was the institutional allogeneic transfusion rate. RESULTS: One hospital withdrew consent after randomization, resulting in 14 hospitals assigned to BCA and 15 to UC. In the BCA arm, the institutional rates of rHuEPO use and ABD participation were 20.1 and 27.1 percent compared to 0.6 and 25.8 percent, respectively, in the UC arm. The allogeneic transfusion rate was substantially reduced in hospitals assigned to the BCA group (p = 0.02; absolute risk reduction, 9.6% [26.1% UC vs. 16.5% BCA]). Multivariate analysis of patient-level data showed that assignment to the UC arm was an independent risk factor for allogeneic transfusion (p = 0.037; odds ratio, 1.8; 95% confidence interval, 1.0-3.1) when adjusted for other prognostic factors. No differences were observed in the use of autologous blood. CONCLUSION: A comprehensive approach to blood conservation was superior to UC for reducing allogeneic transfusion in patients undergoing THJA.

Hi James, I concur with Roger Brighton. Elderly patients recover poorly after THR/TKR with Hb<10. Co-morbidities unmasked are also a worry.Intuitively I feel this will be the same as the debate over safety of bilateral simultaneous vs staged THR/TKRs. Many papers claiming no statistical difference until eventually a study with adequate power showed there was."Evidence" concerns me as there are so many different sorts. "Science" with no beta error would be better, Cheers, Bill

Orthopaedic viewpoint:

Age, debility and demand are major considerations in my practice:

• I am commonly dealing with 70s, often 80s.
• Many have chronic cardiac and/or respiratory dysfunction concurrent with hip and knee arthritis.
• The operation causes some metabolic stress, but we also get them up within 24h (to avoid skin, lung, DVT problems)

Whilst I try to comply with guidelines not to transfuse  Hb< 8, I sometimes vary this because of the above, especially if tachycardic or consisitently hypotensive (particularly in octagenarians when the cutoff is usually 10) - .  They do seem to handle rehab. better above 8.

RogerB

Dr Speiss I read your comments with interest. Isn't there a danger that the pendulum swings too far the other way and we start under transfusing patients and then see complications from undertransfusing ie increased patient mortality. If there is not a trigger and we can't accurately measure oxygenation in cells, then what do we do?? What is your advice as we need a way forward on this now.

I had no idea the current evidence for the transfusion practice we are measuring in NSW public hospitals is so weak. I strongly urge all  clinicians to review their practice we have provided every AHS with a relative use data base which has linked the Patient information data with Blood Bank and Pathology data. This data base which has now been produced for the second year allows you to compare hospitals and DRGs to look at how you compare to other clinicians treating similar groups of patients. For some DRGs clinicians are prescribing blood up to 40% above the state average. In hosptials where the blood bank is actively involved in providing advice to clinicians with expert haematolgists available to consult with, they have much lower usage rates and generally only give one unit and then retest. If you are not sure how to get further advice please contact the chair of your AHS Transfusion Committee or contact the Director of Clinical Governance in your AHS who can get you this information.   

For the haemodynamically-stable post-operative patient, the question is not what the numbers say, but what the patient says.  We are all aware of scales of impairment, frailty etc, but what about assessing vigour in these patients.  Some with a Hb of 85g/L will have more vigour than others with a Hb of 95g/L.

The physiological cost of blood transfusion needs to be factored in to the analysis.  Given the rigidity of the red cells, their depleted statre of NO and the fact that they will need 24 hours to re-gain their full oxygen-carrying capacity, a blood transfusion is not necessarily the best way to restore vigour.  If their iron stores are sufficient (ferritin greater than 100mcg/L), I favour a shot of EPO.

Regarding the paper on effects of Hb at discharge on quality of life after hip replacement (Conlon 2008), I don't find the data all that convincing.  Although there was a statistically significant correlation, it was an observational study and the scatter was large.  Looking at the figure in the paper (and concentrating on the quality of life scores), above a discharge Hb of about 90 there appears to be no correlation between Hb and outcome. The number of patients below 90 are not large but they were indeed much more likely to have had a decline in quality of life scores.

It seems to me this question is crying out for a randomised multi-centre prospective trial. Many practitioners believe transfusion aids recovery - the onus is probably on them to prove the advantage with proper data.

Taking up your point regarding the effect of volume replacement when giving blood in the post operative setting James.  I have also had this thought and wonder as to the placebo effect of having a blood transfusion in the post-op setting?  

At least 2 factors known to enhance the placebo effect are present  - IV administration and colour (red is good).  On top of this patients are usually told how much better they will feel after having the blood and they KNOW how good blood is for them.  We re-enforce the mysterious power of blood by fussing over the patients, measuring P/BP/Temp etc every 30mins, checking ID details etc.  I've often wondered what the outcome would be if blood vs red colloid were given in a double blinded trial to post op patients.

Certainly patients do often claim to feel better after a blood transfusion and many doctors use this as a reason to give them one.  They might not feel quite so good if they were aware of some of the evidence regarding the potential deleterious effects of blood.

Seriously though there are studies (Postoperative Anemia and Quality of Life After Primary Hip Arthroplasty in Patients Over 65 Years Old.  Anesth Analg 2008;106:1056 –61.                                      Influence of Hemoglobin Levels on Inpatient Rehabilitation Outcomes after Total Knee Arthroplasty.  The Journal of Arthroplasty Vol. 21 No. 5 2006) suggesting that a higher post operative haemoglobin can influence rehabilitation

Could Bruce and James comment on this concept?

Thanks Bruce,

I think your historical perspective is most valuable and illustrates how clinical practices become "standard of care" for dunious reasons. Times and opinions are changing and we can only hope that a sound understanding of physiology and pathophysiology come back into clinical decision making.

Transfusion triggers are very interesting if one looks at both history and physiology. We know form a wide range of animal basic science that if one controls volume very carefully critical oxygen delivery occurs at around3-3.5gm/dl (30-35 gm/L) Hgb.  that has been seen in rats, rabbits, swine, dogs, and humans. Critical oxygen delivery point is of key importance. It is that point at which cells switch from aerobic to anaerobic glycolysis. As a physiologist and researcher, to me it is the true definition of shock-the point where cells switch their metabolism and cellular demand for oxygen outsrips supply.

Clearly we do not wish to make a physiologic prep out of every patient and the truth is that rarely are such anemic patients trlu euvolemic! Most of the time in medicine we really do not know where we are in volume status. So, in hypovolemic patients tolerance of anemia is far less. Also we have little data with regards to flow limiting lesions such as atherosclerosis in a large artery or small aretial disease and endothelial disease such as occurs in hypertension and diabetes. So the tendency and assumption is that critical oxygen delivery comes at a higher Hgb in those situations. Truth be told, we have no idea! It could be that it is exactly the same.

That being said a transfusion trigger fo 3-3-5gm/dl was the original one chosen at the turn of the twentieth century because that is where physicians saw cardiac failure. Then around thw World Wars it crept up to 5-7gm/dl because people wished to use blood (now more readily available from refrigerated storage) not as a rescue therapy but as a prophylaxis. It was out of those World Wars and a great deal of public relations hype saying that donation was the patriotic thing to do- save a life- etc that we have essentially preached ourselves into believing that transfusion is good. The 10gm/dl-100g/L religious transfusion trigger came from the Mayo clinic where in John Lundy, MD a very famous anesthesiologist kept recorda baout how often surgeons transfsused in the 1930-40s. His conclusion was that since surgeons seemed to transfuse around 10gm/dl-100gm/L (and they were surgeons/GODS) they must be right. Some sort of circular reasoning there. It was only in 1986 with the AIDS crisis etc that we started to look critically at the issues of oxygen delivery and what is the "best trigger".

But here is the problem, if you transfuse you deliver blood with red cell storage lesions. Taht means that it carries and relaeases O2 in a dysfunctional way. It clogs up the microcirculation and contributes to all types of inflammatory reactions. Low and behold, if you give a unit of red cells fromt he bank the critical oxygen level (Hgb) actually rises. That means that shock comes at a higher hemoglobin. It means that by transfusiong you have made shock actually more likely (by the true physiologic definition). but you have to transfuse somewhere and you cannot let people just bleed to 30-35gm/L and die.  Well the idea ot create a "trigger" that fits a large part of the population is itlsef completely fraught with assumptions and probably no matter what number you pick it is arbitrary. It therefore is worng for far more people than it, as a number is right, for any one individual. What is truly needed is improved monitoring so that we can reall yknow the status of oxygen demand in critical tissues and then target therpay just so for those events. We are not there yet but people are at least beginning to ask the right questions.

Bruce Spiess, MD

Excuse the formating problems in my contribution above. I pasted it as Word text and the formating turned into garbage!

The science is interesting.  The evidence suggests that a Haemoglobin of 50g/L. is the minimum and harm occurs below this number. Above this level it appears that in euvolaemic patients, there is little  physiological adaption until haemoglobin falls to 70g/L.  Keep this in the back of your mind when seeing a patient and consider other reasons such as drugs and iron deficiency if the patient is symptomatic above these levels.

In stable adult patients post op what would constitute an 'appropriate' blood transfusion is it just clinical signs or should Hb level be a consideration as well or is the Hb level irrelevant?