Kalpana R. Prakasa, MBBS, MS

Search Strategy

We searched MEDLINE, EMBASE, and SCOPUS through August 2011, CINAHL, International Pharmaceutical Abstracts, clinicaltrial.gov, and the Cochrane Library through August 2012. We developed a search strategy based on medical subject headings (MeSH) terms and text words of key articles that we identified a priori[9] (see the Appendix for search strategy details).

Study Selection

We reviewed titles followed by abstracts to identify randomized controlled trials (RCTs) or observational studies with comparison groups reporting on the effectiveness or safety of VTE prevention in our populations. Two investigators independently reviewed abstracts, and we excluded the abstracts if both investigators agreed that the article met 1 or more of the exclusion criteria. We included only English‐language articles that evaluated the effectiveness of pharmacological or mechanical interventions that have been approved for clinical use in the United States. To be eligible, the studies must have addressed relevant key questions in the population of our interest. We resolved disagreements by consensus. We used DistillerSR (Evidence Partners Inc., Ottawa, Ontario, Canada), a Web‐based database management program to manage the review process. Two investigators assessed the risk of bias in each study independently, using the Downs and Black instrument for observational studies and trials.[10]

Data Synthesis

For each select population, we created detailed evidence tables containing the information abstracted from the eligible studies. After synthesizing the evidence, we graded the quantity, quality, and consistency of the best available evidence for each select population by adapting an evidence‐grading scheme recommended in the Methods Guide for Conducting Comparative Effectiveness Reviews.[7]

Obese Patients

We found 1 subgroup analysis of an RCT (total 3706 patients, 2563 nonobese and 1118 obese patients) that reported on the comparative effectiveness and safety of fixed low‐dose dalteparin 5000 IU/day compared to placebo among 1118 hospitalized medically ill patients with body mass indices (BMI) greater than 30 kg/m².¹¹ Neither group received additional concurrent prophylactic therapies. The 3 most prevalent medical diagnoses prompting hospitalization were congestive heart failure, respiratory failure, and infectious diseases. Compression ultrasound was performed in all patients by day 21 of hospitalization. The primary end point was the composite of VTE, fatal PE, and sudden death, and secondary end points included DVT, bleeding, and thrombocytopenia by day 21 (Table 1). In obese patients, the primary end point occurred in 2.8% (95% confidence interval [CI]: 1.34.3) of the dalteparin group and in 4.3% (95% CI: 2.56.2) of the placebo group (relative risk [RR]: 0.64; 95% CI: 0.32‐1.28). In nonobese patients, the primary end point occurred in 2.8% (95% CI: 1.8‐3.8) and 5.2% (95% CI: 3.9‐6.6) of the dalteparin and placebo groups, respectively (RR: 0.53; 95% CI: 0.34‐0.82). When weight was modeled as a continuous variable, no statistically significant interaction between weight and dalteparin efficacy was observed (P=0.97). The authors calculated the RR in predefined BMI subgroups and found that dalteparin was effective in reducing VTE in patients with BMIs up to 40, with RRs of <1.0 for all (approximate range, 0.20.8). However, a fixed dose of dalteparin 5000 IU/day was not better than placebo for individuals with BMI >40 kg/m². There was no significant difference in mortality or major hemorrhage by day 21 between treatment and placebo groups.

Freeman and colleagues prospectively assigned 31 medically ill patients with extreme obesity (BMI >40 kg/m²) to 1 of 3 dosing regimens of enoxaparin: a fixed dose of 40 mg daily enoxaparin (control group, n=11), enoxaparin at 0.4 mg/kg (n=9), or enoxaparin at 0.5 mg/kg (n=11).[12] The average BMI of the entire cohort was 62.1 kg/m² (range, 40.582.4). All patients had anti‐factor Xa levels drawn on the day of enrollment and daily for 3 days (Table 2). The relationship between anti‐factor Xa levels and clinical efficacy of low‐molecular weight heparin (LMWH) in VTE prophylaxis is still unclear; however, an anti‐factor Xa level of 0.2 to 0.5 IU/mL, measured 4 hours after the fourth dose of LMWH, is the target level recommended for VTE prophylaxis.[13] Patients who received weight‐based enoxaparin at 0.5mg/kg achieved target anti‐factor Xa level 86% of the time compared to 32% of the time in those receiving 0.4 mg/kg and 19% of the time for those in the fixed‐dose group (P<0.001). No clinical outcomes were reported in this study.

Patients on Antiplatelet Drugs

We did not find studies that directly looked at the comparative effectiveness of VTE prophylaxis in patients who were on antiplatelet drugs including aspirin. However, there were 2 studies that looked at the risk of bleeding in patients who received VTE pharmacologic prophylaxis while concurrently taking antiplatelet agents including aspirin. Both studies used pooled data from large phase III trials.

The study by Eriksson et al. used data from the RECORD (Regulation of Coagulation in Orthopedic Surgery to Prevent Deep Venous Thrombosis and Pulmonary Embolism) trial where over 12,000 patients undergoing elective total knee or hip replacement were randomized to receive VTE prophylaxis with oral rivaroxaban or subcutaneous enoxaparin.[14] Nine percent of participants in each arm (563 in rivaroxaban and 526 in enoxaparin/placebo) were concomitantly using antiplatelet agents or aspirin at least once during the at risk period, defined as starting at day 1 of surgery up to 2 days after the last intake of the study drug. The only end point evaluated was bleeding, and the authors found no statistically significant bleeding difference among the 2 arms (Table 1). Any bleeding event in the rivaroxaban with antiplatelets or aspirin arm was found in 20 (3.6%) patients, whereas in those on enoxaparin/placebo with antiplatelets or aspirin arm it was 17 (3.2%). The relative rate of bleeding among users versus nonusers of antiplatelet drugs or aspirin was 1.32 (95% CI: 0.85‐2.05) in the rivaroxaban group and 1.40 (95% CI: 0.87‐2.25) in the enoxaparin arm (Table 1).

Friedman et al. used pooled data from the RE‐MODEL, RENOVATE, and REMOBILIZE trials, where patients who were undergoing hip or knee arthroplasty were randomized to 220 mg of dabigatran once daily, 150 mg of dabigatran once daily (we focused on this lower dosage as this is the only available dose used in the US), 40 mg of enoxaparin once daily, or 30 mg of enoxaparin twice a day.[15] Of the 8135 patients, 4.7% were on concomitant aspirin. The baseline characteristics of those on aspirin were similar to the other enrollees. The primary outcome was major bleeding events requiring transfusion, symptomatic internal bleeding, or bleeding requiring surgery. Among patients receiving 150 mg of dabigatran, bleeding events with and without concomitant aspirin occurred in 1.6% and 1.0%, respectively (odds ratio [OR]: 1.64; 95% CI: 0.36‐7.49; P=0.523). The percentages of participants with bleeding who received enoxaparin, with and without aspirin, were 3.0% and 1.2%, respectively (OR: 2.57; 95% CI: 0.83‐7.94; P=0.101). The RR of bleeding on dabigatran compared to enoxaparin with and without aspirin therapy was 0.55 (95% CI: 0.11‐2.78) and 0.82 (95% CI: 0.37‐1.84), respectively (Table 1).

Patients With Renal Insufficiency

We found 5 studies that evaluated the comparative effectiveness and safety of pharmacologic prophylaxis for prevention of VTE in patients with acute kidney injury, moderate renal insufficiency, severe renal insufficiency not undergoing dialysis, or patients receiving dialysis. Four studies were RCTs,[16, 17, 18, 19] and 1 used a cohort design assessing separate cohorts before and after a quality improvement intervention.[20] Bauersachs and colleagues conducted an RCT comparing unfractionated heparin at 5000 IU, 3 times daily to certoparin, which is not approved in the United States and is not further discussed here.[16] The rate of DVT among patients treated with unfractionated heparin in patients with a glomerular filtration rate >30 mL/min was marginally lower than those with severe renal dysfunction (10.3 vs 11.1%) (Table 1).

Patients with severe renal dysfunction who received 5000 IU of unfractionated heparin 3 times a day were at increased risk of all bleeds (RR: 3.4; 95% CI: 2.05.9), major bleeds (RR: 7.3; 95% CI: 3.316), and minor bleeds (RR: 2.6; 95% CI: 1.4‐4.9) compared to patients treated with unfractionated heparin without severe renal dysfunction.[16]

A randomized trial by Mah and colleagues compared drug accumulation and anti‐Xa activity in elderly patients with renal dysfunction (defined as a glomerular filtration rate of 20 to 50 mL/min) who received either tinzaparin at 4500 IU once daily or enoxaparin at 4000 IU once daily.[17] Enoxaparin accumulated to a greater extent from day 1 to day 8 than did tinzaparin; the ratio of maximum concentration on day 8 compared to day 1 was 1.22 for enoxaparin and 1.05 for tinzaparin (P=0.016). No VTE events were reported in patients who received tinzaparin or enoxaparin. There was no statistical difference in the incidence of bleeding events between patients receiving tinzaparin (5, including 2 major events) and enoxaparin (4, including 3 major events, P=0.67) (Table 1).

The trial by Dahl and colleagues randomly assigned patients who were over 75 years of age and/or who had moderate renal dysfunction (defined as creatinine clearance between 30 and 49 mL/min) to receive enoxaparin 40 mg daily or dabigatran 150 mg daily.[18] There was no significant difference in the rate of major VTE events between patients receiving dabigatran (4.3%) and enoxaparin (9%) (OR: 0.48; 95% CI: 0.13‐1.73; P=0.271) (Table 1). The rate of major bleeding was significantly higher among patients randomly assigned to receive enoxaparin (4.7%) versus dabigatran (0%) (P=0.039).[18]

Shorr and colleagues published a post hoc subgroup analysis of a multicenter trial in which orthopedic patients were randomly assigned to receive desirudin 15 mg twice daily or enoxaparin 40 mg once daily.[19] Evaluable patients (1565 of the 2079 patients randomized in the trial) receiving desirudin experienced a significantly lower rate of major VTE compared with patients receiving enoxaparin (4.9% vs 7.6%, P=0.019). This relationship was particularly pronounced for evaluable patients whose creatinine clearance was between 30 and 44 mL/min. In evaluable patients with this degree of renal dysfunction, 11% of patients taking enoxaparin compared to 3.4% of those taking desirudin had a major VTE (OR: 3.52; 95% CI: 1.48‐8.4; P=0.004). There was no significant difference in the rates of major bleeding among a subset of patients assessed for safety outcomes (2078 of the 2079 patients randomized in the trial) who received desirudin (0.8%) or enoxaparin (0.2%) (Table 1).

Elsaid and Collins assessed VTE and bleeding events associated with the use of unfractionated heparin 5000U either 2 or 3 times daily and enoxaparin 30 mg once or twice daily across patients stratified by renal function (creatinine clearance <30, 3059, and 60 mL/min). The investigators made assessments before and after a quality improvement intervention that was designed to eliminate the use of enoxaparin in patients whose creatinine clearance was <30 mL/min. No VTE events were reported. Patients receiving enoxaparin were significantly more likely to experience a major bleeding episode compared with patients receiving unfractionated heparin (overall rates for all levels of renal function: 13.5% vs 4. 2%; RR: 3.2; 95% CI: 1.47.3) (Table 2). This association was largely driven by the subgroup of patients with a creatinine clearance <30 mL/min. For this subgroup with severe renal insufficiency, patients receiving enoxaparin were significantly more likely to have a bleed compared with patients receiving unfractionated heparin (18.9% vs 4.1%; RR: 4.68; 95% CI: 1.120.6) (Tables 1 and 2). There was no difference in the bleeding rates for patients whose creatinine clearances were >60 mL/min.[20]

Strength of Evidence

Search Strategy

We searched MEDLINE, EMBASE, and SCOPUS through August 2011, CINAHL, International Pharmaceutical Abstracts, clinicaltrial.gov, and the Cochrane Library through August 2012. We developed a search strategy based on medical subject headings (MeSH) terms and text words of key articles that we identified a priori[9] (see the Appendix for search strategy details).

Study Selection

We reviewed titles followed by abstracts to identify randomized controlled trials (RCTs) or observational studies with comparison groups reporting on the effectiveness or safety of VTE prevention in our populations. Two investigators independently reviewed abstracts, and we excluded the abstracts if both investigators agreed that the article met 1 or more of the exclusion criteria. We included only English‐language articles that evaluated the effectiveness of pharmacological or mechanical interventions that have been approved for clinical use in the United States. To be eligible, the studies must have addressed relevant key questions in the population of our interest. We resolved disagreements by consensus. We used DistillerSR (Evidence Partners Inc., Ottawa, Ontario, Canada), a Web‐based database management program to manage the review process. Two investigators assessed the risk of bias in each study independently, using the Downs and Black instrument for observational studies and trials.[10]

Data Synthesis

For each select population, we created detailed evidence tables containing the information abstracted from the eligible studies. After synthesizing the evidence, we graded the quantity, quality, and consistency of the best available evidence for each select population by adapting an evidence‐grading scheme recommended in the Methods Guide for Conducting Comparative Effectiveness Reviews.[7]

Obese Patients

We found 1 subgroup analysis of an RCT (total 3706 patients, 2563 nonobese and 1118 obese patients) that reported on the comparative effectiveness and safety of fixed low‐dose dalteparin 5000 IU/day compared to placebo among 1118 hospitalized medically ill patients with body mass indices (BMI) greater than 30 kg/m².¹¹ Neither group received additional concurrent prophylactic therapies. The 3 most prevalent medical diagnoses prompting hospitalization were congestive heart failure, respiratory failure, and infectious diseases. Compression ultrasound was performed in all patients by day 21 of hospitalization. The primary end point was the composite of VTE, fatal PE, and sudden death, and secondary end points included DVT, bleeding, and thrombocytopenia by day 21 (Table 1). In obese patients, the primary end point occurred in 2.8% (95% confidence interval [CI]: 1.34.3) of the dalteparin group and in 4.3% (95% CI: 2.56.2) of the placebo group (relative risk [RR]: 0.64; 95% CI: 0.32‐1.28). In nonobese patients, the primary end point occurred in 2.8% (95% CI: 1.8‐3.8) and 5.2% (95% CI: 3.9‐6.6) of the dalteparin and placebo groups, respectively (RR: 0.53; 95% CI: 0.34‐0.82). When weight was modeled as a continuous variable, no statistically significant interaction between weight and dalteparin efficacy was observed (P=0.97). The authors calculated the RR in predefined BMI subgroups and found that dalteparin was effective in reducing VTE in patients with BMIs up to 40, with RRs of <1.0 for all (approximate range, 0.20.8). However, a fixed dose of dalteparin 5000 IU/day was not better than placebo for individuals with BMI >40 kg/m². There was no significant difference in mortality or major hemorrhage by day 21 between treatment and placebo groups.

Freeman and colleagues prospectively assigned 31 medically ill patients with extreme obesity (BMI >40 kg/m²) to 1 of 3 dosing regimens of enoxaparin: a fixed dose of 40 mg daily enoxaparin (control group, n=11), enoxaparin at 0.4 mg/kg (n=9), or enoxaparin at 0.5 mg/kg (n=11).[12] The average BMI of the entire cohort was 62.1 kg/m² (range, 40.582.4). All patients had anti‐factor Xa levels drawn on the day of enrollment and daily for 3 days (Table 2). The relationship between anti‐factor Xa levels and clinical efficacy of low‐molecular weight heparin (LMWH) in VTE prophylaxis is still unclear; however, an anti‐factor Xa level of 0.2 to 0.5 IU/mL, measured 4 hours after the fourth dose of LMWH, is the target level recommended for VTE prophylaxis.[13] Patients who received weight‐based enoxaparin at 0.5mg/kg achieved target anti‐factor Xa level 86% of the time compared to 32% of the time in those receiving 0.4 mg/kg and 19% of the time for those in the fixed‐dose group (P<0.001). No clinical outcomes were reported in this study.

Patients on Antiplatelet Drugs

We did not find studies that directly looked at the comparative effectiveness of VTE prophylaxis in patients who were on antiplatelet drugs including aspirin. However, there were 2 studies that looked at the risk of bleeding in patients who received VTE pharmacologic prophylaxis while concurrently taking antiplatelet agents including aspirin. Both studies used pooled data from large phase III trials.

The study by Eriksson et al. used data from the RECORD (Regulation of Coagulation in Orthopedic Surgery to Prevent Deep Venous Thrombosis and Pulmonary Embolism) trial where over 12,000 patients undergoing elective total knee or hip replacement were randomized to receive VTE prophylaxis with oral rivaroxaban or subcutaneous enoxaparin.[14] Nine percent of participants in each arm (563 in rivaroxaban and 526 in enoxaparin/placebo) were concomitantly using antiplatelet agents or aspirin at least once during the at risk period, defined as starting at day 1 of surgery up to 2 days after the last intake of the study drug. The only end point evaluated was bleeding, and the authors found no statistically significant bleeding difference among the 2 arms (Table 1). Any bleeding event in the rivaroxaban with antiplatelets or aspirin arm was found in 20 (3.6%) patients, whereas in those on enoxaparin/placebo with antiplatelets or aspirin arm it was 17 (3.2%). The relative rate of bleeding among users versus nonusers of antiplatelet drugs or aspirin was 1.32 (95% CI: 0.85‐2.05) in the rivaroxaban group and 1.40 (95% CI: 0.87‐2.25) in the enoxaparin arm (Table 1).

Friedman et al. used pooled data from the RE‐MODEL, RENOVATE, and REMOBILIZE trials, where patients who were undergoing hip or knee arthroplasty were randomized to 220 mg of dabigatran once daily, 150 mg of dabigatran once daily (we focused on this lower dosage as this is the only available dose used in the US), 40 mg of enoxaparin once daily, or 30 mg of enoxaparin twice a day.[15] Of the 8135 patients, 4.7% were on concomitant aspirin. The baseline characteristics of those on aspirin were similar to the other enrollees. The primary outcome was major bleeding events requiring transfusion, symptomatic internal bleeding, or bleeding requiring surgery. Among patients receiving 150 mg of dabigatran, bleeding events with and without concomitant aspirin occurred in 1.6% and 1.0%, respectively (odds ratio [OR]: 1.64; 95% CI: 0.36‐7.49; P=0.523). The percentages of participants with bleeding who received enoxaparin, with and without aspirin, were 3.0% and 1.2%, respectively (OR: 2.57; 95% CI: 0.83‐7.94; P=0.101). The RR of bleeding on dabigatran compared to enoxaparin with and without aspirin therapy was 0.55 (95% CI: 0.11‐2.78) and 0.82 (95% CI: 0.37‐1.84), respectively (Table 1).

Patients With Renal Insufficiency

We found 5 studies that evaluated the comparative effectiveness and safety of pharmacologic prophylaxis for prevention of VTE in patients with acute kidney injury, moderate renal insufficiency, severe renal insufficiency not undergoing dialysis, or patients receiving dialysis. Four studies were RCTs,[16, 17, 18, 19] and 1 used a cohort design assessing separate cohorts before and after a quality improvement intervention.[20] Bauersachs and colleagues conducted an RCT comparing unfractionated heparin at 5000 IU, 3 times daily to certoparin, which is not approved in the United States and is not further discussed here.[16] The rate of DVT among patients treated with unfractionated heparin in patients with a glomerular filtration rate >30 mL/min was marginally lower than those with severe renal dysfunction (10.3 vs 11.1%) (Table 1).

Patients with severe renal dysfunction who received 5000 IU of unfractionated heparin 3 times a day were at increased risk of all bleeds (RR: 3.4; 95% CI: 2.05.9), major bleeds (RR: 7.3; 95% CI: 3.316), and minor bleeds (RR: 2.6; 95% CI: 1.4‐4.9) compared to patients treated with unfractionated heparin without severe renal dysfunction.[16]

A randomized trial by Mah and colleagues compared drug accumulation and anti‐Xa activity in elderly patients with renal dysfunction (defined as a glomerular filtration rate of 20 to 50 mL/min) who received either tinzaparin at 4500 IU once daily or enoxaparin at 4000 IU once daily.[17] Enoxaparin accumulated to a greater extent from day 1 to day 8 than did tinzaparin; the ratio of maximum concentration on day 8 compared to day 1 was 1.22 for enoxaparin and 1.05 for tinzaparin (P=0.016). No VTE events were reported in patients who received tinzaparin or enoxaparin. There was no statistical difference in the incidence of bleeding events between patients receiving tinzaparin (5, including 2 major events) and enoxaparin (4, including 3 major events, P=0.67) (Table 1).

The trial by Dahl and colleagues randomly assigned patients who were over 75 years of age and/or who had moderate renal dysfunction (defined as creatinine clearance between 30 and 49 mL/min) to receive enoxaparin 40 mg daily or dabigatran 150 mg daily.[18] There was no significant difference in the rate of major VTE events between patients receiving dabigatran (4.3%) and enoxaparin (9%) (OR: 0.48; 95% CI: 0.13‐1.73; P=0.271) (Table 1). The rate of major bleeding was significantly higher among patients randomly assigned to receive enoxaparin (4.7%) versus dabigatran (0%) (P=0.039).[18]

Shorr and colleagues published a post hoc subgroup analysis of a multicenter trial in which orthopedic patients were randomly assigned to receive desirudin 15 mg twice daily or enoxaparin 40 mg once daily.[19] Evaluable patients (1565 of the 2079 patients randomized in the trial) receiving desirudin experienced a significantly lower rate of major VTE compared with patients receiving enoxaparin (4.9% vs 7.6%, P=0.019). This relationship was particularly pronounced for evaluable patients whose creatinine clearance was between 30 and 44 mL/min. In evaluable patients with this degree of renal dysfunction, 11% of patients taking enoxaparin compared to 3.4% of those taking desirudin had a major VTE (OR: 3.52; 95% CI: 1.48‐8.4; P=0.004). There was no significant difference in the rates of major bleeding among a subset of patients assessed for safety outcomes (2078 of the 2079 patients randomized in the trial) who received desirudin (0.8%) or enoxaparin (0.2%) (Table 1).

Elsaid and Collins assessed VTE and bleeding events associated with the use of unfractionated heparin 5000U either 2 or 3 times daily and enoxaparin 30 mg once or twice daily across patients stratified by renal function (creatinine clearance <30, 3059, and 60 mL/min). The investigators made assessments before and after a quality improvement intervention that was designed to eliminate the use of enoxaparin in patients whose creatinine clearance was <30 mL/min. No VTE events were reported. Patients receiving enoxaparin were significantly more likely to experience a major bleeding episode compared with patients receiving unfractionated heparin (overall rates for all levels of renal function: 13.5% vs 4. 2%; RR: 3.2; 95% CI: 1.47.3) (Table 2). This association was largely driven by the subgroup of patients with a creatinine clearance <30 mL/min. For this subgroup with severe renal insufficiency, patients receiving enoxaparin were significantly more likely to have a bleed compared with patients receiving unfractionated heparin (18.9% vs 4.1%; RR: 4.68; 95% CI: 1.120.6) (Tables 1 and 2). There was no difference in the bleeding rates for patients whose creatinine clearances were >60 mL/min.[20]

Strength of Evidence

Search Strategy

We searched MEDLINE, EMBASE, and SCOPUS through August 2011, CINAHL, International Pharmaceutical Abstracts, clinicaltrial.gov, and the Cochrane Library through August 2012. We developed a search strategy based on medical subject headings (MeSH) terms and text words of key articles that we identified a priori[9] (see the Appendix for search strategy details).

Study Selection

We reviewed titles followed by abstracts to identify randomized controlled trials (RCTs) or observational studies with comparison groups reporting on the effectiveness or safety of VTE prevention in our populations. Two investigators independently reviewed abstracts, and we excluded the abstracts if both investigators agreed that the article met 1 or more of the exclusion criteria. We included only English‐language articles that evaluated the effectiveness of pharmacological or mechanical interventions that have been approved for clinical use in the United States. To be eligible, the studies must have addressed relevant key questions in the population of our interest. We resolved disagreements by consensus. We used DistillerSR (Evidence Partners Inc., Ottawa, Ontario, Canada), a Web‐based database management program to manage the review process. Two investigators assessed the risk of bias in each study independently, using the Downs and Black instrument for observational studies and trials.[10]

Data Synthesis

For each select population, we created detailed evidence tables containing the information abstracted from the eligible studies. After synthesizing the evidence, we graded the quantity, quality, and consistency of the best available evidence for each select population by adapting an evidence‐grading scheme recommended in the Methods Guide for Conducting Comparative Effectiveness Reviews.[7]

Obese Patients

We found 1 subgroup analysis of an RCT (total 3706 patients, 2563 nonobese and 1118 obese patients) that reported on the comparative effectiveness and safety of fixed low‐dose dalteparin 5000 IU/day compared to placebo among 1118 hospitalized medically ill patients with body mass indices (BMI) greater than 30 kg/m².¹¹ Neither group received additional concurrent prophylactic therapies. The 3 most prevalent medical diagnoses prompting hospitalization were congestive heart failure, respiratory failure, and infectious diseases. Compression ultrasound was performed in all patients by day 21 of hospitalization. The primary end point was the composite of VTE, fatal PE, and sudden death, and secondary end points included DVT, bleeding, and thrombocytopenia by day 21 (Table 1). In obese patients, the primary end point occurred in 2.8% (95% confidence interval [CI]: 1.34.3) of the dalteparin group and in 4.3% (95% CI: 2.56.2) of the placebo group (relative risk [RR]: 0.64; 95% CI: 0.32‐1.28). In nonobese patients, the primary end point occurred in 2.8% (95% CI: 1.8‐3.8) and 5.2% (95% CI: 3.9‐6.6) of the dalteparin and placebo groups, respectively (RR: 0.53; 95% CI: 0.34‐0.82). When weight was modeled as a continuous variable, no statistically significant interaction between weight and dalteparin efficacy was observed (P=0.97). The authors calculated the RR in predefined BMI subgroups and found that dalteparin was effective in reducing VTE in patients with BMIs up to 40, with RRs of <1.0 for all (approximate range, 0.20.8). However, a fixed dose of dalteparin 5000 IU/day was not better than placebo for individuals with BMI >40 kg/m². There was no significant difference in mortality or major hemorrhage by day 21 between treatment and placebo groups.

Freeman and colleagues prospectively assigned 31 medically ill patients with extreme obesity (BMI >40 kg/m²) to 1 of 3 dosing regimens of enoxaparin: a fixed dose of 40 mg daily enoxaparin (control group, n=11), enoxaparin at 0.4 mg/kg (n=9), or enoxaparin at 0.5 mg/kg (n=11).[12] The average BMI of the entire cohort was 62.1 kg/m² (range, 40.582.4). All patients had anti‐factor Xa levels drawn on the day of enrollment and daily for 3 days (Table 2). The relationship between anti‐factor Xa levels and clinical efficacy of low‐molecular weight heparin (LMWH) in VTE prophylaxis is still unclear; however, an anti‐factor Xa level of 0.2 to 0.5 IU/mL, measured 4 hours after the fourth dose of LMWH, is the target level recommended for VTE prophylaxis.[13] Patients who received weight‐based enoxaparin at 0.5mg/kg achieved target anti‐factor Xa level 86% of the time compared to 32% of the time in those receiving 0.4 mg/kg and 19% of the time for those in the fixed‐dose group (P<0.001). No clinical outcomes were reported in this study.

Patients on Antiplatelet Drugs

We did not find studies that directly looked at the comparative effectiveness of VTE prophylaxis in patients who were on antiplatelet drugs including aspirin. However, there were 2 studies that looked at the risk of bleeding in patients who received VTE pharmacologic prophylaxis while concurrently taking antiplatelet agents including aspirin. Both studies used pooled data from large phase III trials.

The study by Eriksson et al. used data from the RECORD (Regulation of Coagulation in Orthopedic Surgery to Prevent Deep Venous Thrombosis and Pulmonary Embolism) trial where over 12,000 patients undergoing elective total knee or hip replacement were randomized to receive VTE prophylaxis with oral rivaroxaban or subcutaneous enoxaparin.[14] Nine percent of participants in each arm (563 in rivaroxaban and 526 in enoxaparin/placebo) were concomitantly using antiplatelet agents or aspirin at least once during the at risk period, defined as starting at day 1 of surgery up to 2 days after the last intake of the study drug. The only end point evaluated was bleeding, and the authors found no statistically significant bleeding difference among the 2 arms (Table 1). Any bleeding event in the rivaroxaban with antiplatelets or aspirin arm was found in 20 (3.6%) patients, whereas in those on enoxaparin/placebo with antiplatelets or aspirin arm it was 17 (3.2%). The relative rate of bleeding among users versus nonusers of antiplatelet drugs or aspirin was 1.32 (95% CI: 0.85‐2.05) in the rivaroxaban group and 1.40 (95% CI: 0.87‐2.25) in the enoxaparin arm (Table 1).

Friedman et al. used pooled data from the RE‐MODEL, RENOVATE, and REMOBILIZE trials, where patients who were undergoing hip or knee arthroplasty were randomized to 220 mg of dabigatran once daily, 150 mg of dabigatran once daily (we focused on this lower dosage as this is the only available dose used in the US), 40 mg of enoxaparin once daily, or 30 mg of enoxaparin twice a day.[15] Of the 8135 patients, 4.7% were on concomitant aspirin. The baseline characteristics of those on aspirin were similar to the other enrollees. The primary outcome was major bleeding events requiring transfusion, symptomatic internal bleeding, or bleeding requiring surgery. Among patients receiving 150 mg of dabigatran, bleeding events with and without concomitant aspirin occurred in 1.6% and 1.0%, respectively (odds ratio [OR]: 1.64; 95% CI: 0.36‐7.49; P=0.523). The percentages of participants with bleeding who received enoxaparin, with and without aspirin, were 3.0% and 1.2%, respectively (OR: 2.57; 95% CI: 0.83‐7.94; P=0.101). The RR of bleeding on dabigatran compared to enoxaparin with and without aspirin therapy was 0.55 (95% CI: 0.11‐2.78) and 0.82 (95% CI: 0.37‐1.84), respectively (Table 1).

Patients With Renal Insufficiency

We found 5 studies that evaluated the comparative effectiveness and safety of pharmacologic prophylaxis for prevention of VTE in patients with acute kidney injury, moderate renal insufficiency, severe renal insufficiency not undergoing dialysis, or patients receiving dialysis. Four studies were RCTs,[16, 17, 18, 19] and 1 used a cohort design assessing separate cohorts before and after a quality improvement intervention.[20] Bauersachs and colleagues conducted an RCT comparing unfractionated heparin at 5000 IU, 3 times daily to certoparin, which is not approved in the United States and is not further discussed here.[16] The rate of DVT among patients treated with unfractionated heparin in patients with a glomerular filtration rate >30 mL/min was marginally lower than those with severe renal dysfunction (10.3 vs 11.1%) (Table 1).

Patients with severe renal dysfunction who received 5000 IU of unfractionated heparin 3 times a day were at increased risk of all bleeds (RR: 3.4; 95% CI: 2.05.9), major bleeds (RR: 7.3; 95% CI: 3.316), and minor bleeds (RR: 2.6; 95% CI: 1.4‐4.9) compared to patients treated with unfractionated heparin without severe renal dysfunction.[16]

A randomized trial by Mah and colleagues compared drug accumulation and anti‐Xa activity in elderly patients with renal dysfunction (defined as a glomerular filtration rate of 20 to 50 mL/min) who received either tinzaparin at 4500 IU once daily or enoxaparin at 4000 IU once daily.[17] Enoxaparin accumulated to a greater extent from day 1 to day 8 than did tinzaparin; the ratio of maximum concentration on day 8 compared to day 1 was 1.22 for enoxaparin and 1.05 for tinzaparin (P=0.016). No VTE events were reported in patients who received tinzaparin or enoxaparin. There was no statistical difference in the incidence of bleeding events between patients receiving tinzaparin (5, including 2 major events) and enoxaparin (4, including 3 major events, P=0.67) (Table 1).

The trial by Dahl and colleagues randomly assigned patients who were over 75 years of age and/or who had moderate renal dysfunction (defined as creatinine clearance between 30 and 49 mL/min) to receive enoxaparin 40 mg daily or dabigatran 150 mg daily.[18] There was no significant difference in the rate of major VTE events between patients receiving dabigatran (4.3%) and enoxaparin (9%) (OR: 0.48; 95% CI: 0.13‐1.73; P=0.271) (Table 1). The rate of major bleeding was significantly higher among patients randomly assigned to receive enoxaparin (4.7%) versus dabigatran (0%) (P=0.039).[18]

Shorr and colleagues published a post hoc subgroup analysis of a multicenter trial in which orthopedic patients were randomly assigned to receive desirudin 15 mg twice daily or enoxaparin 40 mg once daily.[19] Evaluable patients (1565 of the 2079 patients randomized in the trial) receiving desirudin experienced a significantly lower rate of major VTE compared with patients receiving enoxaparin (4.9% vs 7.6%, P=0.019). This relationship was particularly pronounced for evaluable patients whose creatinine clearance was between 30 and 44 mL/min. In evaluable patients with this degree of renal dysfunction, 11% of patients taking enoxaparin compared to 3.4% of those taking desirudin had a major VTE (OR: 3.52; 95% CI: 1.48‐8.4; P=0.004). There was no significant difference in the rates of major bleeding among a subset of patients assessed for safety outcomes (2078 of the 2079 patients randomized in the trial) who received desirudin (0.8%) or enoxaparin (0.2%) (Table 1).

Elsaid and Collins assessed VTE and bleeding events associated with the use of unfractionated heparin 5000U either 2 or 3 times daily and enoxaparin 30 mg once or twice daily across patients stratified by renal function (creatinine clearance <30, 3059, and 60 mL/min). The investigators made assessments before and after a quality improvement intervention that was designed to eliminate the use of enoxaparin in patients whose creatinine clearance was <30 mL/min. No VTE events were reported. Patients receiving enoxaparin were significantly more likely to experience a major bleeding episode compared with patients receiving unfractionated heparin (overall rates for all levels of renal function: 13.5% vs 4. 2%; RR: 3.2; 95% CI: 1.47.3) (Table 2). This association was largely driven by the subgroup of patients with a creatinine clearance <30 mL/min. For this subgroup with severe renal insufficiency, patients receiving enoxaparin were significantly more likely to have a bleed compared with patients receiving unfractionated heparin (18.9% vs 4.1%; RR: 4.68; 95% CI: 1.120.6) (Tables 1 and 2). There was no difference in the bleeding rates for patients whose creatinine clearances were >60 mL/min.[20]

Strength of Evidence