Importance
Many anticoagulant strategies are available for the treatment of acute venous thromboembolism, yet little guidance exists regarding which drug is most effective and safe.
Objective
To summarize and compare the efficacy and safety outcomes associated with 8 anticoagulation options (unfractionated heparin [UFH], low-molecular-weight heparin [LMWH], or fondaparinux in combination with vitamin K antagonists); LMWH with dabigatran or edoxaban; rivaroxaban; apixaban; and LMWH alone) for treatment of venous thromboembolism.
Data Sources
A systematic literature search was conducted using MEDLINE, EMBASE, and the evidence-based medicine reviews from inception through February 28, 2014.
Study Selection
Eligible studies were randomized trials reporting rates of recurrent venous thromboembolism and major bleeding in patients with acute venous thromboembolism. Of the 1197 studies identified, 45 trials including 44 989 patients were included in the analyses.
Data Extraction and Synthesis
Two reviewers independently extracted trial-level data including number of patients, duration of follow-up, and outcomes. The data were pooled using network meta-analysis.
Main Outcomes and Measures
The primary clinical and safety outcomes were recurrent venous thromboembolism and major bleeding, respectively.
Results
Compared with the LMWH–vitamin K antagonist combination, a treatment strategy using the UFH–vitamin K antagonist combination was associated with an increased risk of recurrent venous thromboembolism (hazard ratio [HR], 1.42; 95% credible interval [CrI], 1.15-1.79). The proportion of patients experiencing recurrent venous thromboembolism during 3 months of treatment were 1.84% (95% CrI, 1.33%-2.51%) for the UFH–vitamin K antagonist combination and 1.30% (95% CrI, 1.02%-1.62%) for the LMWH–vitamin K antagonist combination. Rivaroxaban (HR, 0.55; 95% CrI, 0.35-0.89) and apixaban (HR, 0.31; 95% CrI, 0.15-0.62) were associated with a lower risk of bleeding than was the LMWH–vitamin K antagonist combination, with a lower proportion of patients experiencing a major bleeding event during 3 months of anticoagulation: 0.49% (95% CrI, 0.29%-0.85%) for rivaroxaban, 0.28% (95% CrI, 0.14%-0.50%) for apixaban, and 0.89% (95% CrI, 0.66%-1.16%) for the LMWH–vitamin K antagonist combination.
Conclusions and Relevance
Using meta-analytic pooling, there were no statistically significant differences for efficacy and safety associated with most treatment strategies used to treat acute venous thromboembolism compared with the LMWH–vitamin K antagonist combination. However, findings suggest that the UFH–vitamin K antagonist combination is associated with the least effective strategy and that rivaroxaban and apixaban may be associated with the lowest risk for bleeding.
Venous thromboembolism, manifested as deep vein thrombosis or pulmonary embolism, is a common medical condition and is the third leading cause of cardiovascular mortality.1-7 Untreated acute pulmonary embolism is associated with a mortality rate of up to 25%.8 Anticoagulant therapy is effective at treating acute venous thromboembolism and preventing recurrent events and death associated with recurrent venous thromboembolism.9
The traditional immediate treatment of acute venous thromboembolism involves the initial use of parenteral anticoagulants (unfractionated heparin [UFH], low-molecular-weight-heparin [LMWH], or fondaparinux) with transition to the vitamin K antagonist. Parenteral anticoagulant is continued for a minimum of 5 days and until the international normalized ratio is 2.0 or higher for at least 24 hours.10 In patients with acute deep vein thrombosis or pulmonary embolism, LMWH or fondaparinux is recommended over UFH.10
The traditional approach to acute venous thromboembolism may be cumbersome for patients given that the initial phase is administered parenterally and that the use of vitamin K antagonist requires frequent laboratory monitoring and food and drug restrictions due to adverse interactions.11 Direct oral anticoagulants including direct Xa inhibitors (rivaroxaban, apixaban, and edoxaban) and a direct thrombin inhibitor (dabigatran) may offer an attractive alternative to the traditional treatment of acute venous thromboembolism. Recent studies have evaluated the efficacy and safety of direct oral anticoagulants (with and without initial parenteral anticoagulant therapy) for the treatment of acute venous thromboembolism.
Clinicians have many potential treatment options regarding management of acute venous thromboembolism and little guidance exists about which drug is most effective yet safe. Although individual studies have shown promising results, several therapies have been assessed in only a single trial, and direct comparisons are rarely available. Therefore, we sought to summarize and compare the clinical outcomes and safety associated with various management options (UFH–vitamin K antagonist, LMWH–vitamin K antagonist, fondaparinux–vitamin K antagonist, LMWH–dabigatran, and LMWH–edoxaban combinations; rivaroxaban; apixaban; and LMWH alone) for treating acute venous thromboembolism using network meta-analysis.
Data Sources and Searches
A systematic search of the literature was conducted on MEDLINE (1946 to February 28, 2014) and EMBASE (1947 to February 28, 2014) and, using the OVID interface, to search for evidence-based medicine reviews: Cochrane Database of Systematic Review (2005-January 2014); ACP Journal Club (1981-February 2014); Database of Abstract of Reviews of Effects (first quarter, 2014); Cochrane Central Register of Controlled Trials (January 2014); Cochrane Methodology Register (third quarter, 2012); Health Technology Assessment (first quarter, 2014); and the National Health Service (NHS) Economic Evaluation (first quarter, 2014). References of included studies and narrative reviews were considered for additional potential studies. The protocol and systematic search strategy of the review is documented online (PROSPERO registry-CRD42014008671). The systematic search strategy is documented in eAppendix 1 in the Supplement. There were no language restrictions.
Potentially relevant articles were reviewed in full to ensure that they satisfied 3 criteria: (1) the study prospectively enrolled patients who had objectively confirmed symptomatic acute venous thromboembolism (lower extremity deep vein thrombosis, pulmonary embolism, or both) and who had qualifying recurrent venous thromboembolism events that were symptomatic and objectively confirmed; (2) patients were randomized to receive the UFH–vitamin K antagonist, LMWH–vitamin K antagonist, fondaparinux–vitamin K antagonist, LMWH–dabigatran, and LMWH–edoxaban combinations; rivaroxaban; apixaban; or LMWH alone and the study compared 2 different anticoagulation strategies; and (3) 1 or more of the primary or secondary outcomes were reported. Studies were excluded if (1) patients were randomized to placebo or observation; (2) patients were randomized to idraparinux or ximelagatran; (3) only patients with cancer-associated thrombosis were included; (4) study design was phase 1 or 2; and (5) trials evaluated extended venous thromboembolism treatment for secondary prevention.
The primary outcome measures were recurrent venous thromboembolism and major bleeding episodes. Recurrent venous thromboembolism was defined as a new noncompressible segment found on compression leg vein ultrasound imaging, new intraluminal filling defect on venography, newly abnormal impedance plethysmography test, new high-probability ventilation-perfusion scan, or new pulmonary artery filling defect on computed tomography or pulmonary angiography. A major bleeding episode was defined as clinically overt bleeding and associated with at least 1 of the following: (1) a decrease in hemoglobin levels of at least 2 g/dL; (2) transfusion of 2 or more units of packed red blood cells; (3) bleeding was intracranial or retroperitoneal or involved a body cavity; (4) death; or (5) a major bleeding episode as defined by the investigators of each individual study.12
Secondary outcomes included fatal recurrent venous thromboembolism and fatal bleeding episodes. Fatal recurrent venous thromboembolism was diagnosed by autopsy, a high-probability ventilation-perfusion scan or new intraluminal filling defect detected on computed tomography or venography prior to death, or a high clinical suspicion of fatal pulmonary embolism as defined by the investigators of each individual study. A fatal bleeding episode was defined as a major bleeding episode directly leading to death or a death that was reported as associated with bleeding.
Data Extraction and Quality Assessment
Two reviewers (L.A.C., M.C.) independently assessed eligibility of articles identified in the initial search strategy for inclusion in the review and discussed those deemed potentially eligible; independently extracted data (baseline characteristics, definition of outcomes, numbers of events) using a standardized data abstraction form; and assessed studies’ methodological quality using the Risk of Bias Assessment Tool from the Cochrane Handbook for randomized trials.13 Corresponding authors of manuscripts were contacted if primary or secondary outcomes could not be extracted from the original article.
Data Synthesis and Analysis
Outcomes were allocated according to the intention-to-treat principle. Bayesian network meta-analyses and direct frequentist pairwise meta-analyses were conducted for all outcomes. Bayesian network meta-analyses were conducted using a Poisson likelihood model.14 The different treatment strategies were treated as separate nodes (UFH–vitamin K antagonist, LMWH–vitamin K antagonist, fondaparinux–vitamin K antagonist, LMWH–dabigatran, or LMWH–edoxaban combinations; rivaroxaban; apixaban; or LMWH alone; Figure 1). All networks were constructed using NodeXL.15
Hazard ratios (HRs) and 95% credible intervals (CrIs) were modeled using Markov Chain Monte Carlo methods. The proportion of patients expected to have an event during use of each treatment (over 3 and 6 months) was calculated assuming a rate of 5.3 per 100 patient-years ( ≈ 2.6% over 6 months) and 3.6 per 100 patient-years ( ≈ 1.8% over 6 months) for recurrent venous thromboembolism and major bleeding respectively, for patients receiving the LMWH–vitamin K antagonist combination.16,17 The probability that each drug was associated with being the most efficacious treatment was calculated by counting the proportion of iterations of the Markov chain in which each drug had the highest HR.14 Random-effects network meta-analyses with vague priors were conducted for the analyses.14 Fixed-effects models were also conducted, along with an analysis using a binomial likelihood model and random-effects model with informative priors.18,19 Assessment of model fit was based on comparison of residual deviance to the number of unconstrained data points, assessment of the deviance information criterion, and between-study standard deviation.14,20,21 To ensure convergence was reached, trace plots and the Brooks-Gelman-Rubin statistic were assessed.22 Analyses were performed using WinBUGS version 1.4.3 (MRC Biostatistics Unit).
A priori subgroup analysis was conducted to adjust for index acute venous thromboembolism (deep vein thrombosis or pulmonary embolism) and for differences in study duration. We assessed inconsistency by comparing the deviance and deviance information criterion statistics in fitted consistency and inconsistency models.23 We also plotted the posterior mean deviance of the individual data points (each group of the trial) in the inconsistency model against their posterior mean deviance in the consistency model.23 Results from our network meta-analysis were also qualitatively compared with direct frequentist pairwise estimates. Frequentist pairwise meta-analyses were conducted using Comprehensive Meta-Analysis Version 2.2.064 (Biostat).
Of the 1197 citations identified, 45 articles (44 989 patients) were included in the analyses (Figure 2 and eAppendix 1 in the Supplement).17,24-67 Twenty-two clinical trials compared the UFH–vitamin K antagonist combination (357 events, 7133 patients) with the LMWH–vitamin K antagonist combination (533 events, 20 246 patients),24-45 12 compared the UFH–vitamin K antagonist combination (357 events, 7133 patients) with LMWH alone (81 events, 1897 patients),46-57 3 compared the LMWH–vitamin K antagonist combination (533 events, 20 246 patients) with LMWH alone (81 events, 1897 patients),59-61 2 compared the fondaparinux–vitamin K antagonist combination (85 events, 2201 patients) with the LMWH–vitamin K antagonist combination (533 events, 20 246 patients) or with the UFH–vitamin K antagonist (357 events, 7133 patients),58,62 and 6 compared the LMWH–vitamin K antagonist combination (533 events, 20 246 patients) with 1 of the direct oral anticoagulants: 2 dabigatran (60 events, 2553 patients), 1 apixaban (59 events, 2691 patients), 1 edoxaban (66 events, 4118 patients), and 2 rivaroxaban (86 events, 4150 patients).17,63-67 Two of the included studies compared 3 anticoagulation regimens: the UFH–vitamin K antagonist combination (357 events, 7133 patients), the LMWH–vitamin K antagonist combination (533 events, 20 246 patients), and LMWH alone (81 events, 1897 patients).26,27 The baseline characteristics of the included studies are shown in Table 1, Table 2, Table 3, and Table 4 and eAppendix 2 in the Supplement. The sample size ranged from 60 to 8240 patients, with a median sample size of 298. The median follow-up period was 3 months (range, 3-8.2 months). Eleven randomized clinical trials (RCTs) evaluated treatment of patients with any acute venous thromboembolism (both deep vein thrombosis and pulmonary embolism),17,26,37,39,43-45,57,63-65 7 studies assessed patients with pulmonary embolism,33,38,41,52,53,58,67 and the remaining 27 studies evaluated treatment of patients with deep vein thrombosis.24,25,27-32,34-36,40,42,46-51,54-56,59-62,66 Quality assessment is reported in eAppendix 3 in the Supplement.
All studies were included in the evaluation for recurrent venous thromboembolism. Results of Bayesian network meta-analyses are presented in Table 5 and Figure 3. In comparison with the LMWH–vitamin K antagonist combination, all treatment strategies except the UFH–vitamin K antagonist combination were associated with a lower rate of recurrent venous thromboembolism events. Compared with the LMWH–vitamin K antagonist combination, the UFH–vitamin K antagonist combination was associated with an increased rate of recurrent venous thromboembolism with an HR of 1.42 (95% CrI, 1.15-1.79). When the UFH–vitamin K antagonist combination was used as the comparator, the LMWH–vitamin K antagonist combination and LMWH alone were associated with a reduction in recurrent venous thromboembolism event rates (Table 5). Stepwise comparison of all remaining treatment strategies did not reveal significant differences in rates of recurrent venous thromboembolism (Table 5). The LMWH–edoxaban combination and apixaban were associated with the greatest probability of being the best therapy among all treatment strategies evaluated, 33.1% and 31.6 %, respectively. (For details, see eAppendix 4 in the Supplement.)
The UFH–vitamin K antagonist combination was associated with a higher percentage of patients experiencing recurrent venous thromboembolism during 3 months of treatment (1.84%; 95% CrI, 1.33%-2.51%) than patients taking the LMWH–vitamin K antagonist combination (1.30%; 95% CrI, 1.02%-1.62%; eAppendix 5 in the Supplement). Compared with the LMWH–vitamin K antagonist combination, the number needed to treat to harm using the UFH–vitamin K antagonist combination was 188 patients at 3 months (eAppendix 5 in the Supplement).
Comparison of random-effects Bayesian network meta-analyses with direct frequentist meta-analysis pairwise comparisons were similar in magnitude and direction of effect estimates for recurrent venous thromboembolism rates, with the exception of the fondaparinux–vitamin K antagonist combination and LMWH alone. The UFH–vitamin K antagonist combination was associated with lower efficacy than was the LMWH–vitamin K antagonist combination (eAppendix 6 in the Supplement). Subgroup analyses were conducted to evaluate recurrent venous thromboembolism event rates for patients presenting with index deep vein thrombosis and index pulmonary embolism (Table 5). Compared with the LMWH–vitamin K antagonist combination, use of the UFH–vitamin K antagonist combination in patients with index deep vein thrombosis was associated with the lowest efficacy and was associated with an increased risk of recurrent venous thromboembolism (HR, 1.74; 95% CrI, 1.27-2.44). All remaining treatment regimens were not associated with differences in outcomes from the LMWH–vitamin K antagonist combination in this population. An additional sensitivity analysis was performed for variation in study treatment duration (eAppendix 7 in the Supplement). Estimates from the sensitivity analyses were similar to the primary analysis except for the UFH–vitamin K antagonist combination, which was associated with a higher rate of recurrent venous thromboembolism (HR, 1.88; 95% CrI, 1.13-3.38) than in the primary analysis (HR, 1.42; 95% CrI, 1.15-1.80).
The analysis of major bleeding events included 42 trials that involved 44 434 patients (Figure 1 and eAppendix 2 in the Supplement). Twenty-two clinical trials compared the UFH–vitamin K antagonist combination (169 events, 6975 patients) with the LMWH–vitamin K antagonist combination (361 events, 20 124 patients),24-45 10 compared the UFH–vitamin K antagonist combination (169 events, 6975 patients) with LMWH alone (28 events, 1622 patients),46,49-57 2 compared the LMWH–vitamin K antagonist combination (361 events, 20 124 patients) with LMWH alone (28 events, 1622 patients),59,60 2 compared the fondaparinux–vitamin K antagonist combination (50 events, 2201 patients) with the LMWH–vitamin K antagonist combination (361 events, 20 124 patients) or with the UFH–vitamin K antagonist combination (169 events, 6975 patients),58,62 and 6 compared the LMWH–vitamin K antagonist combination (361 events, 20 124 patients) with 1 of the direct oral anticoagulants: 2 dabigatran (35 events, 2553 patients), 1 apixaban (15 events, 2691 patients), 1 edoxaban (56 events, 4118 patients), and 2 rivaroxaban (40 events, 4150 patients).17,63-67 Two of the included studies compared the UFH–vitamin K antagonist combination (169 events, 6975 patients) with the LMWH–vitamin K antagonist combination (361 events, 20 124 patients) and with LMWH alone (28 events, 1622 patients).26,27 The results of network meta-analysis pairwise comparisons are available in Table 5. Compared with the LMWH–vitamin K antagonist combination, rivaroxaban (HR, 0.55; 95% CrI, 0.35-0.89) and apixaban (HR, 0.31; 95% CrI, 0.15-0.62) were associated with the lowest bleeding risk. All other treatment regimens were associated with bleeding risks that did not differ from the LMWH–vitamin K antagonist combination. Additional pairwise comparisons showed that rivaroxaban, apixaban, or both were associated with the lowest bleeding rates compared with the UFH–vitamin K antagonist combination (HR, 0.47; 95% CrI, 0.27-0.80 for rivaroxaban and HR, 0.26; 95% CrI, 0.12-0.54 for apixaban); the fondaparinux–vitamin K antagonist combination (HR, 0.30; 95% CrI, 0.12-0.68 for apixaban); the LMWH–dabigatran combination (HR, 0.42; 95% CrI, 0.17-0.99 for apixaban); and the LMWH–edoxaban combination (HR, 0.37; 95% CrI, 0.15-0.89) for apixaban.
For 3 months’ duration of therapy, rivaroxaban (HR, 0.49%; 95% CrI, 0.29%-0.85%) and apixaban (HR, 0.28%; 95% CrI, 0.14%-0.50%) were associated with a lower proportion of patients experiencing a major bleeding event than patients receiving the LMWH–vitamin K antagonist combination (0.89%; 95% CrI, 0.66%-1.16%; eAppendix 5 in the Supplement). Compared with LMWH–vitamin K antagonist combination, rivaroxaban was associated with the number needed to treat to benefit of 258 patients at 3 months. Apixaban was associated with the number needed to treat to benefit of 165 patients at 3 months (eAppendix 5 in the Supplement). Apixaban was associated with the greatest probability of being the least harmful therapy (88.9%) among all treatment regimens assessed (eAppendix 4 in the Supplement).
Network meta-analysis using both random-effects and a fixed-effects model were performed and yielded similar results, albeit with wider CrIs in the random-effects model (eAppendix 8 in the Supplement). Hazard ratios also align with odds ratios reported in the binomial likelihood model (eAppendix 11 in the Supplement). The network meta-analysis for major bleeding yielded a reasonable fit to the data, even though there were a few studies that did not fit particularly well for recurrent venous thromboembolism (eAppendix 10 in the Supplement). Exclusion of these studies did not alter our main findings, although it did improve model fit and reduce between-study standard deviation (eAppendix 10 in the Supplement). We found no evidence of inconsistency (eAppendix 9 in the Supplement).
Fatal events were rare. One hundred sixty-five patients (0.37%) experienced fatal recurrent venous thromboembolism and 64 (0.14%), fatal bleeding events.
To our knowledge, this network meta-analysis is the largest review, including nearly 45 000 patients, assessing the clinical outcomes and safety associated with different anticoagulation strategies for the treatment of acute venous thromboembolism. We provide estimates on symptomatic recurrent venous thromboembolism and major bleeding outcomes (both patient-important outcomes), which are clinically relevant and are what clinical practice guideline recommendations are based on.68 All management options, with the exception of the UFH–vitamin K antagonist combination, were associated with similar clinical outcomes compared with a management strategy using the LMWH–vitamin K antagonist combination. Treatment using the UFH–vitamin K antagonist combination was associated with a higher risk of recurrent venous thromboembolism during the follow-up period. Despite this, there are clinical circumstances that necessitate the use of UFH including for patients with severe renal insufficiency and those with massive or submassive pulmonary embolism who are potential candidates for thrombolysis or thrombectomy.
Management strategies using rivaroxaban or apixaban, without preceding parenteral anticoagulation, appear to be associated with a reduction in the risk of major bleeding episodes compared with the LMWH–vitamin K antagonist combination. Apixaban was associated with a significant reduction in the risk of major bleeding episodes compared with all the other treatment options except rivaroxaban and LMWH alone. These risks for recurrent venous thromboembolism and major bleeding events are important to help clinicians assess the benefit-harm balance of the various treatment options and tailor their therapeutic approaches accordingly.
The decision to use the LMWH–vitamin K antagonist combination as the comparator for the network meta-analysis was based on the current clinical practice guidelines recommending LMWH or fondaparinux over UFH for the initial treatment of acute deep vein thrombosis or pulmonary embolism.10 Our results are consistent with these recommendations. Both treatment strategies using the LMWH– and fondaparinux–vitamin K antagonist combinations were associated with lower risk of recurrent venous thromboembolism and major bleeding episodes (Table 5). The most recent version of the clinical practice guidelines was published before the results of several clinical trials assessing the efficacy and safety of the direct oral anticoagulants for the treatment of acute venous thromboembolism were available. Therefore, there is very little guidance for clinicians on the use of these new oral anticoagulants for this patient population. A number of direct oral anticoagulants are now approved therapies for the indication of acute deep vein thrombosis and pulmonary embolism (rivaroxaban, dabigatran, and apixaban in the United States). Both dabigatran (150 mg taken orally twice a day) and edoxaban (60 mg taken orally once a day) were assessed following an initial 5 days of LMWH, whereas apixaban (10 mg taken orally twice a day for 7 days, then 5 mg taken orally twice a day) and rivaroxaban (15 mg taken orally twice a day for 21 days, then 20 mg taken orally daily) were evaluated without the use of initial LMWH.17,63-67 Our network meta-analysis demonstrated that treatment options using direct oral anticoagulants are associated with similar clinical outcomes as the more traditional strategies using the LMWH–vitamin K antagonist combination. As previously mentioned, strategies using direct oral anticoagulants seem to be associated with a lower risk of major bleeding episodes, which is most pronounced for rivaroxaban and apixaban. Apixaban was also associated with significantly less bleeding when compared with the LMWH–dabigatran and LMWH–edoxaban combinations (Table 5).
Our study included trials assessing patients with acute venous thromboembolism only and those evaluating use of anticoagulation for secondary prevention of venous thromboembolism were not included in this analysis. A prior systematic review and network meta-analysis on the use of long-term oral anticoagulation for the secondary prevention of venous thromboembolism has been published.69 Our findings are consistent with the results of 2 recently published systematic reviews and meta-analyses on the treatment of acute venous thromboembolism.70,71 Compared with these prior reports, our review is more comprehensive because it is not limited to the inclusion of the studies assessing the direct oral anticoagulants and because included studies assessed the UFH–vitamin K antagonist, LMWH–vitamin K antagonist, and fondaparinux–vitamin K antagonist combinations, thereby providing clinicians with an overall appraisal of all possible therapies. We also report the proportions of patients experiencing an event (and differences between treatments) after using each treatment for a period of 3 and 6 months for both outcomes in eAppendix 5 in the Supplement.
It is important to note the limitations of our study. First, the length of follow-up varied across studies. Trials assessing direct oral anticoagulants were longer in duration than those assessing the UFH–vitamin K antagonist and fondaparinux–vitamin K antagonist combinations or LMWH alone, resulting in potential variations in event rates in the LMWH–vitamin K antagonist combination groups (eAppendix 7, eFigures 7A and 7B in the Supplement). We conducted a sensitivity analysis restricting studies to those that had a minimum of 6 months of treatment (14 studies including 30 027 patients) (eAppendix 7: eFigure 7C in the Supplement). Results of the subgroup analysis aligned with the primary analysis (eAppendix 7: eFigure 7C and eTable 7 in the Supplement), suggesting minimal effect of treatment duration across studies. Second, the proportion of patients experiencing an outcome event while taking anticoagulants for 3 and 6 months (eAppendix 5 in the Supplement) were generated assuming a rate of 5.3 per 100 patient-years ( ≈ 2.6% over 6 months) for recurrent venous thromboembolism and 3.6 events per 100 patient-years ( ≈ 1.8% over 6 months) for major bleeding in the LMWH–vitamin K antagonist combination group.16,17 Although there was considerable variation in event rates in the LMWH–vitamin K antagonist combination groups when all studies are considered, the outcome rates were similar when only studies at least 6 months in duration were considered. The use of another trial to derive baseline rates for the LMWH–vitamin K antagonist combination is unlikely to substantially alter estimates on differences in the proportion of patients experiencing an event at 3 and 6 months (eAppendix 7: eFigures 7A and 7C in the Supplement). Third, patients treated with LMWH alone are a heterogeneous group as LMWH dosing varied across studies. Fourth, patient-level longitudinal data are required to establish more robust conclusions in specific patient populations. Other risk factors including severity of the index event (eg, segmental pulmonary embolism vs massive pulmonary embolism), presence of risk factors for major bleeding (eg, recent bleeding or surgery), or other ongoing risk factors for acute venous thromboembolism (eg, cancer, body mass index, etc.) may influence the risk of recurrent venous thromboembolism and bleeding events, as well as choice of treatment. Fifth, no trials have directly compared the different direct oral anticoagulants, for all trials assessing the efficacy of direct oral anticoagulants were compared with the LMWH–vitamin K antagonist combination. Future direct comparison trials, patient-level network meta-analyses, or high-quality nonrandomized studies are required to confirm our findings. Sixth, our analysis does not incorporate costs or preference for the different outcomes. Future studies need to be conducted to determine the most cost-effective management option. Seventh, the data included in the network meta-analysis was extracted from randomized trials and results are not generalizable to all patients with acute venous thromboembolism. Eighth, our search ended in March 2014.
Using meta-analytic pooling, there were no statistically significant differences in clinical and safety outcomes associated with most treatment strategies when compared with the LMWH–vitamin K antagonist combination. However, findings suggest that the UFH–vitamin K antagonist combination is associated with the least effective strategy and that rivaroxaban and apixaban may be associated with the lowest risk of bleeding.
Corresponding Author: Marc Carrier, MD, MSc, The Ottawa Hospital General Campus, Department of Medicine, 501 Smyth Rd, Box 201A, Ottawa, ON, Canada K1H 8L6 (mcarrier@ottawahospital.on.ca).
Author Contributions: Dr Carrier had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Drs Castellucci and Cameron contributed equally to this article.
Study concept and design: Castellucci, Cameron, Coyle, Carrier.
Acquisition, analysis, or interpretation of data: Castellucci, Cameron, Le Gal, Rodger, P.S. Wells, Clifford, Gandara, G. Wells, Carrier.
Drafting of the manuscript: Castellucci, Cameron, Rodger, Gandara, Carrier.
Critical revision of the manuscript for important intellectual content: All Authors.
Statistical analysis: Cameron, Coyle, Gandara, G. Wells, Carrier.
Administrative, technical, or material support: Clifford, Carrier.
Study supervision: Le Gal, Coyle, Clifford, G. Wells, Carrier.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Le Gal reports receiving payment for lectures from Biomerieux, Bayer, Pfizer, and Sanofi. Dr Wells reports receiving payments for speaker fees from Bayer, Boehringer Ingelheim, Biomerieux, and Bristol-Myers Squibb/Pfizer. Dr Rodger reports receiving research grants from Biomerieux. Dr Carrier reports receiving research grants from Leo Pharma and Bristol-Myers Squibb. Drs Castellucci, Cameron, George Wells, Coyle, Gandara, and Clifford had no financial disclosures to report.
Funding/Support: Dr Carrier is a recipient of a New Investigator Award from the Heart and Stroke Foundation of Canada and holds a T2 Research Chair in Cancer and Thrombosis form the University of Ottawa. Dr Cameron is a recipient of grant CGV 121171. a Vanier Canada Graduate Scholarship from the Canadian Institutes of Health Research and has received funding from Canadian Network and Centre for Trials Internationally (CANNeCTIN). He is also a trainee on Network Meta-analysis Team grant (116573) through the CIHR Drug Safety and Effectiveness Network. Dr Rodger is the recipient of a Career Scientist Award from the Heart and Stroke Foundation of Ontario. Dr Philip S. Wells is a recipient of a Canada Research Chair in Venous Thromboembolism.
Role of Funder/Sponsor: The funding organizations did not have any role in the design and conduct of the study; collection, management, analysis and interpretation of the data; and preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
1.Cohen
AT, Agnelli
G, Anderson
FA,
et al; VTE Impact Assessment Group in Europe (VITAE). Venous thromboembolism (VTE) in Europe: the number of VTE events and associated morbidity and mortality.
Thromb Haemost. 2007;98(4):756-764.
PubMedGoogle Scholar 2.Ageno
W, Prandoni
P, Romualdi
E,
et al. The metabolic syndrome and the risk of venous thrombosis: a case-control study.
J Thromb Haemost. 2006;4(9):1914-1918.
PubMedGoogle ScholarCrossref 3.Heit
JA. The epidemiology of venous thromboembolism in the community: implications for prevention and management.
J Thromb Thrombolysis. 2006;21(1):23-29.
PubMedGoogle ScholarCrossref 4.White
RH. The epidemiology of venous thromboembolism.
Circulation. 2003;107(23)(suppl 1):I4-I8.
PubMedGoogle Scholar 5.Prins
MH, Marchiori
A. Risk of recurrent venous thromboembolism—expanding the frontier.
Thromb Haemost. 2002;87(1):1-3.
PubMedGoogle Scholar 6.Silverstein
MD, Heit
JA, Mohr
DN, Petterson
TM, O’Fallon
WM, Melton
LJ
III. Trends in the incidence of deep vein thrombosis and pulmonary embolism: a 25-year population-based study.
Arch Intern Med. 1998;158(6):585-593.
PubMedGoogle ScholarCrossref 7.Anderson
FA
Jr, Wheeler
HB, Goldberg
RJ,
et al. A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism: the Worcester DVT Study.
Arch Intern Med. 1991;151(5):933-938.
PubMedGoogle ScholarCrossref 8.Barritt
DW, Jordan
SC. Anticoagulant drugs in the treatment of pulmonary embolism: a controlled trial.
Lancet. 1960;1(7138):1309-1312.
PubMedGoogle ScholarCrossref 9.Carrier
M, Le Gal
G, Wells
PS, Rodger
MA. Systematic review: case-fatality rates of recurrent venous thromboembolism and major bleeding events among patients treated for venous thromboembolism.
Ann Intern Med. 2010;152(9):578-589.
PubMedGoogle ScholarCrossref 10.Kearon
C, Akl
EA, Comerota
AJ,
et al; American College of Chest Physicians. Antithrombotic therapy for VTE disease: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines.
Chest. 2012;141(2)(suppl):e419S-e494S.
PubMedGoogle Scholar 11.Ansell
JE, Hirsh
J, Hylek
E, Jacobson
A, Crowther
M, Palareti
G. Pharmacology and management of the vitamin K antagonists: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition).
Chest. 2008;133(6)(suppl):160S-198S.
PubMedGoogle ScholarCrossref 12.Schulman
S, Kearon
C; Subcommittee on Control of Anticoagulation of the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis. Definition of major bleeding in clinical investigations of antihemostatic medicinal products in non-surgical patients.
J Thromb Haemost. 2005;3(4):692-694.
PubMedGoogle ScholarCrossref 15.Hansen
D, Shneiderman
B, Smith
MA. Analyzing Social Media Networks with NodeXL: Insights from a Connected World. Burlington, MA: Elsevier Inc; 2010.
16.Fleurence
RL, Hollenbeak
CS. Rates and probabilities in economic modelling: transformation, translation and appropriate application.
Pharmacoeconomics. 2007;25(1):3-6.
PubMedGoogle ScholarCrossref 17.Agnelli
G, Buller
HR, Cohen
A,
et al; AMPLIFY Investigators. Oral apixaban for the treatment of acute venous thromboembolism.
N Engl J Med. 2013;369(9):799-808.
PubMedGoogle ScholarCrossref 18.Turner
RM, Davey
J, Clarke
MJ, Thompson
SG, Higgins
JP. Predicting the extent of heterogeneity in meta-analysis, using empirical data from the Cochrane Database of Systematic Reviews.
Int J Epidemiol. 2012;41(3):818-827.
PubMedGoogle ScholarCrossref 19.Thorlund
K, Thabane
L, Mills
EJ. Modelling heterogeneity variances in multiple treatment comparison meta-analysis—are informative priors the better solution?
BMC Med Res Methodol. 2013;13:2.
PubMedGoogle ScholarCrossref 20.Cipriani
A, Furukawa
TA, Salanti
G,
et al. Comparative efficacy and acceptability of 12 new-generation antidepressants: a multiple-treatments meta-analysis.
Lancet. 2009;373(9665):746-758.
PubMedGoogle ScholarCrossref 21.Spiegelhalter
DJ, Best
NG, Carlin
BP, Van der Linde
A. Bayesian measures of model complexity and fit.
J R Stat Soc Series B Stat Methodol. 2002;64:583-639.
Google ScholarCrossref 22.Lunn
DJ, Thomas
A, Best
N, Spiegelhalter
D. WinBUGS—a Bayesian modelling framework: concepts, structure, and extensibility.
Stat Comput. 2000;10:325-337.
Google ScholarCrossref 23.Dias
S, Welton
NJ, Sutton
AJ, Caldwell
DM, Lu
G, Ades
AE. Evidence synthesis for decision making 4: inconsistency in networks of evidence based on randomized controlled trials.
Med Decis Making. 2013;33(5):641-656.
PubMedGoogle ScholarCrossref 24.Kirchmaier
CM, Wolf
H, Schäfer
H, Ehlers
B, Breddin
HK; Certoparin-Study Group. Efficacy of a low molecular weight heparin administered intravenously or subcutaneously in comparison with intravenous unfractionated heparin in the treatment of deep venous thrombosis.
Int Angiol. 1998;17(3):135-145.
PubMedGoogle Scholar 25.Breddin
HK, Hach-Wunderle
V, Nakov
R, Kakkar
VV; CORTES Investigators: Clivarin: Assessment of Regression of Thrombosis, Efficacy, and Safety. Effects of a low-molecular-weight heparin on thrombus regression and recurrent thromboembolism in patients with deep-vein thrombosis.
N Engl J Med. 2001;344(9):626-631.
PubMedGoogle ScholarCrossref 26.Merli
G, Spiro
TE, Olsson
CG,
et al; Enoxaparin Clinical Trial Group. Subcutaneous enoxaparin once or twice daily compared with intravenous unfractionated heparin for treatment of venous thromboembolic disease.
Ann Intern Med. 2001;134(3):191-202.
PubMedGoogle ScholarCrossref 27.Kakkar
VV, Gebska
M, Kadziola
Z, Saba
N, Carrasco
P; Bemiparin Investigators. Low-molecular-weight heparin in the acute and long-term treatment of deep vein thrombosis.
Thromb Haemost. 2003;89(4):674-680.
PubMedGoogle Scholar 28.Duroux
P, Beclere
A. A randomised trial of subcutaneous low molecular weight heparin (CY 216) compared with intravenous unfractionated heparin in the treatment of deep vein thrombosis. A collaborative European multicentre study.
Thromb Haemost. 1991;65(3):251-256.
PubMedGoogle Scholar 29.Hull
RD, Raskob
GE, Pineo
GF,
et al. Subcutaneous low-molecular-weight heparin compared with continuous intravenous heparin in the treatment of proximal-vein thrombosis.
N Engl J Med. 1992;326(15):975-982.
PubMedGoogle ScholarCrossref 30.Lopaciuk
S, Meissner
AJ, Filipecki
S,
et al. Subcutaneous low molecular weight heparin versus subcutaneous unfractionated heparin in the treatment of deep vein thrombosis: a Polish multicenter trial.
Thromb Haemost. 1992;68(1):14-18.
PubMedGoogle Scholar 31.Prandoni
P, Lensing
AWA, Büller
HR,
et al. Comparison of subcutaneous low-molecular-weight heparin with intravenous standard heparin in proximal deep-vein thrombosis.
Lancet. 1992;339(8791):441-445.
PubMedGoogle ScholarCrossref 32.Simonneau
G, Charbonnier
B, Decousus
H,
et al. Subcutaneous low-molecular-weight heparin compared with continuous intravenous unfractionated heparin in the treatment of proximal deep vein thrombosis.
Arch Intern Med. 1993;153(13):1541-1546.
PubMedGoogle ScholarCrossref 33.Meyer
G, Brenot
F, Pacouret
G,
et al. Subcutaneous low-molecular-weight heparin fragmin versus intravenous unfractionated heparin in the treatment of acute non massive pulmonary embolism: an open randomized pilot study.
Thromb Haemost. 1995;74(6):1432-1435.
PubMedGoogle Scholar 34.Levine
M, Gent
M, Hirsh
J,
et al. A comparison of low-molecular-weight heparin administered primarily at home with unfractionated heparin administered in the hospital for proximal deep-vein thrombosis.
N Engl J Med. 1996;334(11):677-681.
PubMedGoogle ScholarCrossref 35.Koopman
MMW, Prandoni
P, Piovella
F,
et al; The Tasman Study Group. Treatment of venous thrombosis with intravenous unfractionated heparin administered in the hospital as compared with subcutaneous low-molecular-weight heparin administered at home.
N Engl J Med. 1996;334(11):682-687.
PubMedGoogle ScholarCrossref 36.Fiessinger
JN, Lopez-Fernandez
M, Gatterer
E,
et al. Once-daily subcutaneous dalteparin, a low molecular weight heparin, for the initial treatment of acute deep vein thrombosis.
Thromb Haemost. 1996;76(2):195-199.
PubMedGoogle Scholar 37.The Columbus Investigators. Low-molecular-weight heparin in the treatment of patients with venous thromboembolism.
N Engl J Med. 1997;337(10):657-662.
PubMedGoogle ScholarCrossref 38.Simonneau
G, Sors
H, Charbonnier
B,
et al; The THESEE Study Group. Tinzaparine ou Heparine Standard: Evaluations dans l’Embolie Pulmonaire. A comparison of low-molecular-weight heparin with unfractionated heparin for acute pulmonary embolism.
N Engl J Med. 1997;337(10):663-669.
PubMedGoogle ScholarCrossref 39.Hull
RD, Raskob
GE, Brant
RF,
et al; American-Canadian Thrombosis Study Group. Low-molecular-weight heparin vs heparin in the treatment of patients with pulmonary embolism.
Arch Intern Med. 2000;160(2):229-236.
PubMedGoogle ScholarCrossref 40.Harenberg
J, Schmidt
JA, Koppenhagen
K, Tolle
A, Huisman
MV, Büller
HR; EASTERN Investigators. Fixed-dose, body weight-independent subcutaneous LMW heparin versus adjusted dose unfractionated intravenous heparin in the initial treatment of proximal venous thrombosis.
Thromb Haemost. 2000;83(5):652-656.
PubMedGoogle Scholar 41.Findik
S, Erkan
ML, Selçuk
MB, Albayrak
S, Atici
AG, Doru
F. Low-molecular-weight heparin versus unfractionated heparin in the treatment of patients with acute pulmonary thromboembolism.
Respiration. 2002;69(5):440-444.
PubMedGoogle ScholarCrossref 42.Riess
H, Koppenhagen
K, Tolle
A,
et al; TH-4 Study Group. Fixed-dose, body weight-independent subcutaneous low molecular weight heparin Certoparin compared with adjusted-dose intravenous unfractionated heparin in patients with proximal deep venous thrombosis.
Thromb Haemost. 2003;90(2):252-259.
PubMedGoogle Scholar 43.Prandoni
P, Carnovali
M, Marchiori
A; Galilei Investigators. Subcutaneous adjusted-dose unfractionated heparin vs fixed-dose low-molecular-weight heparin in the initial treatment of venous thromboembolism.
Arch Intern Med. 2004;164(10):1077-1083.
PubMedGoogle ScholarCrossref 44.Kearon
C, Ginsberg
JS, Julian
JA,
et al; Fixed-Dose Heparin (FIDO) Investigators. Comparison of fixed-dose weight-adjusted unfractionated heparin and low-molecular-weight heparin for acute treatment of venous thromboembolism.
JAMA. 2006;296(8):935-942.
PubMedGoogle ScholarCrossref 45.Leizorovicz
A, Siguret
V, Mottier
D,
et al; Innohep in Renal Insufficiency Study Steering Committee. Safety profile of tinzaparin versus subcutaneous unfractionated heparin in elderly patients with impaired renal function treated for acute deep vein thrombosis: the Innohep in Renal Insufficiency Study (IRIS).
Thromb Res. 2011;128(1):27-34.
PubMedGoogle ScholarCrossref 46.Belcaro
G, Nicolaides
AN, Cesarone
MR,
et al. Comparison of low-molecular-weight heparin, administered primarily at home, with unfractionated heparin, administered in hospital, and subcutaneous heparin, administered at home for deep-vein thrombosis.
Angiology. 1999;50(10):781-787.
PubMedGoogle ScholarCrossref 47.Bratt
G, Aberg
W, Johansson
M, Törnebohm
E, Granqvist
S, Lockner
D. Two daily subcutaneous injections of fragmin as compared with intravenous standard heparin in the treatment of deep venous thrombosis (DVT).
Thromb Haemost. 1990;64(4):506-510.
PubMedGoogle Scholar 48.Lindmarker
P, Holmström
M, Granqvist
S, Johnsson
H, Lockner
D. Comparison of once-daily subcutaneous Fragmin with continuous intravenous unfractionated heparin in the treatment of deep vein thrombosis.
Thromb Haemost. 1994;72(2):186-190.
PubMedGoogle Scholar 49.Pini
M, Aiello
S, Manotti
C,
et al. Low molecular weight heparin versus warfarin in the prevention of recurrences after deep vein thrombosis.
Thromb Haemost. 1994;72(2):191-197.
PubMedGoogle Scholar 50.Gonzalez-Fajardo
JA, Arreba
E, Castrodeza
J,
et al. Venographic comparison of subcutaneous low-molecular weight heparin with oral anticoagulant therapy in the long-term treatment of deep venous thrombosis.
J Vasc Surg. 1999;30(2):283-292.
PubMedGoogle ScholarCrossref 51.Veiga
F, Escribá
A, Maluenda
MP,
et al. Low molecular weight heparin (enoxaparin) versus oral anticoagulant therapy (acenocoumarol) in the long-term treatment of deep venous thrombosis in the elderly: a randomized trial.
Thromb Haemost. 2000;84(4):559-564.
PubMedGoogle Scholar 52.Beckman
JA, Dunn
K, Sasahara
AA, Goldhaber
SZ. Enoxaparin monotherapy without oral anticoagulation to treat acute symptomatic pulmonary embolism.
Thromb Haemost. 2003;89(6):953-958.
PubMedGoogle Scholar 53.Pérez de Llano
LA, Baloira Villar
A, Veres Racamonde
A, Veiga
F, Golpe Gómez
R, Pajuelo Fernández
F. Multicenter, prospective study comparing enoxaparin with unfractionated heparin in the treatment of submassive pulmonary thromboembolism [in Spanish].
Arch Bronconeumol. 2003;39(8):341-345.
PubMedGoogle ScholarCrossref 54.Ramacciotti
E, Araújo
GR, Lastoria
S,
et al; CLETRAT Investigators. An open-label, comparative study of the efficacy and safety of once-daily dose of enoxaparin versus unfractionated heparin in the treatment of proximal lower limb deep-vein thrombosis.
Thromb Res. 2004;114(3):149-153.
PubMedGoogle ScholarCrossref 55.Chong
BH, Brighton
TA, Baker
RI, Thurlow
P, Lee
CH; ASTH DVT Study Group. Once-daily enoxaparin in the outpatient setting versus unfractionated heparin in hospital for the treatment of symptomatic deep-vein thrombosis.
J Thromb Thrombolysis. 2005;19(3):173-181.
PubMedGoogle ScholarCrossref 56.Daskalopoulos
ME, Daskalopoulou
SS, Tzortzis
E,
et al. Long-term treatment of deep venous thrombosis with a low molecular weight heparin (tinzaparin): a prospective randomized trial.
Eur J Vasc Endovasc Surg. 2005;29(6):638-650.
PubMedGoogle ScholarCrossref 57.Hull
RD, Pineo
GF, Brant
RF,
et al; LITE Trial Investigators. Self-managed long-term low-molecular-weight heparin therapy: the balance of benefits and harms.
Am J Med. 2007;120(1):72-82.
PubMedGoogle ScholarCrossref 58.Büller
HR, Davidson
BL, Decousus
H,
et al; Matisse Investigators. Subcutaneous fondaparinux versus intravenous unfractionated heparin in the initial treatment of pulmonary embolism.
N Engl J Med. 2003;349(18):1695-1702.
PubMedGoogle ScholarCrossref 59.Das
SK, Cohen
AT, Edmondson
RA, Melissari
E, Kakkar
VV. Low-molecular-weight heparin versus warfarin for prevention of recurrent venous thromboembolism: a randomized trial.
World J Surg. 1996;20(5):521-526, discussion 526-527.
PubMedGoogle ScholarCrossref 60.Lopaciuk
S, Bielska-Falda
H, Noszczyk
W,
et al. Low molecular weight heparin versus acenocoumarol in the secondary prophylaxis of deep vein thrombosis.
Thromb Haemost. 1999;81(1):26-31.
PubMedGoogle Scholar 61.Romera
A, Cairols
MA, Vila-Coll
R,
et al. A randomised open-label trial comparing long-term sub-cutaneous low-molecular-weight heparin compared with oral-anticoagulant therapy in the treatment of deep venous thrombosis.
Eur J Vasc Endovasc Surg. 2009;37(3):349-356.
PubMedGoogle ScholarCrossref 62.Büller
HR, Davidson
BL, Decousus
H,
et al; Matisse Investigators. Fondaparinux or enoxaparin for the initial treatment of symptomatic deep venous thrombosis: a randomized trial.
Ann Intern Med. 2004;140(11):867-873.
PubMedGoogle ScholarCrossref 63.Schulman
S, Kearon
C, Kakkar
AK,
et al; RE-COVER Study Group. Dabigatran versus warfarin in the treatment of acute venous thromboembolism.
N Engl J Med. 2009;361(24):2342-2352.
PubMedGoogle ScholarCrossref 64.Schulman
S, Kakkar
AK, Goldhaber
SZ,
et al; RE-COVER II Trial Investigators. Treatment of acute venous thromboembolism with dabigatran or warfarin and pooled analysis.
Circulation. 2014;129(7):764-772.
PubMedGoogle ScholarCrossref 65.Büller
HR, Décousus
H, Grosso
MA,
et al; Hokusai-VTE Investigators. Edoxaban versus warfarin for the treatment of symptomatic venous thromboembolism.
N Engl J Med. 2013;369(15):1406-1415.
PubMedGoogle ScholarCrossref 66.Bauersachs
R, Berkowitz
SD, Brenner
B,
et al; EINSTEIN Investigators. Oral rivaroxaban for symptomatic venous thromboembolism.
N Engl J Med. 2010;363(26):2499-2510.
PubMedGoogle ScholarCrossref 67.Büller
HR, Prins
MH, Lensin
AW,
et al; EINSTEIN–PE Investigators. Oral rivaroxaban for the treatment of symptomatic pulmonary embolism.
N Engl J Med. 2012;366(14):1287-1297.
PubMedGoogle ScholarCrossref 68.Guyatt
GH, Eikelboom
JW, Gould
MK,
et al; American College of Chest Physicians. Approach to outcome measurement in the prevention of thrombosis in surgical and medical patients: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines.
Chest. 2012;141(2)(suppl):e185S-e194S.
PubMedGoogle Scholar 69.Castellucci
LA, Cameron
C, Le Gal
G,
et al. Efficacy and safety outcomes of oral anticoagulants and antiplatelet drugs in the secondary prevention of venous thromboembolism: systematic review and network meta-analysis.
BMJ. 2013;347:f5133.
PubMedGoogle ScholarCrossref 70.Fox
BD, Kahn
SR, Langleben
D, Eisenberg
MJ, Shimony
A. Efficacy and safety of novel oral anticoagulants for treatment of acute venous thromboembolism: direct and adjusted indirect meta-analysis of randomised controlled trials.
BMJ. 2012;345:e7498.
PubMedGoogle ScholarCrossref 71.van der Hulle
T, Kooiman
J, den Exter
PL, Dekkers
OM, Klok
FA, Huisman
MV. Effectiveness and safety of novel oral anticoagulants compared with vitamin K antagonists in the treatment of acute symptomatic venous thromboembolism: a systematic review and meta-analysis.
J Thromb Haemost. 2014;12(3):320-328.
PubMedGoogle ScholarCrossref