Skip to main content

Table 2 Table reporting risk of bias in individual study

From: Comparing clinical outcomes of NOACs with warfarin on atrial fibrillation with Valvular heart diseases: a meta-analysis

Bias

Judgement

Support

Article 1

ARISTOTLE [8, 9]

Selection

Low

Quote: “we have randomized 18,206 patients with AF from over 1000 centers in about 40 countries. Eligible subjects were randomly assigned in a 1:1 ratio to receive either apixaban or warfarin …”

Performance

Low

Quote: “To maintain blinding, study medications are packaged using a double-dummy design. The 2 sets of tablets each subject receives are distinguishable by color and size, but active apixaban tablets match placebo apixaban tablets and active warfarin tablets match placebo warfarin tablets to ensure blinding of the patient and investigator.”

Detection

Low

Attrition

Low

The reason of exclusion of patient was clearly stated

Reporting

Low

All pre-specified outcomes were reported as HR and 95%CI with respective number of patients

Article 2

ENGAGE AF-TIMI 48 [10, 11]

Selection

Low

Quote: “Subjects are randomized through an interactive voice/ Web response system (IVRS) ... Approximately 20,500 patients will be enrolled with history of AF documented on an electrical recording within the past 12 months for whom anticoagulation is planned for the duration of the trial. Subjects are randomized 1:1:1 ... Randomization is stratified by CHADS2 score 2 to 3 versus 4 to 6 and drug clearance”

Performance

Low

Quote: “ENGAGE AF–TIMI 48 is a large, multinational, randomized (1:1:1), double-blind, double-dummy ... All subjects are dispensed 2 sets of study drug. The first set, edoxaban or matching placebo ... The second set, warfarin or matching placebo”

Detection

Low

Quote: “INR measurements are performed using a point-of-care device supplied to each study site. The INR results generated by the point-of-care device are masked by a “code number,” which is reported by the investigators to a central IVRS … ”

Attrition

Low

The reason of exclusion of patient was clearly stated

Reporting

Low

All pre-specified outcomes were reported as HR and 95%CI with respective number of patients

Article 3

RE-LY [6]

Selection

Low

Quote: “The patients were randomized by a central randomization service, through an interactive voice response system (IVRS) located at the Coordinating Centre at Population Health Research Institute (PHRI) in Hamilton, Canada”

Performance

High

Open label

Detection

Low

Quote: “RE-LY is a phase 3, multicenter, prospective, open-label, randomized trial with blinded evaluation of all outcomes (PROBE design).”

Attrition

Low

The reason of exclusion of patient was clearly stated

Reporting

Low

All pre-specified outcomes were reported as HR and 95%CI with respective number of patients

Article 4

ROCKET-AF [7]

Selection

Low

Quote: “Over 14,000 patients have been randomized in the ROCKET AF trial at 1100 sites across 45 countries. Patients are allocated to 1 of 2 study regimens: rivaroxaban or warfarin.”

Performance

Low

Quote: “A double-blind design was chosen to minimize bias in cointerventions and interpretation of clinical events. To maintain blinding in ROCKET AF, sham INR results are provided. A point-of-care coagulation testing device displays a code number that, when entered into the Interactive Voice Response System along with the subject’s study identification number, is decoded and generates either the subject’s real INR or a sham INR value, depending on the patient’s blinded treatment.”

Detection

Low

Attrition

Low

The reason of exclusion of patient was clearly stated

Reporting

Low

All pre-specified outcomes were reported as HR and 95%CI with respective number of patients

Article 5

Seeger et. al., 2017 [16]

Selection

High

The allocation of group is not randomised

Performance

High

The patient and the physician was not blinded

Detection

High

Attrition

Low

The reason of exclusion of patient was clearly stated

Reporting

Low

All prespecified outcomes were reported as event rate with respective number of patients

Article 6

Geis et. al., 2018 [17]

Selection

High

The allocation of group is not randomised and the study have retrospective design

Performance

High

The patient and the physician was not blinded

Detection

High

Attrition

Low

The reason of exclusion of patient was clearly stated

Reporting

Low

All prespecified outcomes were reported as event rate with respective number of patients

  1. The Cochrane Collaboration’s tool for assessing risk of bias