Direct thrombin inhibitors and factor Xa inhibitors for acute coronary syndromes: a network meta‐analysis

Objectives

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

To evaluate the acute‐phase efficacy and safety of parenteral direct thrombin inhibitors and factor Xa inhibitors in people with ACS during hospitalisation, with rank ordering by using network meta‐analysis. 

Background

Description of the condition

Cardiovascular diseases account for more than 17 million annual deaths worldwide. Among these diseases, acute coronary syndromes (ACS) are severely critical conditions that require immediate hospital care and often invasive life‐supporting treatment (Roth 2017). The pathogenesis of ACS is characterised by the rapid imbalance of myocardial oxygen consumption and oxygen supply from coronary arteries, which usually results from obstruction of coronary arteries by thrombi following plaque formation and rupture processes (Amsterdam 2014). ACS comprises a range of clinical conditions, including unstable angina, non‐ST‐segment elevation myocardial infarction (NSTEMI) and ST‐segment elevation myocardial infarction (STEMI). Unstable angina and NSTEMI are also classified collectively as non‐ST‐segment elevation ACS (NSTE‐ACS) (O'Gara 2013; Roffi 2015).

Since thrombogenesis in coronary arteries plays a central role in the pathogenesis of ACS, the appropriate management of ACS includes not only invasive cardiac procedures to treat anatomical obstruction, but also medical treatments involving antithrombotic therapy, which is indicated in all phases of ACS (CCSAP 2017). Anticoagulants and antiplatelets are two types of antithrombotic agents, and a Cochrane Review has shown that anticoagulants in combination with antiplatelets reduced recurrent thrombotic events in NSTE‐ACS (Andrade‐Castellanos 2014). In addition, invasive cardiovascular procedures, such as percutaneous coronary intervention (PCI) and other mechanical supports, are highly thrombogenic and require adequate anticoagulation therapy (Zeymer 2016). Therefore, parenteral anticoagulation therapy is a reasonable strategy in the acute phase of ACS.

Heparins (unfractionated heparin (UFH) and the low molecular weight heparins (LMWH)) have remained as the anticoagulants of choice during the acute phase of ACS (Roffi 2015). UFH is a sulphate‐polysaccharide, which functions as an indirect factor Xa and factor IIa (thrombin) inhibitor by converting antithrombin (AT) into an activated form. Bleeding is a major side effect of UFH, which is common in all types of antithrombotic agents. With its relatively large molecule size, UFH interacts with plasma proteins and leads to difficulty in determining the optimal dose as well as adverse events such as heparin‐induced thrombocytopenia (HIT) ‐ this forms the basis to monitor activated partial thromboplastin time (APTT) and platelet counts (Garcia 2012). LMWH also indirectly inhibit factor Xa via AT but have less effect on thrombin. There are several advantages of LMWH over UFH, due to more predictable anticoagulant effects and reduced risks of HIT. Enoxaparin is the primary agent among LMWH recommended in the current guidelines for ACS (Roffi 2015). However, enoxaparin is associated with a higher frequency of minor bleeding, and uncertainty remains regarding its use in obese patients or in patients with significant renal insufficiency. Enoxaparin thus requires special monitoring procedures to achieve adequate dosing. 

Description of the intervention

Other than heparins, which indirectly inhibit both factor IIa (thrombin) and factor Xa activities via antithrombin activation, factor Xa inhibitors and direct thrombin inhibitors have been developed and tested in ACS populations to improve clinical outcomes. Fondaparinux, a factor Xa inhibitor, works as an anticoagulant binding to AT. It has pharmacological aspects comparable to those of LMWH, with no interaction with plasma proteins and complete bioavailability after subcutaneous injection. Another parenteral selective factor Xa inhibitor, otamixaban, which directly and selectively inhibits factor Xa, showed no benefit in NSTE‐ACS compared to UFH plus eptifibatide, a glycoprotein IIb/IIIa inhibitor, and is not currently used in the management of ACS (Sabatine 2009; Steg 2013).

Another choice of parenteral anticoagulants is direct thrombin inhibitors, which selectively bind to thrombin, inhibiting thrombin‐mediated cleaving processes of fibrinogen to fibrin. Bivalirudin, a synthetic polypeptide, is usually available for the management of ACS, and it does not interact with plasma proteins and thus is not associated with risk of HIT. Recent clinical trials have illustrated its efficacy and the current guidelines on ACS recommend the use of bivalirudin (Roffi 2015; Ibanez 2017). Other than bivalirudin, argatroban is a choice of an anticoagulant for patients who present HIT. Trials that investigated the efficacy of argatroban for patients with ACS were conducted mainly before PCI was established, and the data might not reflect current clinical practice (Jang 1999). 

How the intervention might work

In the acute phase of ACS, parenteral anticoagulants are primarily utilised, since quick exhibition of anticoagulant effects is vital. There are no clinical trials evaluating the role of direct oral anticoagulants (DOACs) in the acute management of ACS. Current clinical guidelines discuss the use of four parenteral anticoagulants separately, based on ST‐elevation and concurrent treatments: UFH, LWMH, fondaparinux, and bivalirudin.

In STEMI, primary PCI and fibrinolysis are the two key treatment options in addition to antiplatelet therapy, and anticoagulation could differ, depending on concomitant treatments. During primary PCI, UFH is recommended as class I in ESC guidelines (Ibanez 2017). Bivalirudin has been compared to UFH for primary PCI in detail; early trials showed non‐inferiority of bivalirudin to UFH with reducing major bleeding events, while recent studies have cast a question on the superiority of bivalirudin (Stone 2008; Shahzad 2014). Enoxaparin is another alternative to UFH in STEMI, as a randomised trial showed similar or slightly better clinical outcomes with less bleeding risk (Montalescot 2011). Fondaparinux, however, is not recommended for primary PCI because of potential higher risks of catheter thrombosis (Yusuf 2006). Fibrinolysis is also an important treatment option for STEMI, but PCI provides better clinical outcomes and STEMI patients do not frequently undergo fibrinolysis these days (Keeley 2003). UFH and enoxaparin are the two specified anticoagulants in the ESC Guidelines as class I recommendations for fibrinolysis, while fondaparinux could potentially cause catheter thrombosis during PCI and bivalirudin has not yet been fully examined in fibrinolysis for STEMI (Ibanez 2017).

On the other hand, the current guidelines recommend all four prespecified anticoagulants as class I for NSTE‐ACS (Amsterdam 2014; Roffi 2015). Previously, a meta‐analysis reported that UFH and LMWH improved clinical outcomes in NSTE‐ACS patients than placebo with the concurrent use of aspirin (Eikelboom 2000). Although there have been clinical trials on the use of fondaparinux for NSTE‐ACS, the role of fondaparinux in ACS remains controversial (Steg 2010; Szummer 2015). Fondaparinux is recommended in the ESC guidelines, but it is not approved by the US Food and Drug Administration (FDA) for ACS (Roffi 2015). In the ESC guidelines, bivalirudin received a class IA recommendation for NSTE‐ACS as an alternative to UFH plus glycoprotein IIb/IIIa inhibitors during PCI, based on early trials showing lower bleeding rates (Stone 2006: Kastrati 2011). Other recent studies have been conducted to reveal the efficacy and safety profile of bivalirudin in the management of ACS (Cavender 2014; Valgimigli 2015; Erlinge 2017).

Why it is important to do this review

It is critical to determine which anticoagulant agent should be used for ACS, since this decision could have a direct and significant impact on patients' prognoses. For this particular clinical question, where a number of interventions are available, conducting a network meta‐analysis is ideal since the methodology allows for integration of available direct and indirect comparative evidence on relative treatment effects, as well as ranking of efficacy and safety of available interventions. This is especially important for intervention types where head‐to‐head comparisons are lacking. A comprehensive systematic review, together with a network meta‐analysis that included data from more than 110,000 participants, showed no significant differences in mortality rates, although there was a difference in the odds of events such as major bleeding and myocardial infarction (MI) between anticoagulants (Navarese 2015). However, a definitive answer has not yet been obtained, and there have recently been more trials regarding the safety and effectiveness of anticoagulants for ACS. Our Cochrane Review will inform decision‐making by providing evidence on the efficacy and safety of direct thrombin inhibitors and factor Xa inhibitors.

Objectives

To evaluate the acute‐phase efficacy and safety of parenteral direct thrombin inhibitors and factor Xa inhibitors in people with ACS during hospitalisation, with rank ordering by using network meta‐analysis. 

Methods

Criteria for considering studies for this review

Types of studies

We will include randomised parallel‐group trials, as well as cross‐over trials, in which parenteral direct thrombin inhibitors and factor Xa inhibitors have been compared with other anticoagulant strategies, including control groups. For cross‐over trials, we will incorporate data from the first period only, that is before participants cross over. We will narratively discuss the risk of bias deriving from cross‐over trials, especially selection bias, as appropriate. We will also include cluster‐randomised trials after adjusting for the intra‐class correlations. In view of heterogeneity, we will analyse such trials separately in sensitivity analysis.

Types of participants

We will include adults (aged ≥ 18 years) hospitalised with a clinical diagnosis of unstable angina, STEMI or NSTEMI, according to study investigator definitions. Definitions may include the universal definition of ACS (Thygesen 2012; Thygesen 2018). We will assume that any patient that meets all inclusion criteria is, in principle, equally likely to be randomised to any of the interventions in the synthesis comparator set.

Types of interventions

We will include trials investigating the efficacy and safety of any parenteral direct thrombin inhibitors (e.g. bivalirudin) or factor Xa inhibitors (e.g. fondaparinux) in an acute phase of ACS at any dose in an inpatient setting. Comparator interventions will be another anticoagulant (including UFH and LMWH). We will exclude parenteral direct thrombin inhibitors and factor Xa inhibitors that are not approved by the FDA and the European Medicines Agency (e.g. otamixaban). We will analyse concomitant use of glycoprotein IIb/IIIa inhibitors separately in a subgroup analysis.

Types of outcome measures

Reporting one or more of the outcomes listed here in the trial is not an inclusion criterion for the review. Where a published report does not appear to report one of these outcomes, we will access the trial protocol and contact the trial authors to ascertain whether the outcomes were measured but not reported. We will include relevant trials that measured these outcomes but did not report the data at all, or did not report data in a usable format, in the review as part of the narrative analysis.

If data on individual endpoints are not available, or if only composite endpoints (e.g. a composite of all‐cause death, MI, ischaemic stroke) are available, we will describe these study data narratively. We do not plan to include such data in a meta‐analysis. Given that our intention is to assess the acute effects of parenteral anticoagulants in ACS, we are interested in outcome data measured within a follow‐up duration of 30 days or less from the start of the intervention. We are interested in the number of participants experiencing at least one event, and not the total number of events overall.

If a trial does not have raw data and only reports hazard ratios (HRs) in terms of mortality, we will estimate the outcome data from the Kaplan‐Meier plots using the DigitizeIt software (DigitizeIt 2012), or the R software package IPDfromKM (Liu 2021).

Several bleeding criteria are commonly used in clinical trials, such as the 'Thrombolysis in Myocardial Infarction' (TIMI) bleeding criteria, where major bleeding is defined as any intracranial bleeding (excluding microhaemorrhages < 10 mm evident only on gradient‐echo MRI); clinically overt signs of haemorrhage associated with a drop in haemoglobin of ≥ 5 g/dL; or fatal bleeding (bleeding that directly results in death within seven days). Other bleeding assessment systems include the 'Global Use of Strategies to Open Occluded Arteries' (GUSTO) bleeding criteria and the 'Bleeding Academic Research Consortium' (BARC) definition for bleeding (Mehran 2011). We will consider major and minor bleeding events, as defined by each trialist. 

Primary outcomes

1. All‐cause mortality

2. Cardiovascular mortality

3. Non‐fatal acute myocardial infarction or re‐infarction

4. Major bleeding, as defined by the study investigators, based on the aforementioned clinically validated bleeding criteria

Secondary outcomes

1. Minor bleeding (based on a validated assessment criteria such as TIMI, GUSTO and BARC)

2. Stroke (ischaemic, haemorrhagic or stroke of unknown causes)

3. Stent thrombosis

Search methods for identification of studies

Electronic searches

We will identify trials through systematic searches of the following bibliographic databases:

• Cochrane Central Register of Controlled Trials (CENTRAL) in the Cochrane Library;

• MEDLINE (Ovid, from 1946 onwards);

• Embase (Ovid, from 1980 onwards);

• CPCI‐S on the Web of Science (Clarivate Analytics, from 1990 onwards).

We will adapt the preliminary search strategy for MEDLINE (Ovid) (Appendix 1) for use in the other databases. We will apply the Cochrane sensitivity‐maximising randomised controlled trial (RCT) filter to MEDLINE (Ovid) and adaptations of it to the other databases, except CENTRAL (Lefebvre 2011).

To identify ongoing or unpublished trials, we will also conduct a search of the US National Library of Medicine's trials registry, ClinicalTrials.gov (clinicaltrials.gov), as well as the World Health Organization International Clinical Trials Registry Platform (ICTRP) (apps.who.int/trialsearch).

We will search all databases from their inception to the search date with no restrictions on language of publication or publication status. To reflect current trial reporting status of efficacy and safety of direct thrombin inhibitors and factor Xa inhibitors in the context of ACS management, we will only consider data on adverse effects that are described in the included studies.

Searching other resources

We will check the reference lists of all included studies and any relevant systematic reviews identified for additional references to trials. We will also examine any relevant retraction statements and errata for included studies. To identify unpublished or ongoing studies, we will contact known experts in the field who have recently published related review articles in top‐tiered peer‐reviewed journals in cardiovascular medicine (e.g. Circulation, European Heart Journal, Journal of the American College of Cardiology, Nature Reviews Cardiology). 

Data collection and analysis

Selection of studies

Two review authors (TH, AM) will independently screen titles and abstracts for inclusion of all the potential studies we identify as a result of the comprehensive literature search. We will categorise resulting records either as 'retrieve full texts' (eligible or potentially eligible/unclear) or 'clearly irrelevant'. If there are any disagreements, a third review author (JK) will be asked to arbitrate. We will retrieve the full‐text study articles. The same two review authors (TH, AM) will independently screen the full‐text articles against prespecified eligibility criteria to identify studies for inclusion. We will identify and record the reasons for exclusion of ineligible studies. We will resolve any disagreements through discussion, or, if required, we will consult a third review author (JK). We will remove duplicates and collate multiple reports of the same study so that each study, rather than each report, is the unit of interest in the review. We will record the selection process in sufficient detail to complete a PRISMA flow diagram and 'Characteristics of excluded studies' table (Liberati 2009).

Data extraction and management

We will use a pre‐standardised data extraction form for study characteristics and outcome data, which we will pilot on at least one study in the review. One review author (TH) will extract study characteristics from the included studies. We will extract the following study characteristics.

• Methods: study design, total duration of study, number of study centres and location, study setting, and date of study.

• Participants: N randomised, N lost to follow‐up/withdrawn, N analysed, mean age, age range, gender, severity of condition, diagnostic criteria as applied by the trialists, baseline renal function (likely to be reflected by creatinine clearance levels), past cardiac‐related medical history, study inclusion/exclusion criteria.

• Interventions: treatment intervention type, comparator, concomitant medications, and excluded medications.

• Outcomes: primary and secondary outcomes specified and collected, and time points reported.

• Notes: funding for trial, and notable conflicts of interest of trial authors.

For the purpose of network meta‐analyses, we will collect a list of clinical and methodological characteristics that may pose as potential effect modifiers, such as age, gender, type of index diagnosis (unstable angina, NSTEMI, STEMI), medications at randomisation (e.g. antiplatelet drugs), dosage and route of administration of study intervention. We will inspect and discuss these effort modifiers before conducting the network meta‐analysis as part of our assessment of transitivity (Assessment of heterogeneity).

Two review authors (TH, AM) will independently extract outcome data from the included studies. We will resolve disagreements by consensus or by involving a third review author (JK). One review author (TH) will transfer data into the Review Manager 5 file (Review Manager 2020). We will double‐check that data are entered correctly by comparing the data presented in the systematic review with the data extraction form. A second review author (AM or JK) will spot‐check study characteristics for accuracy against the trial report.

Assessment of risk of bias in included studies

Two review authors (TH, AM) will independently assess risk of bias for each study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017). We will resolve any disagreements by discussion or by involving another review author (JK). 

We will assess the risk of bias according to the following domains:

• random sequence generation;

• allocation concealment;

• blinding of participants and personnel;

• blinding of outcome assessment;

• incomplete outcome data;

• selective outcome reporting.

For cluster‐randomised trials, we will assess the risk of bias according to the following domains (Higgins 2011):

• recruitment bias;

• baseline imbalance;

• loss of clusters;

• incorrect analysis;

• comparability with individually randomised trials.

We will assess whether the primary and secondary outcome measures are prespecified and whether there is consistency in reported outcomes. We will evaluate the reporting bias as: low risk of bias (all outcomes prespecified); unclear risk of bias (insufficient or no information regarding prespecified outcomes); high risk of bias (one or more reported primary outcomes not prespecified).

We will grade each potential source of bias as high, low, or unclear and provide a quote from the study report together with a justification for our judgment in the risk of bias table. We will summarise the risk of bias judgments across different studies for each of the domains listed. Where information on risk of bias relates to unpublished data or correspondence with a trialist, we will note this in the risk of bias table.

When considering treatment effects, we will take into account the risk of bias for the studies that contribute to that outcome.

Measures of treatment effect

We will analyse all dichotomous data such as mortality, infarction or reinfarction, and stroke as odds ratios (ORs) with 95% confidence intervals (CIs). 

For indirect comparisons, we will calculate the surface under the cumulative ranking curve (SUCRA) for all outcomes, which considers the estimated effect sizes and associated level of uncertainty, with summary statistics to illustrate the ranking of the interventions and treatment hierarchy (Salanti 2011). By definition, SUCRA is a value between 0 and 1, and the larger the SUCRA, the higher the treatment in the hierarchy according to the outcome (efficacy or safety), thus a larger SUCRA theoretically implies the intervention is more effective or is safer. The effectiveness of each included intervention will therefore be orderly ranked in accordance with their effect estimates (ORs), and plots of the treatment rank probabilities will be provided to rank the various interventions. 

Unit of analysis issues

The unit of analysis in this review is the individual participant according to the study arm to which the participant was randomly assigned. If an included trial compares more than two intervention groups, we will split the number of participants in the control group into two or more groups in order to avoid double‐counting. Since network meta‐analysis allows consideration of correlation between effect sizes from multi‐arm trials (Franchini 2012), we will take into account the respective treatment effects from the same studies.

Dealing with missing data

We will contact investigators or study sponsors in order to verify key study characteristics and to obtain missing numerical outcome data where possible (e.g. when a study is identified as abstract/conference proceeding only). Where possible, we will use the Review Manager 5 calculator to calculate missing standard deviations using other data from the trial (Review Manager 2020), such as CIs or standard errors. Where this is not possible, and the missing data are thought to introduce serious bias, we will explore the impact of including such studies in the overall assessment of results by conducting a sensitivity analysis (Sensitivity analysis).

Assessment of heterogeneity

We will employ a Bayesian hierarchical randomised consistency model (Dias 2013). If we observe disagreement between direct evidence and indirect evidence (i.e. inconsistency), we will adopt the node‐splitting approach that enumerates all possible comparisons with independent indirect evidence and calculate the P value for each split comparison (van Valkenhoef 2016). We will use the I² statistic to measure heterogeneity among the included trials in both standard pairwise meta‐analyses and network meta‐analyses, but acknowledge that there is substantial uncertainty in the value of I² when there is only a small number of studies. We will also consider the P value from the Chi² test, for which P < 0.1 will be considered for statistical significance (Deeks 2011). If we identify important heterogeneity as per the following thresholds provided in the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2011), we will report it and explore possible causes by prespecified subgroup analysis (Subgroup analysis and investigation of heterogeneity).

• 0% to 40%: might not be important.

• 30% to 60%: may represent moderate heterogeneity.

• 50% to 90%: may represent substantial heterogeneity.

• 75% to 100%: considerable heterogeneity.

We will inspect forest plots visually to consider the direction and magnitude of effects and the degree of overlap between CIs. In the case of considerable heterogeneity, we will not pool the study results but will instead describe them narratively.

For the conduct of network meta‐analyses, our rationale to perform such indirect comparisons is based on the assumption of transitivity relative to clinical and methodological heterogeneity, for which there are no substantial differences between comparisons on effect modifiers other than the type of treatment interventions being compared. In addition to the evaluation of the assumption of homogeneity and transitivity in the network meta‐analysis from a clinical perspective, we will compare the distribution of the potential effect modifiers (Data extraction and management) across the different pairwise comparisons.

Assessment of reporting biases

If we are able to include more than 10 trials, we will create and examine a funnel plot for the standard pairwise meta‐analyses to explore possible small‐study biases for the primary outcomes. We will visually inspect the funnel plot for asymmetry (Egger 1997).

Data synthesis

We will undertake meta‐analyses only if there are sufficient data (with at least two studies) and where this is meaningful, that is if the treatments, participants and the underlying clinical question are similar enough for pooling to make sense (Chaimani 2021).

For standard pairwise comparisons, we will synthesise data and present the comparison results (ORs and 95% CIs) estimated from a random‐effects model using Review Manager 5 (Review Manager 2020). We will use random‐effects model for direct comparison meta‐analysis if the level of heterogeneity is judged to be substantial; otherwise, we will use a fixed‐effect model. In the case of considerable heterogeneity, we will describe the study findings narratively (Assessment of heterogeneity).

We will conduct network meta‐analyses to compare the different direct thrombin inhibitors and factor Xa inhibitors for each of the listed outcome measures. We will use a network plot to visually inspect comparisons between interventions.

We will use non‐informative priors with uniform (0‐1) prior distribution for hyper‐parameters for heterogeneity (Lu 2004). Within the Bayesian hierarchical model methodological framework, we will first perform 50,000 Markov Chain Monte Carlo (MCMC) simulations, and then we will generate an additional 10,000 simulations with three sets of different initial values and shear the first several simulations as the burn‐in period in our model. We will check the convergence of the MCMC by checking the trajectory of each chain and using the Brooks‐Gelman Rubin statistical method for assessing model convergence (Brooks 1998). Based on 50,000 simulations with 50 thin (i.e. extract every 50th value), the point estimate will be adopted as the median of the posterior distribution, and the corresponding 95% credible intervals (CrIs) based on the 2.5th and 97.5th percentiles of the posterior distributions will be interpreted similarly as the conventional 95% CIs. The goodness‐of‐fit will be evaluated by a deviance information criterion (DIC). We will conduct all network meta‐analyses using WinBUGS, Lund 2000, and R software, Rücker 2019.

Subgroup analysis and investigation of heterogeneity

We plan to carry out the following subgroup analyses for our a priori primary outcomes that display considerable heterogeneity (Deeks 2011).

• Age (≥ 60 years versus < 60 years).

• Index diagnosis (unstable angina, STEMI or NSTEMI).

• Undergoing percutaneous coronary intervention (PCI) or coronary artery bypass graft (CABG).

• Glycoprotein IIb/IIIa inhibitor use.

We will use the four primary outcomes in subgroup analyses (Primary outcomes). We will use the formal test for subgroup differences in Review Manager 5 (Review Manager 2020), and base our interpretation on this.

Sensitivity analysis

We plan to carry out the following sensitivity analyses to test whether key methodological factors or decisions have affected the main result.

• Only including studies with a low risk of bias (defined as studies in which at least four key domains, namely random sequence, allocation concealment, incomplete outcome data and selective reporting, are assessed as low risk of bias) (Assessment of risk of bias in included studies).

• Excluding studies with missing data (Dealing with missing data).

• Random‐effects versus fixed‐effect model (Data synthesis).

• Cluster‐RCTs versus non‐cluster‐RCTs.

We will use the four primary outcomes in sensitivity analyses (Primary outcomes).

Summary of findings and assessment of the certainty of the evidence

We will use the methods and recommendations described in the Cochrane Handbook for Systematic Reviews of Interventions (Schünemann 2017), as well as the GRADE four‐step approach to rate the quality of the effect estimates for all comparisons and for each of the aforementioned outcome measures (Puhan 2014; Brignardello‐Petersen 2018), employing GRADEpro GDT software (GRADEpro GDT). We will create a summary of findings table for all of our prespecified primary and secondary outcomes (see Types of outcome measures) based on the format illustrated by Yepes‐Nuñez 2019: 

1. all‐cause mortality;

2. cardiovascular mortality;

3. non‐fatal acute myocardial infarction or re‐infarction;

4. major bleeding;

5. minor bleeding;

6. stroke (ischaemic, haemorrhagic, or stroke of unknown causes);

7. stent thrombosis.

We will assess the certainty of the evidence from direct‐comparison estimates, indirect‐comparison estimates, and network meta‐analysis estimates based on the five GRADE considerations: risk of bias, inconsistency of effect, indirectness, imprecision and publication bias.

We will justify all decisions to downgrade the certainty of evidence using footnotes in the summary of findings table, and will include comments to aid the reader's understanding of our judgments where necessary. Two review authors (TH, AM) will independently assess the certainty of the evidence, with any disagreements resolved by discussion or by involving another review author (JK). We will justify, document and incorporate judgments into the reporting of results for each outcome.

We will extract study data, construct our comparisons in data tables, and prepare a summary of findings table before proceeding to write the results and conclusions of our review.

Appendices

Appendix 1. MEDLINE (Ovid) search strategy

1     Acute Coronary Syndrome/ (17482)

2     acute coronary syndrome*.tw. (33833)

3    ACS.tw. (24382)

4     exp Angina, Unstable/ (11120)

5    (unstable adj2 angina).tw. (13515)

6    ((preinfarct* or pre infarct*) adj2 angina).tw. (333)

7     ST Elevation Myocardial Infarction/ (5280)

8    STEMI.tw. (12282)

9     ST elevat* myocardial infarction*.tw. (9776)

10    Non‐ST Elevated Myocardial Infarction/ (1186)

11    NSTEMI.tw. (2687)

12    non‐ST elevat* myocardial infarction*.tw. (2062)

13    non‐ST‐segment elevation ACS.tw. (219)

14    NSTE‐ACS.tw. (1103)

15     1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 or 12 or 13 or 14 (77175)

16    direct thrombin inhibitor*.tw. (2445)

17     exp Factor Xa Inhibitors/ (8379)

18    factor xa inhibit*.tw. (2680)

19    Fondaparinux/ (1131)

20    Fondaparinux.tw. (1785)

21    bivalirudin.tw. (1549)

22     16 or 17 or 18 or 19 or 20 or 21 (12948)

23     15 and 22 (1257)

24    randomized controlled trial.pt. (542570)

25    controlled clinical trial.pt. (94368)

26    randomized.ab. (532663)

27    placebo.ab. (221147)

28     drug therapy.fs. (2368894)

29    randomly.ab. (365180)

30    trial.ab. (566659)

31    groups.ab. (2242115)

32     24 or 25 or 26 or 27 or 28 or 29 or 30 or 31 (5107910)

33     exp animals/ not humans.sh. (4881960)

34     32 not 33 (4442669)

35     23 and 34 (970)

 

The Cochrane sensitivity‐maximising RCT filter has been applied.

Contributions of authors

Tomoki Hattori: developed and finalised the protocol

Atsushi Mizuno: developed and finalised the protocol

Daisuke Yoneoka: developed and finalised the protocol

Wilson Wai San Tam: developed and finalised the protocol

Joey SW Kwong: conceived the concept; developed and finalised the protocol

Sources of support

Internal sources

• Internal Support, Other

  No internal support provided for the conduct of this Cochrane protocol.

External sources

• National Institute for Health Research (NIHR), UK

  This project was supported by the NIHR, via Cochrane Infrastructure funding to the Heart Group. The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the Systematic Reviews Programme, NIHR, NHS or the Department of Health and Social Care.

Declarations of interest

Tomoki Hattori declares having no conflicts of interest.

Atsushi Mizuno declares having no conflicts of interest.

Daisuke Yoneoka declares having no conflicts of interest.

Wilson Wai San Tam declares having no conflicts of interest.

Joey SW Kwong declares working as a Managing Editor with the Cochrane Central Editorial Service without any involvement in the editorial process of the current manuscript.

New