Transfusion Medicine Reviews xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Transfusion Medicine Reviews journal homepage: www.tmreviews.com

Transfusion Interventions in Critical Bleeding Requiring Massive Transfusion: A Systematic Review Zoe K. McQuilten a,b,⁎, Gemma Crighton b, Sunelle Engelbrecht b, Robert Gotmaker c, Susan J. Brunskill d, Michael F. Murphy d, Erica M. Wood b a

Australia and New Zealand Intensive Care Society (ANZICS) Research Centre, Melbourne, Australia Transfusion Research Unit, Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Australia Department of Anaesthesia and Pain Management, Monash Health, Melbourne, Australia d Systematic Reviews Initiative, NHS Blood and Transplant/Oxford University Hospitals NHS Trust, Oxford, United Kingdom b c

a r t i c l e

i n f o

Available online xxxx Keywords: Blood transfusion Hemorrhage Gastrointestinal hemorrhage Injuries Recombinant FVIIa Prothrombin complex concentrate Fibrinogen Perioperative

a b s t r a c t Critical bleeding (CB) requiring massive transfusion (MT) can occur in a variety of clinical contexts and is associated with substantial mortality and morbidity. In 2011, the Australian National Blood Authority (NBA) published patient blood management guidelines for CB and MT, which found limited high-quality evidence from which only 2 recommendations could be made. The aim of this systematic review (SR) was to update these guidelines and identify evidence gaps still to be addressed. A comprehensive search was performed for randomized controlled trials (RCTs) and SRs using MeSH index and free text terms in MEDLINE, the Cochrane Library (Issue 11, 2012), EMBASE, CINHAL, PUBMED, and the Transfusion Evidence Library up to July 15, 2014. The evidence was grouped according to 4 questions based on the original guideline relating to transfusion interventions: (1) effect of dose, timing, and ratio of red blood cells (RBCs) to component therapy on patient outcomes; (2) effect of RBC transfusion on patient outcomes; (3) effect of fresh frozen plasma, platelet, cryoprecipitate, fibrinogen concentrate, and prothrombin complex concentrate on patient outcomes; and (4) effect of recombinant activated factor VII (rFVIIa) on patient outcomes. From this search, 19 studies were identified: 6 RCTs and 13 SRs. Two of the RCTs were pilot/feasibility studies, 3 were investigating rFVIIa, and 1 compared restrictive versus liberal RBC transfusion in upper gastrointestinal hemorrhage. Overall, limited new evidence was identified and substantial evidence gaps remain, particularly with regard to the effect of component therapies, including ratio of RBC to component therapies, on patient outcomes. Clinical trials to address these questions are required. © 2015 Elsevier Inc. All rights reserved.

Contents Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 Search Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 Selection Criteria. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 Data Extraction and Quality Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 Quality Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 Methodological Quality of Included Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 Effects of the Interventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 In Patients with CB Requiring MT, does the Dose, Timing, and Ratio (Algorithm) of RBCs to Component Therapy (FFP, Platelets, Cryoprecipitate, or Fibrinogen Concentration) Influence Morbidity, Mortality, and Transfusion Rate? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 In Patients with CB Requiring MT, What is the Effect of RBC Transfusion on Patient Outcomes? . . . . . . . . . . . . . . . . . . 0 In Patients with CB Requiring MT, What is the effect of FFP, Cryoprecipitate, Fibrinogen Concentrate, Platelet Transfusion, and/or PCC on Patient Outcomes? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 In Patients with CB Requiring MT, What is the Effect of rFVIIa (Prophylaxis or Treatment) on Morbidity, Mortality, and Transfusion Rate? 0 ⁎ Corresponding author at: Dr. Zoe K McQuilten, ANZICS-RC, Department of Epidemiology and Preventive Medicine, Level 6, The Alfred Centre, 99 Commercial Road, Melbourne, Australia, 3004. E-mail address: [email protected] (Z.K. McQuilten). http://dx.doi.org/10.1016/j.tmrv.2015.01.001 0887-7963/© 2015 Elsevier Inc. All rights reserved.

Please cite this article as: McQuilten ZK, et al, Transfusion Interventions in Critical Bleeding Requiring Massive Transfusion: A Systematic Review, Transfus Med Rev (2015), http://dx.doi.org/10.1016/j.tmrv.2015.01.001

2

Z.K. McQuilten et al. / Transfusion Medicine Reviews xxx (2015) xxx–xxx

Discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dose, Timing, and Ratio (algorithm) of RBCs to Component Therapy . . . . Red Blood Cell Transfusion . . . . . . . . . . . . . . . . . . . . . . . FFP, Cryoprecipitate, Fibrinogen Concentrate, PCC, and/or Platelet Transfusion rFVIIa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Limitations of the Review . . . . . . . . . . . . . . . . . . . . . . . Impact of Methodological Assessment. . . . . . . . . . . . . . . . . . Evidence Gaps . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conflict of Interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Critical bleeding (CB) requiring massive transfusion (MT) can occur in a variety of clinical contexts and is associated with substantial mortality and morbidity. Blood component transfusion is an essential element of the management of patients with CB. As with any therapeutic intervention, the decision to transfuse should be guided by the evidence for efficacy balanced against the potential risks. However, despite the crucial role transfusion plays in the management of patients with CB, there have been few studies to inform their optimal use. In 2011, Australia's National Blood Authority (NBA) published patient blood management (PBM) guidelines for CB and MT [1]. These were developed by clinical experts from 11 colleges and societies, including the Australia and New Zealand Society of Blood Transfusion, and patient representatives. The guideline was developed using the 2007 Australian National Health and Medical Research Council (NHMRC) guideline development methodology [2] and was approved by the NHMRC. The results of the systematic review (SR) indicated a paucity of high-quality evidence and the need for further research. Only 2 recommendations could be made: institutions should develop an MT protocol (MTP) that includes the dose, timing, and ratio of blood component therapy for use in trauma patients with, or at risk of, CB requiring MT; and (2) the routine use of activated recombinant factor VII (rFVIIa) in trauma patients with CB requiring MT was not recommended because of its lack of effect of mortality and variable effect on morbidity. However, no recommendations could be made on the dose, timing, or ratio of blood components or use of individual blood components. The aims of this SR were to identify whether any new studies (randomized controlled trials [RCT] or SRs) had been published addressing the evidence gaps identified in the PBM guidelines for patients with CB and MT, with particular reference to transfusion interventions and to identify which evidence gaps remain.

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

0 0 0 0 0 0 0 0 0 0 0 0

least 1 RCT or be an RCT; (2) include a patient population who had CB and had received, or were anticipated to receive, an MT in any clinical setting; (3) have at least 1 relevant transfusion intervention of interest (red blood cell [RBC], fresh frozen plasma (FFP], platelets, cryoprecipitate, prothrombin complex concentrate (PCC), fibrinogen concentrate, rFVIIa); and (4) measure at least one of the following outcomes—mortality, hospital length of stay (LOS), serious adverse events, transfusion-related adverse events, morbidity, and transfusion rate. Studies were excluded if the patient population was exclusively obstetric or paediatric, or if the clinical setting was planned CB during surgery, as these clinical scenarios were not included in the original PBM guideline [1]. The definition of CB used for this SR was the same as that used in the PBM guideline: a major hemorrhage that is life threatening and likely to result in the need for MT [1]. As in the original guideline, studies on hemorrhage of smaller volume in a critical area or organ were not considered in this SR. Any definition of MT was accepted. No restrictions were placed on language and any papers not published in English were translated.

Data Extraction and Quality Assessment All electronically derived citations and abstracts of papers identified by the review search strategy were screened for relevancy by 2 out of 3 review authors. Full text of identified studies were reviewed by 2 reviewers against the inclusion and exclusion criteria. Potentially relevant trials were then formally assessed for eligibility against the criteria outlined above by 2 review authors. A third review author resolved disagreements or discrepancies.

Methods A study protocol was developed that outlined the participants, interventions, comparators, and outcomes of interest, inclusion and exclusion criteria, and search strategy as outlined below. Search Strategy We searched for RCTs and SRs using MeSH index and free text terms in MEDLINE, the Cochrane Library (Issue 11, 2012), EMBASE, CINHAL, PUBMED, and the Transfusion Evidence Library from the period May 2009 to November 2012. An updated search (from November 12, 2012, to July 15, 2014) was performed prior to publication and involved a review of all RCTs and SRs listed on the UKB SRI Transfusion Evidence Library (www.transfusionevidencelibrary.com). Selection Criteria To be eligible for inclusion in the review, studies were required to meet all of the following criteria: (1) be an SR that had included at

Data Collection A standard data extraction form was developed separately for SRs and RCTs. Data extraction was conducted by 2 review authors, and any disagreements were resolved by consensus or with discussion with a third review author. Details collected from each RCT were as follows: author, citation of paper, secondary citations, objectives of the trial, trial location, number of sites, clinical setting, study population, study design, dates of recruitment, sample size, power calculation, whether stopping rules were applied, numbers of participants randomized and analyzed, inclusion and exclusion criteria, experimental and control interventions, primary and secondary outcomes, co-interventions, compliance with interventions, loss to follow-up, results, statistical analysis, and aspects of study quality. For SRs, data collected included author, citation of paper, study objectives, clinical setting, study population, study design, dates and strategy for the literature search, method used for data screening and extraction, inclusion and exclusion criteria, nature of intervention, comparator population, outcomes examined, and results of any metaanalyses performed.

Please cite this article as: McQuilten ZK, et al, Transfusion Interventions in Critical Bleeding Requiring Massive Transfusion: A Systematic Review, Transfus Med Rev (2015), http://dx.doi.org/10.1016/j.tmrv.2015.01.001

Z.K. McQuilten et al. / Transfusion Medicine Reviews xxx (2015) xxx–xxx

3

fibrinogen concentrate, and PCC on patient outcomes; and (4) effect of rFVIIa on patient outcomes [1].

Quality Assessment Assessment of methodological quality of the study was performed independently by 2 reviewers. Assessment of methodology quality for RCTs utilized the components in the assessment of risk of bias according to Cochrane Handbook for Systematic Reviews of Interventions [3]. The key components of assessment included random sequence generation, allocation concealment, blinding of participants, physicians and outcome assessors, power calculation, sample size achieved, trial analysis, co-interventions, numbers lost to follow-up, any sponsorship, and funding and whether all outcomes were reported. Assessment of the methodological quality of the SRs was based on the assessment of multiple systematic reviews (AMSTAR) tool. Key components considered were the following: whether a comprehensive literature was performed, the period of the search, the characteristics of included studies, whether there was parallel independent study selection and dual data extraction, whether a list of included studies was provided, and how the scientific quality of the included studies was assessed [4].

Analysis Because of the heterogeneous nature of the interventions, and the different study types, no meta-analyses were performed. A qualitative analysis was performed, and included studies were grouped according to the four intervention questions originally included in the PBM guideline: (1) dose, timing, and ratio of RBCs to component therapy; (2) effect of RBC on patient outcomes; (3) effect of FFP, platelet, cryoprecipitate,

Results The original literature search identified 3050 citations, and a further 49 were included after updating the search as described in the methods section, resulting in a total of 3094 citations. After initial screening, 69 full text articles were selected and reviewed. Of these, 50 were excluded for the following reasons: the primary intervention or outcome was not relevant; the study's patient population did not meet the inclusion criteria; the study was not an SR or RCT; and, the SR only included observational studies (Figure). This left a total of 19 studies: 6 RCTs and 13 SRs. The RCTs were on restrictive versus liberal RBC transfusion in upper gastrointestinal (GI) bleeding [5], whole blood (WB) versus component therapy in trauma [6], effect of a fixed ratio of components in trauma [7], and rFVIIa versus placebo in trauma [8–10]. The SRs included restrictive versus liberal RBC transfusion in upper GI bleeding [11,12], FFP in a range of clinical settings [13], fibrinogen concentrate in patients with major bleeding [14,15], rFVIIa for indications other than hemophilia [16–21], management of trauma patients with bleeding [22], and diagnosis and management of trauma-induced coagulopathy [22]. Methodological Quality of Included Studies Tables 1 and 2 summarize the methodological quality of the included RCTs and SRs, respectively.

Records identified through search

Records screened 3099

Records clearly did not meet the inclusion criteria for SR n = 3030 Full text review 69 papers Reasons for exclusion (n = 50) Intervention not relevant (n = 4) Outcome not relevant (n = 7) Included in qualitative analysis 19 papers

Patient population not massive transfusion/critical bleeding (n = 8) SR of observational studies only (n = 15) Not a systematic review or RCT (n = 16)

Included in meta-analysis None

Figure. Flow chart of selected articles.

Please cite this article as: McQuilten ZK, et al, Transfusion Interventions in Critical Bleeding Requiring Massive Transfusion: A Systematic Review, Transfus Med Rev (2015), http://dx.doi.org/10.1016/j.tmrv.2015.01.001

4

Z.K. McQuilten et al. / Transfusion Medicine Reviews xxx (2015) xxx–xxx

Table 1 Quality of included RCTs Article

Allocation Random Blinding Power Sample Trial Co-interventions concealment sequence calculation size analysis Participant Physician Outcome allocation achieved

Boffard 2009

NR

NR

Yesa

Yesa

Yesa

Yes

Yes

NR

Cotton 2013

Yes

NR

No

No

No

Yes

Yes

ITT

Dutton 2011

Yes

Yes

Yes

Yes

Yes

Yes

Nob

ITT

Hauser 2010

Yes

Yes

Yes

Yes

Yes

Yes

Nob

ITT

Nascimento Yes 2013

Yes

No

No

No

No

NA

ITT

Villanueva 2013

Yes

No

No

No

Yes

Yes

ITT

Yes

Specific transfusion guideline in line with study guideline NR

Guidelines on bleeding, transfusion and ventilator management Guidelines on bleeding, transfusion and ventilator management Urgent operation/ angioembolization, Decompressive craniectomy Crystalloid, Colloid, Cryoprecipitate, rFVIIa, Tranexamic acid Endoscopic guidelines and recommendations for PPIs, somatostain, antibiotics. No policy for TXA, cell salvage, surgery, hypotension, acidosis.

Loss to followup

Industry or organization sponsorship/ funding

Balanced Novo Nordisk between groups NR The University of Texas US Army Medical Research and Materiel Command Balanced Novo Nordisk between groups.

Any outcome not reported No

No

NA

Balanced Novo Nordisk between groups.

No

None

Special coagulation assay results

None

Canadian Forces Health Services, Defense Research and Development Canada, NBF and AABB

No

Abbreviations: ITT, intention to treat; NA, not applicable; NR, not reported; TXA, tranexamic acid. a Trial stated to be double-blind but minimal details provided. b Study ceased early.

With regard to the methodological quality of the 6 included RCTs, 5 reported adequate allocation concealment [5,6,9,10], and 1 study did not report method of concealment [8]. Four of the RCTs reported on the method of random sequence allocation [5,7,9,10], while 2 studies did not [6,8], and therefore are at risk of selection bias. Three studies reported blinding of participants, personnel, and outcome assessors [8–10]. The other studies were not blinded [5–7] and are therefore at risk of detection bias. Five of the RCTs reported using intention-totreat analysis [5–7,9,10], and 1 study did not report the method of trial analysis [8]. Two studies reported no loss to follow-up [5,7], 3 reported balanced loss to follow-up across treatment arms [8–10], and loss to follow-up was not clearly reported in 1 study [6]. Three studies received industry support or sponsorship [8–10]. Overall, 1 RCT, which had 2 publications included in this review, had good methodological design in all facets of assessment [9,10]. With regard to the 13 included SRs, the periods included in the literature search were extensive for 8 [11,13,15,19–23] and restricted for 3 [14,16,18]. The start dates were not specified in 2 SRs [12,17] that were updates of previously published SRs. Nine of the 13 studies only included RCTs [11,12,15,17–22]. Eight SRs placed no restriction on the language in which the trial had been published. Five of the SRs reported parallel study selection and duplicated data extraction [11–13,19,21], while 3 did not report on their methods for study selection or data extraction [16,18,20]. Twelve searched multiple biographic databases, while 1 study performed only a MEDLINE search [18]. With the exception of 1 SR [18], all performed a quality

assessment of their included studies. Overall, 4 of the 13 SRs were of high methodological quality [11,15,19,21], and these were all Cochrane Database Systematic Reviews. Effects of the Interventions In Patients with CB Requiring MT, does the Dose, Timing, and Ratio (Algorithm) of RBCs to Component Therapy (FFP, Platelets, Cryoprecipitate, or Fibrinogen Concentration) Influence Morbidity, Mortality, and Transfusion Rate? Two RCTs and 2 SRs were identified that addressed this question. There were a number of SRs specifically addressing the effect of the ratio of RBC to blood component therapy; however, these were excluded because they only included observational studies [24–27]. Randomized Controlled Trials. Nascimento et al reported a feasibility RCT, which included trauma patients expected to require a MT and compared a fixed ratio of RBC to FFP to platelets of 1:1:1 compared with standard practice (laboratory result–guided transfusion protocol) [7]. The trial was able to achieve the 1:1:1 ratio in 57% (21 of 37) of the fixed ratio group compared with 6% (2of 32) in the control group, thus demonstrating the feasibility of the intervention. The study was not powered to detect a difference in mortality; however, the all-cause 28-day mortality by intention-to-treat analysis (relative risk for fixed ratio, 2.27; 95% confidence interval [CI], 0.98-9.63) and by per-protocol

Please cite this article as: McQuilten ZK, et al, Transfusion Interventions in Critical Bleeding Requiring Massive Transfusion: A Systematic Review, Transfus Med Rev (2015), http://dx.doi.org/10.1016/j.tmrv.2015.01.001

Z.K. McQuilten et al. / Transfusion Medicine Reviews xxx (2015) xxx–xxx

5

Table 2 Quality of included SRs Article

Period for search

Type of included study

Aguado 2009

January 1980–July 2008

SR, NR Guideline,RCT and case series

NR

Birchall 2008

Unknown– January 2007 1950 to July 2010

RCT

NR

Yes

English, Spanish, French, Portugese and Tialian NR

Curry 2011 (TMR) Curry 1950 to 2011 July 2010 (critical care)

SR (including observational data) RCT

No

Yes

No

No

Yes

No

Jairath 2010

RCT

Yes

Yes

No

Levi 2010 1996–2008 RCT

NR

NR

NR

Marti1948–Dec Carava- 2011 jal 2012

RCT

Yes

Yes

No

Murad 2010

RCT and controlled observational studies RCT

Yes

Yes

No

NR

NR

NR

1950– March 2010

Inception– August 2009

Nishijima 1966–July 2009 2008

Duplicate Duplicate Language study data restrictions selection extraction

Simpson 2012

1948– March 23, 2011

RCT

Yes

Yes

No

Wang 2013

Unknown– Jan 2013

RCT

Yes

Yes

No

Warmuth 1985–May 2012 2010

Prospective, controlled trials

Yes

No

English or German

Wikkelso 2013

RCTs

Yes

NR

No

1950– August 2013

Literature search strategy

List of Scientific included quality of studies included studies assessed

Quality tool used

Multiple bibliographic databases

NR

Yes

Evaluation criteria of the Critical Appraisal Skills Programme and Jadad scale.

Multiple bibliographic databases Multiple bibliographic databases Multiple bibliographic databases and clinical trials register Multiple bibliographic databases and clinical trials register MEDLINE search of RCT Multiple bibliographic databases and clinical trials register Multiple bibliographic databases

Yes

Yes

Based on criteria from Schulz et al [37] JAMA, 1995 and Jüni et al [38]

Yes

Yes

Assessment of quality: 4 defined SR processes

Yes

Yes

Cochrane Handbook for Systematic Reviews of Interventions [3]

Yes

Yes

Criteria adapted from Juni et al [38] and according to Cochrane Handbook for Systematic Reviews of Interventions [39]

NR

NR

Not reported

Yes

Yes

Cochrane Handbook for Systematic Reviews of Interventions [3]

Yes

Yes

Examples were provided of items related to the methodological quality of RCT and observational studies that were provided

Multiple bibliographic databases and clinical trials register Multiple bibliographic databases and clinical trials register Multiple bibliographic databases Multiple bibliographic databases

Yes

Yes

Published criteria by Guyatt GH et al [40]

Yes

Yes

Cochrane Handbook for Systematic Reviews of Interventions [3]

Yes

Yes

Jadad composite scale [41,42]

Yes

Yes

Multiple bibliographic databases and clinical trials register

Yes

Yes

Cochrane Handbook for Systematic Reviews of Interventions [3] CRD's guidance for undertaking reviews in health care Cochrane Handbook for Systematic Reviews of Interventions [3]

Abbreviations: NR, not reported.

analysis (relative risk for fixed ratio, 3.17; 95% CI, 1.15-18.24) was reported as a safety outcome. There was greater plasma wastage seen in the fixed ratio group compared with the control group (86 [22%] of 390 vs 30 [10%] of 289). In this study, prethawed FFP was not available for either intervention group, and the median time to first FFP transfusion was 89 minutes (interquartile range [IQR], 65-150) in the fixed ratio group and 113 (IQR, 81-165) in the control group. A pilot RCT of WB vs component therapy in trauma patients [6] found no difference in the primary outcome of total transfusion volume between the 2 groups (Table 3). In the intention-to-treat analysis, the median 24-hour transfusion volume was 3680 mL (WB) vs 3920 mL component therapy (P = .46). There was also

no difference in 24-hour RBC, plasma, and platelet transfusions received between groups. Mortality rates at 24 hours and 30 days were similar between the WB and component group; however, the study was a pilot study and not powered to detect a difference in these outcomes. Systematic Reviews. A SR of RCTs on the management of traumatic hemorrhagic [22] did not identify any new studies. An SR of SRs on the management of trauma-induced coagulopathy identified 4 reviews that investigated the effect of RBC to blood component therapy ratios, all of which were based on observational and mostly retrospective data [23].

Please cite this article as: McQuilten ZK, et al, Transfusion Interventions in Critical Bleeding Requiring Massive Transfusion: A Systematic Review, Transfus Med Rev (2015), http://dx.doi.org/10.1016/j.tmrv.2015.01.001

6

Z.K. McQuilten et al. / Transfusion Medicine Reviews xxx (2015) xxx–xxx

Table 3 Summary of findings from RCT Article

Study setting and number sites

Dates of Numbers recruitment analyzed

Effect of component therapy on outcomes Nascimento Single-center June 2009– 2013 October 2011

Cotton 2013

Single center

NR

Effect of RBC on outcomes Villanueva Hospital—single June 2003– 2013 center December (Spain) 2009

Effect of rFVIIa on outcomes Dutton Nonmilitary 2011 trauma

150 hospitals

Length of Inclusion criteria follow-up

Intervention

Primary outcome

203 eligible, 78 randomized, 69 analysed (37 in fixed-ratio and 32 in control group)

28 d

16–90 y with traumatic injuries, Intervention - fixed Feasibility bleeding and expected to require ratio of RBCs, FFP, MT and episode of SBP b90 mm Hg and PLT at 1:1:1 ratio Control—standard transfusion practice guided by laboratory parameters

357 eligible, 107 analysis (55 in WB, 52 in component therapy)

30 d

≥18 y, met highest level trauma activation criteria and active bleeding requiring uncrossmatched blood in ED

Intervention—6 WB units Control—6 RBC units + 6 FFP units Each group received 1 PLT

24-h blood product use

Patients ≥18 y with hematemesis, melena, or both, as confirmed by the hospital staff NB: Exclusions—massive exsanguination

Restrictive: Hb threshold 7 g/dL, with a target range for posttransfusion Hb level 7-9 g/dL Liberal: Hb threshold 9 g/dL, with a target range for the posttransfusion Hb level of 9-11 g/dL

Mortality from any cause in the first 45 d

921 randomized, 45 d 889 analyzed (444 in restrictive, 445 in liberal

560 patients randomized (270 rFVIIa and 290 placebo)

Day 30 As per Hauser 2010 for mortality, MOF. Day 90 for SAE

As per Hauser 2010 As per Hauser 2010

Results

Fixed ratio vs standard practice Transfusion ratio of 1:1:1 achieved in 21 (57%) of 37 of intervention compared with 2 (6%) of 32 in control All cause 28-day mortality (ITT): 32.5% vs 14%, RR 2.27 (95% CI, 0.98-9.63) All cause 28-day mortality (perprotocol): 29.7% vs 9.4%, RR 3.17 (95% CI, 1.15-18.24) Event-free survival 20 (54%) of 37 vs 25 (78%) of 32 in control (P = .053) WB vs component therapy Median 24-h total transfusion volume 3680 vs 3920 mL, P = .462. Median 24-h number units 12 (IQR, 6-24) vs 13 (IQR, 5-29), P = .61 No difference 24-h RBC, plasma or platelet transfusion Mortality 24 h: 11deaths vs 10; P = .83 Mortality 30 d: 22 deaths vs 14 deaths; P = .26 ARDS: 0.0% vs 1.9%; P = .32 Restrictive vs liberal transfusion strategy Mortality at 45 d: HR 0.55 (95% CI, 0.330.92), P = .02 Mean (SD) RBC units transfused 1.5 ± 2.3 vs liberal 3.7 ± 3.8, P b .001 Overall complications rate in 40% (179 patients) vs 48% (214 patients), P = .02 Mean hospital LOS (days) 9.6 (SD, 8.7) vs 11.5 (SD, 12.8), P = .01 rFVIIa vs placebo safety outcomes: All adverse events 240 (89%) vs 256 (88%), P = 0.82 SAE 165 (61%) vs 197 (68%), P = .09 ARDS 8 (3%) vs 21 (7.2%), P = .02 Arterial TEA 16 (6%) vs 12 (4%), P = .33

Please cite this article as: McQuilten ZK, et al, Transfusion Interventions in Critical Bleeding Requiring Massive Transfusion: A Systematic Review, Transfus Med Rev (2015), http://dx.doi.org/10.1016/j.tmrv.2015.01.001

Z.K. McQuilten et al. / Transfusion Medicine Reviews xxx (2015) xxx–xxx

7

Table 3 (continued) Article

Study setting and number sites

Dates of Numbers recruitment analyzed

Length of Inclusion criteria follow-up

Hauser 2010

Nonmilitary trauma

150 hospitals

573 patients randomized, 554 in ITT analysis: Blunt rFVIIa 221, placebo 247 Penetrating rFVIIa 46 placebo 40

Boffard 2009

Nonmilitary trauma

32

Blunt: 143/158 30 d randomized patients analyzed (69 rFVIIa and 74 placebo) Penetrating: 134/ 143 randomized patients analyzed (70 rFVIIa and 64 placebo)

Day 30 for Mortality, MOF. Day 90 for SAE

Blunt and/or penetrating trauma patients; 18 to 70 y with continuing torso and/or proximal lower extremity bleeding after receiving 4 RBC despite standard hemostatic interventions

Received 6 RBC units within a 4 h ≥16 and b65 y

Intervention

Primary outcome

Results

Venous TEA 25 (9%) vs 26 (9%) P = .90 Three intravenous 30-Day rFVIIa vs placebo doses of rFVIIa mortality 30-d mortality: 18% vs (200 g/kg at 0 h, 13%, P = .40 100 g/kg at 1 h and MOF day 30: 23% vs 3 h) or placebo 24%, P = .90 Mean (SD) allogeneic units to 24 h 11.2 (15) vs 16.8 (19.3); P = .09 Mean (SD) RBC units to 24 h. 4.5 (7.3) vs 6.2 (6.5); P = .11 Patients requiring MT 14 (30.4) vs 21 (52.5) P = .04 rFVIIa (blunt trauma) vs placebo 30-d mortality 30 d: 11% vs 11%, P = .93 MOF 30 day: 45% vs 53%, P = .06 Mean (SD) allogeneic units to 24 h. 17.1 (26.8) vs 20.7 (25.7); P = .03 Mean (SD) RBC units to 24 h, 6.9 (10.4) vs 8.1 (10.9) P = .04 Patients requiring MT 111 (50.2) vs 134 (54.3) P = .38 rFVIIa 3 doses (200 Number of Updated results of μg/100 μg/100 μg) RBC units previous publication given after 8th RBC transfused [28]: unit, then 1 and 3 h during 48-h rFVIIa vs placebo vs placebo period Estimated RBC posttrial drug reduction in all patients: Blunt 2.0 (95% CI, 0.0-4.6), P = .07 Penetrating 0.2 (95% CI, −0.9 to 2.4) P = .24

Abbreviations: ARDS, acute respiratory distress syndrome; ITT, intention to treat; MOF, multi-organ failure; PLT, platelet; rFVIIa, recombinant activated factor VII; RBC, red blood cell; RRR, relative risk reduction; SAE, serious adverse events; SD, standard deviation; TEA, thromboembolic adverse event.

In Patients with CB Requiring MT, What is the Effect of RBC Transfusion on Patient Outcomes? One RCT [5] and 2 SRs [11,12] comparing restrictive RBC transfusion with liberal RBC transfusion in adults with upper GI bleeding were identified (see Tables 3 and 4). Randomized Controlled Trials. A single-center RCT by Villanueva et al compared a restrictive transfusion approach using a hemoglobin (Hb) trigger of 7 g/dL and a target Hb of 7 to 9 g/dL with a liberal transfusion strategy using an Hb trigger of 9 g/dL and a Hb target of 9 to 11 g/dL. They reported significantly lower mortality at 45 days, fewer RBC units transfused, and fewer overall complications in the restrictive group (see Table 3) [5]. Although the study included patients with acute upper GI bleeding, it excluded patients with massive exsanguinating bleeding who could not be stabilized using crystalloids or colloids alone and requiring transfusion prior to randomization. These patients constituted 2.4% of the screened patients. Of those patients who were included in the study, 28% in the restrictive arm and 31% in the liberal transfusion group had hypovolemic shock (defined as systolic blood pressure b 100 mm Hg or heart rate greater than 100 beats/min) on

admission. The number of red cells transfused ranged from 0 to 36 U, with a mean of 3.7 ± 3.8 in the liberal group. The study also excluded patients with a recent history of trauma or surgery and lower GI bleeding, among other exclusion criteria. Systematic Reviews. An SR, performed before the Villaneuva et al study was published, identified 3 trials (with 126 patients) on the effects of RBC transfusion in adults with upper GI bleeding [11]. The SR reported a trend to higher mortality (relative risk, 5.4; 95% CI, 0.27-107.9) in the combined transfusion arm, more rebleeding (relative risk, 9.75; 95% CI, 1.33-71.33), and no difference in number of RBC transfusion, LOS, or hematocrit (see Table 4). However, the authors cautioned that small numbers and incomplete data limited their findings. A subsequent SR performed by Wang et al, this time comparing a restrictive transfusion strategy with a liberal transfusion strategy, included the same 3 trials from the previous SR as well as the RCT by Villaneuva [12]. This SR reported reduced mortality (odds ratio [OR], 0.52; 95% CI, 0.31- 0.87), hospital LOS (mean difference [MD], −0.17; 95% CI, −0.30 to −0.04), and amount of blood transfused (MD, −0.74; 95% CI, −1.15 to −0.32) in the restrictive transfusion arms, but no difference in rates of rebleeding (OR, 0.26; 95% CI, 0.03-2.10) [12].

Please cite this article as: McQuilten ZK, et al, Transfusion Interventions in Critical Bleeding Requiring Massive Transfusion: A Systematic Review, Transfus Med Rev (2015), http://dx.doi.org/10.1016/j.tmrv.2015.01.001

8

Z.K. McQuilten et al. / Transfusion Medicine Reviews xxx (2015) xxx–xxx

Table 4 Summary of findings from SRs that have reported meta-analyses Article

Main review question

Intervention

Effect of RBC transfusion on outcomes Wang Restrictive vs liberal Restrictive 2013 RBC transfusion in vs liberal upper GI bleeding RBC transfusion

Jairath 2010

Effects and safety of RBC transfusion in adults with upper GI bleeding

RBC transfusion

Patient population No. of studies No. of individuals in analyses

Primary outcome/s

Results (including meta-analyses)

Patients with upper GI bleeding.

6

Mortality: 966 Rebleed: 939 LOS: 916 RBC: 982

Mortality rebleeding

Patients N 16 with upper GI bleeding.

3

Mortality: 77 RBC and Hct: 93 Rebleeding: 50 LOS: 27

All-cause mortality, and adverse effects or complications of RBC

Comparison restrictive vs liberal transfusion Mortality: OR 0.52 (95% CI, 0.3 to 0.87), P = .01 Re-bleeding: OR 0.26 (95% CI, 0.03 to 2.10), P = .21 Hospital LOS: MD −0.17, (95% CI −0.30 to −0.04), P = .009 Amount of blood transfused: MD −0.74, (95% CI, −1.15 to −0.32), P = .0005 Comparison RBC transfusion vs no transfusion Overall mortality: RR 5.4 (95% CI 0.27 to 107.09). Hospital LOS: MD 0.32 (95% CI, −1.07 to 1.71) Number of RBC: MD: 1.07 (95% CI, −0.34 to 2.48) Haematocrit: MD 3.83 (95% CI, −2.03 to 9.6) Rebleeding: RR 9.75 (95% CI, 1.33 to 71.33)

Effect of FFP on outcomes Murad To review the 2010 benefits/harms of plasma

Plasma vs Adults receiving medical plasma management w/o plasma Lower plasma dose Lower plasma: RBC Effect of fibrinogen concentrate on outcomes Fibrinogen Bleeding patients Wikkelso Benefits/harms of fibrinogen concentrate 2013 concentrate vs placebo or usual treatment for bleeding patients Effect of rFVIIa on outcomes Simpson Effects of rFVIIa 2012 used prophylactically or therapeutically

642-2399 37 (24 RCT) All MT studies observational

Death, MI, stroke, ALI, MOF, blood loss, RBC transfusion requirements.

Mortality in MT Plasma/RBC ratios greater than 1:3 vs plasma/RBC ratio less than 1.3: OR 0.38 (95% CI, 0.24 to 0.60) ALI Plasma vs no plasma: OR 2.92 (95% CI, 1.99 to 4.29) MOF Plasma vs no plasma: OR 0.40 (95% CI, 0.26 to 0.60)

6a

248 (81 for mortality)

Overall mortality

Fibrinogen concentrate vs placebo or usual treatment Mortality: 2.6% vs 9.5%, RR 0.28 (95% CI, 0.03 to 2.33) Allogeneic transfusion: RR 0.47 (95% CI, 0.31 to 0.72) RBC transfusion: RR 0.81 (95% CI, 0.32 to 2.02) Adverse events (thrombotic episodes): RR 1.03 (95% CI, 0.27 to 3.97)

Mortality, Bleeding, RBC transfusion Adverse events

rFVIIa (therapeutic) vs placebo Overall mortality: RR 0.91 (95% CI, 0.78 to 1.06) RBC transfusion: MD −89 mL (95% CI, −264 to 87) rFVIIa (prophylactic) vs placebo Overall mortality: RR 1.04 (95% CI, 0.55 to 1.97) RBC transfusion: MD −261 mL (95% CI, 367 to −154) rFVIIa (all) vs placebo TEA: RR 1.18 (95% CI, 0.94 to 1.48) Arterial TEA (prophylactic and therapeutic) RR 1.45 (95% CI, 1.02 to 2.05)

Mortality Adverse events Quality of life

rFVIIa vs placebo Mortality 5 days: RR 0.95 (95% CI, 0.36 to 2.50), P = .16 Mortality 42 days: RR 1.01 (95% CI, 0.55 to 1.87), P = .14 Adverse events: RR 0.94 (95% CI, 0.84 to 1.04), P = .20 Serious events: RR 0.90 (95% CI, 0.73 to 1.09), P = .20 TEA: RR 0.80 (95% CI, 0.40 to 1.60), P = .20 rFVIIa vs placebo All TEA: OR 1.17 (95% CI, 0.94 to 1.47), P = .16 Arterial TEA: OR 1.68 (95% CI, 1.20 to 2.36), P = .003 Venous TEA: OR 0.93 (95% CI, 0.70 to 1.23), P = .61 Coronary events: OR 2.39 (95% CI, 1.39 to 4.09), P = .002 Cerebrovascular events: OR 1.27 (95% CI, 0.74 to 2.17), P = .39

rFVIIa

Patients without hemophilia

Therapeutic: 13 Prophylactic 16

MartiCaravajal 2012

Beneficial and harmful effects of rFVIIa

rFVIIa

Liver disease and upper GI bleeding

2

Therapeutic Mortality: 2856 RBC transfusion: 911 Prophylactic Mortality: 1219 RBC transfusion: 843 TEA combined: 4032 510

Levi 2010

TEA in rFVIIa

rFVIIa

Off label indications (bleeding patients and healthy volunteers)

35

4119

Arterial or venous TEA

Abbreviations: ALI, acute lung injury; MOF, multi-organ failure; NA, not applicable; rFVIIa, activated recombinant factor VII; risk ratio; TEA, thromboembolic adverse events. a Only studies with surgical patients identified.

Please cite this article as: McQuilten ZK, et al, Transfusion Interventions in Critical Bleeding Requiring Massive Transfusion: A Systematic Review, Transfus Med Rev (2015), http://dx.doi.org/10.1016/j.tmrv.2015.01.001

Z.K. McQuilten et al. / Transfusion Medicine Reviews xxx (2015) xxx–xxx

In Patients with CB Requiring MT, What is the effect of FFP, Cryoprecipitate, Fibrinogen Concentrate, Platelet Transfusion, and/or PCC on Patient Outcomes? Three SRs were identified that addressed this question (see Table 4). No new RCTs were identified. Systematic Reviews. Murad et al performed an SR on the benefits and harm of FFP compared with medical management, lower FFP dose, or ratio of FFP/RBC in adult patients [13]. Although the review identified both RCTs and observational studies, the studies identified that addressed the effect of FFP on mortality in trauma patients requiring MT were all observational. The meta-analysis found that plasma/RBC ratios greater than 1:3 were associated with reduced mortality in MT trauma patients (OR, 0.38; 95% CI, 0.24-0.60); however, because this analysis only included observational studies, the quality of the evidence was very low. Two SRs on the benefits and harm of fibrinogen concentrate in patients with bleeding were identified. Warmuth et al identified 2 RCT, both in surgical patients (urology and cardiac surgery) with a combined total of 41 participants, and 2 noncontrolled studies [14]. Wikkelso et al identified 6 RCTs with a total of 248 participants, all of which included elective surgical participants (cardiac, mixed population of cardiac surgery, major abdominal and spinal surgery, and urological surgery) [15]. The meta-analysis found no significant effect of fibrinogen concentrate on mortality (risk ratio [RR], 0.28; 95% CI, 0.03-2.33), a reduction in allogeneic transfusion (RR, 0.47; 95% CI, 0.31-0.72), and no effect on thrombotic episodes (RR, 1.03; 95% CI, 0.27-3.97) [15]. However, the authors cautioned that the included trials were of low quality with high risk of bias and underpowered to detect a difference in mortality or harm. Furthermore, there were no studies in bleeding patients in a nonelective surgical setting. In Patients with CB Requiring MT, What is the Effect of rFVIIa (Prophylaxis or Treatment) on Morbidity, Mortality, and Transfusion Rate? Three RCTs on rFVIIa, 2 of which were reporting results of the same RCT, and 6 SRs were identified. Randomized Controlled Trials. An analysis of an RCT first published in 2005 [28] of rFVIIa vs placebo in nonmilitary trauma patients with a primary outcome of number of RBC units transfused during the 48-hour period after the first dose of trial product was identified. Boffard et al reported no difference in RBC transfusion in either blunt (estimated reduction, 2.0; 95% CI, 0.0-4.6) or penetrating trauma (estimated reduction, 0.2; 95% CI, −0.9 to 2.4) [8]. Hauser et al reported a randomized study of rFVIIa vs placebo in civilian trauma with a primary outcome of 30-day mortality (the CONTROL trial) [9]. This study was ceased early (573 of a planned 1502 patients) due to a lower than expected mortality rate and high likelihood of futility in demonstrating the primary end point in the blunt trauma population. The analysis of the 573 patients randomized found no difference in 30day mortality, a reduction in number of allogeneic, RBC and FFP transfusions in blunt trauma patients in the rFVIIa arm, and a reduction in FFP transfusion but not other products in the penetrating trauma group. Dutton et al separately published detailed safety outcomes of the CONTROL trial and reported no difference in rates of serious adverse events or thromboembolic events, however a reduction in ARDS in the rFVIIa treatment group (3% vs 7.2%; P = .02) [10]. Systematic Reviews. The most recent SR and meta-analysis (Simpson et al, 2012) identified included 13 trials involving 2929 participants on therapeutic rFVIIa to treat established bleeding in a variety of clinical settings (including severe trauma, upper GI bleeding, dengue hemorrhagic fever, bleeding after hematopoietic stem cell transplant, and spontaneous and traumatic intracranial hemorrhage) [21]. The metaanalyses found no observed benefit of rFVIIa in terms of mortality or increased risk of thromboembolic events. However, when 16 trials examining the prophylactic use of rFVIIa were pooled with the 13

9

therapeutic rFVIIa trials, a significant increase in total arterial events was observed (relative risk, 1.46; 95% CI, 1.02-2.05) [21]. An SR on the use of rFVIIa in patients with upper GI bleeding, which included 2 trials with 510 participants, did not report any difference in mortality, adverse events, or thromboembolic events [19]. An SR that focused on the risk of venous and arterial thromboembolic events is shown in Table 4. The 3 other earlier SRs were descriptive reviews and do not provide any additional data. Discussion In the previously published Australian PBM guidelines for CB and MT, few recommendations could be made on the use of RBC and blood components due to the lack of high-quality evidence [1]. Although some additional studies were identified in this current review, particularly on the use of RBC in GI hemorrhage, overall little new evidence was found on transfusion interventions that could be used to update previous recommendations or inform practice. Dose, Timing, and Ratio (algorithm) of RBCs to Component Therapy In the PBM guideline, there was insufficient evidence to support or refute the use of specific ratios for RBCs to blood components [1]. However, the use of an MTP, which specifies ratios of components, was recommended. We did not identify any new evidence to support particular ratios of component therapy. One feasibility study was identified, which reported a trend to higher mortality in the fixed ratio group; however, it was underpowered for patient outcomes (including mortality), and, given the small numbers, its results should be interpreted with caution and not used to support changes in practice [7]. In addition, the lack of access to prethawed FFP, and therefore the delay in FFP transfusion in both intervention arms, also makes the study less generalizable to institutions with readily accessible prethawed FFP. The only SRs on the effects of ratios of RBC to components on patient outcomes included exclusively observational, and mostly retrospective, studies, thus limiting any conclusions that can be made [24–27]. Observational studies on ratios of components in trauma are at risk of survivor bias (ie, patients who survive longer are able to receive more component therapy compared with patients who die earlier) [29] and confounding. In addition, the majority of the observational studies are derived from military or civilian trauma settings, and therefore may not be applicable to other CB settings, such as surgical or GI bleeding. For example, major trauma is often accompanied by an acute coagulopathy [30], and therefore, transfusion requirements may differ from patients requiring MT in other settings. A larger trial powered to assess the efficacy and safety of specific ratios of component therapy in trauma (Pragmatic, Randomized Optimal Platelet and Plasma Ratios [PROPPR] trial) [31] has been completed and will provide more evidence on this question for trauma patients. Red Blood Cell Transfusion A restrictive transfusion practice has been shown to be at least as effective as standard practice in stable, nonbleeding critically ill patients [32], and a recent SR supported restrictive transfusion triggers in clinical settings other than CB [33]. A restrictive transfusion practice in patients with upper GI hemorrhage was associated with reduced mortality, RBC transfusion, and complications in RCT and SR [5,12]. However, the RCT was a single-center, unblinded study, and the exclusion of patients undergoing massive exsanguination (which included patients who despite crystalloid and colloid fluids required transfusion prior to randomization) limits the generalizability of the findings to patients with CB requiring MT. In this RCT, all patients underwent emergency endoscopic evaluation within 6 hours, with therapy instituted based on these findings. It is unclear whether the benefits of a restrictive transfusion strategy would be evident if patients had not received early endoscopy. Also of note, there was increased protocol violation in the restrictive transfusion group [5]. Furthermore, the vast majority of patients in the SR were

Please cite this article as: McQuilten ZK, et al, Transfusion Interventions in Critical Bleeding Requiring Massive Transfusion: A Systematic Review, Transfus Med Rev (2015), http://dx.doi.org/10.1016/j.tmrv.2015.01.001

10

Z.K. McQuilten et al. / Transfusion Medicine Reviews xxx (2015) xxx–xxx

from the single RCT, and the other 3 studies were of small size and low quality with a transfusion threshold that varied across studies. Therefore, although these results support the use of a restrictive transfusion approach in patients with upper GI bleeding (but not massive exsanguination), further studies are required to confirm these results, including the optimal transfusion trigger. A feasibility cluster randomized controlled study (the TRIGGER trial) has recently been completed [34] and may lead to a more definitive trial in this area.

participants and staff to the intervention status. This is often difficult to mitigate; however, if possible, outcome assessors should be blinded to participant intervention status. It is also worth noting that 3 of the RCTs received industry support or sponsorship [8–10], which may have implications on study design, how the study is conducted, and how results are published and distributed.

FFP, Cryoprecipitate, Fibrinogen Concentrate, PCC, and/or Platelet Transfusion

Despite the major role of transfusion in patients with CB and serious hemorrhage, there remains a lack of evidence from high-quality clinical studies to inform practice, particularly with regard to the effect of RBCs and component therapies on patient outcomes, and optimal component ratios. The CRASH2 trial demonstrated that large RCTs can be successfully undertaken in trauma patients, and the feasibility study of fixed ratio compared with standard therapy identified in our review [7], and completion of the PROPPR trial [31], further supports the notion that large prospective studies are possible. Similarly, in patients with upper GI bleeding, large RCTs have been shown to be feasible [5,34]. Further clinical trials to address these gaps in patients with CB requiring MT across different clinical contexts are needed.

Evidence on the use of FFP on outcomes, as for the ratios of blood component therapy, was of very low quality, and therefore, results should be interpreted with caution. No studies on platelet transfusion or PCC were identified. There were no studies on fibrinogen replacement with cryoprecipitate, and although 2 SRs on the use of fibrinogen concentrate in bleeding were identified, they only included trials in elective surgical patients. Low fibrinogen has been shown to be independently associated with higher mortality in trauma patients [35]; however, there are no RCTs on fibrinogen supplementation, type of product (cryoprecipitate, fibrinogen concentrate), or optimal dose in CB, including trauma. rFVIIa The available evidence confirms that the off-label use of rFVIIa in CB or trauma confers no benefit to mortality outcomes. In the SR by Simpson et al, there was a modest reduction in red cell transfusion requirements and blood loss; however, this effect may have been overestimated as some of the negatively weighted studies were not able to be incorporated into the meta-analysis. This possible benefit was offset by a trend toward an increased risk of thromboembolic events, and a significantly increased risk of arterial thromboembolic events when both prophylactic and therapeutic studies were considered. At present, the evidence does not support the routine use of rFVIIa as part of the treatment algorithm in the management of CB or as part of a MTP. Limitations of the Review Although the original PBM guideline addressed the use of MTP and nontransfusion interventions, our review was focused only on transfusion interventions. We did not include pharmacological agents, such as tranexamic acid, which has been shown in the large CRASH-2 RCT to reduce mortality in trauma patients [36]. We also did not include studies that used different methods to direct transfusion interventions, such as thromboelastography. Although a comprehensive literature search strategy was used for the review, an updated search prior to publication using the Transfusion Evidence Library was performed, and therefore we cannot exclude the possibility that we may have inadvertently missed relevant studies published more recently. Impact of Methodological Assessment Four of the SRs, which addressed RBC triggers in GI bleeding [11,12] and rFVIIa use [19,21], were assessed as being of high methodological quality, meaning that the conclusions derived from these reviews were reliable. Two of the SRs were of relatively poor quality [16,18], whereas the remaining SRs were limited by the inclusion of non-RCTs [13,23] or lack of RCTs relevant to the clinical questions for this SR [14,15]. One of the RCTs (which contributed 2 published studies) had low risk of bias in all facets of assessment, which would mean that the findings from these trials were reliable [9,10]. Other studies were limited by a lack of blinding of study personnel, study participants, and outcome assessors [5–7]. This lack of blinding may lead to increased risk of detection bias. Clinical trials in transfusion medicine often face challenges with trial design with respect to blinding, as it is often not possible to blind

Evidence Gaps

Conclusion This updated review of CB requiring MT identified 19 new studies (6 RCTs and 13 SRs) addressing 4 main transfusion interventions. These studies provided limited new evidence that can be used to inform practice in this area and substantial evidence gaps remain, particularly with regard to the effect of component therapies, including ratio of RBC to component therapies, on patient outcomes. Clinical trials to address these questions are required. Conflict of Interest Zoe McQuilten and Erica Wood are employed by Monash University, whose Transfusion Research Unit has received financial support from Alexion, Amgen, Bayer, Celgene, CSL Behring, Janssen-Cilag, Takeda, Novartis, Australian Red Cross Blood Service, New Zealand Blood Service, Department of Health Victoria (Australia), NBA (Australia) and Myeloma Foundation of Australia. None of these funding sources had any involvement the design or conduct of this review. Acknowledgments We thank Jennifer Roberts from the NBA and the authors of the NBA Patient Blood Management Guidelines: Module 1—Critical Bleeding/Massive Transfusion. We thank Dr Carolyn Doree, Information Specialist, NHS Blood and Transplant Systematic Reviews Initiative, for assistance with the literature search. ZM received funding support from the Australian NHMRC Centre of Research Excellence for Patient Blood Management in Critical Illness and Trauma (APP1049071). References [1] National Blood Authority (NBA). Patient blood management guidelines: module 1—critical bleeding/massive transfusion. Canberra, Australia: NBA; 2011[Available at: http://www.blood.gov.au/pbm-module-1. Accessed 17 October 2014]. [2] National Health and Medical Research Council (NHMRC). A guideline to the development, evaluation and implementation of clinical practice guidelines. Canberra, Australia: NHMRC; 1999[Available at: https://www.nhmrc.gov.au/guidelines/publications/cp30. Accessed 17 October 2014]. [3] Higgins JPT, Green S, editors. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration; 2011 [Available from www.cochrane-handbook.org]. [4] Shea BJ, Grimshaw JM, Wells GA, et al. Development of AMSTAR: a measurement tool to assess the methodological quality of systematic reviews. BMC Med Res Methodol 2007;7:10. [5] Villanueva C, Colomo A, Bosch A, et al. Transfusion strategies for acute upper gastrointestinal bleeding. N Engl J Med 2013;368(1):11–21.

Please cite this article as: McQuilten ZK, et al, Transfusion Interventions in Critical Bleeding Requiring Massive Transfusion: A Systematic Review, Transfus Med Rev (2015), http://dx.doi.org/10.1016/j.tmrv.2015.01.001

Z.K. McQuilten et al. / Transfusion Medicine Reviews xxx (2015) xxx–xxx [6] Cotton BA, Podbielski J, Camp E, et al. A randomized controlled pilot trial of modified whole blood versus component therapy in severely injured patients requiring large volume transfusions. Ann Surg 2013;258(4):527–32 [discussion 32–3]. [7] Nascimento B, Callum J, Tien H, et al. Effect of a fixed-ratio (1:1:1) transfusion protocol versus laboratory-results-guided transfusion in patients with severe trauma: a randomized feasibility trial. CMAJ 2013;185(12):E583–9. [8] Boffard KD, Choong PI, Kluger Y, et al. The treatment of bleeding is to stop the bleeding! Treatment of trauma-related hemorrhage. Transfusion 2009;49(Suppl 5):240S–7S. [9] Hauser CJ, Boffard K, Dutton R, et al. Results of the CONTROL trial: efficacy and safety of recombinant activated Factor VII in the management of refractory traumatic hemorrhage. J Trauma 2010;69(3):489–500. [10] Dutton RP, Parr M, Tortella BJ, et al. Recombinant activated factor VII safety in trauma patients: results from the CONTROL trial. J Trauma 2011;71(1):12–9. [11] Jairath V, Hearnshaw S, Brunskill SJ, et al. Red cell transfusion for the management of upper gastrointestinal haemorrhage. Cochrane Database Syst Rev 2010;9:CD006613. [12] Wang J, Bao YX, Bai M, et al. Restrictive vs liberal transfusion for upper gastrointestinal bleeding: a meta-analysis of randomized controlled trials. World J Gastroenterol 2013;19(40):6919–27. [13] Murad MH, Stubbs JR, Gandhi MJ, et al. The effect of plasma transfusion on morbidity and mortality: a systematic review and meta-analysis. Transfusion 2010;50(6): 1370–83. [14] Warmuth M, Mad P, Wild C. Systematic review of the efficacy and safety of fibrinogen concentrate substitution in adults. Acta Anaesthesiol Scand 2012;56(5):539–48. [15] Wikkelso A, Lunde J, Johansen M, et al. Fibrinogen concentrate in bleeding patients. Cochrane Database Syst Rev 2013;8:CD008864. [16] Aguado Romeo MJ, Molina Linde JM, Villegas Portero R. Appropriate use of recombinant activated factor VII in nonhaemophilic patients. Informes de Evaluacion de Tecnologias Sanitarias AETSA 2007/04. Seville: Andalusian Agency for Health Technology Assessment (AETSA); 2009. [17] Birchall J, Stanworth SJ, Duffy MR, et al. Evidence for the use of recombinant factor VIIa in the prevention and treatment of bleeding in patients without hemophilia. Transfus Med Rev 2008;22(3):177–87. [18] Levi M, Levy JH, Andersen HF, et al. Safety of recombinant activated factor VII in randomized clinical trials. N Engl J Med 2010;363(19):1791–800. [19] Marti-Carvajal AJ, Karakitsiou DE, Salanti G. Human recombinant activated factor VII for upper gastrointestinal bleeding in patients with liver diseases. Cochrane Database Syst Rev 2012;3:CD004887. [20] Nishijima DK, Zehtabchi S. Evidence-based emergency medicine/critically appraised topic. The efficacy of recombinant activated factor VII in severe trauma. Ann Emerg Med 2009;54(5):737–744.e1. [21] Simpson E, Lin Y, Stanworth S, et al. Recombinant factor VIIa for the prevention and treatment of bleeding in patients without haemophilia. Cochrane Database Syst Rev 2012;3:CD005011. [22] Curry N, Hopewell S, Doree C, et al. The acute management of trauma hemorrhage: a systematic review of randomized controlled trials. Crit Care 2011;15(2):R92. [23] Curry N, Stanworth S, Hopewell S, et al. Trauma-induced coagulopathy—a review of the systematic reviews: is there sufficient evidence to guide clinical transfusion practice? Transfus Med Rev 2011;25(3):217–231.e2. [24] Zehtabchi S, Nishijima DK. Impact of transfusion of fresh-frozen plasma and packed red blood cells in a 1:1 ratio on survival of emergency department patients with severe trauma. Acad Emerg Med 2009;16(5):371–8.

11

[25] Phan HH, Wisner DH. Should we increase the ratio of plasma/platelets to red blood cells in massive transfusion: what is the evidence? Vox Sang 2010;98(3 Pt 2):395–402. [26] Rajasekhar A, Gowing R, Zarychanski R, et al. Survival of trauma patients after massive red blood cell transfusion using a high or low red blood cell to plasma transfusion ratio. Crit Care Med 2011;39(6):1507–13. [27] Johansson PI, Oliveri RS, Ostrowski SR. Hemostatic resuscitation with plasma and platelets in trauma. J Emerg Trauma Shock 2012;5(2):120–5. [28] Boffard KD, Riou B, Warren B, et al. Recombinant factor VIIa as adjunctive therapy for bleeding control in severely injured trauma patients: two parallel randomized, placebo-controlled, double-blind clinical trials. J Trauma 2005;59(1):8–15 [discussion 15–8]. [29] Snyder CW, Weinberg JA, McGwin Jr G, et al. The relationship of blood product ratio to mortality: survival benefit or survival bias? J Trauma 2009;66(2):358–62 [discussion 62–4]. [30] Frith D, Davenport R, Brohi K. Acute traumatic coagulopathy. Curr Opin Anaesthesiol 2012;25(2):229–34. [31] Baraniuk S, Tilley BC, Del Junco DJ, et al. Pragmatic Randomized Optimal Platelet and Plasma Ratios (PROPPR) Trial: design, rationale and implementation. Injury 2014; 45(9):1287–95. [32] Hebert PC, Wells G, Blajchman MA, et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. Transfusion Requirements in Critical Care Investigators, Canadian Critical Care Trials Group. N Engl J Med 1999;340(6): 409–17. [33] Carson JL, Carless PA, Hebert PC. Transfusion thresholds and other strategies for guiding allogeneic red blood cell transfusion. Cochrane Database Syst Rev 2012;4: CD002042. [34] Gray A, Jairath V, Kahan B, et al. Restrictive versus liberal blood transfusion for acute upper gastrointestinal bleeding (trigger): pragmatic, cluster randomised, feasibility trial. Emerg Med J 2014;31(9):780. [35] Rourke C, Curry N, Khan S, et al. Fibrinogen levels during trauma hemorrhage, response to replacement therapy, and association with patient outcomes. J Thromb Haemost 2012;10(7):1342–51. [36] Shakur H, Roberts I, Bautista R, et al. Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage (CRASH-2): a randomised, placebo-controlled trial. Lancet 2010;376(9734):23–32. [37] Schulz KF, Chalmers I, Hayes RJ, et al. Empirical evidence of bias. Dimensions of methodological quality associated with estimates of treatment effects in controlled trials. JAMA 1995;273(5):408–12. [38] Juni P, Altman DG, Egger M. Systematic reviews in health care: assessing the quality of controlled clinical trials. BMJ 2001;323(7303):42–6. [39] Higgins JPT, Green S, editors. Cochrane Handbook for Systematic Reviews of Interventions 4.2.6 [updated September 2006]; 2006 [Available from http://www.cochrane. org/sites/default/files/uploads/Handbook4.2.6Sep2006.pdf Accessed 17 October 2014]. [40] Guyatt GH, Sackett DL, Cook DJ. Users' guides to the medical literature. II. How to use an article about therapy or prevention. A. Are the results of the study valid? Evidence-Based Medicine Working Group. JAMA 1993;270(21):2598–601. [41] Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996;17(1):1–12. [42] Moher D, Pham B, Jones A, et al. Does quality of reports of randomised trials affect estimates of intervention efficacy reported in meta-analyses? Lancet 1998; 352(9128):609–13.

Please cite this article as: McQuilten ZK, et al, Transfusion Interventions in Critical Bleeding Requiring Massive Transfusion: A Systematic Review, Transfus Med Rev (2015), http://dx.doi.org/10.1016/j.tmrv.2015.01.001