Table 1.
Main characteristics of NOACs.
Figure 1.
Flow of study selection.
Figure 2.
Scheme of the strict, stepwise, fixed-effects meta-analysis with predefined heterogeneity quality criteria.
This stepwise statistical approach was conducted for the identification of groupings that are justified to generate a common estimate (CE) based on the predefined quality criteria of low heterogeneity. If these heterogeneity criteria were not met no CE was generated. Legend: BID, twice-daily dosing; CE, common estimate; I2, Higgins' I2 [8]; QD, once-daily dosing; Qpeto, χ 2-distributed, Cochran's Q [7].
Table 2.
Reported efficacy outcomes (HRs [95% CI] vs warfarin) of the respective NOACs in the phase 3 trials [1]–[5] in the intent-to-treat analysis.
Table 3.
Reported safety outcomes (HRs [95% CI] vs warfarin) of the respective NOACs in the phase 3 trials [1]–[5] in the safety analysis sets.
Table 4.
Analysis of results heterogeneity of NOACs vs warfarin.
Figure 3.
Common estimates where justified and indirect comparisons of all BID or QD dosing regimens of NOACs.
Results are expressed from the respective main dose results of the phase 3 trials [1]–[5] in the intent-to-treat analysis for efficacy (Stroke and systemic embolism, ischemic stroke) and in the safety analysis for intracranial hemorrhage. AP, apixaban; BID, twice-daily dosing; CE, common estimate; CI, confidence interval; DE, dabigatran etexilate; EDOX, edoxaban; HR, hazard ratio; QD, once-daily dosing; RIVA, rivaroxaban; SE, systemic embolism; W, warfarin; aIn the ROCKET-AF trial, only ischemic strokes, excluding unspecified strokes, are reported. Note: bold and italic font marks significantly superior results.
Figure 4.
Indirect comparison of BID vs QD with the CEs and results from respective NOACs.
The comparison is conducted based on the results from the generation of CE for the respective end points. If no CE was generated (according to the strict heterogeneity criteria) a comparison of the respective result of the specific NOAC was applied. Legend: AP, apixaban; BID, twice-daily dosing; CE, common estimate; CE BID, common estimate generated by meta-analysis of DE 150 mg BID and AP 5/2.5 mg BID; CE QD, common estimate generated by meta-analysis of RIVA 20/10 mg QD and EDOX 60/30 mg QD; DE, dabigatran etexilate; EDOX, edoxaban; HR (95% CI), hazard ratio (95% confidence interval); ITT, intention-to-treat analysis; NOACs, novel oral anticoagulants; QD, once-daily dosing; RIVA, rivaroxaban; safety, safety set analysis.