The 3 × 3 table on the left reports RRs from unadjusted network meta-analysis, and the 3 × 3 table on the right is the RRs from an adjusted network meta-analysis, in which the 6-month risks of recurrent VTE and major bleeding were calibrated to 10% and 5%, respectively. The RRs (95% CI) of recurrent VTE and major bleeding for each respective comparison can be found within the rectangle defined by the two anticoagulant strategies. The upper of the two results pertains to recurrent VTE, whereas the lower result pertains to major bleeding. The reference category (ie, denominator) for the relative risk is always higher to the left. For example, in the left 3 × 3 table, a rectangle defined by LMWH and DOAC includes 2 results, namely 1.08 (0.59-1.59) and 0.67 (0.31-1.46). The upper of these two (ie, 1.08) is the RR of recurrent VTE for DOAC vs LMWH (with LMWH as the denominator, ie, a 1.08-fold higher risk of recurrent VTE with DOAC compared with LMWH), whereas the lower of these 2 results (ie, 0.67) is the RR of major bleeding for DOAC vs LMWH (ie, 0.67-fold lower risk with DOAC compared with LMWH). Reprinted with permission from Elsevier/Thrombosis Research.³⁷ 

*Estimates from the unadjusted network meta-analysis.

†Estimates from a network meta-analysis adjusted to a 10% 6-month risk of recurrent VTE in the VKA arm (efficacy) and a 5% 6-month risk of major bleeding in the VKA arm (safety).