Plasma transfusion strategies for critically ill patients (Review) Karam O, Tucci M, Combescure C, Lacroix J, Rimensberger PC
This is a reprint of a Cochrane review, prepared and maintained by The Cochrane Collaboration and published in The Cochrane Library 2013, Issue 12 http://www.thecochranelibrary.com
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
TABLE OF CONTENTS HEADER . . . . . . . . . . . . . . . . . . ABSTRACT . . . . . . . . . . . . . . . . . PLAIN LANGUAGE SUMMARY . . . . . . . . . BACKGROUND . . . . . . . . . . . . . . . OBJECTIVES . . . . . . . . . . . . . . . . METHODS . . . . . . . . . . . . . . . . . RESULTS . . . . . . . . . . . . . . . . . . Figure 1. . . . . . . . . . . . . . . . . DISCUSSION . . . . . . . . . . . . . . . . AUTHORS’ CONCLUSIONS . . . . . . . . . . ACKNOWLEDGEMENTS . . . . . . . . . . . REFERENCES . . . . . . . . . . . . . . . . CHARACTERISTICS OF STUDIES . . . . . . . . DATA AND ANALYSES . . . . . . . . . . . . . CONTRIBUTIONS OF AUTHORS . . . . . . . . DECLARATIONS OF INTEREST . . . . . . . . . DIFFERENCES BETWEEN PROTOCOL AND REVIEW INDEX TERMS . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
1 1 2 2 3 3 6 7 8 8 9 9 10 12 12 12 12 12
i
[Intervention Review]
Plasma transfusion strategies for critically ill patients Oliver Karam1 , Marisa Tucci2 , Christophe Combescure3 , Jacques Lacroix2 , Peter C Rimensberger1 1 Pediatric
Critical Care Unit, Geneva University Hospital, Geneva, Switzerland. 2 Pediatric Intensive Care Unit, CHU Sainte Justine, University of Montreal, Montreal, Canada. 3 Center for Clinical Research, Geneva University Hospital, Geneva, Switzerland
Contact address: Oliver Karam, Pediatric Critical Care Unit, Geneva University Hospital, 6 rue Willy Donzé, Geneva, 1211, Switzerland. [email protected]. Editorial group: Cochrane Injuries Group. Publication status and date: New, published in Issue 12, 2013. Review content assessed as up-to-date: 15 August 2013. Citation: Karam O, Tucci M, Combescure C, Lacroix J, Rimensberger PC. Plasma transfusion strategies for critically ill patients. Cochrane Database of Systematic Reviews 2013, Issue 12. Art. No.: CD010654. DOI: 10.1002/14651858.CD010654.pub2. Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
ABSTRACT Background Although plasma transfusions are frequently prescribed for critically ill patients, most clinical uses of plasma are not supported by evidence. Plasma transfusions do not seem to correct mild coagulation abnormalities based on international normalised ratio (INR) testing, but they seem to be independently associated with worse clinical outcomes in non-massively bleeding patients. Current recommendations on plasma transfusion strategies advocate limiting plasma transfusions to patients who are actively bleeding or who are at risk of bleeding and concomitantly have moderately abnormal coagulation tests. Objectives To determine whether use of a restrictive versus a liberal plasma transfusion threshold affects mortality or morbidity in critically ill patients, and to assess the clinical effects of different plasma transfusion thresholds in critically ill patients. Search methods A search for studies was run on 15 August 2013. We searched the Cochrane Injuries Group Specialised Register, the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE(R) Ovid, MEDLINE(R) Ovid In-Process & Other Non-Indexed Citations, MEDLINE(R) Ovid Daily and OLDMEDLINE(R) Ovid, EMBASE Classic + EMBASE (Ovid SP), reference lists, related websites and trial registries and checked lists of references. Selection criteria Randomised clinical trials that assessed the effects of two plasma transfusion strategies, using a restrictive and a liberal threshold of at least one coagulation test, in critically ill participants. Data collection and analysis Two review authors independently extracted data and assessed trial quality using the standard methods of the Cochrane Handbook for Systematic Reviews of Interventions. Main results Of 843 references identified by our search, none of the trials satisfied our predefined inclusion criteria. No studies are included in this review. Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
1
Authors’ conclusions This review highlights the lack of evidence that is available to guide plasma transfusions in critically ill patients. Randomised controlled trials are needed to determine the appropriate plasma transfusion strategy in critically ill patients.
PLAIN LANGUAGE SUMMARY Plasma transfusion strategies for critically ill patients Plasma transfusions are a frequently used treatment for critically ill patients, and they are usually prescribed to correct abnormal coagulation tests and to prevent or stop bleeding. Although plasma transfusions are frequently prescribed for critically ill patients, some of the reasons for their use are not supported by evidence from medical research. Some research has found an association of plasma transfusions with worse outcomes, and other studies have suggested that plasma transfusions do not help to return blood to its normal thickness. It is important to identify which patients, if any, may benefit from plasma transfusions. We searched the medical literature for reports of randomised controlled trials of plasma transfusion in people with a critical illness. Upon reading 843 study records, we found that currently no randomised trials have compared different plasma transfusion strategies. Such studies are needed to help physicians prescribe plasma transfusion only for patients who will benefit from them. Two main obstacles are associated with designing a randomised controlled trial on this topic, but they can be overcome. Currently, no consensus has been reached on the best way to determine whether a patient requires a plasma transfusion. Another obstacle is the difficulty involved in identifying important bleeding, as one must consider not only the amount of bleeding but also its location, the body’s own response to the bleeding (haemodynamic response) and the need for red blood cell transfusions.
BACKGROUND Plasma transfusions have been used since 1941 (Schmidt 2012). In 2008, 4,484,000 plasma units were transfused in patients in the United States (National Survey 2009). More than 10% of critically ill patients, both adults and children, receive a plasma transfusion during their hospital stay, making plasma transfusion a frequently used treatment modality (Luk 2002; Puetz 2012; Stanworth 2011). In current practice, plasma transfusions are widely used in critical care; they are administered most often to correct abnormal coagulation tests or to prevent bleeding (Vlaar 2009). In situations in which active bleeding is attributable to a coagulation factor deficiency, plasma transfusions can constitute a life-saving intervention by improving coagulation factor deficit (Stanworth 2007), especially in patients requiring massive transfusion (Zink 2009). However, in less dramatic situations, some researchers have shown that transfusing plasma often fails to correct mild coagulation abnormalities (Abdel-Wahab 2006; Holland 2006). Furthermore, currently available evidence derived from several retrospective studies in adults suggests that plasma transfusions are associated with increased odds of acute lung injury (Dara 2005; Khan 2007; van Stein 2010), nosocomial infection (Sarani 2008)
and transfusion-associated circulatory overload (Narick 2012). In critically ill adults with haemorrhagic shock, a prospective study reported an independent association between plasma transfusions and worse outcomes such as multiple organ failure and acute respiratory distress syndrome (ARDS) (Watson 2009). In critically ill children, a prospective observational study also reported an independent association between plasma transfusions and multiple organ failure and nosocomial infection (Karam 2013).
Description of the intervention Transfusion strategies can be viewed as liberal or restrictive. A liberal transfusion strategy implies that blood products are transfused for lower thresholds than with a restrictive transfusion strategy. Transfusion strategies have been evaluated for red blood cells (RBCs) (Carson 2012). RBC transfusions are a common treatment in both adult and paediatric critical care settings (Bateman 2008), as it is hypothesised that critically ill patients could benefit from RBC transfusions to enhance oxygen delivery. Nevertheless, as for plasma, it has been recognised that RBC transfusions are associated with increased morbidity (Bateman 2008; Gong 2005) and mortality (Kneyber 2007). Adequate RBC transfusion strate-
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
2
gies have been evaluated in critically ill patients, in whom a restrictive haemoglobin strategy, as compared with a liberal strategy, decreased transfusion requirements without increasing adverse outcomes (Hébert 1999; Lacroix 2007; Villanueva 2013).
Why it is important to do this review Current recommendations on plasma transfusion strategies advocate limiting plasma transfusions to patients who are actively bleeding or are at risk of bleeding and who concomitantly have moderately abnormal coagulation tests (Crosby 1997; Liumbruno 2009; O’Shaughnessy 2004). However, these recommendations fail to cite randomised controlled trials (RCTs). Plasma transfusions are a frequent treatment in critical care settings, although they seem to be associated with unfavourable outcomes if not administered wisely. No systematic review of RCTs has yet been published to guide physicians in selecting appropriate plasma transfusion strategies for use in critically ill patients.
Types of interventions The intervention in the experimental group should follow a protocol with one defined threshold required for plasma transfusion. The intervention in the control group should consist of another threshold or standard treatment. We planned to extract the numbers of plasma transfusions from the original studies to describe the intervention in experimental and control groups. By definition, we named “liberal” a transfusion strategy based on a threshold allowing for more transfusions, and “restrictive” a transfusion strategy based on a threshold allowing for fewer transfusions. We planned to extract data on co-interventions (i.e. transfusion of red blood cells and platelets, fibrinogen or other specific coagulation factors, steroids, need for surgery and need for extracorporeal life support) related to the main outcome.
Types of outcome measures
Primary outcomes
• All-cause mortality, at the end of the follow-up period in each trial.
OBJECTIVES To determine whether use of a restrictive versus a liberal plasma transfusion threshold affects mortality or morbidity in critically ill patients, and to assess the clinical effects of different plasma transfusion thresholds in critically ill patients.
Secondary outcomes
• • • • •
Nosocomial infections. Multiple-organ dysfunction (new or progressive). Volume of blood lost. Transfusion of RBCs and platelets. Transfusion reactions.
METHODS Search methods for identification of studies Criteria for considering studies for this review
The search was not restricted by date, language or publication status.
Types of studies
Electronic searches
We considered all randomised clinical trials for inclusion (irrespective of language, blinding, publication status or sample size) if two thresholds (restrictive and liberal) and at least one coagulation test (international normalised ratio (INR), activated partial thromboplastin time (aPTT), fibrinogen or thromboelastography) were compared. Trials on plasmapheresis and plasma exchanges were not included.
The Cochrane Injuries Group Trials Search Co-ordinator searched the following. • Cochrane Injuries Group Specialised Register (15 August 2013). • Cochrane Central Register of Controlled Trials (CENTRAL) (2013, Issue 7 of 12). • MEDLINE(R) Ovid, MEDLINE(R) Ovid In-Process & Other Non-Indexed Citations, MEDLINE(R) Ovid Daily and OLDMEDLINE(R) Ovid (1946 to 15 August 2013). • EMBASE Classic + EMBASE (Ovid SP) (1947 to 2013 week 32).
Types of participants All critically ill patients of all ages (neonates, children and adults) requiring plasma transfusion.
Search strategies are listed in Appendix 1.
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
3
Searching other resources On 20 August 2013, we searched the following databases for ongoing and unpublished trials. • The Clinical Trials Search Portal of the World Health Organization (http://apps.who.int/trialsearch/). • The Clinical Trials Registry of the U.S. National Institutes of Health (http://clinicaltrials.gov). We searched the following websites on 20 August 2013. • Transfusion Evidence Library (http:// transfusionguidelines.org). • http://www.evidence.nhs.uk/. • American Association of Blood Banks (http:// www.aabb.org). In addition, we checked the reference lists of identified material for further relevant trials.
Data collection and analysis
Selection of studies Two review authors (OK and MT) independently screened titles and abstracts in the literature search. Authors of primary studies that seemed eligible for inclusion would have been contacted for clarification of data and to obtain missing information. Study eligibility was assessed against defined inclusion criteria. We planned to assess separately methodological quality, including participant randomisation; blinding of participants, investigators and outcome assessors; intention-to-treat analysis; and completeness of follow-up. We planned to request the consensus of an additional review author (JL) to resolve any disagreement between the primary review authors. Data extraction and management Two review authors (OK and MT) would have independently extracted data from the selected trials using a standardised data extraction form. We planned to consult JL in the event of any disagreements, which we would have resolved through discussion and consensus. We planned to independently extract the following data. • Year and language of publication. • Country in which the trial was conducted. • Year of conduct of the trial. • Inclusion and exclusion criteria. • Sample size. • Plasma transfusion thresholds (INR, aPTT, fibrinogen, thromboelastography) for the restrictive and liberal strategy groups. • Outcomes (in each group).
• Methodological quality. We planned to seek any unclear or missing information by contacting the authors of the individual trials. Had there been any doubt whether trial reports shared the same participants-completely or partially (by identifying common authors and centres)-the authors of the trials would have been contacted to clarify whether the trial had been duplicated. We planned to resolve any differences in opinion through discussion or arbitration with a third review author (JL).
Assessment of risk of bias in included studies We planned that two review authors (OK and MT) would independently assess each trial for risk of bias. In cases of disagreement, consensus would have been reached by discussion with a third review author (JL). Risk of bias for RCTs that met eligibility criteria would have been assessed using The Cochrane Collaboration tool for assessment of risk of bias provided in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We planned to assess the following areas. Was the allocation sequence adequately generated? • Criteria for a judgement of ’low risk for bias’: Random sequence generation was performed by using a random number table, a computer random number generator or other accepted method. • Criteria for the judgement of ’unclear risk of bias’: The study was described as randomised, but the method for sequence generation was not described, or the method for sequence generation was inadequately described to permit assessment of risk of bias. • Criteria for the judgement of ’high risk of bias’: The method for sequence generation was described but was not random. Quasi-randomised trials, which base allocation on factors such as date of birth, day of the week or medical record number, will be considered as having a high risk of bias. Was allocation adequately concealed? • Criteria for a judgement of ’low risk for bias’: Allocation concealment was described and would not allow investigators enrolling participants to foresee participant assignment. Acceptable methods of allocation concealment include central allocation and sequentially numbered, opaque, sealed envelopes. • Criteria for the judgement of ’unclear risk of bias’: The method for allocation concealment was not described, or the method for allocation concealment was inadequately described to permit assessment of risk of bias. • Criteria for the judgement of ’high risk of bias’: The method for allocation concealment was described but might allow investigators to foresee participant assignment. Studies using an open random allocation schedule, assignment envelopes without appropriate safeguards and alternation or rotation or any
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
4
other unconcealed method would be considered as having a high risk of bias. Was knowledge of the allocated intervention concealed from outcome assessors? • Criteria for a judgement of ’low risk for bias’: Outcome assessment was not blinded, but the outcome and the outcome measurement were not likely to be influenced by lack of blinding; or outcome assessment was blinded. • Criteria for the judgement of ’unclear risk of bias’: The study contained insufficient information to permit judgement of the presence or absence of blinding. • Criteria for the judgement of ’high risk of bias’: Outcome assessment was not blinded, and the outcome or the outcome measurement was likely to be influenced by lack of blinding. Given that blinding of study physicians to the transfusion strategy would not be possible, we do not anticipate that any studies will include blinding of study physicians. Blinding of study participants is likely to occur because of the severity of their injuries and their physiological derangement, but this is unlikely to influence outcomes. This potential source of bias will be assessed for each outcome. Were incomplete outcome data adequately addressed? • Criteria for a judgement of ’low risk of bias’: No participants were lost to follow-up or withdrew from the study post-randomisation; loss to follow-up or withdrawal from the study was balanced across groups and was likely unrelated to true outcomes. • Criteria for the judgement of ’unclear risk of bias’: Reporting of attrition or exclusions was insufficient to permit judgement regarding the risk of bias. • Criteria for the judgement of ’high risk of bias’: Participants were lost to follow-up or withdrew from the study for reasons likely related to true outcomes; outcomes were not assessed using an intention-to-treat analysis and with substantial departure of the intervention received from that assigned at randomisation; the number of participants lost to follow-up might induce clinically relevant bias in the effect estimate or the effect size. This potential source of bias will be assessed for each outcome. Are reports of the study free of the suggestion of selective outcome reporting? • Criteria for a judgement of ’low risk of bias’: In studies with an available study protocol, all of the study’s prespecified outcomes were reported in the prespecified manner; in studies for which the study protocol is not available, published reports included all expected outcomes, including those that were prespecified (convincing text of this nature may be uncommon). • Criteria for the judgement of ’unclear risk of bias’: Information was insufficient to allow assessment of whether the risk of selective outcome reporting was present. • Criteria for the judgement of ’high risk of bias’: In studies with an available study protocol, not all of the study’s
prespecified outcomes were reported, or outcomes were not reported in the prespecified manner, or outcomes were changed during the course of the study. Was the study free of other significant sources of bias? • Criteria for a judgement of ’low risk of bias’: Studies appeared to be free of other sources of bias. • Criteria for the judgement of ’unclear risk of bias’: Risk of bias may be present, information was insufficient to allow assessment of whether an important risk of bias exists, or the rationale or evidence that an identified problem will introduce bias was insufficient. • Criteria for the judgement of ’high risk of bias’: At least one important source of bias was present, such as early stoppage of the study due to a data-dependent process, extreme baseline imbalance across study groups or potential bias due to source of study funding.
Measures of treatment effect The meta-analysis was planned to study the relationship between a plasma transfusion threshold and an outcome. The arms of the RCTs would have been labelled “Restrictive threshold” and “Liberal threshold”. For dichotomous outcomes, we planned to express results as risk ratios (RRs) with 95% confidence intervals (CIs). For continuous outcomes, we planned to express results as mean differences (MDs) (or standardised mean differences (SMDs) if different scales of measurement had been used) with 95% CIs.
Dealing with missing data If the main data (primary and secondary outcomes of the metaanalysis) were not provided in the report of a study, we planned to contact the study authors. If data could not have been obtained, a sensitivity analysis would have been performed by including the studies with missing data and extrapolating the data considering extreme scenarios. The purpose of this sensitivity analysis would have been to check the robustness of the conclusion while taking into account missing data.
Assessment of heterogeneity We planned to assess the clinical diversity (clinical heterogeneity) of the included studies. We would have undertaken subgroup analysis to examine possible clinical variability when the I2 statistic was 50% or less but heterogeneity remained statistically significant. We planned to analyse statistical heterogeneity using the Chi2 test on N-1 degrees of freedom, with an alpha of 0.1 used for statistical significance and with the I2 statistic (Higgins 2011). I2 ranges between 0 to 25%, 26 to 50%, 51 to 75%, and 76 to 100% corresponded to categories of low, medium, high and very high levels of statistical heterogeneity, respectively. We would have set an I2 threshold of greater than 76% to indicate substantial variation
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
5
across trials due to heterogeneity. We would have used a fixedeffect model if we would had found low to high heterogeneity between trials. We would have used a random-effects model if very high heterogeneity had existed between trials (Higgins 2011). We planned to test for homogeneity between trials for each outcome using Cochran’s Q statistic, with P value less than or equal to 0.10.
To assess clinical heterogeneity, we planned to perform the following subgroup analyses. • Coagulation tests: INR, aPTT, fibrinogen, thromboelastography. • Admission categories: cardiac surgery participants, trauma participants, burn participants. • Severity of the condition: severity at admission, severity of bleeding. • Age of the participant.
Assessment of reporting biases We would have analysed possible publication bias using a funnel plot if at least 10 studies were available for each outcome.
Data synthesis As plasma transfusion strategies are likely to be different among neonates, children and adults, meta-analysis would have been performed separately for each of these age groups. For the primary outcome, as well as for each of the secondary outcomes, we planned to present in plain words the number of trials reporting the outcome under evaluation, the effect of the intervention and the heterogeneity noted among those trials. We planned to analyse dichotomous data using an RR and to analyse continuous data using weighted MDs (or SMDs accordingly). A Mantel-Haenszel model would have been used, and pooled estimates would have been reported with a 95% CI. We planned to present the intervention effect using a forest plot to display effect estimates and confidence intervals for individual studies.
Subgroup analysis and investigation of heterogeneity
Sensitivity analysis A sensitivity analysis was planned to check the robustness of the conclusions with regard to missing data. We had also planned to compare trials with low risk of bias (adequate generation of allocation sequence and follow-up) versus trials with high risk of bias (one or more of the two components inadequate or unclear). If a study would have been visibly different to others, we would have conducted another sensitivity analysis by temporarily removing that study to check whether the conclusions were similar.
RESULTS
Description of studies We identified no published or ongoing trials that met our inclusion criteria. Results of the search The search identified 843 records (Figure 1).
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
6
Figure 1. Study flow diagram.
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
7
Included studies We identified no RCTs that evaluated plasma transfusion strategies according to predetermined coagulation test thresholds.
Excluded studies We identified the protocol for an RCT on plasma transfusion strategies, as designed by Müller et al (Müller 2011). The study authors described a non-inferiority randomised trial in critically ill adults with abnormal coagulation tests (INR 1.5 to 3.0) requiring an invasive procedure. In the control group, a prophylactic plasma transfusion was administered before the invasive procedure was performed, whereas the plasma transfusion was omitted in the intervention group. The study authors planned to enrol 200 participants per treatment arm (400 participants in total). Although this study randomly assigned two plasma transfusion strategies (prophylactic plasma transfusion vs no plasma transfusion), it did not randomly assign two strategies based on coagulation test results. Therefore, it did not meet our inclusion criteria.
nor on non-bleeding patients who do not need the invasive procedure, although these individuals represent 33% of patients receiving plasma transfusions (Lauzier 2007). Research is needed to inform treatment decisions in patients with these characteristics. Despite the paucity of evidence to guide their use, plasma transfusions are nonetheless a common treatment in critically ill patients, as more than 10% of critically ill patients receive a plasma transfusion during their hospital stay (Luk 2002; Puetz 2012; Stanworth 2011). Obervational data, however, suggest that plasma transfusions fail to correct mildly abnormal coagulation tests (Abdel-Wahab 2006; Holland 2006) and are associated with worse clinical outcomes in non-massively bleeding patients (Dara 2005; Karam 2013; Khan 2007; Narick 2012; Sarani 2008; van Stein 2010; Watson 2009). However, these observational studies might suffer from a “confounding by indication” bias, as prognostic factors perceived by the attending physician may influence treatment decisions. Therefore, in the absence of randomised controlled trials assessing appropriate transfusion strategies, the decision to proceed with plasma transfusion must be based on individualised indications, with balance of the risks and benefits.
Risk of bias in included studies No studies are included in this review.
AUTHORS’ CONCLUSIONS Implications for practice
Effects of interventions
We found no randomised controlled trials assessing the effects of different plasma transfusion strategies in critically ill participants.
No studies are included in this review.
Implications for research
This systematic review shows that, in critically ill patients, no randomised controlled trials have evaluated plasma transfusion strategies based on coagulation test thresholds.
Given the frequency of plasma transfusions and the association between this intervention and worse clinical outcomes in nonmassively bleeding patients, trials to determine which patients will benefit from this intervention are warranted. More than 60 years after the introduction of plasma transfusions, no such trials have been published. Randomised controlled trials on plasma transfusion strategies are necessary.
We identified only one protocol of a trial on plasma transfusions (Müller 2011). In critically ill adults with mildly abnormal coagulation tests (INR 1.5 to 3.0) who need invasive procedures (insertion of a central venous catheter, thoracocentesis, etc.), the study authors planned to randomly assign a prophylactic plasma transfusion versus no transfusion and to assess the occurrence of significant bleeding during the procedure. This is an important question, as one study suggests that 17% of plasma transfusions are given to non-bleeding patients with abnormal coagulation tests who need a procedure (Lauzier 2007). However, this trial will not provide information on similar patients with higher INR levels,
The paucity of randomised controlled trials on plasma transfusion strategies might be explained by the difficulties involved in designing such trials. In contrast to trials on red blood cell transfusion strategies, in which the thresholds are haemoglobin levels, no consensus has been reached on the appropriate trigger for plasma transfusions. Some authors suggest using INR (Abdel-Wahab 2006; Lauzier 2007), whereas others suggest using prothrombin time or activated partial thromboplastin time ratios (Liumbruno 2009) or thromboelastography (Afshari 2011), which assesses the viscoelastic property of clot formation under low shear conditions. Another obstacle to designing a randomised controlled trial is the dif-
DISCUSSION
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
8
ficulty involved in assessing clinically significant bleeding, as one must incorporate not only the volume but also its localisation, the haemodynamic response to bleeding and the need for red blood cell transfusions.
ACKNOWLEDGEMENTS We would like to thank Dr Serge Grazioli and Dr Sonia Labarinas for their input on this review.
REFERENCES
References to studies excluded from this review Müller 2011 {published data only (unpublished sought but not used)} Müller MC, de Jonge E, Arbous MS, Spoelstra-de Man AM, Karakus A, Vroom MB, et al.Transfusion of fresh frozen plasma in non-bleeding ICU patients-TOPIC trial: study protocol for a randomized controlled trial. Trials 2011;12(266):1–7. [PUBMED: 22196464]
Additional references Abdel-Wahab 2006 Abdel-Wahab OI, Healy B, Dzik WH. Effect of fresh-frozen plasma transfusion on prothrombin time and bleeding in patients with mild coagulation abnormalities. Transfusion 2006;46(8):1279–85. [PUBMED: 16934060] Afshari 2011 Afshari A, Wikkelsø A, Brok J, Møller AM, Wetterslev J. Thrombelastography (TEG) or thromboelastometry (ROTEM) to monitor haemotherapy versus usual care in patients with massive transfusion. Cochrane Database of Systematic Reviews 2011, Issue 3. [DOI: 10.1002/ 14651858.CD007871.pub2; PUBMED: 21412912] Bateman 2008 Bateman ST, Lacroix J, Boven K, Forbes P, Barton R, Thomas NJ, et al.Anemia, blood loss, and blood transfusions in North American children in the intensive care unit. American Journal of Respiratory and Critical Care Medicine 2008;178(1):26–33. [PUBMED: 18420962] Carson 2012 Carson JL, Carless PA, Hebert PC. Transfusion thresholds and other strategies for guiding allogeneic red blood cell transfusion. Cochrane Database of Systematic Reviews 2012, Issue 4. [DOI: 10.1002/14651858.CD002042.pub3] Crosby 1997 Expert Working Group. Guidelines for red blood cell and plasma transfusion for adults and children. Canadian Medical Association Journal 1997;156(11 suppl):S1–24. Dara 2005 Dara SI, Rana R, Afessa B, Moore SB, Gajic O. Fresh frozen plasma transfusion in critically ill medical patients
with coagulopathy. Critical Care Medicine 2005;33(11): 2667–71. [PUBMED: 16276195] Gong 2005 Gong MN, Thompson BT, Williams P, Pothier L, Boyce PD, Christiani DC. Clinical predictors of and mortality in acute respiratory distress syndrome: potential role of red cell transfusion. Critical Care Medicine 2005;33(6):1191–8. [PUBMED: 15942330] Higgins 2011 Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. www.cochrane-handbook.org. Holland 2006 Holland LL, Brooks JP. Toward rational fresh frozen plasma transfusion: the effect of plasma transfusion on coagulation test results. American Journal of Clinical Pathology 2006;126 (1):133–9. [PUBMED: 16753596] Hébert 1999 Hébert PC, Wells G, Blajchman MA, Marshall J, Martin C, Pagliarello G, et al.A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. New England Journal of Medicine 1999;340(6):409–17. [PUBMED: 9971864] Karam 2013 Karam O, Lacroix J, Robitaille N, Rimensberger PC, Tucci M. Association between plasma transfusions and clinical outcome in critically ill children: a prospective observational study. Vox Sanguinis 2013 Mai;104(4):342–9. [PUBMED: 23294337] Karam 2013a Karam O, Tucci M, Lacroix J, Rimensberger PC. International survey on plasma transfusion practices in critically ill children. Transfusion. In press. Khan 2007 Khan H, Belsher J, Yilmaz M, Afessa B, Winters JL, Moore SB, et al.Fresh-frozen plasma and platelet transfusions are associated with development of acute lung injury in critically ill medical patients. Chest 2007;131(5):1308–14. [PUBMED: 17400669]
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
9
Kneyber 2007 Kneyber MC, Hersi MI, Twisk JW, Markhorst DG, Plötz FB. Red blood cell transfusion in critically ill children is independently associated with increased mortality. Intensive Care Medicine 2007;33(8):1414–22. [PUBMED: 17572875] Lacroix 2007 Lacroix J, Hébert PC, Hutchison JS, Hume HA, Tucci M, Ducruet T, et al.Transfusion strategies for patients in pediatric intensive care units. New England Journal of Medicine 2007;356(16):1609–19. [PUBMED: 17442904] Lauzier 2007 Lauzier F, Cook D, Griffith L, Upton J, Crowther M. Fresh frozen plasma transfusion in critically ill patients. Crt Care Med 2007;35(7):1655–9. [PUBMED: 17522577] Liumbruno 2009 Liumbruno G, Bennardello F, Lattanzio A, Piccoli P, Rossetti G. Recommendations for the transfusion of plasma and platelets. Blood Transfusion 2009;7(2):132–50. [PUBMED: 19503635] Luk 2002 Luk C, Eckert KM, Barr RM, Chin-Yee IH. Prospective audit of the use of fresh-frozen plasma, based on Canadian Medical Association transfusion guidelines. Canadian Medical Association Journal 2002;166(12):1539–40. [PUBMED: 12074120] Narick 2012 Narick C, Triulzi DJ, Yazer MH. Transfusion-associated circulatory overload after plasma transfusion. Transfusion 2012;52(1):160–5. [PUBMED: 21762464] National Survey 2009 Office of the Assistant Secretary for Health. The 2009 National Blood Collection and Utilization Survey Report. Washington, DC: US Department of Health and Human Services, 2011. [: 9781563953286]
ill surgical patients is associated with an increased risk of infection. Critical Care Medicine 2008;36(4):1114–8. [PUBMED: 18379235] Schmidt 2012 Schmidt PJ. The plasma wars: a history. Transfusion 2012; 52(S1):2S–4S. [PUBMED: 22578368] Stanworth 2007 Stanworth SJ, Hyde CJ, Murphy MF. Evidence for indications of fresh frozen plasma. Transfusion Clinique et Biologique 2007;14(6):551–6. [PUBMED: 18430602] Stanworth 2011 Stanworth SJ, Walsh TS, Prescott RJ, Lee RJ, Watson DM, Wyncoll D. A national study of plasma use in critical care: clinical indications, dose and effect on prothrombin time. Critical Care 2011;15(2):R108. [PUBMED: 21466676] van Stein 2010 van Stein D, Beckers EA, Sintnicolaas K, Porcelijn L, Danovic F, Wollersheim JA, et al.Transfusion-related acute lung injury reports in the Netherlands: an observational study. Transfusion 2010;50(1):213–20. [PUBMED: 19694998] Villanueva 2013 Villanueva C, Colomo A, Bosch A, Concepción M, Hernandez-Gea V, Aracil C, et al.Transfusion strategies for acute upper gastrointestinal bleeding. New England Journal of Medicine 2013;368(1):11–21. [PUBMED: 23281973] Vlaar 2009 Vlaar AP, in der Maur AL, Binnekade JM, Schultz MJ, Juffermans NP. A survey of physicians’ reasons to transfuse plasma and platelets in the critically ill: a prospective singlecentre cohort study. Transfusion Medicine 2009;19(4): 207–12. [PUBMED: 19706138]
O’Shaughnessy 2004 O’Shaughnessy DF, Atterbury C, Bolton Maggs P, Murphy M, Thomas D, Yates S, et al.Guidelines for the use of freshfrozen plasma, cryoprecipitate and cryosupernatant. British Journal of Haematology 2004;126(1):11–28. [PUBMED: 15198728]
Watson 2009 Watson GA, Sperry JL, Rosengart MR, Minei JP, Harbrecht BG, Moore EE, et al.Fresh frozen plasma is independently associated with a higher risk of multiple organ failure and acute respiratory distress syndrome. Journal of Trauma 2009;67(2):221–7. [PUBMED: 19667872]
Puetz 2012 Puetz J, Witmer C, Huang Y, Raffini L. Widespread use of fresh frozen plasma in US children’s hospitals despite limited evidence demonstrating a beneficial effect. Journal of Pediatrics 2012;160(2):210–5. [PUBMED: 21924435]
Zink 2009 Zink KA, Sambasivan CN, Holcomb JB, Chisholm G, Schreiber MA. A high ratio of plasma and platelets to packed red blood cells in the first 6 hours of massive transfusion improves outcomes in a large multicenter study. American Journal of Surgery 2009;197(5):565–70. [PUBMED: 19393349] ∗ Indicates the major publication for the study
Sarani 2008 Sarani B, Dunkman WJ, Dean L, Sonnad S, Rohrbach JI, Gracias VH. Transfusion of fresh frozen plasma in critically
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
10
CHARACTERISTICS OF STUDIES
Characteristics of excluded studies [ordered by study ID]
Study
Reason for exclusion
Müller 2011
Randomly assigns two plasma transfusion strategies (prophylactic plasma transfusion vs no plasma transfusion) but does not randomly assign two strategies based on coagulation test thresholds
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
11
DATA AND ANALYSES This review has no analyses.
CONTRIBUTIONS OF AUTHORS All authors contributed to the protocol and to the systematic review.
DECLARATIONS OF INTEREST JL: All grants that were received to conduct studies in the field of transfusion medicine came from public agencies. JL also received royalties for books that he edited completely or in part. None of these sources of revenue are in conflict with this review. All other review authors: none known.
DIFFERENCES BETWEEN PROTOCOL AND REVIEW We have removed the criteria that “Studies deemed to have a high risk of bias in any domain will not be included in data synthesis”, as we will conduct a sensitivity analysis on study quality, and it is unlikely that many studies will be judged to have low risk of bias in every domain.
INDEX TERMS Medical Subject Headings (MeSH) ∗ Critical
Illness; Blood Transfusion [∗ methods]
MeSH check words Humans
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
12
This is a reprint of a Cochrane review, prepared and maintained by The Cochrane Collaboration and published in The Cochrane Library 2013, Issue 12 http://www.thecochranelibrary.com
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
TABLE OF CONTENTS HEADER . . . . . . . . . . . . . . . . . . ABSTRACT . . . . . . . . . . . . . . . . . PLAIN LANGUAGE SUMMARY . . . . . . . . . BACKGROUND . . . . . . . . . . . . . . . OBJECTIVES . . . . . . . . . . . . . . . . METHODS . . . . . . . . . . . . . . . . . RESULTS . . . . . . . . . . . . . . . . . . Figure 1. . . . . . . . . . . . . . . . . DISCUSSION . . . . . . . . . . . . . . . . AUTHORS’ CONCLUSIONS . . . . . . . . . . ACKNOWLEDGEMENTS . . . . . . . . . . . REFERENCES . . . . . . . . . . . . . . . . CHARACTERISTICS OF STUDIES . . . . . . . . DATA AND ANALYSES . . . . . . . . . . . . . CONTRIBUTIONS OF AUTHORS . . . . . . . . DECLARATIONS OF INTEREST . . . . . . . . . DIFFERENCES BETWEEN PROTOCOL AND REVIEW INDEX TERMS . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
1 1 2 2 3 3 6 7 8 8 9 9 10 12 12 12 12 12
i
[Intervention Review]
Plasma transfusion strategies for critically ill patients Oliver Karam1 , Marisa Tucci2 , Christophe Combescure3 , Jacques Lacroix2 , Peter C Rimensberger1 1 Pediatric
Critical Care Unit, Geneva University Hospital, Geneva, Switzerland. 2 Pediatric Intensive Care Unit, CHU Sainte Justine, University of Montreal, Montreal, Canada. 3 Center for Clinical Research, Geneva University Hospital, Geneva, Switzerland
Contact address: Oliver Karam, Pediatric Critical Care Unit, Geneva University Hospital, 6 rue Willy Donzé, Geneva, 1211, Switzerland. [email protected]. Editorial group: Cochrane Injuries Group. Publication status and date: New, published in Issue 12, 2013. Review content assessed as up-to-date: 15 August 2013. Citation: Karam O, Tucci M, Combescure C, Lacroix J, Rimensberger PC. Plasma transfusion strategies for critically ill patients. Cochrane Database of Systematic Reviews 2013, Issue 12. Art. No.: CD010654. DOI: 10.1002/14651858.CD010654.pub2. Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
ABSTRACT Background Although plasma transfusions are frequently prescribed for critically ill patients, most clinical uses of plasma are not supported by evidence. Plasma transfusions do not seem to correct mild coagulation abnormalities based on international normalised ratio (INR) testing, but they seem to be independently associated with worse clinical outcomes in non-massively bleeding patients. Current recommendations on plasma transfusion strategies advocate limiting plasma transfusions to patients who are actively bleeding or who are at risk of bleeding and concomitantly have moderately abnormal coagulation tests. Objectives To determine whether use of a restrictive versus a liberal plasma transfusion threshold affects mortality or morbidity in critically ill patients, and to assess the clinical effects of different plasma transfusion thresholds in critically ill patients. Search methods A search for studies was run on 15 August 2013. We searched the Cochrane Injuries Group Specialised Register, the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE(R) Ovid, MEDLINE(R) Ovid In-Process & Other Non-Indexed Citations, MEDLINE(R) Ovid Daily and OLDMEDLINE(R) Ovid, EMBASE Classic + EMBASE (Ovid SP), reference lists, related websites and trial registries and checked lists of references. Selection criteria Randomised clinical trials that assessed the effects of two plasma transfusion strategies, using a restrictive and a liberal threshold of at least one coagulation test, in critically ill participants. Data collection and analysis Two review authors independently extracted data and assessed trial quality using the standard methods of the Cochrane Handbook for Systematic Reviews of Interventions. Main results Of 843 references identified by our search, none of the trials satisfied our predefined inclusion criteria. No studies are included in this review. Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
1
Authors’ conclusions This review highlights the lack of evidence that is available to guide plasma transfusions in critically ill patients. Randomised controlled trials are needed to determine the appropriate plasma transfusion strategy in critically ill patients.
PLAIN LANGUAGE SUMMARY Plasma transfusion strategies for critically ill patients Plasma transfusions are a frequently used treatment for critically ill patients, and they are usually prescribed to correct abnormal coagulation tests and to prevent or stop bleeding. Although plasma transfusions are frequently prescribed for critically ill patients, some of the reasons for their use are not supported by evidence from medical research. Some research has found an association of plasma transfusions with worse outcomes, and other studies have suggested that plasma transfusions do not help to return blood to its normal thickness. It is important to identify which patients, if any, may benefit from plasma transfusions. We searched the medical literature for reports of randomised controlled trials of plasma transfusion in people with a critical illness. Upon reading 843 study records, we found that currently no randomised trials have compared different plasma transfusion strategies. Such studies are needed to help physicians prescribe plasma transfusion only for patients who will benefit from them. Two main obstacles are associated with designing a randomised controlled trial on this topic, but they can be overcome. Currently, no consensus has been reached on the best way to determine whether a patient requires a plasma transfusion. Another obstacle is the difficulty involved in identifying important bleeding, as one must consider not only the amount of bleeding but also its location, the body’s own response to the bleeding (haemodynamic response) and the need for red blood cell transfusions.
BACKGROUND Plasma transfusions have been used since 1941 (Schmidt 2012). In 2008, 4,484,000 plasma units were transfused in patients in the United States (National Survey 2009). More than 10% of critically ill patients, both adults and children, receive a plasma transfusion during their hospital stay, making plasma transfusion a frequently used treatment modality (Luk 2002; Puetz 2012; Stanworth 2011). In current practice, plasma transfusions are widely used in critical care; they are administered most often to correct abnormal coagulation tests or to prevent bleeding (Vlaar 2009). In situations in which active bleeding is attributable to a coagulation factor deficiency, plasma transfusions can constitute a life-saving intervention by improving coagulation factor deficit (Stanworth 2007), especially in patients requiring massive transfusion (Zink 2009). However, in less dramatic situations, some researchers have shown that transfusing plasma often fails to correct mild coagulation abnormalities (Abdel-Wahab 2006; Holland 2006). Furthermore, currently available evidence derived from several retrospective studies in adults suggests that plasma transfusions are associated with increased odds of acute lung injury (Dara 2005; Khan 2007; van Stein 2010), nosocomial infection (Sarani 2008)
and transfusion-associated circulatory overload (Narick 2012). In critically ill adults with haemorrhagic shock, a prospective study reported an independent association between plasma transfusions and worse outcomes such as multiple organ failure and acute respiratory distress syndrome (ARDS) (Watson 2009). In critically ill children, a prospective observational study also reported an independent association between plasma transfusions and multiple organ failure and nosocomial infection (Karam 2013).
Description of the intervention Transfusion strategies can be viewed as liberal or restrictive. A liberal transfusion strategy implies that blood products are transfused for lower thresholds than with a restrictive transfusion strategy. Transfusion strategies have been evaluated for red blood cells (RBCs) (Carson 2012). RBC transfusions are a common treatment in both adult and paediatric critical care settings (Bateman 2008), as it is hypothesised that critically ill patients could benefit from RBC transfusions to enhance oxygen delivery. Nevertheless, as for plasma, it has been recognised that RBC transfusions are associated with increased morbidity (Bateman 2008; Gong 2005) and mortality (Kneyber 2007). Adequate RBC transfusion strate-
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
2
gies have been evaluated in critically ill patients, in whom a restrictive haemoglobin strategy, as compared with a liberal strategy, decreased transfusion requirements without increasing adverse outcomes (Hébert 1999; Lacroix 2007; Villanueva 2013).
Why it is important to do this review Current recommendations on plasma transfusion strategies advocate limiting plasma transfusions to patients who are actively bleeding or are at risk of bleeding and who concomitantly have moderately abnormal coagulation tests (Crosby 1997; Liumbruno 2009; O’Shaughnessy 2004). However, these recommendations fail to cite randomised controlled trials (RCTs). Plasma transfusions are a frequent treatment in critical care settings, although they seem to be associated with unfavourable outcomes if not administered wisely. No systematic review of RCTs has yet been published to guide physicians in selecting appropriate plasma transfusion strategies for use in critically ill patients.
Types of interventions The intervention in the experimental group should follow a protocol with one defined threshold required for plasma transfusion. The intervention in the control group should consist of another threshold or standard treatment. We planned to extract the numbers of plasma transfusions from the original studies to describe the intervention in experimental and control groups. By definition, we named “liberal” a transfusion strategy based on a threshold allowing for more transfusions, and “restrictive” a transfusion strategy based on a threshold allowing for fewer transfusions. We planned to extract data on co-interventions (i.e. transfusion of red blood cells and platelets, fibrinogen or other specific coagulation factors, steroids, need for surgery and need for extracorporeal life support) related to the main outcome.
Types of outcome measures
Primary outcomes
• All-cause mortality, at the end of the follow-up period in each trial.
OBJECTIVES To determine whether use of a restrictive versus a liberal plasma transfusion threshold affects mortality or morbidity in critically ill patients, and to assess the clinical effects of different plasma transfusion thresholds in critically ill patients.
Secondary outcomes
• • • • •
Nosocomial infections. Multiple-organ dysfunction (new or progressive). Volume of blood lost. Transfusion of RBCs and platelets. Transfusion reactions.
METHODS Search methods for identification of studies Criteria for considering studies for this review
The search was not restricted by date, language or publication status.
Types of studies
Electronic searches
We considered all randomised clinical trials for inclusion (irrespective of language, blinding, publication status or sample size) if two thresholds (restrictive and liberal) and at least one coagulation test (international normalised ratio (INR), activated partial thromboplastin time (aPTT), fibrinogen or thromboelastography) were compared. Trials on plasmapheresis and plasma exchanges were not included.
The Cochrane Injuries Group Trials Search Co-ordinator searched the following. • Cochrane Injuries Group Specialised Register (15 August 2013). • Cochrane Central Register of Controlled Trials (CENTRAL) (2013, Issue 7 of 12). • MEDLINE(R) Ovid, MEDLINE(R) Ovid In-Process & Other Non-Indexed Citations, MEDLINE(R) Ovid Daily and OLDMEDLINE(R) Ovid (1946 to 15 August 2013). • EMBASE Classic + EMBASE (Ovid SP) (1947 to 2013 week 32).
Types of participants All critically ill patients of all ages (neonates, children and adults) requiring plasma transfusion.
Search strategies are listed in Appendix 1.
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
3
Searching other resources On 20 August 2013, we searched the following databases for ongoing and unpublished trials. • The Clinical Trials Search Portal of the World Health Organization (http://apps.who.int/trialsearch/). • The Clinical Trials Registry of the U.S. National Institutes of Health (http://clinicaltrials.gov). We searched the following websites on 20 August 2013. • Transfusion Evidence Library (http:// transfusionguidelines.org). • http://www.evidence.nhs.uk/. • American Association of Blood Banks (http:// www.aabb.org). In addition, we checked the reference lists of identified material for further relevant trials.
Data collection and analysis
Selection of studies Two review authors (OK and MT) independently screened titles and abstracts in the literature search. Authors of primary studies that seemed eligible for inclusion would have been contacted for clarification of data and to obtain missing information. Study eligibility was assessed against defined inclusion criteria. We planned to assess separately methodological quality, including participant randomisation; blinding of participants, investigators and outcome assessors; intention-to-treat analysis; and completeness of follow-up. We planned to request the consensus of an additional review author (JL) to resolve any disagreement between the primary review authors. Data extraction and management Two review authors (OK and MT) would have independently extracted data from the selected trials using a standardised data extraction form. We planned to consult JL in the event of any disagreements, which we would have resolved through discussion and consensus. We planned to independently extract the following data. • Year and language of publication. • Country in which the trial was conducted. • Year of conduct of the trial. • Inclusion and exclusion criteria. • Sample size. • Plasma transfusion thresholds (INR, aPTT, fibrinogen, thromboelastography) for the restrictive and liberal strategy groups. • Outcomes (in each group).
• Methodological quality. We planned to seek any unclear or missing information by contacting the authors of the individual trials. Had there been any doubt whether trial reports shared the same participants-completely or partially (by identifying common authors and centres)-the authors of the trials would have been contacted to clarify whether the trial had been duplicated. We planned to resolve any differences in opinion through discussion or arbitration with a third review author (JL).
Assessment of risk of bias in included studies We planned that two review authors (OK and MT) would independently assess each trial for risk of bias. In cases of disagreement, consensus would have been reached by discussion with a third review author (JL). Risk of bias for RCTs that met eligibility criteria would have been assessed using The Cochrane Collaboration tool for assessment of risk of bias provided in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We planned to assess the following areas. Was the allocation sequence adequately generated? • Criteria for a judgement of ’low risk for bias’: Random sequence generation was performed by using a random number table, a computer random number generator or other accepted method. • Criteria for the judgement of ’unclear risk of bias’: The study was described as randomised, but the method for sequence generation was not described, or the method for sequence generation was inadequately described to permit assessment of risk of bias. • Criteria for the judgement of ’high risk of bias’: The method for sequence generation was described but was not random. Quasi-randomised trials, which base allocation on factors such as date of birth, day of the week or medical record number, will be considered as having a high risk of bias. Was allocation adequately concealed? • Criteria for a judgement of ’low risk for bias’: Allocation concealment was described and would not allow investigators enrolling participants to foresee participant assignment. Acceptable methods of allocation concealment include central allocation and sequentially numbered, opaque, sealed envelopes. • Criteria for the judgement of ’unclear risk of bias’: The method for allocation concealment was not described, or the method for allocation concealment was inadequately described to permit assessment of risk of bias. • Criteria for the judgement of ’high risk of bias’: The method for allocation concealment was described but might allow investigators to foresee participant assignment. Studies using an open random allocation schedule, assignment envelopes without appropriate safeguards and alternation or rotation or any
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
4
other unconcealed method would be considered as having a high risk of bias. Was knowledge of the allocated intervention concealed from outcome assessors? • Criteria for a judgement of ’low risk for bias’: Outcome assessment was not blinded, but the outcome and the outcome measurement were not likely to be influenced by lack of blinding; or outcome assessment was blinded. • Criteria for the judgement of ’unclear risk of bias’: The study contained insufficient information to permit judgement of the presence or absence of blinding. • Criteria for the judgement of ’high risk of bias’: Outcome assessment was not blinded, and the outcome or the outcome measurement was likely to be influenced by lack of blinding. Given that blinding of study physicians to the transfusion strategy would not be possible, we do not anticipate that any studies will include blinding of study physicians. Blinding of study participants is likely to occur because of the severity of their injuries and their physiological derangement, but this is unlikely to influence outcomes. This potential source of bias will be assessed for each outcome. Were incomplete outcome data adequately addressed? • Criteria for a judgement of ’low risk of bias’: No participants were lost to follow-up or withdrew from the study post-randomisation; loss to follow-up or withdrawal from the study was balanced across groups and was likely unrelated to true outcomes. • Criteria for the judgement of ’unclear risk of bias’: Reporting of attrition or exclusions was insufficient to permit judgement regarding the risk of bias. • Criteria for the judgement of ’high risk of bias’: Participants were lost to follow-up or withdrew from the study for reasons likely related to true outcomes; outcomes were not assessed using an intention-to-treat analysis and with substantial departure of the intervention received from that assigned at randomisation; the number of participants lost to follow-up might induce clinically relevant bias in the effect estimate or the effect size. This potential source of bias will be assessed for each outcome. Are reports of the study free of the suggestion of selective outcome reporting? • Criteria for a judgement of ’low risk of bias’: In studies with an available study protocol, all of the study’s prespecified outcomes were reported in the prespecified manner; in studies for which the study protocol is not available, published reports included all expected outcomes, including those that were prespecified (convincing text of this nature may be uncommon). • Criteria for the judgement of ’unclear risk of bias’: Information was insufficient to allow assessment of whether the risk of selective outcome reporting was present. • Criteria for the judgement of ’high risk of bias’: In studies with an available study protocol, not all of the study’s
prespecified outcomes were reported, or outcomes were not reported in the prespecified manner, or outcomes were changed during the course of the study. Was the study free of other significant sources of bias? • Criteria for a judgement of ’low risk of bias’: Studies appeared to be free of other sources of bias. • Criteria for the judgement of ’unclear risk of bias’: Risk of bias may be present, information was insufficient to allow assessment of whether an important risk of bias exists, or the rationale or evidence that an identified problem will introduce bias was insufficient. • Criteria for the judgement of ’high risk of bias’: At least one important source of bias was present, such as early stoppage of the study due to a data-dependent process, extreme baseline imbalance across study groups or potential bias due to source of study funding.
Measures of treatment effect The meta-analysis was planned to study the relationship between a plasma transfusion threshold and an outcome. The arms of the RCTs would have been labelled “Restrictive threshold” and “Liberal threshold”. For dichotomous outcomes, we planned to express results as risk ratios (RRs) with 95% confidence intervals (CIs). For continuous outcomes, we planned to express results as mean differences (MDs) (or standardised mean differences (SMDs) if different scales of measurement had been used) with 95% CIs.
Dealing with missing data If the main data (primary and secondary outcomes of the metaanalysis) were not provided in the report of a study, we planned to contact the study authors. If data could not have been obtained, a sensitivity analysis would have been performed by including the studies with missing data and extrapolating the data considering extreme scenarios. The purpose of this sensitivity analysis would have been to check the robustness of the conclusion while taking into account missing data.
Assessment of heterogeneity We planned to assess the clinical diversity (clinical heterogeneity) of the included studies. We would have undertaken subgroup analysis to examine possible clinical variability when the I2 statistic was 50% or less but heterogeneity remained statistically significant. We planned to analyse statistical heterogeneity using the Chi2 test on N-1 degrees of freedom, with an alpha of 0.1 used for statistical significance and with the I2 statistic (Higgins 2011). I2 ranges between 0 to 25%, 26 to 50%, 51 to 75%, and 76 to 100% corresponded to categories of low, medium, high and very high levels of statistical heterogeneity, respectively. We would have set an I2 threshold of greater than 76% to indicate substantial variation
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
5
across trials due to heterogeneity. We would have used a fixedeffect model if we would had found low to high heterogeneity between trials. We would have used a random-effects model if very high heterogeneity had existed between trials (Higgins 2011). We planned to test for homogeneity between trials for each outcome using Cochran’s Q statistic, with P value less than or equal to 0.10.
To assess clinical heterogeneity, we planned to perform the following subgroup analyses. • Coagulation tests: INR, aPTT, fibrinogen, thromboelastography. • Admission categories: cardiac surgery participants, trauma participants, burn participants. • Severity of the condition: severity at admission, severity of bleeding. • Age of the participant.
Assessment of reporting biases We would have analysed possible publication bias using a funnel plot if at least 10 studies were available for each outcome.
Data synthesis As plasma transfusion strategies are likely to be different among neonates, children and adults, meta-analysis would have been performed separately for each of these age groups. For the primary outcome, as well as for each of the secondary outcomes, we planned to present in plain words the number of trials reporting the outcome under evaluation, the effect of the intervention and the heterogeneity noted among those trials. We planned to analyse dichotomous data using an RR and to analyse continuous data using weighted MDs (or SMDs accordingly). A Mantel-Haenszel model would have been used, and pooled estimates would have been reported with a 95% CI. We planned to present the intervention effect using a forest plot to display effect estimates and confidence intervals for individual studies.
Subgroup analysis and investigation of heterogeneity
Sensitivity analysis A sensitivity analysis was planned to check the robustness of the conclusions with regard to missing data. We had also planned to compare trials with low risk of bias (adequate generation of allocation sequence and follow-up) versus trials with high risk of bias (one or more of the two components inadequate or unclear). If a study would have been visibly different to others, we would have conducted another sensitivity analysis by temporarily removing that study to check whether the conclusions were similar.
RESULTS
Description of studies We identified no published or ongoing trials that met our inclusion criteria. Results of the search The search identified 843 records (Figure 1).
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
6
Figure 1. Study flow diagram.
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
7
Included studies We identified no RCTs that evaluated plasma transfusion strategies according to predetermined coagulation test thresholds.
Excluded studies We identified the protocol for an RCT on plasma transfusion strategies, as designed by Müller et al (Müller 2011). The study authors described a non-inferiority randomised trial in critically ill adults with abnormal coagulation tests (INR 1.5 to 3.0) requiring an invasive procedure. In the control group, a prophylactic plasma transfusion was administered before the invasive procedure was performed, whereas the plasma transfusion was omitted in the intervention group. The study authors planned to enrol 200 participants per treatment arm (400 participants in total). Although this study randomly assigned two plasma transfusion strategies (prophylactic plasma transfusion vs no plasma transfusion), it did not randomly assign two strategies based on coagulation test results. Therefore, it did not meet our inclusion criteria.
nor on non-bleeding patients who do not need the invasive procedure, although these individuals represent 33% of patients receiving plasma transfusions (Lauzier 2007). Research is needed to inform treatment decisions in patients with these characteristics. Despite the paucity of evidence to guide their use, plasma transfusions are nonetheless a common treatment in critically ill patients, as more than 10% of critically ill patients receive a plasma transfusion during their hospital stay (Luk 2002; Puetz 2012; Stanworth 2011). Obervational data, however, suggest that plasma transfusions fail to correct mildly abnormal coagulation tests (Abdel-Wahab 2006; Holland 2006) and are associated with worse clinical outcomes in non-massively bleeding patients (Dara 2005; Karam 2013; Khan 2007; Narick 2012; Sarani 2008; van Stein 2010; Watson 2009). However, these observational studies might suffer from a “confounding by indication” bias, as prognostic factors perceived by the attending physician may influence treatment decisions. Therefore, in the absence of randomised controlled trials assessing appropriate transfusion strategies, the decision to proceed with plasma transfusion must be based on individualised indications, with balance of the risks and benefits.
Risk of bias in included studies No studies are included in this review.
AUTHORS’ CONCLUSIONS Implications for practice
Effects of interventions
We found no randomised controlled trials assessing the effects of different plasma transfusion strategies in critically ill participants.
No studies are included in this review.
Implications for research
This systematic review shows that, in critically ill patients, no randomised controlled trials have evaluated plasma transfusion strategies based on coagulation test thresholds.
Given the frequency of plasma transfusions and the association between this intervention and worse clinical outcomes in nonmassively bleeding patients, trials to determine which patients will benefit from this intervention are warranted. More than 60 years after the introduction of plasma transfusions, no such trials have been published. Randomised controlled trials on plasma transfusion strategies are necessary.
We identified only one protocol of a trial on plasma transfusions (Müller 2011). In critically ill adults with mildly abnormal coagulation tests (INR 1.5 to 3.0) who need invasive procedures (insertion of a central venous catheter, thoracocentesis, etc.), the study authors planned to randomly assign a prophylactic plasma transfusion versus no transfusion and to assess the occurrence of significant bleeding during the procedure. This is an important question, as one study suggests that 17% of plasma transfusions are given to non-bleeding patients with abnormal coagulation tests who need a procedure (Lauzier 2007). However, this trial will not provide information on similar patients with higher INR levels,
The paucity of randomised controlled trials on plasma transfusion strategies might be explained by the difficulties involved in designing such trials. In contrast to trials on red blood cell transfusion strategies, in which the thresholds are haemoglobin levels, no consensus has been reached on the appropriate trigger for plasma transfusions. Some authors suggest using INR (Abdel-Wahab 2006; Lauzier 2007), whereas others suggest using prothrombin time or activated partial thromboplastin time ratios (Liumbruno 2009) or thromboelastography (Afshari 2011), which assesses the viscoelastic property of clot formation under low shear conditions. Another obstacle to designing a randomised controlled trial is the dif-
DISCUSSION
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
8
ficulty involved in assessing clinically significant bleeding, as one must incorporate not only the volume but also its localisation, the haemodynamic response to bleeding and the need for red blood cell transfusions.
ACKNOWLEDGEMENTS We would like to thank Dr Serge Grazioli and Dr Sonia Labarinas for their input on this review.
REFERENCES
References to studies excluded from this review Müller 2011 {published data only (unpublished sought but not used)} Müller MC, de Jonge E, Arbous MS, Spoelstra-de Man AM, Karakus A, Vroom MB, et al.Transfusion of fresh frozen plasma in non-bleeding ICU patients-TOPIC trial: study protocol for a randomized controlled trial. Trials 2011;12(266):1–7. [PUBMED: 22196464]
Additional references Abdel-Wahab 2006 Abdel-Wahab OI, Healy B, Dzik WH. Effect of fresh-frozen plasma transfusion on prothrombin time and bleeding in patients with mild coagulation abnormalities. Transfusion 2006;46(8):1279–85. [PUBMED: 16934060] Afshari 2011 Afshari A, Wikkelsø A, Brok J, Møller AM, Wetterslev J. Thrombelastography (TEG) or thromboelastometry (ROTEM) to monitor haemotherapy versus usual care in patients with massive transfusion. Cochrane Database of Systematic Reviews 2011, Issue 3. [DOI: 10.1002/ 14651858.CD007871.pub2; PUBMED: 21412912] Bateman 2008 Bateman ST, Lacroix J, Boven K, Forbes P, Barton R, Thomas NJ, et al.Anemia, blood loss, and blood transfusions in North American children in the intensive care unit. American Journal of Respiratory and Critical Care Medicine 2008;178(1):26–33. [PUBMED: 18420962] Carson 2012 Carson JL, Carless PA, Hebert PC. Transfusion thresholds and other strategies for guiding allogeneic red blood cell transfusion. Cochrane Database of Systematic Reviews 2012, Issue 4. [DOI: 10.1002/14651858.CD002042.pub3] Crosby 1997 Expert Working Group. Guidelines for red blood cell and plasma transfusion for adults and children. Canadian Medical Association Journal 1997;156(11 suppl):S1–24. Dara 2005 Dara SI, Rana R, Afessa B, Moore SB, Gajic O. Fresh frozen plasma transfusion in critically ill medical patients
with coagulopathy. Critical Care Medicine 2005;33(11): 2667–71. [PUBMED: 16276195] Gong 2005 Gong MN, Thompson BT, Williams P, Pothier L, Boyce PD, Christiani DC. Clinical predictors of and mortality in acute respiratory distress syndrome: potential role of red cell transfusion. Critical Care Medicine 2005;33(6):1191–8. [PUBMED: 15942330] Higgins 2011 Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. www.cochrane-handbook.org. Holland 2006 Holland LL, Brooks JP. Toward rational fresh frozen plasma transfusion: the effect of plasma transfusion on coagulation test results. American Journal of Clinical Pathology 2006;126 (1):133–9. [PUBMED: 16753596] Hébert 1999 Hébert PC, Wells G, Blajchman MA, Marshall J, Martin C, Pagliarello G, et al.A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. New England Journal of Medicine 1999;340(6):409–17. [PUBMED: 9971864] Karam 2013 Karam O, Lacroix J, Robitaille N, Rimensberger PC, Tucci M. Association between plasma transfusions and clinical outcome in critically ill children: a prospective observational study. Vox Sanguinis 2013 Mai;104(4):342–9. [PUBMED: 23294337] Karam 2013a Karam O, Tucci M, Lacroix J, Rimensberger PC. International survey on plasma transfusion practices in critically ill children. Transfusion. In press. Khan 2007 Khan H, Belsher J, Yilmaz M, Afessa B, Winters JL, Moore SB, et al.Fresh-frozen plasma and platelet transfusions are associated with development of acute lung injury in critically ill medical patients. Chest 2007;131(5):1308–14. [PUBMED: 17400669]
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
9
Kneyber 2007 Kneyber MC, Hersi MI, Twisk JW, Markhorst DG, Plötz FB. Red blood cell transfusion in critically ill children is independently associated with increased mortality. Intensive Care Medicine 2007;33(8):1414–22. [PUBMED: 17572875] Lacroix 2007 Lacroix J, Hébert PC, Hutchison JS, Hume HA, Tucci M, Ducruet T, et al.Transfusion strategies for patients in pediatric intensive care units. New England Journal of Medicine 2007;356(16):1609–19. [PUBMED: 17442904] Lauzier 2007 Lauzier F, Cook D, Griffith L, Upton J, Crowther M. Fresh frozen plasma transfusion in critically ill patients. Crt Care Med 2007;35(7):1655–9. [PUBMED: 17522577] Liumbruno 2009 Liumbruno G, Bennardello F, Lattanzio A, Piccoli P, Rossetti G. Recommendations for the transfusion of plasma and platelets. Blood Transfusion 2009;7(2):132–50. [PUBMED: 19503635] Luk 2002 Luk C, Eckert KM, Barr RM, Chin-Yee IH. Prospective audit of the use of fresh-frozen plasma, based on Canadian Medical Association transfusion guidelines. Canadian Medical Association Journal 2002;166(12):1539–40. [PUBMED: 12074120] Narick 2012 Narick C, Triulzi DJ, Yazer MH. Transfusion-associated circulatory overload after plasma transfusion. Transfusion 2012;52(1):160–5. [PUBMED: 21762464] National Survey 2009 Office of the Assistant Secretary for Health. The 2009 National Blood Collection and Utilization Survey Report. Washington, DC: US Department of Health and Human Services, 2011. [: 9781563953286]
ill surgical patients is associated with an increased risk of infection. Critical Care Medicine 2008;36(4):1114–8. [PUBMED: 18379235] Schmidt 2012 Schmidt PJ. The plasma wars: a history. Transfusion 2012; 52(S1):2S–4S. [PUBMED: 22578368] Stanworth 2007 Stanworth SJ, Hyde CJ, Murphy MF. Evidence for indications of fresh frozen plasma. Transfusion Clinique et Biologique 2007;14(6):551–6. [PUBMED: 18430602] Stanworth 2011 Stanworth SJ, Walsh TS, Prescott RJ, Lee RJ, Watson DM, Wyncoll D. A national study of plasma use in critical care: clinical indications, dose and effect on prothrombin time. Critical Care 2011;15(2):R108. [PUBMED: 21466676] van Stein 2010 van Stein D, Beckers EA, Sintnicolaas K, Porcelijn L, Danovic F, Wollersheim JA, et al.Transfusion-related acute lung injury reports in the Netherlands: an observational study. Transfusion 2010;50(1):213–20. [PUBMED: 19694998] Villanueva 2013 Villanueva C, Colomo A, Bosch A, Concepción M, Hernandez-Gea V, Aracil C, et al.Transfusion strategies for acute upper gastrointestinal bleeding. New England Journal of Medicine 2013;368(1):11–21. [PUBMED: 23281973] Vlaar 2009 Vlaar AP, in der Maur AL, Binnekade JM, Schultz MJ, Juffermans NP. A survey of physicians’ reasons to transfuse plasma and platelets in the critically ill: a prospective singlecentre cohort study. Transfusion Medicine 2009;19(4): 207–12. [PUBMED: 19706138]
O’Shaughnessy 2004 O’Shaughnessy DF, Atterbury C, Bolton Maggs P, Murphy M, Thomas D, Yates S, et al.Guidelines for the use of freshfrozen plasma, cryoprecipitate and cryosupernatant. British Journal of Haematology 2004;126(1):11–28. [PUBMED: 15198728]
Watson 2009 Watson GA, Sperry JL, Rosengart MR, Minei JP, Harbrecht BG, Moore EE, et al.Fresh frozen plasma is independently associated with a higher risk of multiple organ failure and acute respiratory distress syndrome. Journal of Trauma 2009;67(2):221–7. [PUBMED: 19667872]
Puetz 2012 Puetz J, Witmer C, Huang Y, Raffini L. Widespread use of fresh frozen plasma in US children’s hospitals despite limited evidence demonstrating a beneficial effect. Journal of Pediatrics 2012;160(2):210–5. [PUBMED: 21924435]
Zink 2009 Zink KA, Sambasivan CN, Holcomb JB, Chisholm G, Schreiber MA. A high ratio of plasma and platelets to packed red blood cells in the first 6 hours of massive transfusion improves outcomes in a large multicenter study. American Journal of Surgery 2009;197(5):565–70. [PUBMED: 19393349] ∗ Indicates the major publication for the study
Sarani 2008 Sarani B, Dunkman WJ, Dean L, Sonnad S, Rohrbach JI, Gracias VH. Transfusion of fresh frozen plasma in critically
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
10
CHARACTERISTICS OF STUDIES
Characteristics of excluded studies [ordered by study ID]
Study
Reason for exclusion
Müller 2011
Randomly assigns two plasma transfusion strategies (prophylactic plasma transfusion vs no plasma transfusion) but does not randomly assign two strategies based on coagulation test thresholds
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
11
DATA AND ANALYSES This review has no analyses.
CONTRIBUTIONS OF AUTHORS All authors contributed to the protocol and to the systematic review.
DECLARATIONS OF INTEREST JL: All grants that were received to conduct studies in the field of transfusion medicine came from public agencies. JL also received royalties for books that he edited completely or in part. None of these sources of revenue are in conflict with this review. All other review authors: none known.
DIFFERENCES BETWEEN PROTOCOL AND REVIEW We have removed the criteria that “Studies deemed to have a high risk of bias in any domain will not be included in data synthesis”, as we will conduct a sensitivity analysis on study quality, and it is unlikely that many studies will be judged to have low risk of bias in every domain.
INDEX TERMS Medical Subject Headings (MeSH) ∗ Critical
Illness; Blood Transfusion [∗ methods]
MeSH check words Humans
Plasma transfusion strategies for critically ill patients (Review) Copyright © 2013 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
12
