Venous Thromboembolism Prophylaxis After Major Orthopaedic Surgery: A Pooled Analysis of Randomized Controlled Trials

Transcription

1 Winner of the AAHKS Award Venous Thromboembolism Prophylaxis After Major Orthopaedic Surgery: A Pooled Analysis of Randomized Controlled Trials Greg A. Brown, MD, PhD The Journal of Arthroplasty Vol. 24 No. 6 Suppl Abstract: The use of aspirin for venous thromboembolism (VTE) prophylaxis after major orthopaedic surgery is controversial. The hypothesis of the present study is that aspirin will decrease the rate of operative site bleeding without increasing thromboembolic events when aspirin is used for VTE prophylaxis after major orthopaedic surgery. A pooled analysis of 14 randomized controlled trials (RCTs) cited by the American College of Chest Physicians (ACCP) guidelines was performed to determine pooled rates of symptomatic deep vein thromboses, pulmonary emboli (PE), fatal PE, and operative site bleeding rates. The VTE rates with aspirin were not significantly different than the rates for vitamin K antagonists (VKA), low molecular weight heparins (LMWH), and pentasaccharides. The operative site bleeding relative risks of VKA, LMWH, and pentasaccharides versus aspirin, are 4.9, 6.4, and 4.2, respectively. A pooled analysis of RCTs supports the use of aspirin for VTE prophylaxis after major orthopaedic surgery. Keywords: venous thromboembolism, orthopaedic surgery, pooled rates Published by Elsevier Inc. Guideline compliance is currently being used as a quality measure of health care in the United Statess. The American College of Chest Physicians (ACCP) has published guidelines for venous thromboembolism (VTE) prevention for multiple medical conditions and surgical procedures that place patients at high risk for venous thromboembolic events, including orthopaedic procedures for total hip arthroplasty (THA), total knee arthroplasty (TKA), and hip fracture surgery (HFS) [1]. Following identification of these VTE high-risk conditions and procedures, ACCP made treatment recommendations for HFS, THA, and TKA. Orthopaedic surgeons have raised numerous concerns about bleeding risks and wound adverse events for THA, TKA, or HFS patients when using ACCP-recommended guidelines, but these concerns have been ignored [2]. From the Department of Orthopaedic Surgery, Park Nicollet Health Services, St. Louis Park, Minnesota; and Department of Orthopaedic Surgery, University of Minnesota, Minneapolis, Minnesota. Submitted January 29, 2009; accepted June 2, No benefits or funds were received in support of the study. Reprint requests: Greg A. Brown, MD, PhD, Department of Orthopaedic Surgery, Park Nicollet Clinic Meadowbrook, Suite E400, 6490 Excelsior Blvd, St. Louis Park, MN Published by Elsevier Inc /09/ $36.00/0 doi: /j.arth The ACCP guidelines purport to use evidence-based medicine principles when evaluating venous thromboembolism prophylaxis after major orthopaedic procedures. However, the methodological flaws incorporated in the ACCP guidelines include: (1) exclusion of randomized controlled trials (RCTs) without venographic outcome assessment; (2) incomplete outcome measures without inclusion criteria requiring measurement of symptomatic deep vein thromboses (DVT), pulmonary emboli (PE), fatal pulmonary emboli, major operative site bleeding complications, and major non-operative site bleeding complications; (3) no quantitative analyses (meta-analysis or pooled analysis) of randomized controlled trials to estimate incidence of symptomatic DVTs, PEs, fatal PEs, major operative site bleeding complications, and major non-operative site bleeding complications; and (4) potential conflicts of interest for multiple members of the guideline drafting committee [3]. The controversy has worsened because the Surgical Care Improvement Project (SCIP) has adopted VTE prophylaxis recommendations for selected surgical procedures (including elective total hip arthroplasty, elective total knee arthroplasty, and hip fracture surgery) from the 2004 ACCP guidelines [4]. The hospital pay-forperformance (P4P) program developed by the Centers for Medicare and Medicaid Services requires hospitals to 77

2 78 The Journal of Arthroplasty Vol. 24 No. 6 Suppl. 1 September 2009 Table 1. Abstracted Data Elements Stratified by VTE Prophylactic Treatment Study/Year Procedure THA TKA Hip Fx Prophylaxis Number of Patients Number of Patients Tested Percent Tested Bauer et al/ Pentasaccharide % Lassen et al/ Pentasaccharide % Turpie et al/ Pentasaccharide % Eriksson et al/ Pentasaccharide % Friedman et al/ LMWH qd % Friedman et al/ LMWH bid % Francis et al/ LMWH % Hull et al/ LMWH % Hamulyak et al/ LMWH % Hull et al/ LMWH preop % Hull et al/ LMWH postop % Leclerc et al/ LMWH % Heit et al/ LMWH low dose % Heit et al/ LMWH mid dose % Heit et al/ LMWH high dose % Fitzgerald et al/ LMWH % Colwell et al/ LMWH Bauer et al/ LMWH % Lassen et al/ LMWH % Turpie et al/ LMWH % Eriksson et al/ LMWH % Friedman et al/ Warfarin % Francis et al/ Warfarin % Hull et al/ Warfarin % Hamulyak et al/ Warfarin % Hull et al/ Warfarin % Leclerc et al/ Warfarin % Heit et al/ Warfarin % Fitzgerald et al/ Warfarin % Colwell et al/ Warfarin Rodgers et al/ Aspirin Rodgers et al/ Aspirin Rodgers et al/ Placebo Rodgers et al/ Placebo follow the SCIP guidelines. The American Academy of Orthopaedic Surgeons (AAOS) has developed VTE guidelines based on pulmonary embolus and bleeding risk stratification [5]. The AAOS guidelines are compatible with the SCIP guidelines if a patient's increased risk for bleeding is documented in the medical record and mechanical prophylaxis measures are used in conjunction with aspirin [6]. Fourteen randomized controlled trials were cited by the 2004 ACCP guidelines. Four classes of pharmacologic agents were studied: (1) aspirin, (2) vitamin K antagonists (VKA), (3) low molecular weight heparins (LMWH), and (4) pentasaccharides. The 14 RCTs evaluated the use of pharmacologic agents for venous thromboembolism prophylaxis after major orthopaedic surgery, including total hip arthroplasty (THA), total knee arthroplasty (TKA), and hip fracture surgery (HFS). The hypothesis of the present study is that aspirin will decrease the rate of operative site bleeding adverse events without increasing the rate of thromboembolic events when

3 Venous Thromboembolism Prophylaxis Brown 79 Positive Venogram Symptomatic DVT Fatal Major Operative Site Major Non-Operative Site Total Major % 0.94% 0.80% 0.30% 1.91% 0.47% 2.38% % 1.28% 0.45% 0.13% 2.92% 1.33% 3.66% % 2.01% 0.40% 0.04% 2.24% 1.49% 2.57% % 0.62% 0.22% 0.46% 2.57% 3.03% % 1.02% 0.52% 0.47% 1.79% 2.26% used for venous thromboembolism prophylaxis after major orthopaedic surgery. Materials and Methods A quantitative systematic review (pooled analysis) of all randomized control trials cited by the ACCP guidelines for VTE prophylaxis after major orthopaedic surgery (THA, TKA, and HFS) was conducted. Nine RCTs compared pharmacologic VTE prophylaxis treatments for total hip arthroplasty [7-15]. Eight RCTs compared pharmacologic VTE prophylaxis treatments for total knee arthroplasty [9,10,13,14,16-19]. Two RCTs compared pharmacologic VTE prophylaxis treatments for hip fracture surgery [13,20]. Three randomized controlled trials included total hip arthroplasty and total knee arthroplasty [9,10,14]. One randomized controlled trial included hip fracture surgery, total hip arthroplasty, and total knee arthroplasty [13]. Fourteen randomized controlled trials were included in the systematic review. Two randomized controlled trials discussed in the ACCP guidelines (but

4 80 The Journal of Arthroplasty Vol. 24 No. 6 Suppl. 1 September 2009 Table 2. Pooled Rates Stratified by VTE Prophylactic Treatment Total Pooled Subjects Venographic Deep Vein Thrombosis Symptomatic Deep Vein Thrombosis Fatal Operative Site Non-Operative Site Total Pentasaccharide % 0.94% 0.80% 0.30% 1.91% 0.47% 2.38% LMWH % 1.28% 0.45% 0.13% 2.92% 1.33% 3.66% Warfarin % 2.01% 0.40% 0.04% 2.24% 1.49% 2.57% Aspirin % 0.62% 0.22% 0.46% 2.57% 3.03% Placebo % 1.02% 0.52% 0.47% 1.79% 2.26% excluded from the ACCP guidelines because the studies did not include venography) were included in the systematic review [7,13]. Each randomized controlled trial was reviewed and data abstracted for the following data elements: (1) screened patients, (2) procedure type (THA, TKA, or HFS), (3) cancelled procedures, (4) pharmacologic VTE prophylactic agent, (5) number of subjects in each subgroup, (6) duration of VTE prophylaxis, (7) number of subjects using graduated compression stockings, (8) use of pneumatic compression devices permitted or number of subjects using pneumatic compression devices, (9) number of subjects using non-steroidal anti-inflammatory drugs (NSAIDs), (10) number of subjects studied with venograms, (11) mean time to venogram (if applicable), (12) positive venograms, (13) positive non-invasive studies, (14) symptomatic deep vein thrombosis (DVTs), (15) pulmonary embolus (PE), (16) fatal pulmonary embolus, (17) major operative site bleeding complications, (18) major non-operative site bleeding complications, and (19) total major bleeding events (Table 5, available online). A pooled data analysis was performed stratified by VTE prophylaxis agent (Table 1). The rates of venographic DVT, symptomatic DVT, PE, fatal PE, major operative site bleeding events, and major non-operative site bleeding events were calculated based on the pooled data (Table 2). Chi-square tests were used to compare binomial pooled rates. Chi-square 2 2 tests require all cells to be 5 or greater. If a cell value was less than 5, Fisher's exact test was used. Significance level was Fifteen statistical comparisons were conducted (aspirin versus pentasaccharides, LMWH, and VKA across five rates: symptomatic DVT, PE, fatal PE, major operative site bleeding, and major non-operative site bleeding) (Table 3). A Bonferroni correction for statistical significance was performed. The corrected significance level is 0.05/15 = Binomial relative risks with confidence intervals (CI) were calculated for each comparison (Table 4). Results The abstracted data is stratified by randomized controlled trial (Table electronically published). The abstracted data is stratified by VTE prophylactic agent (Table 1). The pooled incidence of venographic deep vein thromboses, symptomatic deep vein thromboses, pulmonary emboli, fatal pulmonary emboli, major operative site bleeding complications, and major non-operative site bleeding complications are listed in Table 2. Chi-square tests for binomial variables (2 2) were used to statistically compare rates. The p-values for each statistical comparison are reported in Table 3. The following significant differences were observed: (1) symptomatic DVT rates aspirin versus VKA (P = ); and (2) operative site bleeding/wound adverse events aspirin versus VKA (P = ), aspirin versus LMWH (P = 1.0v10-35 ), aspirin versus pentasaccharides (P = ). With the Bonferroni correction (α = ), the difference in fatal pulmonary embolus rates between aspirin and VKA is not statistically significant (Chi-square P =.017, Fisher's exact test P =.019). The non-operative bleeding event rate comparisons are included in Table 3 for scientific completeness, but the statistical comparisons are not valid (see Discussion section). The two proportions relative risks and confidence intervals for symptomatic deep vein thromboses, pulmonary emboli, fatal pulmonary emboli, major operative site bleeding complications, and non-operative bleeding complications comparing (1) pentasaccharide/aspirin; (2) LMWH/aspirin; and (3) warfarin/aspirin are reported in Table 4. The non-operative bleeding event relative risks and confidence intervals are included in Table 4 for scientific completeness, but the statistical comparisons are not valid (see Discussion section). Discussion This quantitative systemic review (pooled analysis) of fourteen randomized controlled trials for venous Table 3. Chi-Square P-Values for Pooled Rates Comparing Aspirin to Pentasaccharides, LMWH, and Warfarin (Bonferroni correction 0.05/15 = ) Symptomatic Deep Vein Thrombosis Fatal Operative Site Non-Operative Site Pentasaccharide vs ASA E E-14 LMWH vs ASA E E-08 Warfarin vs ASA 3.5E E

5 Venous Thromboembolism Prophylaxis Brown 81 Table 4. Binomial s and s of Pentasaccharides, LMWH, and Warfarin Compared to Aspirin Symptomatic Deep Vein Thrombosis Fatal Operative Site Non- Operative Site Pentasaccharide vs ASA LMWH vs ASA Warfarin vs ASA thromboembolism (VTE) prophylaxis after major orthopaedic surgery finds no significant difference in clinically-relevant VTE outcomes, except vitamin K antagonists have a higher rate of symptomatic DVTs compared to aspirin. Inclusion of the Embolism Prevention (PEP) trial [13] negates the ACCP guideline's statement that there is no evidence that aspirin is effective in reducing VTE events after major orthopaedic surgery. In fact, the use of vitamin K antagonists, low-molecular-weight heparins, and pentasaccharides after major orthopaedic surgery significantly increase the risk of operative site bleeding events, without reducing clinically-relevant symptomatic deep vein thrombosis, pulmonary embolism, and fatal pulmonary embolism rates. This pooled analysis is transparent, quantitative, and includes complete clinical outcomes. However, pooled analysis results have certain limitations and biases. Because each RCT study design may differ, the outcomes are not always comparable. The PEP trial did not delineate major or minor bleeding events. However, 182 of 224 non-operative site bleeding events did not require a transfusion (placebo 122 of 154). All bleeding events were included in the pooled analysis because events could not be differentiated between major and minor. Also, the PEP trial is the only RCT comparing an active comparator to placebo. The remaining 13 RCTs compare two active comparators. The generalizability of the pooled analysis is limited by the inclusion/exclusion criteria of each RCT. However, all trials are prospective and randomized, so a selection bias should not be present. Although pooled analyses have limitations, a meta-analysis could not be performed because there are no RCTs comparing aspirin to VKAs, LMWHs, or pentasaccharides. Evidence-based medicine requires inclusion of all clinically-relevant outcomes. The ACCP guidelines minimize the importance of symptomatic DVTs, PEs, fatal PEs, and operative site bleeding complications. Level I evidence for therapeutic studies requires randomized controlled trials or systematic reviews of randomized controlled trials with homogeneous studies. There is no requirement for specific measurement techniques, such as venography. In fact, venographic testing over-estimated the effectiveness of LMWH as noted by O'Donnell et al [21]. The ACCP guidelines note we place a relatively low value on the prevention of venographic thrombosis, and a relatively high value on minimizing bleeding complications [1]. However, because the ACCP guidelines use narrative review of the RCTs, the process by which VTE and bleeding events are weighted is not transparent. The ACCP guidelines do not appear to be consistent with their noted relative value weighting. No clinical evidence is presented in the ACCP guidelines that post-thrombotic syndrome is a significant clinical problem associated with THA, TKA, or HFS. The ACCP inclusion criteria specify DVT outcome assessment with contrast venography or duplex ultrasonography (DUS). The basis for these strict inclusion criteria is the hypothesized link between asymptomatic DVTs and post-thrombotic syndrome with its associated morbidity. Without a Level 1 prognostic outcome study documenting post-thrombotic syndrome after THA, TKA, and HFS, the ACCP drafting committee has reduced their guidelines to expert opinion. The Embolism Prevention (PEP) trial is the only adequately-powered randomized controlled trial to observe a 1% difference (1% vs 2%) in symptomatic DVTs after major orthopaedic surgery (each subgroup requires 2316 subjects). After excluding the PEP trial, the ACCP guidelines recommend against the use of aspirin alone (Grade 1A). The recommendation against aspirin use is incorrect because the PEP trial is the highestpowered randomized controlled trial studying VTE prophylaxis after major orthopaedic surgery. The ACCP systematic review lacks scientific rigor. The ACCP narrative review argues that the PEP trial has methodological flaws. The ACCP review does not define an RCT quality assessment method. Without a formal study quality assessment, there is no basis for excluding the PEP trial. Cochrane Reviews have greater methodological rigor than other reviews [22]. The Cochrane Review for VTE prophylaxis after hip fractures noted that prior research in the prophylaxis of DVT and thromboembolism using interventions under review has many shortcomings. many of the above [study design

6 82 The Journal of Arthroplasty Vol. 24 No. 6 Suppl. 1 September 2009 problems] are avoided by the recent PEP trial, which can be considered a good example to follow [23]. The Scottish Intercollegiate Guidelines Network endorses the PEP study findings and recommends All patients with hip fracture should receive aspirin (150 mg orally, started on admission and continued for 35 days) unless contraindicated [24]. Other guideline organizations endorse the PEP study findings. An additional inconsistency of the ACCP guidelines is that no RCTs using vitamin K antagonists for hip fracture surgery are cited, but the guidelines support the use of VKAs for hip fracture surgery while dismissing the highest-powered randomized controlled trial of aspirin (PEP trial) and excluding the use of aspirin for hip fractures. Evidence-based medicine is intended to empower the patient's informed preference [25]. Therefore, EBM guidelines must include a transparent risk/benefit analysis of the clinical risks and benefits of any intervention. Without a meta-analysis or pooled-analysis, there is no quantitative comparison of the relative risks of aspirin, warfarin, LMWH, or pentasaccharides and the patient cannot express an informed preference. The non-operative site bleeding event rates are not comparable. Aspirin is the only pharmacologic agent compared to placebo [13]. The non-operative site bleeding event rate for aspirin was 2.57%. The nonoperative site bleeding event rate for placebo was 1.79%. The non-operative site bleeding event rate attributable to aspirin is 0.78%. This rate is lower than the nonoperative site bleeding event rates for LMWH and VKA and higher than pentasaccharides. Additionally, 182 of the 218 (83.4%) non-operative site bleeding events were gastrointestinal (GI) bleeding events that did not require a transfusion so the clinical severity of the GI bleeding events is unknown. Therefore, the differences between non-operative site bleeding event rates are reported for scientific completeness, but the significance of these findings is questionable. The application of evidence-based medicine to an individual patient requires not only a review of the clinical evidence, but an assessment of the generalizability of the clinical evidence to the individual patient. The AAOS guidelines risk stratify the patients according to VTE risks and bleeding risks to individualize evidencebased medical care [5], but the ACCP guidelines provide not risk stratification recommendations. Six of seven members of the ACCP drafting committee have listed potential conflicts of interest with pharmaceutical companies [3]. When CMS adopted the SCIP guidelines for their pay-for-performance program, CMS had an obligation to perform due diligence before accepting the ACCP guidelines and SCIP guidelines for implementation. Given the potential conflicts of interest of the ACCP drafting committee, the acceptance of the PEP trial as a model trial for VTE prophylaxis, and the cost-effectiveness of aspirin, CMS must reassess the SCIP guidelines. John Abramson has described the commercial bias on the scientific evidence' that informs medical decisions [26]. National guidelines, such as the SCIP guidelines, require a drafting committee without potential conflicts of interest. Multimodal VTE prophylaxis is safe and effective. The ACCP guidelines recommend against the use of aspirin alone (Grade 1A) [1]. However, orthopaedic surgeons are not using aspirin alone, but using aspirin as part of multimodal VTE prophylaxis including intermittent pneumatic compression devices and early mobilization [27]. Multimodal VTE prophylaxis allows the minimization of bleeding risks by using aspirin in lower VTE risk patients. The ACCP and SCIP guidelines do not endorse multimodal VTE prophylaxis. Multimodal VTE prophylaxis with regional anesthesia, pneumatic compression, and aspirin is safer than potent anticoagulants. Sharrock et al [28] conducted a metaanalysis of VTE prophylaxis evaluating three month allcause mortality comparing potent anticoagulants (Group A: low-molecular-weight heparin, ximelagatran, fondaparinux, or rivaroxaban), multimodal prophylaxis (Group B: regional anesthesia, pneumatic compression, and aspirin), and warfarin (Group C). The random effects meta-analysis found that the relative risk of three month all-cause mortality comparing Group A to Group B was 2.48 ( confidence interval , P b.01) and comparing Group C to Group B was 2.29 ( confidence interval , P =.03). Alternatively stated, the three month all-cause mortality of patients receiving potent anticoagulants or warfarin was more than two-fold higher than patients receiving multimodal VTE prophylaxis with aspirin. If evidence-based medicine is to be credible, a systematic review must be transparent, quantitative, and include all evidence. The present study supports the use of aspirin for venous thromboembolism prophylaxis after major orthopaedic surgery. Appendix A. Supplementary Data Supplementary data associated with this article can be found, in the online version, at doi: /j.arth References 1. Geerts WH, Pineo GF, Heit JA, et al. Prevention of venous thromboembolism: The seventh ACCP conference on antithrombotic and thrombolytic therapy. Chest 2004; 126:338S. 2. Callaghan JJ, et al. Prophylaxis for Thromboembolic Disease: Recommendations from the American College of Chest Physicians - Are They Appropriate for Orthopaedic Surgery? J Arthroplast 2005;20: American College of Chest Physicians. Financial disclosures. Chest 2004;126:167S. 4. Medicare Quality Improvement Community. Surgical care improvement project (SCIP) VTE prophylaxis options for surgery. [Edited]

7 Venous Thromboembolism Prophylaxis Brown American Academy of Orthopaedic Surgeons. American Academy of Orthopaedic Surgeons clinical guideline on prevention of symptomatic pulmonary embolism in patients undergoing total hip or knee arthroplasty. [Edited] www. aaos.org/research/guidelines/pe_summary.pdf, Haralson RH. Clearing up the confusion with VTE and/or PE prophylaxis guidelines. AAOS Now; p. 11. [Edited, 11]. 7. Colwell CW, Collis DK, Paulson R, et al. Comparison of enoxaparin and warfarin for the prevention of venous thromboembolic disease after total hip arthroplasty. J Bone Jt Surg Am 1999;81-A: Francis CW, et al. Prevention of deep-vein thrombosis after total hip arthroplasty. J Bone Jt Surg Am 1997; 79-A: Hamulyak K, Lensing AWA, van der Meer J, et al. Subcutaneous low-molecular-weight heparin or oral anticoagulants for the prevention of deep-vein thrombosis in elective hip and knee replacement? Thromb Haemost 1994; 74: Hull R, et al. A comparison of subcutaneous low-molecularweight heparin with warfarin sodium for prophylaxis against deep-vein thrombosis after hip or knee implantation. N Engl J Med 1993;329: Hull RD, et al. Low-molecular-weight heparin prophylaxis using dalteparin in close proximity to surgery vs warfarin in hip arthroplasty patients: A double-blind, randomized comparison. Arch Int Med 2000;160: Lassen MR, Bauer KA, Eriksson BI, et al. Postoperative fondaparinux versus preoperative enoxaparin for prevention of venous thromboembolism in elective hip-replacement surgery: A randomised double-blind comparison. Lancet 2002;359: Embolism Prevention (PEP) Trial Collaborative Group. Prevention of pulmonary embolism and deep vein thrombosis with low dose aspirin: Embolism Prevention (PEP) trial. Lancet 2000;355: RD Heparin Arthroplasty Group. RD heparin compared with warfarin for prevention of venous thromboembolic disease following total hip or knee arthroplasty. J Bone Jt Surg Am 1994;76-A: Turpie AGG, Bauer KA, Eriksson BI, et al. Postoperative fondaparinux versus postoperative enoxaparin for prevention of venous thromboembolism after elective hip-replacement surgery: A randomised double-blind trial. Lancet 2002;359: Bauer KA, Eriksson BI, Lassen MR, et al. Fondaparinux compared with enoxaparin for the prevention of venous thromboembolism after elective major knee surgery. N Engl J Med 2001;345: Fitzgerald RH, Spiro TE, Trowbridge AA, et al. Prevention of venous thromboembolic disease following primary total knee arthroplasty: A randomized, multicenter, open-label, parallel-group comparison of enoxaparin and warfarin. J Bone Jt Surg Am 2001;83-A: Heit JA, Berkowitz SD, Bona R, et al. Efficacy and safety of low molecular weight heparin (ardeparin sodium) compared to warfarin for the prevention of venous thromboembolism after total knee replacement surgery: A double-blind, dose-ranging study. Thromb Haemost 1997; 77: Leclerc JR, et al. Prevention of venous thromboembolism after knee arthroplasty: A randomized, double-blind trial comparing enoxaparin with warfarin. Ann Int Med 1996; 124: Eriksson BI, Bauer KA, Lassen MR, et al. Fondaparinux compared with enoxaparin for the prevention of venous thromboembolism after hip-fracture surgery. N Engl J Med 2001;345: O'Donnell M, Linkins L-A, Kearon C, et al. Reduction of out-of-hospital symptomatic venous thromboembolism by extended thromboprophylaxis with low-molecular-weight heparin following elective hip arthroplasty: A systematic review. Arch Int Med 2003;163: Jadad AR, Cook DJ, Jones A, et al. Methodology and reports of systematic reviews and meta-analyses: A comparison of Cochrane reviews with articles published in paper-based journals. JAMA 1998;280: Handol HHG, Farrar MJ, McBirnie J, et al. Heparin, low molecular weight heparin and physical methods for preventing deep vein thrombosis and pulmonary embolism following surgery for hip fractures. [Edited]Cochrane Database Syst Rev Scottish Intercollegiate Guidelines Network (SIGN). Prevention and Management of Hip Fracture in Older People. Edinburgh: SIGN Executive, Royal College of Physicians; p. 14. [Edited]. 25. Fisher CG, Wood KB. Introduction to and techniques of evidence-based medicine. Spine 2007;32:S Abramson J. Overdosed America: The Broken Promise of American Medicine. New York: HarperCollins; 2004 [Edited]. 27. Dorr LD, Gendelman V, Maheshwari AV, et al. Multimodal thromboprophylaxis for total hip and knee arthroplasty based on risk assessment. J Bone Jt Surg Am 2007; 89-A: Sharrock NE, Della Valle AG, Go G, et al. Potent anticoagulants are associated with a higher all-cause mortality rate after hip and knee arthroplasty. Clin Orthop Relat Res 2008;466:714.