doi:10.1093/bja/aeu379
CRITICAL CARE
Department of Anesthesiology, Magee Womens Hospital, University of Pittsburgh Medical Center, Pittsburgh, USA
E-mail: watejh@upmc.edu
From the experience gained during the Iraq and Afghanistan support hospital. Massive transfusion was defined as .10
conflicts, a concept of damage control resuscitation has RBC units in 24 h. They separated 246 patients into three
arisen. Within the context of damage control resuscitation is groups, depending on the ratio of RBCs to plasma. High
the incorporation of the massive transfusion protocol, which plasma patients who received high ratios of plasma units to
dictates that blood is delivered expeditiously to the bleeding RBCs had higher survival rates compared with those who
patient, generally in a fixed ratio. The protocol is intended received a low ratio of plasma units to RBCs. From this non-
to provide blood to the patient bedside in a rapid fashion. randomized, retrospective analysis, the authors concluded
At the author’s institution for example, 10 units of O+ uncross- that a higher ratio of plasma to erythrocytes contributed to a
matched packed red blood cell (RBC) units, 6 units of AB better outcome. What seems to have been neglected in the
plasma, and 2 bags of platelets are provided. analysis was that the patients who received the high ratio
The importance of the ratio of these products is where had lower rates of thoracic and head injuries; had haemoglobin
controversy arises. Some evidence, which will be discussed, concentrations on presentation to the combat hospital that
implies that these products should be administered in a were 1.5 gm dl21 higher than the low ratio group; had
fixed ratio of 1 unit of packed erythrocytes to 1 unit of reduced base deficit (8 compared with the low ratio group
plasma to 1 unit of platelets (generally termed the 1:1:1 with a base deficit of 13); and had a lower International nor-
transfusion ratio). Most of these studies focus primarily on malized ratio (INR) of 1.54 compared with the low ratio group
the erythrocyte to plasma ratio so in reality, this discussion whose INR was 1.78. So, the low ratio group was more acidotic,
primarily entails a 1:1 erythrocyte:plasma transfusion ratio. more coagulopathic, and more anaemic than the high ratio
In 1982, the phrase ‘viscious circle of trauma’ was coined, group. In addition, for the patients who died in the low ratio
which refers to the acidosis, hypothermia, and coagulopathy group, they did so at an average of 2 h from arrival into the
of trauma.1 The intent of aggressive plasma use is to address combat facility; whereas, the patients in the high ratio group
the coagulopathy of trauma at an early stage. The following dis- died on average at 37 h and died more commonly from
cussion is an unsystematic review of literature related to the sepsis, multi-organ failure, and central nervous system injury.
fixed ratio plasma to red cell transfusion strategy. A number of studies similar in study design followed the
Borgman study (see Table 1). In 2008, a retrospective civilian
Fixed ratio transfusion study involving 16 level I trauma centres and 467 patients
The evidence for a 1:1 transfusion ratio derives from a number was published,5 which also found improved survival with
of studies. The landmark study that started this practice was by higher ratios of plasma and platelets. This study appeared to
Borgman and colleagues in 2007.2 This was a retrospective be the catalyst for many trauma surgeons to adopt the 1:1
chart review of massive transfusion at a US Army combat transfusion strategy. Like the Borgman study, there were
& The Author 2014. Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved.
For Permissions, please email: journals.permissions@oup.com
BJA Waters
stark differences in time-to-death between the patients that Part of this minimal effect relates to the variable pro-
received high ratios of plasma and platelets. If the time of coagulant potency of transfused plasma. The INR of trans-
death and the average number of blood products administered fused plasma ranges from ,1 to .1 based on the donor
to the patients who died in each group are used to estimate and the effect of storage. Some plasma can have an INR as
rates of blood product administration between ratio groups, high as 1.3.23 So, if an INR level is ,1.3, there is a good
stark differences in patient populations appear to exist in this chance that plasma transfusion will raise the INR. In addition,
study also. Patients who were categorized in a low plasma, correction of an INR is not linear and at low INR values,
low platelet group, received erythrocytes at a rate of 5.25 plasma has little effect.24 25 If plasma is given to raise
units per hour before death whereas for the high plasma, coagulation factor concentrations by 10%, the effect will be
high platelet group erythrocytes were given at a rate of 0.63 very different depending upon the starting INR. For instance,
units per hour (Table 2). Additionally, the mechanism of if the INR is 1.5 and enough plasma is given to raise factor
injury was truncal and head injury in 58% of the low ratio concentrations by 10%, then the result will be to correct it
patients and 23% in the high ratio patients. While this study to 1.4. A similar change in 10% when the INR is 3.0 will
has been highly influential, caution would be suggested in result in a new INR of 2.3, so the effect is much greater at
translating these findings to civilian trauma. higher INRs. Recent evidence suggests that conventional la-
Each of the studies outlined in Table 1 has flaws associated boratory tests of coagulation are insensitive in detection
with their retrospective study design. Most prominently, sur- acquired coagulopathy as in trauma, or in guiding procoagu-
vivor bias has been a repetitive source of error in these 1:1 lant therapy.26 27
ratio studies. Survivor bias is associated with having survived With the understanding that varying effects of plasma
ii4
MTP in haemorrhagic shock
Table 1 Studies advocating 1:1 transfusion ratios. PRBC, packed red blood cells. NR, not reported; FFP, plasma; PRBC, erythrocytes; TEP, transfusion exsanguination protocol; MT, massive transfusion;
MTP, massive transfusion protocol; LOS, length of stay; MOF, multiple organ failure; ARDS, acute respiratory distress syndrome.
BJA
ii5
Table 2: Blood component transfusions by plasma and platelet ratios. Reproduced with permission from Holcomb and colleagues.2
RBC = erythrocytes; FFP = Plasma; rFVIIa = recombinant factor VIIa
High plasma, High High plasma, low Low plasma, high Low plasma, low P
Platelets platelets platelets platelets
Plasma (units) 17 (12) 16 (10) 7 (5) 6 (6) ,0.001
Platelets (units) 20 (16) 5 (6) 18 (10) 4 (6) ,0.001
RBC (units) 22 (17) 21 (12) 21 (11) 21 (12)
Crystalloid (L) 14 (10) 13 (7) 17 (12) 11(10) ,0.001
FFP:RBC ratio 0.8 (0.3) 0.8 (0.3) 0.3 (0.1) 0.2 (0.1) ,0.001
Platelet:RBC ratio 0.9 (0.4) 0.2 (0.2) 0.9 (0.4) 0.1 (0.2) ,0.001
Crystalloid:RBC 0.8 (0.5) 0.8 (0.6) 0.9 (0.6) 0.6 (0.5) ,0.001
ratio
Received rFVIIa (%) 34 (22) 11 (11) 21 (25) 15 (12) 0.006
Time Time
Fig 1 Survivor bias. (A) a theoretical patient receives 4 erythrocyte units followed by a plasma unit and subsequently dies, so a 4:1 ratio results. (B), a
theoretical patient receives 4 erythrocyte units followed by 4 plasma units but doesn’t die resulting in a 1:1 ratio. The therapy is identical with the
exception being the death of the patient.
Table 3: Group sizes* and numbers of deaths in low (,1:2) and high (.1:2) FFP:PRBC ratio groups by time interval of hospitalization. Reproduced
with permission from Snyder and colleagues.12
≤30 >30 –60 >60 – 90 >90 – >2 –3 >3 –4 >4–5 >5–6 >6–12 >12 – 18 >18– 24 >24 Total
min. min. min. 120 min. hrs. hrs. hrs. hrs. hrs. hrs. hrs. hrs.
,1:2 (low ratio)
No. pts in group 86 102 103 102 95 91 85 87 74 73 74 74 74
No. deaths 0 0 8 10 9 3 2 1 2 2 0 6 43
during interval
.1:2 (high ratio)
No. pts in group 1 6 13 20 34 39 46 46 60 61 60 60 60
No. deaths 0 0 0 0 0 2 1 0 7 1 0 13 24
during interval
plasma transfusion increased the risk of death in a cohort of colleagues30 found an increased risk of multiple organ
patients that had a predicted risk of death ,20%. Similarly, failure when patients were transfused aggressively with
Inaba and colleagues29 found an association between plasma. These authors suggested that caution is warranted
plasma use and overall complications in non-massively trans- when empirically transfusing plasma. These three studies
fused patients (defined as patients receiving ,10 units of suggest that in the bleeding patient, more plasma might
PRBCs during the first 24 h of their hospitalization). In a retro- have unpredictable results when given using a fixed ratio
spective review of 1440 trauma patients, Johnson and strategy.
ii6
MTP in haemorrhagic shock BJA
Laboratory-guided plasma use risks increase with escalating plasma exposure, thus affecting
patient safety. Assessing the extent of coagulopathy through
In order to avoid excessive or overzealous administration of
conventional laboratory testing cannot help guide bleeding
blood products, several investigators now suggest that transfu-
management, but minor derangements in routine coagulation
sion be guided by laboratory testing. In a randomized study,
tests can be used to avoid excessive allogeneic blood transfu-
Nascimento and colleagues31 compared a fixed transfusion
sions including plasma and its associated complications.
ratio with transfusion guided by laboratory data every 2 h.
Future prospective research is warranted to assess the poten-
This study enrolled 78 patients with the primary aim being to
tial benefits of rapid delivery of plasma in a fixed ratio to
assess the feasibility of doing a larger study of this topic. They
RBC concentrates compared with laboratory testing guided
found a trend towards worse outcomes with the fixed ratio
plasma without a fixed ratio to other blood products compared
strategy of care compared with laboratory-guided plasma
with individualized coagulation management based on point-
and platelet use, but concluded that a larger study size was
of-care coagulation monitoring.
needed in order to show statistical significance. Regardless,
the authors make a significant leap in sophistication from the
recipe of giving blood products in a fixed ratio to one that is
Declaration of interest
guided by laboratory data. None declared.
Traditional laboratory testing provides slow information
by which to make a care decision but the concept of point- References
of-care testing allows for rapid testing and resulting of coagu-
1 Kashuk JL, Moore EE, Millikan JS, Moore JB. Major abdominal vascu-
ii7
BJA Waters
14 Teixeira PG, Inaba K, Shulman I, et al. Impact of plasma transfu- 29 Inaba K, Branco BC, Rhee P, et al. Impact of plasma transfusion in
sion in massively transfused trauma patients. J Trauma 2009; trauma patients who do not require massive transfusion. J Am
66: 693 – 7 Coll Surg 2010; 210: 957– 65
15 McVay PA, Toy PT. Lack of increased bleeding after liver biopsy in 30 Johnson JL, Moore EE, Kashuk JL, et al. Effect of blood products
patients with mild hemostatic abnormalities. Am J Clin Pathol transfusion on the development of postinjury multiple organ
1990; 94: 747–53 failure. Arch Surg 2010; 145: 973–7
16 Ewe K. Bleeding after liver biopsy does not correlate with indices of 31 Nascimento B, Callum J, Tien H, et al. Effect of a fixed-ratio (1:1:1)
peripheral coagulation. Dig Dis Sci 1981; 26: 388– 93 transfusion protocol versus laboratory-results-guided transfusion
17 Counts RB, Haisch C, Simon TL, et al. Hemostasis in massively trans- in patients with severe trauma: a randomized feasibility trial.
fused trauma patients. Ann Surg 1979; 190: 91– 9 Can Med Anaesth J 2013; 185: E583–9
18 Yazer MH. The how’s and why’s of evidence based plasma therapy. 32 Carroll RC, Craft RM, Langdon RJ, et al. Early evaluation of acute trau-
Korean J Hematol 2010; 45: 152–7 matic coagulopathy by thrombelastography. Transl Res 2009; 154:
19 Domen RE, Hoeltge GA. Allergic transfusion reactions: an evalu- 34–9
ation of 273 consecutive reactions. Arch Pathol Lab Med 2003; 33 Kashuk JL, Moore EE, Wohlauer M, et al. Initial experiences
127: 316– 20 with point-of-care rapid thrombelastography for management of
20 Kleinman S, Caulfield T, Chan P, et al. Toward an understanding of life-threatening postinjury coagulopathy. Transfusion 2012; 52:
transfusion-related acute lung injury: statement of a consensus 23–33
panel. Transfusion 2004; 44: 1774– 89 34 Johansson PI, Sorensen AM, Larsen CF, et al. Low hemorrhage-
21 Bux J. Transfusion-related acute lung injury (TRALI): a serious related mortality in trauma patients in a Level I trauma center
adverse event of blood transfusion. Vox Sang 2005; 89: 1 –10 employing transfusion packages and early thromboelastography-
directed hemostatic resuscitation with plasma and platelets.
ii8