Study reference

Study characteristics

Patient characteristics ²

Intervention (I)

Comparison / control (C) ³

Follow-up

Outcome measures and effect size ⁴

Comments

Cohen,

2016 (RAPS study)

Type of study: randomised, controlled, open-label, phase 2/3,

non-inferiority clinical trial

Setting and country: specialist

haematology and rheumatology clinics at University

College London Hospitals and Guy’s and St Thomas’

Hospitals NHS Foundation Trusts, London, UK, between June 5, 2013, and

Nov 11, 2014.

Funding and conflicts of interest:

-Funding: Arthritis Research UK, Comprehensive Clinical Trials Unit at UCL, LUPUS UK, Bayer, National Institute for Health Research Biomedical Research Centre.

-HC has received institutional research support from Bayer, with

honoraria for lectures and participation on an advisory board being

diverted to a local charity. DRJA has received honoraria from Bayer for

participation in an international meeting. The other authors declare no

competing interests.

Inclusion criteria:

-thrombotic APS (all patients enrolled met the international consensus criteria for APS)

-at least one venous thrombo- embolism when taking no or subtherapeutic

anticoagulant therapy

-taking standard-intensity warfarin (target international

normalised ratio (INR) 2·5) for at least 3 months since the last venous thromboembolic event.

-Women had to be using adequate contraception unless they were postmenopausal or had undergone sterilisation

Exclusion criteria:

-previous arterial thrombotic events due to APS

-recurrent venous thromboembolism when taking

warfarin at a therapeutic INR of 2·0–3·0

-younger than 18 years

-pregnancy or lactation

-severe renal impairment

-alanine aminotransferase more

than twice the upper limit of normal

-Child-Pugh class B or C cirrhosis -thrombocytopenia (platelets <75 × 10⁹/L)

-non-adherence to warfarin regimen (based on clinical

assessment)

-taking azole class antifungals, protease inhibitors (eg, ritonavir) for HIV, strong CYP3A4 inducers

(eg, rifampicin, phenytoin, carbamazepine, phenobarbital,

or St John’s wort), or dronedarone

N total at baseline:

N=116

Intervention: 57

Control: 59

Important prognostic factors²:

Age (years), mean ± sd

I: 47 ±17

C: 50 ±14

Male, n(%)

I: 15 (26%)

C: 17 (29%)

SLE (stratification variable), n(%)

I: 11 (19%)

C: 11 (19%)

Rivaroxaban dose, n(%)

20 mg once daily

I: 55 (96%)

15 mg once daily

I: 2 (4%)

Thrombin generation

ETP (nmol/L per min)

I: 555 (497–619)

C: 542 (469–626)

Groups comparable at baseline?

Baseline characteristics were similar in the two groups.

Describe intervention (treatment/procedure/test):

Rivaroxaban

20 mg oral rivaroxaban once daily (or 15 mg once daily depending on local clinical care and following the summary of product characteristics in

patients with creatinine clearance 30–49 mL/min) for 180 days.

Describe control (treatment/procedure/test):

Warfarin

Standard intensity warfarin with target INR 2·5 (range 2·0–3·0).

Length of follow-up

210 days

Loss-to-follow-up

Of the 116 patients, six (5%) did not contribute data for the primary outcome (ETP).

Intervention:

3/57 (5.3%)

Reasons (describe): 3 patients excluded

3 results <LLOD at

baseline

Control:

3/59 (5.1%)

Reasons (describe): 3 patients excluded

-1 results <LLOD at

baseline and day 42

-1 results <LLOD at

day 42

-1 withdrew before

day 42

Non-inferiority threshold

Threshold for non-inferiority of rivaroxaban: less than 20% difference from warfarin in the mean percentage change for the primary outcome (percentage change in ETP from randomisation to day 42 (first trial visit after

randomisation).

Statistical analyses

The authors describe in their Methods section that they used a modified intention-to-treat approach to include all randomised patients with assessable data in all analyses.

Thrombin generation at day 42: endogenous thrombin potential (ETP) (nmol/L per min)

I: 1086 (957 to 1233)

C: 548 (484 to 621)

Treatment effect (95% CI): 2·0 (1·7 to 2·4)

P<0·0001

New thrombotic events at day 210, n (%)

Deep vein thrombosis

I: 0/57 (0)

C: 0/58 (0)

Pulmonary embolism

I: 0/57 (0)

C: 0/58 (0)

Arterial thrombosis

I: 0/57 (0)

C: 0/58 (0)

Other

I: 0/57 (0)

C: 0/58 (0)

Any combination

I: 0/57 (0)

C: 0/58 (0)

Bleeding events at day 210, n(%)

Major

I: 0/57 (0)

C: 0/58 (0)

Clinically relevant

I: 3/57 (5)

C: 2/55 (4)

Treatment effect 1·7 (–5·9 to 9·3)

Minor

I: 10/57 (18)

C: 8/55 (15)

3·0 (–10·5 to 16·5)

Unclassified, insufficient information

I: 1/57 (2)

C: 0/58 (0)

Treatment effect 1·8 (–1·7 to 5·3)

Adverse events at day 210, n(%)

Serious adverse events

I: 4/57 (7)

C: 3/55 (5)

Treatment effect 1·5 (–7·5 to 10·5)

Serious adverse reactions

I: 0/57 (0)

C: 1/55 (2)

Treatment effect –1·8 (–5·3 to 1·7)

Suspected unexpected serious adverse reaction

I: 0/57 (0)

C: 0/58 (0)

Authors’ conclusion

Our findings suggest that in patients with APS who have had previous venous

thromboembolism and need standard-intensity anticoagulation (i.e, target INR 2·5) the overall thrombotic risk is not increased with rivaroxaban compared with that related to warfarin. The absence of new thrombosis or

major bleeding and low rate of clinically relevant bleeding supports this conclusion.

Overall, rivaroxaban seems efficacious and safe, and might offer a

convenient alternative to warfarin in this subgroup of patients with antiphospholipid syndrome.

Goldhaber, 2016

Type of study:

post hoc subgroup analyses of pooled

data from the RE-COVER and RE-COVER II trials, and

of data from the RE-MEDY trial

Setting and country: not reported by Goldhaber.

Funding and conflicts of interest:

-The RE-COVER and RE-MEDY studies were funded by Boehringer

Ingelheim.

-Consulting activities for Bayer and Boehringer Ingelheim amongst other things.

Inclusion criteria:

RE-COVER and RE-COVER II

-patients aged ⩾ 18 years

-objectively confirmed, symptomatic, proximal deep vein thrombosis or pulmonary embolism.

RE-MEDY

-patients aged ⩾ 18 years

-objectively confirmed, symptomatic, proximal deep vein thrombosis or pulmonary embolism.

-had been treated with an approved anticoagulant for 3–12

months, or with dabigatran in RE-COVER or RE-COVER II

Testing for thrombophilia and APS was not mandatory in the RE-COVER, RE-COVER II, or RE-MEDY trials. However, when tests were performed, results were recorded.

APS was defined as at least one positive test for lupus anticoagulant and/or for anticardiolipin antibodies combined with symptomatic, objectively verified venous thromboembolism.

Exclusion criteria:

Not reported by Goldhaber 2016.

N total at baseline:

To avoid double counting of patients, criteria were established to account for rollover patients who were treated in RE-COVER or

RE-COVER II and then enrolled in RE-MEDY.

N=151

Intervention: 71

Control: 80

The APS analysis was based on pooled data from 6822 treated patients from the RE-COVER, RE-COVER II, and RE-MEDY trials.

Important prognostic factors²:

Age (years), mean ± sd

I: 47.8 ±14.9

C: 47.4±16.8

Male, n (%)

I: 47 (66.2)

C: 49 (61.3)

Groups comparable at baseline?

Not reported for APS patients receiving dabigatran etexilate versus warfarin. Patients with APS were slightly younger than those who had no APS or had not been tested.

RE-COVER and RE-COVER II

warfarin–placebo+ parenteral anticoagulation for ⩾ 5 days. After discontinuing parenteral therapy,

patients continued warfarin (therapeutic INR range 2.0–

3.0) or received dabigatran etexilate 150 mg twice daily

for 6 months (double-dummy, oral-only treatment period).

RE-MEDY

dabigatran etexilate 150 mg twice daily for 6–36 months.

RE-COVER and RE-COVER II

warfarin

RE-MEDY

warfarin (INR range 2.0–3.0) for 6–36

months.

Length of follow-up

RE-COVER and RE-COVER II

From randomization to the end of

the prespecified treatment period: 6 months.

RE-MEDY

From randomization to the end of the planned treatment period: 6–36 months.

Incomplete outcome data

Intervention: 1/71 (1.4%)

Control: 3/80 (3.8%)

The overall proportion of patients with known APS in the pooled RE-COVER/RE-COVER II and RE-MEDY analysis was 2.2%. Antiphospholipid syndrome comprised 21% of the known thrombophilias.

Major bleeding events, n (%)

According to the International Society on Thrombosis and Haemostasis (ISTH) criteria

I: 1/70 (1.4)

C: 2/77 (2.6)

HR 0.46 (0.04, 5.43)

Major bleeding events/Clinically relevant bleeding events, n (%)

I: 6/70 (8.6)

C: 14/77 (18.2)

HR 0.53 (0.20, 1.41)

Any bleeding event, n (%)

I: 14/70 (20.0)

C: 31/77 (40.3)

HR 0.50 (0.26, 0.95)

Venous thromboembolism/venous thromboembolism-related deaths, n (%)

I: 3/71 (4.2)

C: 4/80 (5.0)

HR 0.43 (0.08, 2.38)

Remarks

-not all patients were tested for APS. Therefore, our results do not reflect the true number of patients with this conditions

-the diagnosis of APS did not follow the revised Sapporo criteria with two tests at least

12 weeks apart, and it was not known whether all titers of antibodies reached the required criteria. Without the

Sapporo criteria being met, perhaps these patients could be considered as having the lupus anticoagulant disorder rather than APS.

-Limitations of a post hoc analysis preclude from providing precision in patients with the most severe thrombophilic disorders.

Authors’ conclusions

The results suggest that

the efficacy of dabigatran etexilate remains similar to that

of warfarin for the treatment and prevention of venous thromboembolism in

patients with APS. Rates of bleeding events tended to be lower with dabigatran etexilate than with warfarin in

patients with APS, reaching statistical significance for any

bleeding events.

Ordi-Ros, 2019

Type of study: open-label, phase 3, randomized noninferiority trial.

Setting and country: recruitment between March 2013 to December 2014, 6 university hospitals in Spain

Funding and conflicts of interest:

-Financial support: Bayer Hispania.

-Drs. Ordi-Ros and Corte´ s-Hernández report institutional support from Bayer Hispania to conduct the study. Dr.

Ortiz-Santamaria reports personal fees and nonfinancial support from Pfizer, Janssen, Lilly, GSK, Novartis, Roche, and Abbott outside the submitted work. Authors not named here have disclosed no conflicts of interest.

Inclusion criteria:

-aged 18 to 75 years

-APS by international consensus

criteria

-objectively confirmed arterial or venous thrombosis and a positive result on aPL testing on 2 occasions at least 3 months apart.

Exclusion criteria:

-clinically significant bleeding diathesis (such as refractory thrombocytopenia with a platelet count ≤50 × 109 cells/L)

-intracranial hemorrhage, stroke, or gastrointestinal bleeding within the previous 3 months

-pregnant or lactating;

-severe renal impairment

-an alanine aminotransferase level more than twice the upper limit of normal

-Child–Pugh class B or C cirrhosis -nonadherent to their warfarin regimen

-receiving cytochrome P450 3A4 inducers.

N total at baseline:

N=190

Intervention: 95

Control: 95

Important prognostic factors²:

Age (years), median (IQR)

I: 47 (40–55)

C: 51 (38–63)

Female, n(%)

I: 61 (64.2)

C: 60 (63.2)

APS category, n (%)

Primary

I: 64 (67.4)

C: 68 (71.6)

Secondary

I: 31 (32.6)

C: 27 (28.4)

Clinical criteria for initial anticoagulation, n (%)

Venous thrombosis

I: 69 (72.6)

C: 70 (73.7)

Arterial thrombosis

I: 37 (38.9)

C: 34 (35.8)

Both arterial and venous thrombosis

I: 11 (11.6)

C: 9 (9.5)

Rivaroxaban dose, n(%)

20 mg once daily

I: 90 (94.7)

15 mg once daily

I: 5 (5.2)

Aspirin use, n (%)

I: 13 (13.7)

C: 11 (11.6)

Groups comparable at baseline?

Baseline variables were evenly distributed, except for a higher proportion of patients with migraine, livedo racemosa, recurrent thrombosis, and older age in the rivaroxaban group.

This imbalance occurred by chance, and the analysis

adjusted for these factors.

rivaroxaban

rivaroxaban (20 mg/d, or 15 mg/d for patients with a creatinine clearance of 30

to 49 mL/min/1.73 m2)

During the study, the rivaroxaban dosage could be modified in relation to changes in creatinine clearance.

The use of additional aspirin, antimalarial drugs, or

immunesuppressive agents was allowed at the investigator's discretion

vitamin K antagonists (VKA)

adjusted dosage of VKAs (target INR, 2.0 to 3.0, or 3.1

to 4.0 for those with a history of recurrent thrombosis)

The use of additional aspirin, antimalarial drugs, or

immunosuppressive agents was allowed at the investigator's discretion

Length of follow-up

Patients were evaluated monthly for the first 3 months and every 3 months thereafter. Follow-up was 36 months, regardless of events.

Loss to follow-up

Intervention:

11/95 (11.6%)

Reasons (describe):

-Death (n = 5)

-Early discontinuation (n = 6)

Lupus nephritis flare: 1

Adverse events: 5

Control:

9/95 (9.5%)

Reasons (describe):

-Death (n = 3)

-Early discontinuation (n = 6)

Lupus nephritis flare: 1

Adverse events: 1

Irregular INR control: 4

Non-inferiority threshold

The noninferiority margin was set to correspond to the preservation of 50% of the warfarin versus placebo effect.

Statistical analyses

Because this was a noninferiority study, the primary efficacy analysis was prespecified to be performed in the per protocol population, which included all patients receiving at least 1 dose of the study drug and being followed for events occurring while receiving treatment

or within 7 days after discontinuation. An intention-to treat (ITT) analysis was also performed. Follow-up in the ITT population continued until notification of study termination. The safety population comprised all patients receiving at least 1 dose of study medication.

Intention to treat

All thrombotic events, n(%)

I: 12/95 (12.6)

C: 6/95 (6.3)

RR 2.00 (0.78–5.11)

P=0.150

HR 2.10 (0.79–5.59)

P=0.140

Arterial events

I: 11/95 (11.6)

C: 3/95 (3.2)

RR 3.67 (1.06–12.73)

P=0.040

HR 3.84 (1.07–13.76)

P=0.040

Venous events

I: 2/95 (2.1)

C: 3/95 (3.2)

RR 0.67 (0.11–3.90)

P=0.65

HR 0.70 (0.12–4.18)

P=0.69

Stroke

I: 10/95 (10.5)

C: 0/95 (0)

RR 21.00 (1.25–353.3)

P=0.001

HR 20.01 (1.12–431.8)

P=0.040

Per protocol, as treated

All thrombotic events, n(%)

I: 11/95 (11.6)

C: 6/95 (6.3)

RR 1.83 (0.71–4.76)

P=0.21

HR 1.94 (0.72–5.24)

P=0.190

Arterial events

I: 10/95 (10.5)

C: 3/95 (3.2)

RR 3.33 (0.95–11.73)

P=0.060

HR 3.52 (0.97–12.79)

P=0.060

Venous events

I: 2/95 (2.1)

C: 3/95 (3.2)

RR 0.67 (0.11–3.90)

P=0.65

HR 0.70 (0.12–4.21)

P=0.70

Stroke

I: 9/95 (9.5)

C: 0/95 (0)

RR 19.00 (1.12–321.9)

P<0.001

HR 19.97 (1.00–400.0)

P=0.050

As treated safety population

Any bleeding

I: 31/95 (32.6)

C: 26/95 (27.4)

RR 1.19 (0.77–1.85)

P=0.43

HR 1.27 (0.75–2.15)

P=0.37

Major bleeding

I: 6/95 (6.3)

C: 7/95 (7.4)

RR 0.86 (0.30–2.46)

P=0.77

HR 0.88 (0.30–2.63)

P=0.82

Nonmajor clinically relevant bleeding

I: 9/95 (9.5)

C: 5/95 (5.3)

RR 1.80 (0.63–5.17)

P=0.28

HR 2.35 (0.72–7.61)

P=0.160

Minor bleeding

I: 16/95 (16.8)

C: 14/95 (14.7)

RR 1.14 (0.59–2.21)

P=0.69

HR 1.16 (0.57–2.37)

P=0.69

Death

I: 5/95 (5.2)

C: 3/95 (3.2)

RR 1.67 (0.41-6.78)

P=0.47

Authors’ conclusions

Rivaroxaban did not demonstrate noninferiority to dose-adjusted VKAs for secondary thromboprophylaxis in patients with thrombotic APS. We found an increased risk for recurrent thrombosis in the rivaroxaban group, with a predominance of arterial thrombotic events and stroke in particular.

Pengo, 2018 (TRAPS study)

Type of study:

randomized open-label multicenter noninferiority study

Setting and country: trial enrollment began on 2 November 2014, 14 centers in Italy

Funding and conflicts of interest:

-Funding: this work was supported by funding from Bayer S.p.A. (V.P.).

-The authors declare no competing financial interests.

Inclusion criteria:

-between ages of 18 and 75 years

-confirmed diagnosis of APS and positive for all 3 aPL tests in the

last blood sampling (triple positivity)

-history of thrombosis (objectively proven arterial, venous, and/or biopsyproven microthrombosis)

(APS patients at high-risk for thromboembolic recurrence).

Exclusion criteria:

-Severe hypersensitivity reaction to rivaroxaban

-Calculated CLCR <30 mL/min at the screening visit

-Current pregnancy or breast feeding

-Concomitant treatment with other anticoagulants

-Patients taking interfering medications: pharmacologic interactions may occur with strong inhibitors of p-glycoprotein and of

CYP3A4

-Hemorrhage risk-related criteria

-Known liver cirrhosis or ALT above three times the upper normal

value.

N total at baseline:

N=120

Intervention: 59

Control: 61

Important prognostic factors²:

Age (years), mean ± sd

I: 46.5 ± 10.2

C: 46.1 ± 13.2

Female, n(%)

I: 39 (66)

C: 38 (62)

Rivaroxaban dose, n(%)

20 mg once daily

I: 57/59 (96.6%)

15 mg once daily

I: 2/59 (3.4%)

Autoimmune disease, n (%)

SLE

I: 10/59 (16.9%)

C: 15/61 (24.6%)

Other autoimmune disease

I: 14/59 (23.7%)

C: 10/61 (16.4%)

Groups comparable at baseline?

Randomization produced a satisfactory balance between groups for the considered variables. Risk factors for thrombosis were also comparable between the 2 study groups.

Rivaroxaban

20 mg once daily

if they had a creatinine clearance (CrCl).50 mL/min, as determined by the Cockroft–Gault method. In case of CrCl between 30 and 50 mL/min, 15 mg of rivaroxaban was administered once

daily.

The use of aspirin in

addition to anticoagulant drugs was allowed at the investigator’s discretion.

Warfarin

The initial warfarin dose is determined using the protocol endorsed by the

Italian Federation of Anticoagulation Clinics (FCSA). The subsequent warfarin doses are

adjusted to maintain the INR within the range of

2.0–3.0, aiming for a target of 2.5.

The use of aspirin in

addition to anticoagulant drugs was allowed at the investigator’s discretion.

Length of follow-up:

The mean follow-up period was 569 days for the “as treated” cohort and 611 days for the ITT cohort.

Loss-to-follow-up:

Intervention:

9/59 (15%)

Reasons (describe):

-Withdrawal of

informed consent: 2

-Clinically relevant

nonmajor bleeding: 2

-Planned pregnancy: 3

-Other: 2

Control:

3/61 (5%)

Reasons (describe):

-Withdrawal of

informed consent: 1

-Clinically relevant

nonmajor bleeding: 0

-Planned pregnancy: 1

-Other: 1

Non-inferiority threshold

The noninferiority margin was set to correspond to the preservation of 50% of the warfarin effect, as adopted in the other rivaroxaban registration trials.

Statistical analyses

The primary outcome was analyzed according to “as treated” and ITT principle; “as treated” analysis included patients who completed the study on the assigned treatment on randomization, whereas ITT included all patients who underwent randomization.

ITT analysis

Thromboembolic events, major bleeding, and

vascular death, n (%)

I: 13/59 (22)

C: 2/61 (3)

HR 7.4 (1.7-32.9)

P=.008

Arterial thrombosis, n(%)

I: 7/59 (12)

C: 0/61 (0)

Ischemic stroke

I: 4/59 (7)

C: 0/61 (0)

Myocardial infarction

I: 3/59 (5)

C: 0/61 (0)

Venous thromboembolism

I: 1/59 (2)

C: 0/61 (0)

Major bleeding

I: 4/59 (7)

C: 2/61 (3)

HR 2.3 (0.4-12.5)

P=.3

Death

I: 1/59 (2)

C: 0/61 (0)

“As treated” analysis

Thromboembolic events, major bleeding, and

vascular death, n (%)

I: 11/59 (19)

C: 2/61 (3)

HR 6.7 (1.5-30.5)

P=.01

Arterial thrombosis, n(%)

I: 7/59 (12)

C: 0/61 (0)

Ischemic stroke

I: 4/59 (7)

C: 0/61 (0)

Myocardial infarction

I: 3/59 (5)

C: 0/61 (0)

Venous thromboembolism

I: 0/59 (0)

C: 0/61 (0)

Major bleeding

I: 4/59 (7)

C: 2/61 (3)

HR 2.5 (0.5-13.6)

P=.3

Death

I: 0/59 (0)

C: 0/61 (0)

Remarks:

The trial was terminated prematurely after the enrolment of 120 patients (59 randomized to rivaroxaban and 61 to warfarin) because of an excess of events among patients in the rivaroxaban arm. All participating patients discontinued the

assigned study drug and switched to a nonstudy vitamin K antagonist.

Authors’ conclusion

Rivaroxaban in high-risk patients with APS was

associated with an excess of events compared with warfarin.

Study reference

(first author, publication year)

Describe method of randomisation¹

Bias due to inadequate concealment of allocation?²

(unlikely/likely/unclear)

Bias due to inadequate blinding of participants to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of care providers to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of outcome assessors to treatment allocation?³

(unlikely/likely/unclear)

Bias due to selective outcome reporting on basis of the results?⁴

(unlikely/likely/unclear)

Bias due to loss to follow-up?⁵

(unlikely/likely/unclear)

Bias due to violation of

intention to treat analysis?⁶

(unlikely/likely/unclear)

Cohen, 2016

Randomisation was performed by a web-based independent randomisation service (Sealed Envelope,

London, UK) to ensure allocation concealment. The

schedule was created using permuted blocks with arandom block length, stratified by centre and patient

type (with versus without systemic lupus erythematosus).

Unlikely

Randomisation was performed by a web-based independent randomisation service.

Unlikely

The trial was open label to ensure optimum warfarin dosing: anticoagulant monitoring is necessary. Additionally, the management of bleeding events differs between patients receiving warfarin and

rivaroxaban.

However, it is not likely that this results in risk of bias for the outcome measures (i.e., thrombotic events, bleeding events).

Unclear

The trial was open label.

Unlikely

Masking of treatment allocation was also not

possible in the RAPS central laboratory because different

tests were needed for the two anticoagulants, and samples

taken at baseline and day 42 were tested simultaneously to

minimise variability between assays.

Reports of SAEs, serious adverse reactions, and suspected

unexpected serious adverse reactions were reviewed by external, independent, medically qualified staff.

Clinically relevant and minor bleeding events across all sites were

pseudoanonymised and reviewed by one investigator (DAI) to remove the potential bias of interoperator

variation. The classification of bleeding events as

clinically relevant or minor, as per the protocol, was checked and changed if appropriate.

Unlikely

Outcomes in the protocol and published report are comparable.

Unlikely

3/57 (5.3%) patients in the intervention group and 3/59 (5.1%) patients in the control group did not contribute data to the primary outcome.

Unclear

A modified intention-to-treat approach was used to include all randomised patients with assessable data in

all analyses.

However, because this was a noninferiority trial, per protocol analysis might be preferable.

Ordi-Ros, 2019

The randomized list, stratified by center and presence of systemic lupus erythematosus, was created at VHH by using computer-generated random-number sequences (C4-Study design pack software (GlaxoSmithKline)) in blocks of 10. Sequentially numbered,

concealed envelopes containing group assignments

were provided to the investigators.

Unlikely

Computer-generated random-number sequences were used.

Unlikely

The trial was open label to ensure optimum VKA dosing and monitoring, and because bleeding management differs between VKAs and rivaroxaban.

However, it is not likely that this results in risk of bias for the outcome measures (i.e., thrombotic events, bleeding events).

Unclear

The trial was open label.

Unlikely

An independent committee blinded to the clinical end points applied protocol definitions to adjudicate suspected cases of thrombosis, death, and bleeding events that contributed to the prespecified end points.

Unlikely

Outcomes in the Methods’ section and Result’s section were comparable.

Unlikely

In each group, 6.3% of

patients permanently stopped their assigned therapy before a thrombotic event and before the end date.

Unlikely

Because this was a noninferiority study, the primary efficacy analysis was prespecified to be performed in the per protocol population. An intention-to treat analysis was also performed.

Pengo, 2018

After signing an informed consent form, patients underwent Web-based randomization using random block sizes

of 2, 4, and 6 and stratified based on sex and the presence or

absence of an associated autoimmune disease. In this way, 4

strata were constructed: females with or without associated

autoimmune disease and males with or without associated autoimmune disease.

Unlikely

Web-based randomization.

Unlikely

Open-label trial.

However, it is not likely that this results in risk of bias for the outcome measures (i.e., thrombotic events, bleeding events).

Unclear

Open-label trial.

Unclear

Outcome assessors not described.

Unlikely

Outcomes in the study protocol and published report were compared. Primary outcomes were comparable. Minor bleeding was not reported (secondary outcome).

However, the trial was terminated prematurely because of an excess of events in the intervention group. Therefore, selective reporting was unlikely.

Unclear

12 patients exited the study prematurely. 9/59 (15%) of the intervention group and 3/61 (5%) of the control group. Reasons were reported.

The trial was terminated prematurely after the enrolment of 120 patients because of an excess of events among patients in the rivaroxaban arm. The prespecified sample size was 536 patients.

Unlikely

The primary outcome was analyzed according to “as treated” and intention to treat (ITT) principle.