Hi, my name is Marian Mellis and I thank you for the opportunity today to speak to you on behalf of the entire Taxi Cab group to discuss both the methodology and a general overview of the recent plasma and platelet transfusion guidance that we created for critically ill children. Before I begin, I want to thank both the NIH for supporting our work as well as the executive committee of Taxi Cab which was composed of Robert Parker from Stony Brook, Stacey Valentine from University of Massachusetts, as well as Scott Bateman from University of Massachusetts, who both spearheaded the Taxi Initiative, as well as Oliver Karam from Virginia Commonwealth University. We will start with some background. Many of you may be familiar with the Taxi Initiative that was published in 2018 in Pediatric Critical Care Medicine and described recommendations for red blood cell transfusion in critically ill children. I do not want to spend a lot of time going through the specifics of their decision tree, but our methodology was very similar to the Taxi methodology and we tried to produce decision trees in a similar way. As many of you may know, plasma and platelet transfusions have been associated with increased morbidity and mortality in critically ill children. In 2013, the group from St. Justine published a prospective observational study of 831 children who were consecutively admitted to their PICU and found that the receipt of a plasma transfusion was independently associated with new or progressive multi-organ dysfunction syndrome, nosocomial infections, and an increased PICU length of stay. Similarly, in a point prevalence study published in 2019, looking at the use of platelet transfusions in critically ill children, in which over 500 children were enrolled from 86 PICUs, we found that the total administered dose of platelets was independently associated with an increased risk of ICU mortality, even when adjusted for all of the factors listed here. Despite their associated morbidities, very little guidance exists to direct plasma and platelet transfusions in critically ill children. This is a table specific to platelet transfusions published in the AABB guide, providing transfusion guidelines for platelet transfusions in both neonates and older children. And as you can see from this table, apart from a threshold of 5,000 to 10,000 with the failure of platelet production and some guidance given for children with qualitative platelet defects and on ECMO, all of which are based on expert opinion, there's very little other guidance to situations that may be specific to the ICU. For example, there's no addressing if a child is febrile, if they might have clinically relevant bleeding, if they have hyperleukocytosis, receiving anticoagulation, or if they're undergoing procedures within the ICU and what associated thresholds may be used. Because of the lack of guidance, clinical practice therefore varies. In a point prevalent study of plasma transfusions, very similar to the one described for platelet transfusions, in which over 400 children were enrolled in over 100 PICUs, Oliver Karam and colleagues found that a third of plasma transfusions are given to children who have no bleeding and no procedure upcoming. Similarly, from the platelet prevalence study that was previously mentioned, we found that the majority of platelet transfusions are given for prophylaxis and the thresholds that are used to guide the platelet transfusions are extremely varied. Now to discuss the objectives of the Taxi-Cab Initiative. Similar to the Taxi Initiative, the Taxi-Cab Initiative sought to develop evidence-based and, when evidence was lacking, expert-based consensus statements to guide the transfusion and blood management practices of both plasma and platelet transfusions in critically ill children. We also sought to create an implementation initiative as well as develop future research priorities. Now to move on to our methods. We assembled an expert panel with the goal of having a group of people diverse in background, research interests, and specialization. In order to be considered a member of the expert panel, an individual had to have published in pediatric chemostasis and or plasma and platelet transfusion. As you can see, we had a large group of pediatric intensivists, some of whom also had specialization in pediatric cardiology, adult intensive care medicine, as well as pediatric hematology oncology. We also recruited pediatric anesthesiologists, some of whom have expertise in pediatric cardiac surgery. In addition, we included pediatric surgeons, pediatric cardiothoracic surgeons, a pediatric neurosurgeon, pediatric emergency medicine specialists, a group of specialists in pediatric transfusion medicine, many of whom are double-boarded in pediatric hematology oncology, as well as our librarians at Cornell to help us with the systematic review. At the beginning of the process, each expert declared his or her conflicts, which were then reviewed by an outside independent conflicts office, as we wanted to make sure that no one was conflicted in developing the recommendations. We then involved a group of international organizations to be involved both within the both within the process of developing the recommendations as well endorsement once the recommendations were made. We included societies involved in critical care medicine, research organizations, transfusion medicine societies, as well as societies representing patients who are at high risk of receiving plasma and or platelet transfusions. It should be noted that the recommendations are still under review by the ICTMG for endorsement. We then defined subgroups and split up the experts into areas of their particular expertise to develop recommendations within each category. Categories included laboratory measures used to guide transfusion, products processing and characteristics, children with traumatic brain injury and or intracranial hemorrhage, pediatric trauma, children following cardiopulmonary bypass surgery, children being supported by ECMO, children with oncologic diagnoses and or those having received hemopoietic stem cell transplants, children with liver failure and or requiring liver transplantation, children with DIC and or sepsis, children following non-cardiac surgery, and children undergoing invasive procedures outside of the operating room. With the help of our librarians from Weill Cornell, we conducted a large systematic review. We included searches within Medline, Embase, and Cochrane with a combination of relevant subject headings and text words that represented plasma and platelet transfusions in critically ill pediatric patients. For the search, we did not apply any language publication date or article type restrictions. Once the abstracts were pulled, we then applied inclusion and exclusion criteria. We chose to include critically ill children who were term infants starting at one month of age up to 18 years of age. Some subgroups such as those looking at cardiopulmonary bypass in whom the neonatal population is very common did include neonates. Each study needed to evaluate plasma and or platelet transfusion strategies in either prophylactic or therapeutic means to prevent or treat bleeding. The following exclusion criteria were applied. We excluded studies that were not related to indications for plasma and or platelet transfusions within each group's disease of interest. We excluded studies that were focused on plasmapheresis and those who included an adult-only study population. We excluded those who were focused on animals as well as those that did not contain original data which includes meta-analyses. We excluded case reports or case series when the sample size was less than or equal to 10. We excluded abstract only and we excluded studies that were not published in English. The following diagram reviews our process. After the abstracts were reviewed by the executive committee and many excluded based on the exclusion criteria mentioned, they were then passed on to each subgroup where they were reviewed for that particular subgroup. The groups chose their articles, extracted data, they performed bias assessments, and then from the articles from which they had chosen we then referred to their references and went through a process of snowballing. Based on the literature, each group developed recommendations. We conducted a virtual meeting in 2020 where each group presented their recommendations to the large group for discussion. We then had an online voting in which each recommendation was voted upon by all of the experts using a Likert scale with seven to nine considered agreement with the recommendation. Eighty percent of the group had to agree for the recommendation to pass. For those recommendations that did not pass, each group had the opportunity to either revise or discard. We had a second round of voting in which the new and revised recommendations were voted upon using the same process at which point there were no further recommendations that the groups wanted to make. We then met again virtually and reviewed all of the recommendations, many of which were expert panel consensus statements which I will review in a moment. We reviewed decision trees that we had constructed and we discussed research priorities. One word on the wording of our recommendations. As I mentioned, our goal was to develop graded recommendations where there was evidence. However, where evidence was lacking, we sought to develop expert consensus statements. We also developed good practice statements to point out important points of management that will not be tested. An example of this outside of transfusion medicine is measuring blood pressure in patients with septic shock. The language that we used was based on grade methodology, and so when evidence was not strong, we used words such as reasonable or can be considered as suggested by the grade methodology. In order to make sure there was consistency among all of the subgroups, we had to consider two different concepts in how we would address those. The first is that of bleeding. There are not many definitions of bleeding that are specific to critically ill children, and so the groups agreed to use the basic definition that was developed with specificity for critically ill children when thinking through bleeding and split bleeding into severe, moderate, severe, moderate, and minimal. The WHO classification of bleeding has been validated in children with oncologic diagnoses, and therefore the group focused on oncology and hematopoietic stem cell transplantation also considered the WHO classification of bleeding. When thinking through coagulation tests, though all groups recognize that the INR was not an ideal test to predict the risk of bleeding in any patient population of critically ill children, we knew that it was the test that was used and reported the most. In addition, prothrombin ratios are used in Europe, and therefore we used INR and PTR interchangeably. Now to move on to our results. Our initial systematic review resulted in over 22,000 abstracts imported for screening, of which over 5,000 duplicates were removed. The executive committee, and primarily Bob Parker, screened over 17,000 abstracts, the majority of which were removed as irrelevant, either based on the fact that they were non-English or they were not involving plasma and or platelet transfusions in critically ill children, of which we had 1,255 abstracts remaining that were then passed on to each of the subgroups for review for pertinence to their subgroups. After the process which I previously described, the group came to consensus on four recommendations, five good practice statements, and 44 expert consensus-based statements. I want to review the four graded clinical recommendations that were made, as they were not based on thresholds and did not appear within our decision trees. The first two were in regard to product selection and processing. The first, which is a strong recommendation, based on low-quality pediatric evidence, is that in critically ill neonatal and pediatric patients, the use of leukocyte-reduced cellular blood components is recommended. The second, again a strong recommendation, with low-quality pediatric evidence, is that in critically ill RHD-negative neonates, infants, or children in need of a platelet transfusion, RHD-positive platelets should be used if an RHD-negative platelet component is not available. Two graded recommendations were also made in relation to children undergoing cardiopulmonary bypass surgery. The first, which is a weak recommendation, with moderate quality level of evidence, is that in neonatal and pediatric patients undergoing cardiac surgery with cardiopulmonary bypass, viscoelastic monitoring might be considered as an adjunct to standard hemostatic testing, both in the operating room after re-warming and prodomain, and in intensive care units, to inform decisions regarding prophylactic and therapeutic transfusions of plasma and or platelets. Second, also a weak recommendation, with moderate quality level of evidence, in neonatal and pediatric patients undergoing cardiac surgery with cardiopulmonary bypass, the development of institution-specific transfusion algorithms for post-bypass, again after re-warming and protomine, and in the intensive care unit, might be considered to reduce individual case and global overall blood product usage. As mentioned, the expert panel developed good practice statements. Three of the five good practice statements were general and can be applied to all critically ill children. When deciding to transfuse plasma and or platelets to a critically ill pediatric patients, we suggest monitoring not only hemostasis, such as the coagulation dysfunction and or platelet count, but also the overall clinical context, such as the symptoms, signs of bleeding, physiologic markers, other laboratory results like platelet dysfunction, and the risk, benefits, and alternatives to plasma and platelet transfusion. In critically ill pediatric patients, the use of plasma as a volume expander may not be beneficial. And last, in critically ill pediatric patients, taking measures to minimize the number of donor exposures might be considered. The group also developed decision trees, one for platelet transfusions, which is displayed here, and one for plasma transfusions that will be displayed in a moment. We use coloring similar to that of the taxi recommendations. So where you see green, the transfusion is recommended. Yellow, the transfusion can be considered, and red do not transfuse. We broke down the transfusions into those given for prophylaxis, where no or minimal bleeding is present on the left, and those that are given therapeutically for either moderate to severe bleeding on the right, and massive hemorrhage on the far right. As you can note, there are a lot of gray bars, which represents where clinical judgment should be used, where there was neither strong evidence nor expert consensus. I'm not going to go through the entire chart, but you can see that it's broken down by patient diagnoses, and there are thresholds recommended for some conditions. And this is the plasma transfusion decision tree. Again, with the same color coding, the same breakdown into both prophylactic and therapeutic transfusions, and in this case, INR and or PTR are used as recommended thresholds. As you can see, again, especially for the use of therapeutic plasma, there is very little evidence, and clinical judgment should be used in the majority of cases. Following the recommendation development, the group focused on the development of research priorities and publication in the January edition of Pediatric Critical Care Medicine, as well as dissemination, and we also have been working on a plan for implementation. Ken Remy will be giving a discussion of all of these. And with that, I'd like to thank you for your time and listening. I'd like to thank all of the expert panel members, all of whom started this process at the beginning of 2020, when we had no idea what was before us, and as many of you, we were all also taking care of both adults and children sick in the coronavirus pandemic. I would also like to mention that this process will be updated every five years, and so if you're interested in joining our future efforts or nominating someone to join our future efforts, please feel free to reach out to me at the email address listed. Thank you again.

plasma transfusion

platelet transfusion

critically ill children

Taxi Cab group

blood management practices

transfusion guidance