Page 3 - Rivaroxaban or Enoxaparin in Nonmajor Orthopedic Surgery

Rivaroxaban or Enoxaparin in Nonmajor Orthopedic Surgery

per 24 hours). Randomization (in randomly per- after the end of treatment to evaluate the occur-
muted blocks of four) was conducted within 10 rence of venous thromboembolic events.
hours after surgery and was performed centrally Symptomatic venous thromboembolic events
in a 1:1 ratio with the use of an interactive Web- had to be confirmed by objective tests — that is,
response system (ClinInfo) that assigned a unique compression ultrasonography for deep-vein throm-
randomization number to each eligible patient. bosis and computed tomographic pulmonary an-
Randomization was stratified according to cen- giography, ventilation–perfusion lung scanning,
ter and intended treatment duration (2 weeks to or pulmonary angiography for pulmonary embo-
1 month, >1 month to 2 months, or >2 months). lism. Fatal pulmonary embolism was confirmed
The intended duration of treatment was based on by means of autopsy or was imputed in cases of
medical judgment and corresponded to the planned unexplained death when pulmonary embolism
duration of immobilization (plaster cast or no- could not be ruled out.
weight-bearing recommendation or partial weight
bearing) and country-specific recommendations Outcomes
for the prevention of venous thromboembolism in The primary efficacy outcome of major venous
adults undergoing orthopedic surgery. 2,4 thromboembolism was a composite of symptom-
Patients who were randomly assigned to the atic distal or proximal deep-vein thrombosis,
rivaroxaban group were to receive 10 mg of riva- pulmonary embolism, or venous thromboembo-
roxaban orally and a subcutaneous injection of lism–related death during the treatment period
placebo (in lieu of enoxaparin); patients who were or asymptomatic proximal deep-vein thrombosis
randomly assigned to the enoxaparin group were at the end of treatment. Prespecified secondary
to receive a subcutaneous injection of enoxaparin outcomes were safety outcomes of major bleed-
(at a dose of 40 mg [4000 IU of anti-Xa activity] ing (fatal, critical, or clinically overt bleeding or
in 0.4 ml of diluent) and an oral tablet of pla- bleeding at the surgical site leading to interven-
11
cebo (in lieu of rivaroxaban) (Fig. S1 in the tion ), nonmajor clinically relevant bleeding,
Supplementary Appendix). The trial drug and overt thrombocytopenia, and death from any
matching placebo were administered once daily cause. Full definitions are provided in the End
every 24 hours within a window of ±2 hours. Points section in the Supplementary Appendix.
Provided that hemostasis had been established, All suspected thrombotic or bleeding events and
the first dose of the trial drug was administered deaths were adjudicated by a central independent
between 6 and 10 hours after surgery if it could committee whose members were unaware of the
be given before 10 p.m. and at least 24 hours after treatment assignments. An additional post hoc
any preoperative administration of low-molecular- analysis compared the composite of venous
weight heparin. If the first dose could not be thromboembolism or major bleeding between
given by 10 p.m., one postoperative dose of low- groups (referred to as “net clinical benefit”).
molecular-weight heparin was allowed, and ad-
ministration of the first dose of the trial drug Statistical Analysis
was postponed until the following day. Lists of To determine the noninferiority of rivaroxaban to
concomitant medications that were, or were not, enoxaparin, the primary analysis was performed
permitted during the trial are provided in the in the intention-to-treat population (all the pa-
Supplementary Appendix. tients who underwent randomization) and in the
At discharge, patients were provided with per-protocol population (all patients meeting the
sufficient trial drugs for the intended treatment inclusion criteria who underwent surgery, received
duration (i.e., until the end of immobilization). at least one dose of trial medication, and had no
All the patients underwent systematic compres- major protocol violations). In the primary inten-
sion ultrasonography at the end of immobiliza- tion-to-treat analysis, we used multiple imputa-
tion (i.e., between 15 days and 3 months after tion to account for missing data as described in
randomization) in order to detect asymptomatic the Supplementary Appendix. The risk ratio and
proximal deep-vein thrombosis (see the Compres- its 95% confidence interval were estimated with
sion Ultrasonography Assessment section in the the use of a logistic-regression model. The non-
Supplementary Appendix). Patients were contacted inferiority margin for the upper limit of the 95%
by telephone 30 days (within a window of ±7 days) confidence interval of the risk ratio comparing

n engl j med nejm.org 3
The New England Journal of Medicine
Downloaded from nejm.org by Obai Mahmoud on March 31, 2020. For personal use only. No other uses without permission.
Copyright © 2020 Massachusetts Medical Society. All rights reserved.