Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Protocol - Intervention

Prophylactic platelet transfusions prior to surgery for people with a low platelet count

Authors' declarations of interest

Version published: 02 September 2017 Version history

https://doi.org/10.1002/14651858.CD012779

This is not the most recent version

view the current version 17 September 2018
Collapse all Expand all

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

To determine the clinical effectiveness and safety of prophylactic platelet transfusions prior to surgery for people with a low platelet count or platelet dysfunction (inherited or acquired).

Background

Description of the condition

Platelets are an essential component in the formation of a blood clot (BCSH 2003). A low platelet count can lead to a range of bleeding symptoms such as bruising, nosebleeds and, rarely, life‐threatening or fatal bleeding.

Thrombocytopenia is defined as a platelet count less than 150 x 109/L (BCSH 2003). When this is dilutional, associated with an expanded blood volume, the drop is mild and rarely clinically significant. Severe thrombocytopenia is defined as a platelet count less than 50 x 109/L (BCSH 2003). Thrombocytopenia can be caused by: reduced platelet production in the bone marrow often as a result of chemotherapy or a haematological malignancy (blood cancer) (Leguit 2010; Weinzierl 2013); increased platelet consumption as occurs in bleeding or disseminated intravascular coagulation (DIC) (Levi 2009); increased platelet destruction such as immune thrombocytopenia (Neunert 2013; Pacheco 2011; Provan 2010); or a combination of these conditions .
Mild, dilutional thrombocytopenia is common in pregnancy (7% to 12% of pregnancies), but severe thrombocytopenia (platelet count less than 50 x 109/L) is much less common (0.05% to 1% of pregnancies) and is a sign of complications (Burrows 1990; Nisha 2012; Sainio 2000). A platelet count less than 150 x 109/L is very common in individuals with chronic liver disease (up to 76%) (Afdhal 2008), and people who are critically ill (up to 68%) (Hui 2011). A large United Kingdom (UK) study of patients admitted to the intensive care unit (ICU) reported that 9% developed severe thrombocytopenia (Stanworth 2013). Thrombocytopenia is also frequent in people with haematological malignancies (Leguit 2010; Weinzierl 2013), and most platelet transfusions are used in individuals with haematological disorders (Cameron 2007; Greeno 2007; Pendry 2011).

People with thrombocytopenia often require a surgical procedure. A low platelet count is a relative contraindication to surgery due to the risk of bleeding (Estcourt 2017; Kaufman 2015; NICE 2015). Platelet transfusions are one of a number of interventions used in modern clinical practice to prevent and treat bleeding in people with thrombocytopenia.

Description of the intervention

Platelet concentrates are the second most frequently used blood component (Bolton‐Maggs 2016). Approximately 2.2 million platelet units are transfused annually in the USA (Whitaker 2013). Seventy‐four per cent of platelet transfusions are given prophylactically to non‐bleeding thrombocytopenic people and 15% are given to prevent bleeding prior to surgery or a procedure in people with haematological malignancies. In many cases platelet transfusions are given at platelet counts higher than the recommended triggers (Estcourt 2012; Greeno 2007).

Unlike other blood components, platelets must be kept on a shaker at room temperature, limiting the shelf life of platelet units to five to seven days. This makes it difficult for hospitals to manage their platelet stock (Fuller 2011).

Current practice in many countries is to correct thrombocytopenia with platelet transfusions prior to surgery. Guidelines often recommend a platelet count threshold of 50 x 109/L prior to major surgery and 100 x 109/L prior to surgery involving the brain or eyes (Estcourt 2017; Kaufman 2015; NICE 2015). Guidelines often do not go into further detail about risks for different types of surgery. Some low‐risk surgery may not require platelet transfusions at all, other procedures may be higher risk and the risk may also be dependent on patient co‐morbidities.

Platelet transfusions are not risk‐free. In 2014, 34% of all transfusion‐related adverse events reported to the UK national reporting system (Serious Hazards of Transfusion (SHOT)) were due to platelet components. The most common adverse events due to platelet components were febrile and allergic reactions (Birchall 2015). Although most of these reactions are not life‐threatening they can be extremely distressing for the person and time consuming for health professionals to investigate and exclude a more serious cause. Rarer, but more serious sequelae, include: anaphylaxis (life‐threatening allergic reaction), transfusion‐transmitted infections (TTI) and transfusion‐related acute lung injury (TRALI) (Blumberg 2010; Chapman 2015; Kaufman 2015; Slichter 2007; Vlaar 2013). Platelets units are stored at room temperature on a shaker, which increases the risk of bacterial growth (1:2000 to 1:3000) (Jacobs 2011). In 2015, there were four near miss incidents (three in platelets) reported to the unit between 2011 and 2015 and a total of 37/44 bacterial transfusion‐transmissions to individual recipients (34 incidents) were caused by the transfusion of platelets (Serious Hazards of Transfusion (SHOT) 2015).

A recent prospective multicentre cohort study concluded that in critically ill people, transfusion of platelets, but not of red blood cells and plasma, is an independent risk factor for acquiring a nosocomial infection (Engele 2016).

Alternative agents which could replace or reduce platelet transfusions may be more effective than platelet transfusions at controlling bleeding and will have a different side‐effect profile. Alternatives include artificial platelet substitutes, cryosupernatant, recombinant factor VIIa (rFVIIa), fibrinogen, recombinant factor XIII (rFXIII), thrombopoietin (TPO) mimetics and antifibrinolytic drugs.

How the intervention might work

Platelet transfusions

The premise for pre‐procedure intervention with platelet transfusion is as follows: thrombocytopenia increases the risk of bleeding, platelet transfusion corrects thrombocytopenia, a higher platelet count prevents bleeding and overall there is benefit to the patient. This presumption is however over simplistic.

In a small randomised controlled trial (RCT) of only 23 participants with thrombocytopenia who required 35 procedures and 84 teeth removed, bleeding complications were minimal without blood product support (Perdigão 2012).

One study including a total of 1720 patients with thrombocytopenia undergoing coronary artery bypass graft (CABG) surgery study pooled individual patient data from one pilot study and six RCTs. Platelet transfusion compared with no platelet transfusion was associated with a significant increase in mortality among patients undergoing CABG surgery (odds ratio (OR), 4.76; 95% confidence interval (CI), 1.65 to 13.73; P = 0.009). Although the authors used propensity score analysis, it is not clear if the increased mortality was due to platelet transfusion or because people who were more unwell received platelet transfusions (Spiess 2004).

Alternatives to platelet transfusions

Alternatives to platelet transfusion either simulate the effects of platelets (artificial platelet substitutes), stimulate additional fibrin formation (cryosupernatant, rFVIIa and fibrinogen), promote von Willebrand factor release and platelet function (desmopressin), increase platelet production (TPO mimetics), strengthen clot structure (rFXIII) or decrease clot breakdown (antifibrinolytics). These agents aim to promote haemostasis without the side effects associated with platelet transfusions. Their main adverse effect is excessive clotting and thrombosis.

In this review we will exclude trials that assess the use of: rFVIIa; fibrinogen concentrate; rFXIII; prothrombin complex concentrate; and desmopressin as these are the subject of other Cochrane reviews that compared these interventions to an active comparator in people requiring a surgical procedure (Desborough 2017; Fabes 2013; Simpson 2012).

Artificial platelet substitutes

Artificial platelet substitutes such as microspheres of human albumin coated with fibrinogen, lyophilised platelets, infusible plasma membranes, and liposomes with inserted platelet receptors aim to reproduce the active components of platelets without associated adverse events (Desborough 2016). Artificial platelets are not yet in routine clinical use, so their costs and adverse events are at present unclear.

Cryosupernatant

Cryosupernatant is a source of clotting factors and can be administered intravenously. It is a blood component and is associated with a small risk of transfusion reactions and transfusion‐transmitted infections.

Thrombopoietin (TPO) mimetics

Thrombopoietin (TPO) is made by the liver and is the key regulator of bone marrow platelet production. TPO mimetics have been used in several disease states to promote both an increase in the cells that produce platelets (megakaryopoiesis) and the production of platelets themselves (thrombopoiesis) (Kuter 2014). The two main TPO mimetics in current use are romiplostim (weekly injection) and eltrombopag (daily oral tablet), both of which are recommended by the National Institute for Health and Care Excellence (NICE) for use in adults with immune thrombocytopenia (ITP) who have severe disease and a high risk of bleeding (NICE 2011; NICE 2013). While a systematic review found that these agents improve platelet counts, there was no evidence that they reduced the risk of significant bleeding for people with ITP (Zeng 2011). TPO mimetics are more expensive than platelet transfusions (Joint Formulary Committee 2016). Interleukin 6 and interleukin 11 may also act as stimulants of thrombopoiesis (Gordon 1995; Kurzrock 2011; Tsimberidou 2005). They are not in routine clinical use, so their costs are unclear at present.

Antifibrinolytic drugs

Fibrinolysis is the process by which blood clots are broken down after they have been formed. Anti‐fibrinolytic drugs block this process, resulting in greater clot strength. The three most commonly used antifibrinolytic drugs are tranexamic acid, aprotinin and epsilon‐aminocaproic acid. Other Cochrane systematic reviews have assessed these agents in people undergoing surgical procedures (Henry 2011; McNicol 2016), or in people with haematological disorders (Estcourt 2016a).

Why it is important to do this review

People with a low platelet count often require surgery. Current guidelines are mainly based on expert opinion rather than good evidence and frequently do not go into detail about the risks for different types of surgery or define a specific platelet count threshold. Some low‐ risk surgery, for example dental extraction may not require platelet transfusions at all. Platelet transfusions may cause immediate‐ or longer‐ term harm and delay the start of life‐saving treatments. Alternatives to platelets may be more effective and safer. There is therefore a need to assess the likely benefit of platelet transfusion and their alternatives, in different procedures, against known risks.

In this review we aim to answer the following questions.

Do people require prophylactic platelet transfusion prior to certain types of surgery?

If platelet transfusions are required, which platelet count threshold should be used to trigger the transfusion of prophylactic platelets prior to surgery?

Are prophylactic platelet transfusions superior to other alternative treatments?

Objectives

To determine the clinical effectiveness and safety of prophylactic platelet transfusions prior to surgery for people with a low platelet count or platelet dysfunction (inherited or acquired).

Methods

Criteria for considering studies for this review

Types of studies

We will include randomised controlled trials (RCTs), non‐randomised controlled trials (non‐RCTs) and controlled before‐after studies (CBAs), irrespective of language or publication status. We will exclude uncontrolled studies, cross‐sectional studies and case‐control studies.

We will only include cluster‐RCTs, non‐randomised cluster trials, and CBAs with at least two intervention sites and two control sites. In studies with only one intervention or control site, the intervention (or comparison) is completely confounded by the study site, making it difficult to attribute any observed differences to the intervention rather than to other site‐specific variables.

If there are sufficient data to answer this review's questions using only data from RCTs, we will only report data from RCTs.

Types of participants

People of all ages with a low platelet count who are due to have surgery.

We will exclude studies on people with a low platelet count who are actively bleeding because they will receive platelet transfusions as part of the treatment of bleeding.

Types of interventions

We will include RCTs, non‐RCTs and controlled before‐after studies (CBAs) comparing three types of platelet transfusion regimens.

Comparison 1:. Prophylactic platelet transfusion prior to surgery versus no prophylactic platelet transfusion prior to surgery (placebo or no treatment).

Comparison 2: Prophylactic platelet transfusion prior to surgery versus alternative treatments (cryosupernatant, antifibrinolytics, thrombopoietin (TPO) mimetics). In this review we will exclude trials that assess the use of recombinant factor VIIa (rFVIIa); fibrinogen concentrate, recombinant factor XIII (rFXIII), prothrombin complex concentrate, and desmopressin as these are the subject of other Cochrane reviews that compared these interventions to an active comparator in people requiring a surgical procedure (Desborough 2017; Fabes 2013; Simpson 2012).

Comparison 3: Different platelet count thresholds for administering a prophylactic platelet transfusion prior to surgery.

We will record type of platelet component and dose of platelet component received.

Types of outcome measures

Primary outcomes

  • Mortality (all‐causes, secondary to bleeding, secondary to thromboembolism and secondary to infection) within 30 days and 90 days of surgery.

  • The number of participants with major procedure‐related bleeding within seven days of surgery, defined as:

    • surgical site bleeding requiring a second intervention or reoperation or surgical site bleeding that causes a haematoma or haemarthrosis of sufficient size to delay mobilisation or wound healing;

    • bleeding of sufficient size to cause delayed wound healing, or wound infection or surgical site bleeding that is unexpected and prolonged or causes haemodynamic instability (as defined by the study) that is associated with a 20 g/L drop in haemoglobin (Hb);

    • bleeding that requires two or more units of whole blood/red cells within 24 hours of the bleeding;

    • bleeding defined by the study with no further details.

Secondary outcomes

  • The number of participants with minor procedure‐related bleeding within seven days of surgery (e.g. haematoma, prolonged bleeding at surgical site that does not fulfil the definition for major bleeding).

  • Number of platelet transfusions per participant and number of platelet components per participant.

  • Number of red cell transfusions per participant and number of red cell components per participant.

  • Proportion of participants requiring additional interventions to stop bleeding (surgical, medical e.g. tranexamic acid, other blood products e.g. fresh frozen plasma (FFP), cryoprecipitate, fibrinogen) within seven days from the surgery.

  • Quality of life assessment using validated tools.

  • Serious adverse events due to:

    • transfusion (transfusion reactions, transfusion‐related acute lung injury (TRALI), transfusion related infection, transfusion‐associated circulatory overload (TACO), transfusion‐related dyspnoea) within 24 hours of the transfusion;

    • surgery (e.g. delayed wound healing, infection) within 30 days after the operation.

  • Length of hospital stay and length of intensive therapy unit (ITU) stay.

  • Venous and arterial thromboembolism (including deep vein thrombosis; pulmonary embolism; stroke; myocardial infarction).

Search methods for identification of studies

The Systematic Review Initiative’s Information Specialist (CD) will develop the search strategies in collaboration with the Cochrane Haematological Malignancies Group.

Electronic searches

We will search the following databases.

Bibliographic databases

  • Cochrane Central Register of Controlled Trials (CENTRAL, the Cochrane Library, current issue) (Appendix 1)

  • MEDLINE (OvidSP, Epub Ahead of Print, In‐Process and other Non‐Indexed Citations, and 1946 to present) (Appendix 2)

  • PubMed (for e‐publications ahead or print only) (www.ncbi.nlm.nih.gov/pubmed) (Appendix 3)

  • Embase (OvidSP, 1974 to present) (Appendix 4)

  • CINAHL (EBSCOHost, 1937 to present) (Appendix 5)

  • Transfusion Evidence Library (www.transfusionevidencelibrary.com) (1950 to present ‐ this includes a search of grey literature) (Appendix 6)

  • LILACS (1982 to present) (http://lilacs.bvsalud.org/en/) (Appendix 7)

  • Web of Science: Conference Proceedings Citation Index‐Science (CPCI‐S) (Thomson Reuters, 1990 to present) (Appendix 8)

Online databases of on‐going trials

We will combine searches in MEDLINE and Embase with the recommended Cochrane RCT search filters (Lefebvre 2011), systematic review filters based on those of the Scottish Intercollegiate Guidelines Network (SIGN) (www.sign.ac.uk/methodology/filters.html) and controlled before‐after studies filters based on those used in reviews of the Cochrane Effective Practice and Organisation of Care Group (EPOC 2015) (http://epoc.cochrane.org/). Searches in CINAHL will be combined with the SIGN systematic review and RCT filter and an EPOC‐based filter. We will not limit searches by language, year of publication or publication type.

Once we identify studies for inclusion we will search MEDLINE (OvidSP) for errata or retraction statements for the reports of these studies.

Searching other resources

We will also handsearch the reference lists of included studies and any relevant systematic reviews to identify further relevant studies. We will make contact with lead authors of relevant studies to identify any unpublished material, missing data or information regarding ongoing studies.

Data collection and analysis

We will summarise data in accordance with standard Cochrane methodologies. We will analyse data from different study designs separately.

Selection of studies

We will select studies with reference to the methods outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a). The Systematic Review Initiative’s Information Specialist (CD) will initially screen all search hits for relevance against the eligibility criteria and discard all those that are clearly irrelevant. Thereafter, two review authors (LE, RM) will independently screen all the remaining references for relevance against the full eligibility criteria. Full‐text papers will be retrieved for all references for which a decision on eligibility cannot be made from only screening title and abstract. If necessary additional information will be requested from study authors to assess the eligibility for inclusion of individual studies.The two review authors will discuss the results of study selection and try to resolve any discrepancies between themselves. In the event when it is not possible, the decision of eligibility will be referred to a third review author (MT). The results of study selection will be reported using a PRISMA flow diagram (Moher 2009). We will record the reasons for excluding studies based on full‐text assessment and will add those to the 'Characteristics of excluded studies' table.

Multiple reports of one study will be collated so that the study, and not the report, is the unit of analysis.

Data extraction and management

Two review authors (RM, LE) will independently extract data as recommended by the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a ), using standardised forms available in Covidence software (Covidence 2016). Two different data extraction forms will be piloted for included RCTs and NRS separately. If an agreement cannot be reached, the two review authors will try to come to a consensus; they will seek the advice of a third review author (MT). The review authors will not be blinded to names of authors, institutions, journals or the study outcomes. They will extract the following information for each study.

For randomised controlled trials

  • Source: study ID, report ID, review author ID, date of extraction, ID of author checking extracted data, citation of paper, contact author's details.

  • General study information: publication type, study objectives, funding source, conflict of interest declared, other relevant study publication reviewed.

  • Study details and methods: location, country, clinical setting, number of centres, study design, total study duration, recruitment dates, length of follow‐up, power calculation, primary analysis (and definition), stopping rules, method of sequence generation, allocation concealment, blinding (of clinicians, participants and outcome assessors) and any concerns regarding bias.

  • Characteristics of interventions: number of study arms, description of experimental arm, description of control arm, type of platelet component (e.g. apheresis or pooled), dose of platelet component, thresholds of platelets transfusions, type of surgery.

  • Characteristics of participants: age, gender, primary diagnosis, surgery types procedure (minor, major, surgery to sensitive areas as ocular surgery or neurosurgery), platelet count, coagulation abnormalities, anticoagulant medications, antiplatelet medications.

  • Participant flow: total number screened for inclusion, total number recruited, total number excluded, total number allocated to each study arm, total number analysed (for review outcomes), number of allocated participants who received planned treatment, number of dropouts with reasons (percentage in each arm), protocol violations, missing data.

  • Method of data analyses.

  • Outcomes: mortality (all‐causes, secondary to bleeding, secondary to thromboembolism and secondary to infection) within 30 days and 90 days of surgery; number of participants with major procedure‐related bleeding within seven days of surgery; number of participants with minor procedure‐related bleeding within seven days of surgery; number of platelet transfusions per participant and number of platelet components per participant; number of red cell transfusions per participant and number of red cell components per participant; proportion of participants requiring additional interventions to stop bleeding within seven days from the surgery; quality of life assessment using validated tools; serious adverse events due to transfusion (within 24 hours of the transfusion) or surgery (within 30 days after the operation); length of hospital stay and length of ITU stay, venous and arterial thromboembolism..

For Non‐randomised controlled trials

In addition to all the information listed for RCTs we will extract information on the following.

  • Study design.

  • Method of selecting participants: sample source, sample size, participants eligibility criteria, number of participants at each follow‐up point. and the source of study control group and baseline differences between the two groups.

  • Confounding factors: baseline confounding factors and co‐interventions that might lead potentially to bias are identified in the study and relevant confounding factors and co‐interventions that could introduce bias after the starting of platelets transfusions; the comparability of groups on confounding factors.

  • Method of assigning the intervention.

  • Co‐intervention status: this is in order to document if any other co‐interventions are considered in the study.

  • Method of data analysis: methods used to control for confounding and on multiple effect estimates (both unadjusted and adjusted estimates) as recommended in chapter 13 of theCochrane Handbook of Systematic Reviews of Interventions (Reeves 2011).

Assessment of risk of bias in included studies

Randomised controlled trials (RCTs)

We will assess the risk of bias for all included RCTs using the Cochrane 'Risk of bias' tool according to chapter eight of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011b). Two review authors (LE, RM) will work independently to assess each element of potential bias listed below as 'high', 'low' or 'unclear' risk of bias. We will report a brief description of the judgement statements upon which the review authors have assessed potential bias in the ’Characteristics of included studies’ table. We will ensure that a consensus on the degree of risk of bias is met through comparison of the review authors’ statements and where necessary, through consultation with a third review author (SH). We will use Cochrane's tool for assessing risk of bias, that will include the following domains.

  • Selection bias: we will describe for each included study if and how the allocation sequence was generated and if allocation was adequately concealed prior to assignment. We will also describe the method used to conceal the allocation sequence in detail and determine if intervention allocation could have been foreseen in advance of, or during recruitment, or changed after assignment.

  • Performance bias: we will describe for each included study, where possible, if the study participants and personnel were adequately blinded from knowledge of which intervention a participant received. We will judge studies as low risk of bias if they were blinded, or if we judge that lack of blinding could not have affected the results.

  • Detection bias: was blinding of the outcome assessors effective in preventing systematic differences in the way in which the outcomes were determined?

  • Incomplete outcome data: we will describe for each included study the attrition bias due to amount, nature or handling of incomplete outcome data. We will also try to evaluate whether intention‐to‐treat analysis has been performed or could be performed from published information.

  • Selective outcome reporting or reporting bias: we will describe for each included study the possibility of selective outcome reporting bias.

  • Other bias: was the study apparently free of other problems that could put it at risk of bias?

We will summarise the risk of bias for each key outcome for each included study. We will judge studies with at least one domain of high risk at high risk of bias overall etc.

Non‐randomised controlled trials (non‐RCTs)

We will use ROBINS‐I tool (formerly known as ACROBAT‐NRSI) to rate the quality of non‐randomised controlled trials (non‐RCTs) and controlled before‐after studies (CBAs) studies (Sterne 2016). This tool is based on the Cochrane 'Risk of bias' tool for rating the quality of RCTs (Higgins 2011c). The tool covers seven domains and the quality of evidence is rated 'low', 'moderate', 'serious', 'critical or no information', and the response options are 'yes', 'probably yes', 'no', 'probably no' and 'no information', (see Appendix 11 for a copy of the tool) and uses signalling questions for the assessment of:

  • bias due to confounding;

  • bias in the selection of participants;

  • bias in measurement of interventions;

  • bias due to departure from intended interventions;

  • bias due to missing data;

  • bias in measurement of outcomes;

  • bias in the selection of the reported result.

For 'low risk of bias' the study is judged to be at low risk of bias on all of the tool's seven domains.

For 'moderate risk of bias' the study is judged to be at low to moderate risk of bias in all of the tool's seven domains.

For 'serious risk of bias ' the study is judged to be at serious risk of bias in at least one of the tool's seven domains.

For ''critical risk of bias' to study is judged to be at critical risk of bias in at lease one domain of the tool's seven domains.

For 'no information on bias' when information in one or more key 'Risk of bias' domains are lacking.

Two review authors (LE, RM) will assess independently each domain of potential bias listed and will also tabulate a brief description of the judgement statements upon which the authors have assessed potential bias in the ’Characteristics of included studies’ table. We will ensure that a consensus on the degree of risk of bias is met through comparison of the review authors’ statements and where necessary, through consultation with a third review author (SH). We will highlight the highest quality evidence for each outcome.

We have pre‐specified the following main potential confounding factors.

  • Primary diagnosis of patient (e.g. liver disease; critical illness; pregnancy)

  • Age: variability in the age of patients included, e.g. paediatric (less than 16 years) versus adult (> 16 years) versus older adult (> 60 years)

  • Gender: male to female ratio

  • Previous severe bleeding (e.g. World Health Organization (WHO) grade 3 or 4 or equivalent)

Measures of treatment effect

Randomised controlled trials (RCTs)

For continuous outcomes, we will record the mean, standard deviation and total number of participants in both the treatment and control groups. For dichotomous outcomes we will record the number of events and the total number of participants in both the treatment and control groups.

For continuous outcomes using the same scale, we will perform analyses using the mean difference (MD) with 95% confidence intervals (CIs). If continuous outcomes are reported using different scales, we will use standardised mean difference (SMD).

If available, we will extract and report hazard ratios (HRs) for time‐to‐event‐data (mortality or time in hospital) data. If HRs are not available, we will make every effort to estimate as accurately as possible the HR using the available data and a purpose‐built method based on the Parmar and Tierney approach (Parmar 1998; Tierney 2007). If sufficient studies provide HRs, we will use HRs in favour of risk ratios (RRs) or MDs in a meta‐analysis, but for completeness, we will also perform a separate meta‐analysis of data from studies providing only RRs or MDs for the same outcome.

For dichotomous outcomes, we will report the pooled RR with a 95% CI. (Deeks 2011). Where the number of observed events is small (< 5% of sample per group), and where trials have balanced treatment groups, we will report the Peto’s Odds Ratio (OR) with 95% CI (Deeks 2011).

For cluster‐RCTs, we will extract and report direct estimates of the effect measure (e.g. RR with a 95% CI) from an analysis that accounts for the clustered design. We will obtain statistical advice (MT) to ensure the analysis is appropriate. If appropriate analyses are not available, we will make every effort to approximate the analysis following the recommendations in Chapter 16 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011d).

If data allow, we will undertake quantitative assessments using Review Manager 5 (RevMan 2014).

Non‐randomised studies (Non‐RCTs)

For dichotomous outcomes, if available we will extract and report the RR with a 95% CI from statistical analyses adjusting for baseline differences (such as Poisson regressions or logistic regressions) or the ratio of risk ratios (i.e. the risk ratio post‐intervention/risk ratio pre‐intervention). For continuous variables, if available we will extract and report the absolute change from a statistical analysis adjusting for baseline differences (such as regression models, mixed models or hierarchical models), or the relative change adjusted for baseline differences in the outcome measures (i.e. the absolute post‐intervention difference between the intervention and control groups, as well as the absolute pre‐intervention difference between the intervention and control groups/the post‐intervention level in the control group) (EPOC 2015).

If data allow, we will undertake quantitative assessments using Review Manager 5 (RevMan 2014).

All studies

Where appropriate, we will report the number needed to treat to for an additional beneficial outcome (NNTB) and the number needed to treat for an additional harmful outcome (NNTH) with 95% CIs.

If we cannot report the available data in any of the formats described above, we will perform a narrative report, and if appropriate, we will present the data in tables.

Unit of analysis issues

We do not expect to encounter unit of analysis issues as cluster‐RCTs, cross‐over studies and multiple observations for the same outcome are unlikely to be included in this review. Should any studies of these designs arise, we will treat these in accordance with the advice given in Chapter 16 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011c). If participants are randomised more than once, we will contact the authors of the study to provide us with data associated with the initial randomisation. For studies with multiple treatment groups, two review authors (RM and LE) will exclude subgroups that are considered irrelevant to the analysis. We will tabulate all subgroups in the ’Characteristics of included studies’ table. When appropriate, we will combine groups to create a single pair‐wise comparison. If this is not possible, we will select the most appropriate pair of interventions and exclude the others (Higgins 2011c).

Dealing with missing data

Where we identify data to be missing or unclear in published literature, we will contact study authors directly. If unsuccessful, our analysis will be based on the number reaching follow‐up and we will perform analysis for worse‐ and best‐case scenarios. We will record the number of patients lost to follow‐up for each study. Where possible, we will analyse data by intention‐to‐treat (ITT), but if insufficient data are available, we will present per protocol (PP) analyses (Higgins 2011c).

Assessment of heterogeneity

We will analyse the data in RCTs, non‐RCTs, and CBA studies separately.

If the clinical and methodological characteristics of individual studies are sufficiently homogeneous, we will combine the data and perform a meta‐analysis. We will assess the extent of heterogeneity by both visual inspection of forest plots and utilising statistical methods.

We will assess statistical heterogeneity of treatment effects between studies using a Chi2 test with a significance level at P < 0.1. We will use the I2 statistic to quantify the degree of potential heterogeneity and classify it as low if I2 ≤ 50%, moderate if I2 is 50% to 80% or considerable if I2 is > 80%. We will use the random‐effects model for low to moderate heterogeneity. If statistical heterogeneity is considerable, the overall summary statistic will not be reported.

Potential causes of heterogeneity will be assessed by sensitivity and subgroup analyses (Deeks 2011).

Assessment of reporting biases

We will explore potential publication bias (small‐trial bias) by generating a funnel plot and using a linear regression test if we find at least 10 studies are identified for inclusion in a meta‐analysis, We will consider a P value < 0.1 as significant for this test (Sterne 2011). Data synthesis If studies are sufficiently homogenous in their study design, we will conduct a meta‐analysis according to the recommendations of Cochrane (Deeks 2011).

Data synthesis

If studies are sufficiently homogenous in their study design, we will conduct a meta‐analysis according to the recommendations of Cochrane (Deeks 2011). We will not conduct meta‐analyses that include both RCTs and non‐RCTs. We will conduct separate meta‐analyses for each comparison. Different thresholds within the comparisons will only be grouped together if they are considered to be clinically similar.

Randomised controlled trials (RCTs)

For RCTs where meta‐analysis is feasible, we will use the random‐effects model for pooling the data. For binary outcomes, we will base the estimation of the between‐study variance on the Mantel‐Haenszel estimator. We will use the inverse‐variance method for continuous outcomes, outcomes that include data from cluster‐RCTs, or outcomes where HRs are available. If heterogeneity is found to be above 80%, and we identify a cause for the heterogeneity, we will explore this with subgroup analyses. If we cannot find a cause for the heterogeneity then we will not perform a meta‐analysis, but comment on the results as a narrative with the results from all studies presented in tables.

Non‐randomised studies (non‐RCTs)

If meta‐analysis is feasible for non‐RCTs or CBA studies, we will analyse non‐RCTs and CBA studies separately. We will only analyse outcomes with adjusted effect estimates if these are adjusted for the same factors using the inverse‐variance method as recommended in chapter 13 of the Cochrane Handbook of Systematic Reviews of Interventions (Reeves 2011).

All studies

We will use the random‐effects model for all analyses as we anticipate that true effects will be related but will not be the same for included studies. If we cannot perform a meta‐analysis. we will comment on the results as a narrative with the results from all studies presented in tables.

'Summary of findings' table

We will use the GRADE tool (study limitations, consistency of effect, imprecision, indirectness and publication bias) to assess the quality of evidence for each outcome. We will present a 'Summary of findings' table as suggested in Chapters 11 and 12 of the Cochrane Handbook for Systematic Reviews of Interventions (Schünemann 2011a; Schünemann 2011b).

We will use the GRADE approach to rate the quality of the evidence as ‘high’, ‘moderate’, ‘low’, or ‘very low’ using the five GRADE considerations.

  • Risk of Bias: serious or very serious

  • Inconsistency: serious or very serious

  • Indirectness: serious or very serious

  • Imprecision: serious or very serious

  • Publication bias: likely or very likely

The outcomes we will include are listed below in order of most relevant endpoints for participants.

  • All‐cause mortality

  • Mortality secondary to bleeding

  • Mortality secondary to thromboembolism

  • Mortality secondary to infection

  • Number of participants with major procedure‐related bleeding within seven days of surgery

  • Number of participants with minor procedure‐related bleeding within seven days of surgery

  • Serious adverse events due to platelet transfusions

Subgroup analysis and investigation of heterogeneity

If adequate data are available, we will perform subgroup analyses for each of the following outcomes in order to assess the effect on heterogeneity.

  • Age of participant (neonate, infant, child, adult)

  • Type of surgery: minor or major (cardiac, eye, neurosurgery, dental, orthopaedic, liver, obstetric, gynaecological, plastic, gastrointestinal)

  • Underlying cause of thrombocytopenia (bone marrow failure due to disease or treatment, increased destruction of platelets, or increased consumption of platelets)

  • Dose of platelet component

  • Co‐existing coagulopathy

  • Co‐existing platelet dysfunction (inherited or acquired)

Sensitivity analysis

We will assess the robustness of the results by performing the following sensitivity analyses when possible.

  • Including studies with a ‘low risk of bias’ (e.g.RCTs with methods assessed as low risk for random sequence generation and concealment of treatment allocation).

  • Including studies with less than a 20% dropout.