Losing sight of the precautionary principle

Not being a transfusion medicine specialist, I was asked to write this editorial to bring an “outsider's” perspective to the issue being addressed by Dunbar and Yazer1 in this issue of TRANSFUSION, which, in broad terms, is the continued challenge of ensuring that the blood inventory is sufficient to meet increasing needs. The specific question that Dunbar and Yazer addressed was whether it is safe to use group A plasma to supplement group AB plasma for the initial resuscitation of bleeding trauma patients whose ABO status is unknown. This is a relevant question: owing to the widespread adoption of high plasma-to-RBC transfusion ratios (to better approximate the composition of whole blood) in massive transfusion protocols (MTPs) for trauma, plasma is increasingly being used as part of the initial resuscitation of trauma patients. Ideally, all plasma transfusions should be ABO compatible to avoid hemolytic transfusion reactions. If the ABO group is unknown, as is often the case during the initial phase of trauma resuscitation, however, the next best option is to transfuse AB plasma. AB plasma is considered to be “universal” because it lacks anti-A and anti-B hemagglutinins and will therefore not lead to hemolytic transfusion reactions in patients with any ABO blood group. As a result, it is standard practice for many trauma centers to maintain a ready supply of thawed AB plasma. Since group AB donors comprise only a small fraction (<5%) of the North American donor pool, however, this practice places a substantial strain on the blood inventory. As a potential solution, Isaak and colleagues in 20112 proposed that group A plasma be used to supplement group AB plasma for the initial resuscitation of trauma patients whose ABO status is unknown. To mitigate the risk of hemolytic reactions in group B or AB patients, they stipulated that the amount of transfusion should be limited to 1 L or less and that only units that have “very low” anti-B agglutinin titers should be used. They classified units as very low titer if after a 1-in-16 dilution with saline they did not cause agglutination (at room temperature or at 37°C) when mixed with 4% group B RBCs. They estimated that only 15% of group A plasma donors would meet this criterion. The rationale for their proposal was that patients who are at risk for having hemolytic transfusion reactions to group A plasma (i.e., group B or AB patients) comprise only 15% of the North American population and that there “appears to be” a correlation between the risk of hemolytic transfusion reactions and agglutinin titers. Appreciating the inherent risks of this approach, they concluded that “this new concept of very low titer group A plasma as universal plasma in MTP has not been studied in trauma patients requiring MTP and will require thoughtful study design to test the hypothesis.” Currently, data supporting the safety of this practice are scant, consisting of a few observational studies (cited by Dunbar and Yazer plus another published this year)3 that included approximately 200 group B or AB trauma patients who received group A plasma. Although it is encouraging that none of these studies identified any major safety concerns with this practice, they all had important limitations and the level of evidence is weak. The study by Dunbar and Yazer makes an important contribution by providing data on an additional 354 trauma patients who received incompatible group A plasma. After careful comparisons with 809 patients who received compatible plasma, they found that the use of incompatible group A plasma was not associated with increased mortality or length of hospital stay. While this finding further supports the safety of this practice, given the important limitations of the study, it is by no means conclusive. As noted by the authors, the study's limitations include the retrospective design of the study, the potential for selection bias owing to missing data, inclusion of centers with different transfusion practices, and most importantly, inability to assess the myriad of complications that could result from acute hemolysis, such as acute kidney injury and immune modulation.4 To date, there have been no adequately powered, prospective trials to determine with any degree of confidence if the use of incompatible group A plasma in group B and AB trauma patients is safe. Nevertheless, according to a recent survey,5 most US trauma centers are now maintaining thawed group A plasma to supplement AB plasma for use as part of the initial resuscitation of trauma patients whose ABO status is unknown. Moreover, many of them are doing so without implementing the risk-mitigating strategies that were recommended by Isaak and coworkers (i.e., limiting transfusion volumes or measuring titers).2 As a public health issue, then, it seems that the transfusion community has deemed it appropriate to resolve a blood inventory issue by allowing incompatible plasma to be used for trauma resuscitation, accepting that some patients may develop hemolytic transfusion reactions without conclusive evidence that these patients will not be harmed … which takes us to the precautionary principle. The precautionary principle requires that precautionary measures be taken if an activity raises threats of harm to human health, even if there is uncertainty about the level of risk.6 Does the use of incompatible plasma in the context of uncertain risk abide by this principle? And if not, should we put aside the precautionary principle for this practice, or should we instead put aside this practice? The precautionary principle has been the prevailing paradigm for establishing safety policies in transfusion medicine for more than 20 years.7 Its use, however, has recently come under scrutiny because of instances where the negative consequences of precautionary measures—notably reduced donor pools and increased costs—have outweighed their safety benefits.8 To help guide the appropriate use of the precautionary principle, Wilson and Atkinson8 have proposed the following framework: First, determine if there is uncertainty about the magnitude of the exposure or the risk that it poses, in which case precautionary measures are indicated. Next, determine the type of precautionary measure that should be instituted by balancing the theoretical risks that it is meant to mitigate against all its harms and economic costs. If the balance is clearly favorable, such as when the harms and costs are minimal, then strong measures are indicated. Otherwise, institution of weak or intermediate measures may be more appropriate. Examining the use of group A plasma for the initial resuscitation of trauma patients whose ABO group is unknown within this framework, it is clear that, due to the uncertainty about the magnitude of hemolysis and the risk that it poses, precautionary measures are indicated. This approach broadens the application of the precautionary principle in transfusion medicine, which to date has been focused on addressing infectious disease concerns as it relates to the safety of the blood supply. For this more clinical application, available precautionary measures can be ranked as strong (not supplementing group AB plasma with A plasma), intermediate (using very-low-titer group A plasma units), or weak (limiting transfusion volume of group A plasma). While the strong approach would offer the greatest safety benefit, the other approaches would help address the blood inventory issue. The appropriateness of the strong approach would therefore depend on whether it is possible to resolve the blood inventory issue. One readily available solution is to supplement the AB plasma inventory by using FDA-approved solvent/detergent-treated pooled human plasma products such as Octaplas (OctaPharma). We could also rationalize plasma transfusions by moving away from high plasma-to-RBC transfusion ratios, which tend to lead to plasma overtransfusion,9 toward a more flexible, targeted approach to coagulation using point-of-care hemostatic testing, which has been shown to reduce transfusions in other settings.10 It may also be possible to use coagulation factor concentrates to bridge the gap until patients’ ABO status is determined and group-specific plasma is prepared, but the utility of this option needs to be clinically tested.11 Since these and other options can likely resolve the blood inventory issue, then it follows that the most appropriate course of action is to implement a strong precautionary measure, aiming for zero risk. Within this framework, the answer to the question posed—should we put aside the precautionary principle for this practice, or should we instead put aside this practice?—is evident: Group A plasma should not be used as part of the initial resuscitation of trauma patients with unknown ABO status because it would unnecessarily expose group B and AB patients to the risks of hemolytic transfusions reactions, the clinical significance of which remains undefined. This conclusion applies until such time that we have conclusive evidence that this practice is safe. We are not currently there; the burden of proof rests with the proponents of this practice.6 The author has disclosed no conflicts of interest. Keyvan Karkouti, MD, FRCPC, MSc1,2,3 e-mail: keyvan.karkouti@uhn.ca 1Department of Anesthesia and Pain Management Toronto General Hospital University Health Network University of Toronto 2Institute of Health Policy, Management, and Evaluation, University of Toronto 3Toronto General Hospital Research Institute and Peter Munk Cardiac Centre University Health Network Toronto, Ontario, Canada

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.