Relative Effects of Two Different Enoxaparin Regimens as Comparators Against Newer Oral Anticoagulants. Meta-analysis and Adjusted Indirect Comparison

Transcription

1 CHEST Original Research ANTITHROMBOTIC THERAPY Relative Effects of Two Different Enoxaparin Regimens as Comparators Against Newer Oral Anticoagulants Meta-analysis and Adjusted Indirect Comparison Chun Shing Kwok, MBBS ; Shiva Pradhan, MBBS ; Jessica Ka-yan Yeong, MBBS ; and Yoon K. Loke, MD Background: Two different regimens of enoxaparin (40 mg once daily or 30 mg bid) have been used as control arms in trials of new oral anticoagulants. The choice of enoxaparin comparator may influence the perceived relative efficacy and safety of the newer agents, and we aimed to identify any significant differences between the two enoxaparin regimens. Methods: We searched MEDLINE, EMBASE, and Cochrane Library for randomized controlled trials that compared enoxaparin to oral anticoagulant (apixaban, dabigatran, rivaroxaban) thromboprophylaxis in elective total knee or hip arthroplasty. Total VTE and bleeding events were pooled using fixed-effects meta-analysis and heterogeneity assessed with the I 2 statistic. We conducted adjusted indirect comparisons of bid vs once-daily enoxaparin regimes based on new oral anticoagulants as common comparators. Results: Fourteen randomized controlled trials in hip and knee surgery met the inclusion criteria. Adjusted indirect comparison showed that bid enoxaparin was significantly more effective in preventing VTE than enoxaparin once daily (relative risk [RR], 0.71; 95% CI, ; P,.00001). For major and clinically relevant hemorrhage, adjusted indirect comparison showed that enoxaparin bid was nonsignificantly associated with increased risk of bleeding (RR 1.27; 95% CI, ; P 5.08) above that of enoxaparin once daily. Subgroup analysis limited to total knee arthroplasty trials showed similar results. Conclusions: The use of once-daily enoxaparin regimen as control in clinical trials will lead to more favorable estimates of relative efficacy for the new oral anticoagulants than if enoxaparin 30 mg bid had been chosen as a comparator. CHEST 2013; 144(2): Abbreviations: RCT 5 randomized controlled trial; RR 5 relative risk VTE is a well-recognized complication of lowerlimb orthopedic surgery, and use of thromboprophylactic anticoagulants is recommended in clinical guidelines. A recent review found that low-molecularweight heparin was favored by the American College of Chest Physicians, whereas the National Institute of Clinical Excellence (England) supported the use of newer oral agents as alternatives to low-molecularweight heparin. 1 Current meta-analyses suggest that newer oral agents (such as apixaban, dabigatran, and rivaroxaban) can offer a more favorable benefit to harm profile than established parenteral anticoagulants such as enoxaparin. 2,3 However, robust interpretation of the data from meta-analyses is clouded by the use of two different enoxaparin regimens as controls against the newer oral anticoagulants. The two recognized regimens for VTE prophylaxis in joint surgery are enoxaparin 30 mg bid or 40 mg once daily. 4 Simply pooling all the trials together without considering variation in enoxaparin dosing could lead to heterogeneity and mask important differences in efficacy and safety. At present, there are limited data on bid vs oncedaily enoxaparin in thromboprophylaxis. A retrospective study reviewed 129 patients with spinal cord injury and found no significant difference in DVT, pulmonary embolism, and bleeding complications. 5 Another study of 63 patients found an increased risk of DVT with bid compared with once-daily enoxaparin. 6 journal.publications.chestnet.org CHEST / 144 / 2 / AUGUST

2 There has yet to be a large randomized trial to answer this question. There have been large, high-quality, randomized controlled studies that compare the new oral agents against either once-daily or bid enoxaparin. Through the use of a common comparator, it is possible to conduct an adjusted indirect comparison analysis to determine the relative treatment effect of the bid regimen vs once daily. Hence, the main aim of this study was to evaluate differences in efficacy and safety between bid and once-daily enoxaparin by conducting a systematic review of randomized trials of new oral anticoagulants against either of the enoxaparin regimens. Search Strategy Materials and Methods We conducted a systematic review of the literature to identify knee or hip surgery trials involving the oral anticoagulants dabigatran, rivaroxaban, or apixaban compared with once-daily or bid enoxaparin. A search was conducted of MEDLINE, EMBASE, and Cochrane library. The exact search strategy is shown in e-appendix 1. We also checked the bibliographies of included trials for any relevant studies. Eligibility Criteria The specific inclusion criteria for randomized controlled trials (RCTs) were (1) double-blind parallel group randomized trial; (2) participants undergoing elective knee or hip surgery; (3) intervention of dabigatran, rivaroxaban, or apixaban compared with enoxaparin (40 mg once daily or 30 mg bid) over the same treatment duration; and (4) clear reporting of total venous thromboembolic events and hemorrhagic adverse events. Data Abstraction Three reviewers (C. S. K., S. P., and J. K. Y.) independently and in duplicate assessed study eligibility and extracted numerical outcomes data from the included studies. The data extracted were then checked by Y. K. L. The reviewers obtained full consensus on inclusion of the studies and data extraction after resolving any discrepancies through discussion. For the primary outcome, data were extracted on the composite measure of total VTE events, which consisted of all or a Manuscript received October 27, 2012 ; revision accepted February 1, Affiliations: From the Medical Division (Drs Kwok, Yeong, and Loke), Norfolk and Norwich University Hospitals National Health Service Foundation Trust; Population Health Group (Drs Kwok, Yeong, and Loke), Norwich Medical School, University of East Anglia; and Department of Medicine (Dr Pradhan), James Paget University Hospitals NHS Foundation Trust, Norfolk, England. Funding/Support: The authors have reported to CHEST that no funding was received for this study. Correspondence to: Yoon K. Loke, MD, Norwich Medical School, University of East Anglia, Norwich, Norfolk, NR4 7TJ, England; y.loke@uea.ac.uk 2013 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians. See online for more details. DOI: /chest combination of DVT, nonfatal pulmonary embolism, and all-cause mortality. We assessed the overall risk of hemorrhage by considering the total numbers affected by hemorrhagic adverse events consisting of (1) major bleeds and (2) clinically relevant nonmajor bleeds. Study Characteristics and Quality Assessment Three reviewers (C. S. K., S. P., and J. K. Y.) extracted data on study characteristics, which were then checked by a second reviewer (Y. K. L.). The dose and duration of interventions and comparators was recorded, as well as baseline characteristics of participants in the RCTs. For dose ranging studies involving escalating doses of oral anticoagulants, we extracted data from the arms that used doses similar to that of the subsequent largescale phase 3 trials. Quality assessment was conducted based on the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions, 7 including consideration of randomization sequence generation, allocation concealment, blinding of participants, personnel and outcome, incomplete outcome data, selective outcome reporting, and other sources of bias. Quantitative Data Synthesis First, RevMan (Nordic Cochrane Center) was used to conduct fixed-effect meta-analysis for the pooled relative risks (RRs) from individual trials, with 95% CIs for dichotomous outcomes. The main analysis was on an intention-to-treat basis. Subgroup analysis was performed by considering only total knee arthroplasty studies. Statistical heterogeneity was assessed using I 2 statistic, with I 2 values of 30% to 60% representing a moderate level of heterogeneity.8 Having obtained summary estimates of treatment effect from the preceding meta-analysis, adjusted indirect comparison (Bucher s method) 9 was performed using ITC software (Canadian Health Authority), 10 whereby the treatment effects from the two different enoxaparin regimens could be compared indirectly through common controls (apixaban or dabigatran or rivaroxaban). Finally, the adjusted indirect comparisons arising from each set of oral anticoagulant trials were then pooled using inverse variance meta-analysis to estimate the overall effect. As the validity of the indirect comparison may be affected by differences among the trials (such as operation site and length of therapy), we chose to perform additional sensitivity analyses on more closely matched datasets restricted to a specific site of arthroplasty or to trials of a similar duration. Results Study Selection, Design, and Methodology Fourteen double-blind randomized trials met the inclusion criteria (four dabigatran, six rivaroxaban, and four apixaban trials); the study selection process is shown in Figure 1. Five of the 14 trials used enoxaparin 30 mg bid as the control arm while the remaining nine used once-daily enoxaparin 40 mg. Study methods are shown in Table 1. Quality Assessment The quality of studies is shown in e-table 1. Randomization methods were adequate and sufficiently 594 Original Research

3 Figure 1. Flowchart describing study selection of randomized controlled trials of dabigatran, rivaroxaban, or apixaban in elective orthopedic surgery. described in all but three studies. There was some degree of lost to follow-up and missing data mainly stemming from incomplete venographic assessment of leg veins. VTE and hemorrhagic outcomes were prespecified in all studies, and we did not detect any evidence of selective outcome reporting. Efficacy and Safety Results The results for each trial are shown in e-table 2. Pooled data from meta-analyses of total VTE and hemorrhagic adverse events for trials of new oral anticoagulants vs the different regimens of enoxaparin are Table 1 Study Participant Characteristics for Included RCTs Study Design Type of Surgery Enoxaparin Regimen, d Oral Anticoagulant Regimen, d Venography, d Follow-up, d RE-MODEL 11 Double-blind RCT Total knee RE-NOVATE12 Double-blind RCT Total hip RE-MOBILIZE13 Double-blind RCT Total knee RE-NOVATE II 14 Double-blind RCT Total hip RECORD1 15 Double-blind RCT Total hip RECORD316 Double-blind RCT Total knee RECORD417 Double-blind RCT Total knee ODIXA KNEE 18 Double-blind RCT Total knee ODIXA HIP BD 19 Double-blind RCT Total hip ODIXA HIP OD 20 Double-blind RCT Total hip ADVANCE-1 21 Double-blind RCT Total knee ADVANCE-222 Double-blind RCT Total knee ADVANCE-323 Double-blind RCT Total hip APROPOS24 Double-blind RCT Total knee OD 5 once daily; RCT 5 randomized control trial. 40 mg OD, 6-10 Dabigatran 150 or 220 mg OD, mg OD, Dabigatran 150 or mg OD, mg bid, Dabigatran 150 or mg OD, mg OD, Dabigatran 220 mg OD, mg OD, 35 Rivaroxaban 10 mg OD, mg OD, Rivaroxaban 10 mg OD, mg bid, Rivaroxaban 10 mg OD, mg bid, 5-9 Rivaroxaban 2.5, 5, , 20, or 30 mg bid, 5-9 (we analyzed 5 mg bid arm) 40 mg OD, 5-9 Rivaroxaban 2.5, 5, , 20, or 30 mg bid, 5-9 (we analyzed 5 mg bid arm) 40 mg OD, 5-9 Rivaroxaban 10, 20, or 30 mg OD, 5-9 (we analyzed 10 mg arm) 30 mg bid, 12 Apixaban 2.5 mg bid, mg OD, Apixaban 2.5 mg bid, mg OD, 35 Apixaban 2.5 mg bid, mg bid, or warfarin Apixaban 5, 10 or 20 mg OD, 2.5, 5 or 10 mg bid, (we analyzed 2.5 mg bid arm) journal.publications.chestnet.org CHEST / 144 / 2 / AUGUST

4 shown in Table 2 and e-figures 1-4. We conducted adjusted indirect comparisons of enoxaparin 30 mg bid vs enoxaparin 40 mg once daily with each new oral anticoagulant as common control, with results shown in Table 2. Total VTE Meta-analyses of total VTE adverse events with new oral anticoagulants directly compared against the two different enoxaparin regimen are shown in Table 2. In all instances, the benefits of the new oral anticoagulant were less apparent with the enoxaparin 30 mg bid control, and there was greater perceived efficacy of the new oral anticoagulants when analyzed against the enoxaparin 40 mg once-daily comparator. This is particularly true for dabigatran, which was found to be significantly less effective at preventing VTE if compared with enoxaparin 30 mg bid (RR, 1.26; 95% CI, ; P 5.006) but appeared to be just as good as enoxaparin when the 40-mg once-daily regimen was used as the comparator arm (RR, 1.01; 95% CI, ; P 5.90). Similarly, there was no demonstrable advantage with apixaban over enoxaparin in preventing VTE if compared with enoxaparin 30 mg bid (RR, 0.97; 95% CI, ; P 5.82), whereas apixaban was significantly better (RR, 0.55; 95% CI, ; P,.00001) than enoxaparin 40 mg once daily. Indirect comparison revealed that enoxaparin 30 mg bid was consistently more efficacious than 40 mg once daily in all three datasets, with the pooled overall estimate indicating significant superiority (RR, 0.71; 95% CI, ; P,.00001) for prevention of VTE with 30 mg bid ( Fig 2 ). Major and Clinically Relevant Bleeds Meta-analyses of hemorrhagic adverse events with new oral anticoagulants directly compared against the two different enoxaparin regimens are shown in Table 2. Using enoxaparin 30 mg bid as comparator led to lower levels of perceived harm from the new oral anticoagulants than when enoxaparin 40 mg once daily was used as a control. For example, both apixaban and dabigatran were significantly less likely to cause major or clinically relevant bleeds if compared with enoxaparin 30 mg bid, but neither apixaban nor dabigatran were markedly different in risk of hemorrhage against enoxaparin 40 mg once daily. In the adjusted indirect comparison, enoxaparin 30 mg bid was associated with a greater risk of bleeds than 40 mg once daily, with the pooled overall estimate indicating nonsignificantly greater likelihood of harm (RR, 1.27; 95% CI, ; P 5.08) with 30 mg bid ( Fig 3 ). We conducted sensitivity analysis by restricting the analysis to trials of knee arthroplasty and this yielded similar findings (Table 3 ). Sensitivity Analyses and Restriction to Trials of Knee Arthroplasty Post hoc sensitivity analyses were conducted based on type of surgery, duration of anticoagulant use, and nature of VTE event. We conducted sensitivity analysis by restricting the analysis to eight trials of knee arthroplasty and this yielded similar findings, with the indirect comparison demonstrating a lower risk of total VTE (RR, 0.74; 95% CI, ) for enoxaparin 30 mg bid as compared with enoxaparin 40 mg once daily ( Table 3 ). Table 2 Risk of Venous Thromboembolic Events and Clinically Relevant Bleeding for Enoxaparin and Oral Anticoagulants RR Ratio (95% CI) Outcome/Group vs Enoxaparin 30 mg Bid vs Enoxaparin 40 mg OD Adjusted Indirect Comparison Enoxaparin 30 mg Bid vs Enoxaparin 40 mg OD For total VTE Apixaban 0.97 (0.76, 1.25) 0.55 (0.46, 0.65) 0.57 ( ) Dabigatran 1.26 (1.07, 1.48) 1.01 (0.89, 1.14) 0.80 ( ) Rivaroxaban 0.70 (0.54, 0.91) 0.47 (0.38, 0.58) 0.67 ( ) Pooled estimate using all three adjusted 0.71 ( ) indirect comparisons For major and clinically relevant bleeds Apixaban 0.66 (0.46, 0.95) 0.88 (0.73, 1.07) 1.33 ( ) Dabigatran 0.84 (0.55, 1.28) 1.20 (0.98, 1.47) 1.44 ( ) Rivaroxaban 1.25 (0.82, 1.89) 1.27 (0.98, 1.65) 1.02 ( ) Pooled estimate using all three adjusted indirect comparisons 1.27 ( ) RR 5 relative risk. See Table 1 legend for expansion of other abbreviation. 596 Original Research

5 Figure 2. Risk of total venous thromboembolic events with bid vs once-daily enoxaparin. df 5 degrees of freedom; IV 5 inverse variance. Restriction to Trials of Similar Duration Eleven trials were based on durations of 5 to 14 days, and only four trials used durations of 28 to 35 days. We performed a sensitivity analysis by restricting the adjusted indirect comparison with these 11 shorterduration trials. Enoxaparin 30 mg bid was superior to the 40-mg once-daily regimen in reducing risk of total VTE (RR, 0.66; 95% CI, ). Restriction to Symptomatic VTE Events Overall, symptomatic VTE rates were lower in participants receiving one of the new oral anticoagulants, 0.37% as compared with 0.66% in the enoxaparin 40-mg once-daily control subjects. The pooled-risk difference analysis demonstrates that on average, the new oral anticoagulants would have prevented three symptomatic VTE events (95% CI, one to five events) per 1000 patients treated when compared with enoxaparin 40 mg once daily. In contrast, no meaningful differences were demonstrated in the trials between the new oral anticoagulants and enoxaparin 30 mg bid as illustrated by the very similar symptomatic event rates of 1.02% and 1.07%, respectively. The adjusted indirect comparison based on analysis of symptomatic VTE rates overall shows enoxaparin 30 mg bid to be superior in thromboprophylaxis (RR, 0.55; 95% CI, ) as compared with enoxaparin 40 mg once daily. Although the absolute event rates for symptomatic VTE were generally low, both the absolute and RR data demonstrate that evaluation of the new oral anticoagulants against enoxaparin 30 mg bid will yield less positive impressions of benefit for the new oral anticoagulants than if enoxaparin 40 mg once daily was used as comparator. Discussion Our adjusted indirect comparison indicates that once-daily 40-mg enoxaparin is associated with a significantly inferior effect (about 30% less benefit) for prevention of total VTE and 45% less benefit for symptomatic VTE as compared with bid 30-mg enoxaparin. The direction of effect is consistent, irrespective of whether the adjusted indirect comparison was performed through the apixaban or dabigatran or rivaroxaban trials dataset, or whether we used total VTE or symptomatic VTE as the outcome measure. Given that two different enoxaparin regimens have been deployed as control arms in clinical trials, our results have important implications for clinicians who are attempting to draw meaningful interpretations on comparative effectiveness of the new oral anticoagulants. In particular, positive (but incomplete) judgments may stem from focusing on the trials that used once-daily enoxaparin as control. Figure 3. Risk of major or clinically relevant bleeding events with bid vs once-daily enoxaparin. See Figure 2 legend for expansion of abbreviations. journal.publications.chestnet.org CHEST / 144 / 2 / AUGUST

6 Table 3 Sensitivity Analysis Based on Knee Arthroplasty Trials RR Ratio (95% CI) Outcome/Group vs Enoxaparin 30 mg Bid vs Enoxaparin 40 mg OD Adjusted Indirect Comparison Enoxaparin 30 mg Bid vs Enoxaparin 40 mg OD For total VTE Apixaban 0.97 (0.76, 1.25) 0.61 (0.50, 0.73) 0.63 ( ) Dabigatran 1.26 (1.07, 1.48) 1.03 (0.89, 1.19) 0.82 ( ) Rivaroxaban 0.70 (0.54, 0.91) 0.48 (0.38, 0.63) 0.69 ( ) Pooled estimate using all three adjusted 0.74 ( ) indirect comparisons For major and clinically relevant bleeds Apixaban 0.66 (0.46, 0.95) 0.73 (0.51, 1.04) 1.11 ( ) Dabigatran 0.84 (0.55, 1.28) 1.17 (0.82, 1.68) 1.39 ( ) Rivaroxaban 1.25 (0.82, 1.89) 1.20 (0.76, 1.92) 0.96 ( ) Pooled estimate using all three adjusted indirect comparisons 1.15 ( ) See Table 1 and 2 legends for expansion of abbreviations. Here, we believe that interpretation of the efficacy of the new oral anticoagulants should be stratified according to control enoxaparin regimen. Pooled estimates that combine both enoxaparin regimens as a single control arm will disguise the fact that the efficacy of the new oral anticoagulants is less favorable and make it more difficult to discern any increased risk of bleeding. Additional research has also raised concerns about the pharmacokinetic and pharmacodynamics impact of once-daily enoxaparin. Rutherford et al 25 argued that a once-daily enoxaparin regimen was inadequate and should be abandoned because many of the critically ill patients who received this regimen had antifactor Xa activity below the target range. Conversely, Vincent et al 26 reported that target levels were met in almost all 36 patients in a surgical ICU treated with enoxaparin 30 mg bid. Adjusted indirect comparison has limitations stemming from potential confounding due to differences between trial participants, delivery of the intervention, and measurement of outcomes. Nevertheless, this perceived diminution of effect size with enoxaparin once daily for preventing VTE was apparent in all three datasets analyzed here, be it apixaban, dabigatran, or rivaroxaban. As the data regarding enoxaparin 30 mg bid stems exclusively from knee arthroplasty trials, we performed a sensitivity analysis using a restricted, more closely matched trial dataset of enoxaparin once-daily regimens in knee arthroplasty and found that this yielded very similar results to the overall analysis. Equally, while the trials all reported on somewhat similar categories of major and clinically relevant bleeding, Dahl et al 27 have pointed out that the actual case definitions used in detecting such events has not always been fully consistent. Despite these potential limitations, the validity of our adjusted indirect comparison is supported by the biologically consistent findings regarding the separate outcomes of VTE and hemorrhages, whereby enoxaparin bid was significantly more beneficial, but also potentially more likely to cause hemorrhage than enoxaparin once daily. While it would have been preferable to include direct clinical trial comparisons of bid vs once-daily enoxaparin, the available studies 5,6 were observational in nature and only had small sample sizes. Our findings have a number of clinical implications. Enoxaparin 30 mg bid is approved for use in North America following hip and knee arthroplasty, and our analysis shows no evidence of a superior benefit:harm balance with alternatives such as the new oral anticoagulants or once-daily enoxaparin. Other than convenience, there do not appear to be any compelling clinical reasons for health-care professionals and hospitals who are current users of the enoxaparin 30 mg bid regimen to move away from their usual standard of care. Rivaroxaban was the only oral agent with significantly greater efficacy than bid enoxaparin in VTE prevention, but was associated with the possibility of increased hemorrhage. We believe that any perceived superiority of the newer oral anticoagulants may arise from the use of an inferior once-daily regimen of enoxaparin as the comparator arm in most of the clinical trials, and we advocate caution rather than a rushed adoption of new technologies. We note that the American College of Chest Physicians recommend low-molecular-weight heparins as first-line options in VTE prophylaxis after orthopedic surgery. 28 The evidence from our analysis is consistent with these guidelines and will help clinicians in understanding differences and choosing 598 Original Research

7 between the two enoxaparin regimens. In patients with high risk of VTE and low risk of hemorrhage, enoxaparin 30 mg bid may continue to be a reasonable option. Acknowledgments Author contributions: Dr Loke is the guarantor for the manuscript. Dr Kwok: contributed to the study by performing the search, screening search results and reviewing abstracts for study selection, extracting the data, performing the data analysis, and writing the manuscript. Dr Pradhan: contributed to the study by screening search results and reviewing abstracts for study selection, extracting the data, and writing the manuscript. Dr Yeong: contributed to the study by screening search results and reviewing abstracts for study selection, extracting the data, and writing the manuscript. Dr Loke: contributed to the study by performing the search, checking data, performing the data analysis, and writing the manuscript. Financial/nonfinancial disclosures: The authors have reported to CHEST that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article. Additional information: The e-appendix, e-tables, and e-figures can be found in the Supplemental Materials area of the online article. References 1. Wong JM, Loke YK. Update on thromboprophylaxis in orthopedic surgery and critical appraisal of the role of enoxaparin. Orthopedic Research Reviews ;4: Loke YK, Kwok CS. Dabigatran and rivaroxaban for prevention of venous thromboembolism systematic review and adjusted indirect comparison. J Clin Pharm Ther ;36(1): Neumann I, Rada G, Claro JC, et al. Oral direct factor Xa inhibitors versus low-molecular-weight heparin to prevent venous thromboembolism in patients undergoing total hip or knee : a systematic review and meta-analysis. Ann Intern Med ;156(10): Ginsberg JS. The development and introduction of biosimilar anticoagulants focus on enoxaparin. Biosimilars ;2: Hebbeler SL, Marciniak CM, Crandall S, Chen D, Nussbaum S, Mendelewski S. Daily vs twice daily enoxaparin in the prevention of venous thromboembolic disorders during rehabilitation following acute spinal cord injury. J Spinal Cord Med ;27(3): Riha GM, Van PY, Differding JA, Schreiber MA ; Oregon Health & Science University Trauma Research Group. Incidence of deep vein thrombosis is increased with 30 mg twice daily dosing of enoxaparin compared with 40 mg daily. Am J Surg ;203(5): Higgins JP, Altman DG. Assessing risk of bias in included studies. In: Higgins JP, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions. Cochrane Book Series. Chichester, England : John Wiley & Sons, Ltd ; Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ ;327(7414): Bucher HC, Guyatt GH, Griffith LE, Walter SD. The results of direct and indirect treatment comparisons in meta-analysis of randomized controlled trials. J Clin Epidemiol ;50(6): Wells GA, Sultan SA, Chen L, Khan M, Coyle D. Indirect Treatment Comparison [computer program]. 1.0 ed Canadian Agency for Drugs and Technologies in Health website. Accessed May 7, Eriksson BI, Dahl OE, Rosencher N, et al ; RE-MODEL Study Group. Oral dabigatran etexilate vs. subcutaneous enoxaparin for the prevention of venous thromboembolism after total knee : the RE-MODEL randomized trial. J Thromb Haemost ;5 (11 ): Eriksson BI, Dahl OE, Rosencher N, et al ; RE-NOVATE Study Group. Dabigatran etexilate versus enoxaparin for prevention of venous thromboembolism after total hip : a randomised, double-blind, non-inferiority trial. Lancet ;370 (9591 ): RE-MOBILIZE Writing Committee ; Ginsberg JS, Davidson BL, Comp PC, et al. Oral thrombin inhibitor dabigatran etexilate vs North American enoxaparin regimen for prevention of venous thromboembolism after knee arthroplasty surgery. J Arthroplasty ;24 (1 ): Eriksson BI, Dahl OE, Huo MH, et al ; RE-NOVATE II Study Group. Oral dabigatran versus enoxaparin for thromboprophylaxis after primary total hip arthroplasty (RE-NOVATE II*). A randomised, double-blind, non-inferiority trial. Thromb Haemost ;105 (4 ): Eriksson BI, Borris LC, Friedman RJ, et al ; RECORD1 Study Group. Rivaroxaban versus enoxaparin for thromboprophylaxis after hip arthroplasty. N Engl J Med ;358 (26 ): Lassen MR, Ageno W, Borris LC, et al ; RECORD3 Investigators. Rivaroxaban versus enoxaparin for thromboprophylaxis after total knee arthroplasty. N Engl J Med ;358 (26 ): Turpie AG, Lassen MR, Davidson BL, et al ; RECORD4 Investigators. Rivaroxaban versus enoxaparin for thromboprophylaxis after total knee arthroplasty (RECORD4): a randomised trial. Lancet ;373 (9676 ): Turpie AG, Fisher WD, Bauer KA, et al ; OdiXa-Knee Study Group. BAY : an oral, direct factor Xa inhibitor for the prevention of venous thromboembolism in patients after total knee. A phase II dose-ranging study. J Thromb Haemost ;3 (11 ): Eriksson BI, Borris L, Dahl OE, et al ; ODIXa-HIP Study Investigators. Oral, direct Factor Xa inhibition with BAY for the prevention of venous thromboembolism after total hip. J Thromb Haemost ;4 (1 ): Eriksson BI, Borris LC, Dahl OE, et al ; ODIXa-HIP Study Investigators. A once-daily, oral, direct Factor Xa inhibitor, rivaroxaban (BAY ), for thromboprophylaxis after total hip. Circulation ;114 (22 ): Lassen MR, Raskob GE, Gallus A, Pineo G, Chen D, Portman RJ. Apixaban or enoxaparin for thromboprophylaxis after knee. N Engl J Med ;361 (6 ): Lassen MR, Raskob GE, Gallus A, Pineo G, Chen D, Hornick P ; ADVANCE-2 investigators. Apixaban versus enoxaparin for thromboprophylaxis after knee (ADVANCE-2): a randomised double-blind trial. Lancet ; 375 ( 9717 ): Lassen MR, Gallus A, Raskob GE, Pineo G, Chen D, Ramirez LM ; ADVANCE-3 Investigators. Apixaban versus enoxaparin for thromboprophylaxis after hip. N Engl J Med ;363 (26 ): Lassen MR, Davidson BL, Gallus A, Pineo G, Ansell J, Deitchman D. The efficacy and safety of apixaban, an oral, direct factor Xa inhibitor, as thromboprophylaxis in patients following total knee. J Thromb Haemost ; 5 (12 ): journal.publications.chestnet.org CHEST / 144 / 2 / AUGUST

8 25. Rutherford EJ, Schooler WG, Sredzienski E, Abrams JE, Skeete DA. Optimal dose of enoxaparin in critically ill trauma and surgical patients. J Trauma ;58 (6 ): Vincent PD, Albert M, Champagne MC, et al. Factors influencing enoxaparin anti-xa activity in surgical critically ill patients. J Crit Care ;26 (4 ): Dahl OE, Quinlan DJ, Bergqvist D, Eikelboom JW. A critical appraisal of bleeding events reported in venous thromboem- bolism prevention trials of patients undergoing hip and knee arthroplasty. J Thromb Haemost ;8 (9 ): Falck-Ytter Y, Francis CW, Johanson NA, et al ; American College of Chest Physicians. Prevention of VTE in orthopedic surgery patients: antithrombotic therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians evidencebased clinical practice guidelines. Chest ;141 (2 )(suppl ): e278s-e325s. 600 Original Research