INTRODUCTION — The risk of postoperative venous thromboembolism (VTE; deep venous thrombosis and pulmonary embolism) in orthopedic patients is among the highest of all surgical specialties.

This topic discusses VTE prevention strategies for those undergoing major orthopedic surgery. Our approach is for the most part consistent with the American College of Chest Physicians (ACCP) and the American Society of Hematology (ASH) [1,2]. Prevention of VTE in nonorthopedic surgical patients and in hospitalized medical patients are presented separately. (See "Prevention of venous thromboembolic disease in acutely ill hospitalized medical adults" and "Prevention of venous thromboembolic disease in adult nonorthopedic surgical patients".)

RISK OF THROMBOSIS AND BLEEDING — The risk of postoperative VTE and bleeding in orthopedic patients depends upon both procedure- and patient-related risk factors, all of which need to be assessed prior to surgery. Noteworthy is that orthopedic surgeries can be elective or are trauma-related, and the spectrum of surgeries range from minor to major. Thus, an individual patient’s risk of both thrombosis and bleeding are highly variable, and the preferred method of thromboprophylaxis should, therefore, be individualized.    

Assessing the risk of thrombosis — At baseline, select orthopedic surgeries are considered high risk (hip and knee arthroplasty, hip fracture surgery, pelvic and multiple fractures) and low risk (foot and ankle fractures; tibial, shoulder and elbow surgery; arthroscopy) for VTE with further stratification necessary according to the presence of additional patient- or procedure-related factors. Although the Caprini score (table 1) can be used to stratify the risk of VTE, the score was validated in patients undergoing non-orthopedic surgery [3]. For orthopedic patients, gestalt assessment is often used. (See "Prevention of venous thromboembolic disease in adult nonorthopedic surgical patients", section on 'Thrombosis risk model (Caprini)'.)

Accurate estimates of the baseline VTE risk, in the absence of prophylaxis, have been hampered by factors including the derivation of data from heterogeneous populations, the reporting of both symptomatic and asymptomatic events, and the reduced prevalence of VTE likely due to increasing use of thromboprophylaxis and early ambulation during hospital admission [4-15]. In addition, most of the data that assess the baseline risk of VTE are derived from populations of patients undergoing major orthopedic surgeries (typically hip or knee arthroplasty and hip fracture surgery) and may not reflect the prevalence in patients undergoing minor orthopedic surgery (eg, arthroscopic procedures, foot and ankle surgery, upper extremity surgery), who are typically younger and for whom the risk is presumably lower.

VTE risk is procedure- and patient-related:

●Procedure-related – Several procedure-related factors contribute to the risk of VTE in all surgical patients including the extent and duration of surgery, the type of anesthesia, and likelihood for immobilization and casting postoperatively.

●Patient-related – Several patient-related factors (eg, malignancy, prior VTE) that can contribute to VTE risk in orthopedic patients are listed in the table (table 2). Among these, factors that are specific to patients undergoing major orthopedic surgery that contribute to the increased risk of VTE include older age >75 years (particularly ≥85 years), poor ambulation (prior to surgery), obesity, and cardiovascular disease [16-20]. However, these factors are less prevalent in those undergoing minor orthopedic procedures (eg, arthroscopic procedures, foot and ankle surgery, upper extremity surgery) who are generally younger and more active. (See "Overview of the causes of venous thrombosis".)

High risk: Hip and knee arthroplasty, hip fracture surgery, pelvic and multiple fractures — In general, VTE risk is highest in those undergoing major orthopedic surgery including (see 'Total hip or knee arthroplasty and hip fracture surgery' below):

●Total hip arthroplasty (THA)

●Total knee arthroplasty (TKA)  

●Hip fracture surgery (HFS) and hip fracture when surgery is not feasible

●Pelvic fractures

●Multiple fractures from severe trauma

While in the past, rates of VTE (no prophylaxis) in patients undergoing major orthopedic surgery were reported to be as high as 30 percent, data since then have reported rates less than 5 percent [1,21,22]. The American College of Chest Physicians (ACCP) has estimated that the baseline risk for 35 days beyond surgery is 4.3 percent, with the highest risk occurring in the first 7 to 14 days (1.8 percent for symptomatic deep vein thrombosis [DVT] and 1 percent for pulmonary embolism [PE]); rates fall in the subsequent 15 to 35 days (1 and 0.5 percent for symptomatic DVT and PE, respectively) [1].  

Pathogenetic factors that may contribute to the particularly high risk in these groups include: injury to the deep veins from positioning of the extremity and compression of the femoral vein due to flexion and adduction of the hip during hip surgery; the anatomic position of the extremity and possibly the use of a thigh tourniquet during knee surgery; and performing bilateral as opposed to unilateral arthroplasty [16,23-25].

Low risk: Surgery below the knee; upper extremity surgery; arthroscopy — The risk is lower in other forms of orthopedic surgery (assuming no additional risk factors are present) including lower extremity surgery requiring immobilization or casting (eg, foot and ankle fractures, tibial osteotomy, tendon repair, hallux valgus repair); shoulder, elbow, and hand surgery [26]; and arthroscopic procedures. However, there is a paucity of data estimating the true risk in individual surgeries and the ranges are wide. Extrapolating from randomized trials, most of these patients have a <2 percent risk of VTE in the first three months following surgery; these data are discussed below. (See 'Lower extremity injury requiring immobilization' below and 'Knee arthroscopy' below and 'Others' below.)

Assess the risk of bleeding — For patients in whom pharmacologic VTE prophylaxis is warranted, a full history and examination should be obtained to assess the risk of major bleeding (ie, fatal bleeding, and/or symptomatic bleeding in a critical area or organ, and/or bleeding causing a fall in hemoglobin of ≥2 g/dL or leading to transfusion of two or more units of whole blood or red cells, or bleeding requiring reoperation [27]).

Estimates of the baseline risk of bleeding in the absence of prophylaxis in orthopedic patients have been hampered by factors including derivation of data from heterogeneous populations, increasing use of thromboprophylaxis, and improved surgical techniques. Nonetheless, the ACCP and others have estimated that the average untreated bleeding risk in nontrauma patients is <2 percent [1,28], suggesting that the influence of anticoagulant use has minimal effect on the risk of bleeding. However, estimates are likely biased by the inclusion in trials of only patients with a low risk of bleeding. Most experts consider that major shoulder, hip, knee, and some hand or foot surgeries are at higher risk of bleeding (2 to 4 percent), while minor procedures like arthroscopy are considered as having a lower bleeding risk (<2 percent) (table 3). Bleeding risk is higher for patients with multiple orthopedic fractures, and those with concomitant traumatic injuries. (See "Venous thromboembolism risk and prevention in the severely injured trauma patient".)

Patients with individual risk factors for bleeding include those with contraindications to pharmacologic prophylaxis (eg, active bleeding or intracranial hemorrhage), patients who have underlying bleeding diathesis or thrombocytopenia (eg, platelet count <50,000/microL) (table 4), or patients in whom the risk of bleeding is potentially catastrophic. Epistaxis and menstrual bleeding are not contraindications to pharmacologic thromboprophylaxis.

TOTAL HIP OR KNEE ARTHROPLASTY AND HIP FRACTURE SURGERY

Low bleeding risk: Pharmacologic prophylaxis (up to 10 to 14 days) — For patients undergoing total hip arthroplasty (THA), total knee arthroplasty (TKA), and in whom the risk of bleeding is low:

●We recommend pharmacologic prophylaxis with or without intermittent pneumatic compression devices rather than no prophylaxis. As the initial agent of choice in patients with THA and TKA, we prefer low molecular weight (LMW) heparin or the direct oral anticoagulants (DOACs); among the DOACs, we prefer rivaroxaban or apixaban, rather than dabigatran or edoxaban.

●LMW heparin, fondaparinux and many of the DOACs are renally excreted. Thus, for those with severe renal insufficiency (eg, creatinine clearance <20 to 30 mL/min), unfractionated heparin (UFH) is the preferred alternative to LMW heparin and DOACs (table 5); however, warfarin may be used if heparin injections are undesirable.

●Aspirin should not be used as the sole initial agent for VTE prophylaxis but switching to aspirin following a short course (eg, five days) of rivaroxaban may be suitable for select low risk patients.

●For patients undergoing hip fracture surgery (HFS), we use similar principles for agent selection; however, since DOACs have not been evaluated in patients with HFS, we prefer to avoid their use in this population, until their safety and efficacy are proven.

●We also suggest a similar strategy in patients who have a hip fracture in whom surgery is delayed or not feasible (eg, for a medical reason) based upon the rationale that the fracture itself and associated immobility increase the risk of VTE to a level that warrants pharmacologic or combined methods of prophylaxis.  

In patients undergoing major orthopedic surgery, LMW heparin was traditionally considered the gold standard and has been shown to be more effective than low-dose UFH or therapeutic anticoagulation with warfarin, but less effective than fondaparinux. Bleeding rates appear to be similar among these agents with the exception of fondaparinux, which may be associated with a higher incidence of hemorrhage. On balance, most trials also suggest that prophylactic doses of DOACs, particularly rivaroxaban or apixaban, have similar efficacy and safety compared with LMW heparin. Aspirin is less well supported by data as an initial agent; it is not clear if it is as effective when compared with anticoagulants in the initial high risk period postoperatively (ie, first two weeks); the efficacy of aspirin is best supported as an extended duration prophylactic agent beyond the first five days. Further trial data are forthcoming. (See 'Low molecular weight heparin' below and 'Second line options' below.)

Importantly, the clinician should be aware that VTE prophylaxis reduces but does not eliminate the risk of VTE entirely, has a greater impact on reducing the rate of asymptomatic events rather than symptomatic events, and has minimal effect on mortality. This is particularly true in patients with the highest risk (eg, multiply injured patients). (See "Venous thromboembolism risk and prevention in the severely injured trauma patient", section on 'Pathophysiology and risk factors'.)

Most of the data discussed in this section describe thromboprophylaxis for the first 10 to 14 days, while extended duration prophylaxis beyond 10 to 14 days is discussed below. (See 'Duration' below.)

The value of other anticoagulants in VTE prevention, such as osocimab (factor XIa inhibitor that may have lower bleeding rates than other agents), has been studied in pre-clinical and early phase trials. These data are discussed separately. (See "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects", section on 'Anticoagulants in development'.)

First line options

Low molecular weight heparin — In patients undergoing major orthopedic surgery, several LMW heparin regimens have been shown in randomized trials to be highly effective and associated with a low risk of major bleeding; these features together with the extensive experience with LMW heparin have traditionally made it the agent of choice for those in whom pharmacologic prophylaxis is indicated. However, LMW heparin should be avoided in patients with severe renal insufficiency (eg, creatinine clearance <20 to 30 mL/min).

LMW heparin has been compared with the following:

●Placebo – Several older randomized trials in patients with THA, TKA [29-32] and HFS [33] have reported that compared with placebo, LMW heparin consistently reduces the incidence of asymptomatic VTE by about 50 percent (11 to 30 percent versus 40 to 60 percent). The effect on symptomatic events, pulmonary embolism (PE), and bleeding was inconsistently reported.

●Direct oral anticoagulants – DOACs (direct factor Xa inhibitors [rivaroxaban, apixaban, edoxaban] and the thrombin inhibitor, dabigatran) are being increasingly used by clinicians. Data from randomized trials, systematic reviews, and meta-analyses comparing DOACs with LMW heparin report unchanged or reduced rates of symptomatic deep vein thrombosis (DVT) [34-41]. Similarly, while most analyses report no difference in rates of bleeding, a small proportion report a possible slight increased bleeding risk with DOACs. On balance, the data suggest that the safety and efficacy of DOACs for the prevention of VTE are similar to LMW heparin [36,40]. As examples:

•One 2012 meta-analysis of 22 randomized trials (32,159 patients) compared the factor Xa inhibitors (eg, rivaroxaban, apixaban, edoxaban) with LMW heparin in adults for VTE prevention following hip or knee replacement [36]. Factor Xa inhibitors were associated with a reduced risk of symptomatic DVT (4 fewer events per 1000) without an effect on nonfatal PE or death. However, high doses of factor Xa inhibitors, but not lower doses, increased the risk of bleeding more than the LMW heparins (2 more bleeding events per 1000). Major limitations of this analysis were that many of the included studies reported bleeding as a composite outcome and the outcomes were missing in 3 to 41 percent of patients. In addition, the enoxaparin dosing in most of the included studies was the lower dose of 40 mg daily, (not 30 mg twice-daily), which may have favored LMW heparin and falsely increased bleeding events in the DOAC group.

•Additional systematic reviews and meta-analyses report no difference between LMW heparin and the direct thrombin inhibitor dabigatran in the rates of VTE prevention (relative risk [RR] 0.71, 95% CI 0.23-2.12) or bleeding (RR 1.12, 95% CI 0.94-1.35) [38,40]. Similarly, another meta-analysis reported that patients on DOACs have a lower rate of VTE without any increase in the rate of bleeding [41].

Data supporting individual DOAC agents include the following:

•Rivaroxaban – Several studies in patients with THA and TKA but not HFS have reported similar or superior efficacy with LMW heparin, and while some showed a trend towards an increase in bleeding, it was not consistent or significant [42-49].

•Dabigatran – Several randomized trials (including REMODEL, RENOVATE, REMOBILIZE) and one meta-analysis of three major trials in patients with THA, TKA but not HFS in general report similar efficacy and safety of dabigatran when compared with LMW heparin [50-55]. Extended duration prophylaxis with this agent is discussed separately. (See 'Duration' below.)

•Apixaban – Several randomized trials of patients with THA or TKA report comparable rates of VTE between apixaban and LMW heparin without an increased risk of bleeding [56-59]. Extended duration prophylaxis with this agent is discussed separately. (See 'Duration' below.)

•Edoxaban – Data regarding the use of edoxaban for the prevention of VTE come from two randomized trials in Japanese patients undergoing THA or TKA. These trials (STARS E-3 [60] and STARS J-5 [61]) compared edoxaban with the LMW heparin enoxaparin [62]. A pooled analysis of the results showed that edoxaban resulted in a lower incidence of VTE (5.1 versus 10.7 percent) and a similar safety profile. Edoxaban has not yet received regulatory approval for prevention of VTE in most countries.

●Warfarin – There have been several randomized trials and meta-analyses comparing LMW heparins with therapeutic warfarin in patients undergoing THA and TKA, but not HFS [63-73]. Most trials confirm that LMW heparin reduces the VTE rate by about one-third compared with warfarin. While earlier studies reported a higher number of major bleeding events associated with LMW heparin, later studies that compared these agents as extended forms of prophylaxis did not confirm this increased risk of bleeding. (See 'Timing of initiation' below.)

●Unfractionated heparin – In patients undergoing major orthopedic surgery, several randomized trials and meta-analyses have reported that LMW heparin is more effective than low-dose UFH (typically given as 5000 units twice daily) and has similar bleeding risk [28].

Best illustrating this is a meta-analysis that included 64 trials of mixed medical and surgical populations, in whom >3000 patients had arthroplasty or HFS. LMW heparin was shown to be superior to low dose UFH with a 20 percent relative risk reduction in the rate of asymptomatic DVT without an increase in the risk of major bleeding, an effect that was consistent among the subgroups of THA, TKA, and HFS [28].

LMW heparin also has similar efficacy to higher doses of UFH but may result in lower rates of bleeding. As an example, LMW heparin when compared with UFH given as 7500 units every 12 hours resulted in similar rates of VTE (19 versus 23 percent) but a lower frequency of bleeding complications (5 versus 9 percent) [74]. A comparison of LMW heparin and 8 hourly regimen of low dose UFH has not been adequately studied in orthopedic patients. However, one meta-analysis that included surgical patients, some of whom were orthopedic, showed no difference in the rate of VTE between 8 hourly and 12 hourly regimens of UFH [75]. (See "Prevention of venous thromboembolic disease in adult nonorthopedic surgical patients", section on 'Low-dose unfractionated heparin'.)

●Aspirin – Most studies testing aspirin suggest that it is less effective than LMW heparin [76-79]. Trials have been criticized for using high doses of aspirin resulting in adverse effects while others were potentially flawed. In a meta-analysis of six low quality randomized trials, there was a nonsignificant trend favoring anticoagulation with enoxaparin following hip fracture repair (RR: 1.6) but not knee or hip arthroplasty [79]. The risk of bleeding was lower with aspirin than anticoagulants. Although aspirin and LMW heparin were reported in one extended duration trial to have similar efficacy, this trial was likely flawed since most patients received 10 days of LMW heparin before being randomized to aspirin or LMW heparin; this trial is discussed below. (See 'Duration' below.)

●Fondaparinux – Randomized trials have reported that LMW heparin is less effective at preventing VTE compared with fondaparinux [1,80-86] but LMW heparin remains the drug of preference since fondaparinux is associated with more bleeding events.  

Best illustrating this is a 2016 meta-analysis of 25 studies that included 21,000 mixed surgical and medical patients, most of whom were undergoing major orthopedic surgery, in which pentasaccharides, mostly fondaparinux, were more effective for short-term VTE prevention compared with LMW heparin. Fondaparinux reduced the risk of VTE (RR 0.24, 95% CI 0.15-0.38) [83]. However, fondaparinux also resulted in an increased risk of major bleeding when compared with both placebo and LMW heparin (RR 2.56), although bleeding excess may have been due to initiation of fondaparinux too close to surgery.

●Mechanical methods – LMW heparin has been poorly studied in comparison with mechanical methods (intermittent pneumatic compression [IPC], venous foot pump [VFP] devices, graduated compression stockings [GCS]) [87-91]. As examples:

•In a randomized trial of 1761 patients undergoing TKA, a lower rate of DVT and mortality were reported in those who were treated for seven days with LMW heparin compared with GCS alone (0.9 versus 3.2 percent), although many of the events prevented by LMW heparin were episodes of asymptomatic distal DVT [87].  

•In another randomized trial of 274 patients undergoing THA, patients receiving LMW heparin compared with use of a VFP had a nonsignificant reduction in the overall rate of venographic VTE (13 versus 18 percent) and proximal DVT (9 versus 13 percent) [90]. There were no major bleeding complications but many patients (>20 percent) were intolerant of the VFP.

Direct oral anticoagulants — Several orally active agents that inhibit either factor Xa (rivaroxaban, apixaban) or thrombin (dabigatran) have been approved for VTE prophylaxis by regulatory agencies in the United States and several other countries [92,93]. DOACs, particularly rivaroxaban, are being increasingly used by clinicians as an alternative to LMW heparin. DOACs may also be an option in patients with a history of heparin-induced thrombocytopenia (HIT), for those unwilling to receive injections or who wish to avoid monitoring associated warfarin. They should be avoided in patients with severe renal insufficiency.

Numerous randomized trials and meta-analyses of patients with THA and TKA have compared the efficacy and safety of DOACs with the following:

●LMW heparin [34-40,94,95] (see 'Low molecular weight heparin' above).

●Aspirin (only in extended duration trials) (see 'Duration' below).

●Fondaparinux (see 'Fondaparinux' below)

DOACs have not been compared with warfarin or UFH nor have they been tested in patients with HFS.

Indirect comparisons of the DOACs with each other suggested that rivaroxaban may be more effective for preventing symptomatic DVT (RR 0.50, 95% CI 0.37-0.68) compared with dabigatran or apixaban but was associated with excess bleeding risk (RR 1.14; 95% CI 0.80-1.64) [39,40]. A second indirect comparison reported no difference among them but a slightly lower bleeding risk with apixaban and dabigatran compared with other DOACs [96].

General data regarding DOAC dosing, dosing in renal insufficiency, and adverse effects are discussed separately (see "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects"). The use of DOACs in patients with a history of HIT is discussed separately. (See "Management of heparin-induced thrombocytopenia", section on 'Direct oral anticoagulants'.)

Aspirin — Aspirin has been studied in the setting of extended prophylaxis [97-99], as a single agent or following a short course of anticoagulation. We prefer the use of aspirin be limited to the latter. Further details are provided below. (See 'Duration' below.)

Second line options

Low-dose unfractionated heparin — Low-dose UFH is an alternative to LMW heparin. One meta-analysis of older randomized trials (almost 7000 medical and surgical patients) reported that UFH resulted in a 58 percent risk reduction in the rate of asymptomatic DVT as well as a significant reduction in PE compared with no prophylaxis [28]; subgroup analysis showed similar efficacy in those undergoing THA and HFS. There was trend toward an increase in major bleeding but no effect on mortality. Trials that have shown inferiority of UFH to LMW heparin are described above. (See 'Low molecular weight heparin' above.)

Fondaparinux — Fondaparinux is an alternative to LMW heparin and UFH when these agents are not available or in patients with heparin-induced thrombocytopenia. Several randomized trials and meta-analyses have reported that although fondaparinux is superior to LMW heparin in patients undergoing major orthopedic surgery, it is associated with an increased risk of bleeding. These trials are discussed above (see 'Low molecular weight heparin' above). Extended duration thromboprophylaxis with fondaparinux is discussed below. (See 'Duration' below.)

Fondaparinux has also been compared with DOACs. Retrospective studies report that when compared with fondaparinux, DOACS (rivaroxaban, edoxaban) resulted in lower rates of VTE and either a similar or improved safety profile [100,101].  

Warfarin — Therapeutic anticoagulation with warfarin (usually for four weeks) is an alternative to LMW heparin or low-dose UFH (eg, patients in whom injections are not feasible or are undesirable).

Warfarin has been shown in older trials to reduce the incidence of asymptomatic VTE in those undergoing major hip surgery, particularly HFS, with a relative risk reduction of 55 percent when compared with no prophylaxis [28,102,103]. Major randomized trials and meta-analyses report superiority of LMW heparin compared with warfarin, the details of which are discussed above. (See 'Low molecular weight heparin' above.)

Whether low-intensity regimens are of similar efficacy to standard-intensity regimens is unclear. As an example, one randomized noninferiority trial of older patients (mean age 72 years) undergoing THA or TKA compared a low-intensity regimen of warfarin (target international normalized ratio [INR] 1.8) with a standard-intensity regimen (target INR 2.5) (3 percent noninferiority margin) [104]. Treatment was, on average, for one month, and the target INR was reached in approximately half of each group. Although higher rates of VTE and lower major bleeding were reported with a low-intensity regimen, the difference was not statistically significant (3.8 versus 5.1 percent for VTE; 0.4 versus 0.9 percent for major bleeding). However, the rate of VTE was lower than anticipated, thereby limiting the power of the study. In addition, the study may have been biased due to its open-label design and may not be generalizable to younger patients or minorities, who were under-represented in the trial. Further study is needed to determine whether lower-dose regimens of warfarin are as effective and as safe as standard-dose regimens in patients undergoing THA or TKA and until such data are available, dosing of warfarin should continue to target an INR of 2.5.

In an older study, warfarin was superior to aspirin (20 versus 41 percent) for the prevention of DVT following surgery for hip fracture [103].

Small trials have also compared warfarin with IPC devices with mixed results [105-109]. In a randomized trial of 232 patients undergoing THA, clinically significant (ie, proximal) thrombosis was more common with IPC use than warfarin (12 versus 3 percent).

Genotype-guided dosing does not offer any advantage over routine dosing. (See "Warfarin and other VKAs: Dosing and adverse effects".)

Combined — While many experts combine mechanical and pharmacologic methods in high risk orthopedic patients, few studies have examined the benefits associated with this strategy. In a post-hoc analysis of a randomized trial involving patients undergoing TKA that compared edoxaban and enoxaparin, the overall incidence of VTE was lower in patients who received concomitant GCS compared with those who were anticoagulated without GCS (6 versus 13 percent), but the difference was not statistically significant [110]. In another meta-analysis of 11 studies that included over 7000 surgical patients at high risk of VTE, most of whom underwent major orthopedic surgery, compared with IPC alone, combined prophylactic modalities decreased the incidence of symptomatic PE from about 3 percent to 1 percent and DVT from about 4 percent to 1 percent [111].

High bleeding risk: Mechanical methods — Options for mechanical methods of thromboprophylaxis include IPC, GCS (also known as elastic stockings), and the VFP. Among the available devices, IPC is typically preferred since more robust data is available in orthopedic patients using these devices [1]. For patients with contraindications to pharmacologic prophylaxis (eg, active bleeding, intracranial hemorrhage, bleeding diathesis (table 4)) or patients at high risk of bleeding (eg, trauma, coagulopathy), we suggest mechanical methods, preferably with IPC. Vena cava filters should not be routinely used.

This approach is based upon data which suggest that mechanical devices reduce the risk of asymptomatic DVT by >50 percent compared with placebo (see 'Intermittent pneumatic compression and venous foot pump' below and 'Graduated compression stockings' below). However, mechanical methods are typically less effective than LMW heparin or warfarin but have a lower bleeding risk (see 'Low molecular weight heparin' above and 'Warfarin' above). Thus, switching to or adding a pharmacologic agent, such as LMW heparin, should be done as soon as hemostasis is assessed as adequate, bleeding risk becomes acceptably low, and/or the bleeding diathesis has been reversed. Mechanical methods are also frequently used in combination with pharmacologic methods, although there are limited data to support this practice. (See 'Combined' above.)

One advantage is that mechanical devices can be applied to the contralateral leg to prevent DVT in the non-operated leg (up to 20 percent of DVT occur in the non-operated leg) and may theoretically prevent DVT in the operated leg. Their major disadvantage is poor compliance with one study reporting that properly functioning devices were used in <50 percent of cases [112,113]. However, portable battery powered devices may offer improved compliance [114]. Other adverse effects including skin breakdown and contraindications to their use are discussed separately. (See "Prevention of venous thromboembolic disease in adult nonorthopedic surgical patients", section on 'Intermittent pneumatic compression and venous foot pump'.)

Intermittent pneumatic compression and venous foot pump — Data supporting the use of IPC for the prevention of VTE in orthopedic surgical patients are limited and subject to bias (eg, unblinding, unsealed randomization, small numbers).

Several small randomized trials of patients undergoing THA, TKA, or HFS compared IPC with no prophylaxis and, in general, demonstrate a reduction in the rate of DVT [115-118]. As an example, in a randomized trial of 310 THA patients, IPC use reduced the rate of venographic VTE from 49 percent to 24 percent and proximal DVT from 27 percent to 14 percent [115]. Noteworthy, is that the residual rates were still high despite prophylaxis. Studies that suggest inferiority with LMW heparin or warfarin are discussed above. (See 'Low molecular weight heparin' above and 'Warfarin' above.)

Graduated compression stockings — Evidence to support the use of GCS is derived from indirect data in other surgical and medical patients; outcome associated with GCS use in those populations is mixed and associated with an increase in the risk of local skin complications. Only one small trial in orthopedic patients has been performed showing a similar lack of benefit [119]. Nonetheless, they are frequently used in combination with pharmacologic methods in high risk patients. (See "Prevention and treatment of venous thromboembolism in patients with acute stroke", section on 'Ineffective or unproven treatments'.)

Administration

Timing of initiation — The optimal timing of pharmacologic thromboprophylaxis in patients undergoing major orthopedic surgery is unknown and practice is variable among surgeons. Timing also depends upon the methods chosen:

●Low molecular weight and unfractionated heparin – Most experts agree that the administration of pharmacologic prophylaxis with LMW heparin should not be administered close to surgery (eg, within four hours preoperatively and within four hours postoperatively). Many prefer to administer thromboprophylaxis 12 hours or more preoperatively and/or 12 hours or more postoperatively. This strategy is based upon trials that have suggested an unacceptable increase in the bleeding risk when pharmacologic agents are given perioperatively (ie, within four hours of surgery).

Randomized trials have shown that although perioperative thromboprophylaxis with LMW heparin results in enhanced efficacy, bleeding rates are higher. One randomized trial of patients undergoing THA compared preoperative (12 hours or more), postoperative (12 to 24 hours), and perioperative (within two hours before surgery and within four hours after surgery) administration of LMW heparin (dalteparin) with postoperative warfarin (begun on day 1). Although perioperative dalteparin resulted in improved prevention of VTE, major bleeding rates were 5 to 7 percent, compared with preoperative and postoperative rates of bleeding with warfarin which were about 1 to 3 percent [70]. A systematic review of four randomized trials reported similar rates of efficacy when LMW heparin was initiated postoperatively in close proximity to surgery at half the usual dose that was not associated with an increase in major bleeding rates [120].

While there are no data, we use the same strategy for those treated with low dose UFH.

●Fondaparinux – Fondaparinux is approved to start six or more hours after skin closure consistent with major trials that demonstrated its efficacy. However, many experts administer the first dose 8 to 12 hours postoperatively to mitigate the bleeding risk, a practice that is also consistent with guidelines [1]. (See 'Fondaparinux' above.)

●Mechanical methods – For those in whom mechanical methods (IPC, GCS, VFP) are indicated, devices are typically placed on the patient just prior to the start of surgery and used continuously postoperatively until hospital discharge or ambulation. When mechanical methods are used in patients at high risk of bleeding, pharmacologic agents are started or added postoperatively, as soon as hemostasis is achieved and it is considered safe (eg, 12 to 72 hours).

●Oral agents – Although poorly studied, oral agents, warfarin, aspirin, and DOACs, are generally begun postoperatively 8 to 12 hours or more after surgery, provided the patient can eat. However, specific recommendations for each of the DOACs vary and are listed elsewhere (see 'Dosing' below). If oral intake is expected to be delayed, then LMW heparin (or UFH) should be administered subcutaneously in the interim. Although timing is poorly studied with oral agents, we base this practice upon major trials that demonstrate efficacy with this approach. (See 'Warfarin' above and 'Direct oral anticoagulants' above and 'Aspirin' above.)

Duration — In patients with THA, TKA, or HFS, we recommend administration of pharmacologic prophylaxis for a minimum of 10 to 14 days and we suggest that it be continued for up to 35 days after surgery; however, most clinicians prefer courses within the lower end of that range (eg, 10 to 14 days) in those undergoing TKA with longer courses in the upper end of that range (eg, 30 days) in those undergoing THA. Pharmacologic prophylaxis may be discontinued when patients become fully ambulatory or are discharged to home, provided a 10 to 14 day course has already been administered.

This approach is based upon meta-analyses and randomized trials of pharmacologic agents in those undergoing major orthopedic surgery that report a significant reduction in the rate of VTE without a significant increase in the bleeding rate for the first two weeks, the period with the highest risk of VTE [73,82,121-133]. Some of these data have also shown that compared with placebo, extended prophylaxis reduces the rate of VTE without an excess of major bleeding events; however, it may be associated with an increase in minor bleeding. The duration of extended prophylaxis following orthopedic surgery has varied between studies; data are strongest for those undergoing THA, with limited data in patients undergoing TKA or HFS. However, in one observational study from Denmark the VTE rate was 0.4 percent in over 17,000 fast-track THA or TKA surgeries with in-hospital only thromboprophylaxis when the length of stay was ≤5 days [134]. This strategy requires further study before it can be adopted as routine.

Agent selection — Among the agents, LMW heparin was traditionally the preferred agent based upon experience and trials that have demonstrated efficacy [122,125,127,129,132]. However, DOACs and strategies that switch to aspirin after a short course of anticoagulation are being increasingly used with one well-conducted randomized trial which has reported that a short course of rivaroxaban for five days followed by an extended course of aspirin (9 to 30 days for TKA and THA, respectively) is as effective as an extended course of low dose rivaroxaban [133]. It is likely that this trial is practice-changing for selected, low-risk patients undergoing TKA and THA. Although further study will shed light on this combination in practice, based upon this study, we suggest that rivaroxaban followed by aspirin is an alternative to LMW heparin for select patients undergoing THA and TKA who are at low risk for VTE. Further details regarding this trial are discussed in the aspirin bullet below.

Agents that have been shown to have efficacy in the setting of extended prophylaxis include:

●LMW heparin – Several meta-analyses and randomized trials have shown efficacy of LMW heparin:

•A 2012 meta-analysis of six trials (mostly patients undergoing THA), reported that compared with 7 to 10 days of thromboprophylaxis, extended duration prophylaxis (mostly LMW heparin) resulted in fewer cases of PE (odds ratio [OR], 0.14) and symptomatic DVT (OR, 0.36) [129]. Although there were more minor bleeding events with prolonged prophylaxis (OR, 2.44), there was no increase in major bleeding events.

•In a 2016 meta-analysis of 16 randomized trials (mostly undergoing THA) extended-duration treatment with heparin (LMW heparin or UFH), warfarin, or DOACs resulted in reduction in symptomatic VTE (OR, 0.59 [heparin]; 0.1 [warfarin]; 0.2 [DOACs]), when compared with standard-duration (7 to 14 days) thromboprophylaxis. In general, there was no increased risk of major bleeding, although heparin was associated with increased risk of minor bleeding (OR, 2.0) [132].

•In a randomized trial of 873 patients who underwent THA or TKA, in the 438 patients who underwent elective total knee replacement, compared with placebo, LMW heparin (enoxaparin), did not result in significant benefit (VTE rates 18 versus 20 percent) [125]. In contrast, in those who underwent THA, the rate of VTE was significantly reduced by the administration of enoxaparin (23 versus 8 percent).

●Direct oral anticoagulants – Major randomized trials of rivaroxaban (RECORD2) [43], dabigatran (RENOVATE) [54], and apixaban [59] all reported reductions in the rates of a composite endpoint of total VTE and all-cause mortality without an increase in the risk of major bleeding in patients with THA who received extended thromboprophylaxis for 35 days when compared with conventional 10 to 14 days of prophylaxis. Extended duration rivaroxaban compared with aspirin (EPCAT II) is discussed below [133].

●Aspirin – Several studies have examined the efficacy of aspirin in an extended duration setting [78,133,135-141].

•In one trial (EPCAT), 778 patients who had undergone THA and had received an initial 10 days of LMW heparin (dalteparin), were randomly assigned to either extended prophylaxis with LMW heparin or aspirin [135]. Compared with LMW heparin, aspirin was associated with a similar rate of VTE (1.3 for aspirin versus 0.3 percent for LMW heparin) and bleeding events (0.5 versus 1.3 percent) at 28 days. However, this trial was stopped early due to slow enrollment and the effect size was based on a small number of events, limiting the quality of the data.

•In another trial (EPCAT II) of 3424 patients with THA or TKA, following an initial five days of rivaroxaban, aspirin (81 mg) administered for an additional nine days (TKA) or 30 days (THA) resulted in similar rates of VTE (<1 percent) and major bleeding (<1 percent) when compared with rivaroxaban (10 mg) administered for the same period [133]. Patients included in this trial were those at low risk of VTE and included patients undergoing elective unilateral THA/TKA who were ambulatory within 24 hours after surgery and who did not have additional risk factors for VTE (eg, prior VTE, active cancer, known hereditary or acquired thrombophilia) as well as patients without indications for long term anticoagulation, and without lower limb or hip fracture in the previous three months or expected major surgery in the oncoming three months.

•In a randomized trial of almost 18,000 patients, most of whom had HFS, compared with placebo, aspirin administered for 35 days reduced the incidence of symptomatic DVT without an increase in major bleeding events but had no effect on the rate of clinically significant PE [98].

•In a randomized trial of 275 patients undergoing TKA, when used in combination with pneumatic compression devices, four weeks of aspirin had similar rates of DVT when compared with LMW heparin (18 versus 14 percent) [78]. In another trial in patients with THA, IPC devices in combination with six weeks of aspirin did not result in a significant reduction in rates of VTE compared with aspirin alone [136].

•In a review of the Michigan Arthroplasty registry of over 41,000 patients undergoing TKA, the overall incidence of VTE was 1.38 percent [142]. The incidence was 4.79 percent among those who received no pharmacologic prophylaxis, 1.42 percent for those treated with anticoagulation alone (various agents and combinations), 1.31 percent for those prescribed both anticoagulation and aspirin, and 1.16 percent for those treated with aspirin alone. Bleeding occurred in 1.10 percent of patients overall, and in 1.50, 1.35, 1.14 and 0.90 percent of the no prophylaxis, anticoagulation and aspirin, anticoagulation, and aspirin groups, respectively. Aspirin alone was noninferior for both the composite VTE outcome (unadjusted analysis) and for bleeding complications (unadjusted and adjusted analysis) compared with other nonaspirin treatment. Sensitivity analysis in which patients with a history of VTE were excluded revealed the same composite VTE outcome. The association of individual nonaspirin anticoagulation agents (LMWH, warfarin, direct factor Xa inhibitors, direct thrombin inhibitors, synthetic pentasaccharides) with VTE was not studied. This study addressed a number of potential confounders, including history of VTE, however the authors acknowledged that important unmeasured variables (such as surgeons’ individual practice and determination of patient risk, postoperative rehabilitation protocols) could not be captured. Caution is needed when applying the results of this study to patients with additional risk factors for VTE (eg, prior VTE, active cancer, known hereditary or acquired thrombophilia, delayed mobilization).

•In a meta-analysis if 13 randomized trials, there was no difference between aspirin and other anticoagulants (relative risk 1.12; 95% CI 0.78-1.62) [143]. However, low quality of evidence was included in the analysis and the wide confidence limits suggest a variable response in the efficacy of aspirin.

●Warfarin – One randomized trial of 1279 patients who had undergone THA showed that compared to LMW heparin, six weeks of therapeutic warfarin resulted in a similar rate of VTE (2 versus 3 percent) but at the expense of a higher bleeding rate (6 versus 1 percent) [73].

●Fondaparinux – In a randomized trial of 656 patients undergoing HFS, compared with one week of fondaparinux, one month of fondaparinux reduced the incidence of symptomatic VTE (0.3 versus 2.7 percent) [144]. Although there was a trend towards an increase in major bleeding with fondaparinux, it was not significant.

Dosing — Assuming no extremes in body weight or renal insufficiency, the following dosing regimen is appropriate:

●LMW heparin – Dosing of LMW heparin is based upon a regimen that administers the initial dose ≥12 hours pre- and/or postoperatively (see 'Timing of initiation' above). These regimens differ in detail from product label recommendations but are generally consistent with 2012 American College of Chest Physicians guidelines on antithrombotic therapy and prevention of thrombosis [1]. Additional dosing options including alternatives found in the labeling are listed separately in the drug information monographs included with UpToDate.

Suggested dosing is the following:

•Enoxaparin [145]

-THA: 30 mg subcutaneously every 12 hours OR 40 mg once daily started either ≥12 hours before or ≥12 hours after surgery.

-TKA: 30 mg subcutaneously every 12 hours started either ≥12 hours before or ≥12 hours after surgery.

-HFS: Dosing is typically similar to that in patients undergoing THA.

•Dalteparin [146,147] – THA and TKA (off-label): 5000 units subcutaneously once daily, started either ≥12 hours before or ≥12 hours after surgery. Although not preferred, if the initial dose is administered close to surgery (ie, four to eight hours after THA surgery), the United States product labeling recommends using half the initial dose (2500 units) followed by the usual maintenance dose, thereafter.

•Tinzaparin [148,149] – THA and TKA: 4500 units subcutaneously once daily started either ≥12 hours before or ≥12 hours after surgery. If a weight-adjusted dosing method is warranted, 50 or 75 units/kg can be given once daily; refer to detail in the drug information monograph. Tinzaparin is not available in the United States.

•Nadroparin [150] – THA and TKA: 38 units/kg subcutaneously once daily (maximum: 3800 units) starting either ≥12 hours before or ≥12 hours after surgery; on postoperative day 4 increase the dose to 57 units/kg once daily (maximum: 5700 units). Nadroparin is not available in the United States.

LMW heparins are cleared renally to varying degrees and dose adjustment is advised for patients with renal insufficiency (table 5) and in patients who are obese (table 6). (See "Prevention of venous thromboembolic disease in adult nonorthopedic surgical patients", section on 'Pharmacologic dosing' and "Bariatric operations: Perioperative morbidity and mortality", section on 'Venous thromboembolism'.)

●UFH – 5000 units subcutaneously twice daily (less commonly three times daily). In obese patients the optimal dose is unknown and some experts use 7500 units twice daily, but there are no data to support this regimen [151,152]. No renal adjustment is necessary.

●Fondaparinux – 2.5 mg subcutaneously once daily beginning 8 to ≥12 hours after surgery; fondaparinux is contraindicated in patients who weigh <50 kg and avoided in those with renal insufficiency [153].

●Warfarin – Typical dosing regimens for warfarin start at 5 mg orally once daily; occasionally a lower or higher starting dose is warranted. Warfarin is started 12 to 24 hours after surgery (provided the patient can eat) or the evening before surgery. While some experts adjust thereafter for a therapeutic target international normalized ratio (INR) of 1.5 to 2.5 [154-156], most experts adjusted to target an INR of 2.5 (range 2 to 3). We suggest the use of a validated warfarin dose adjustment nomogram. (See "Warfarin and other VKAs: Dosing and adverse effects", section on 'Initial dosing'.)

●Aspirin – While in the past doses of oral aspirin as high as 500 mg three times a day were used, the adverse effects were unacceptably high such that lower doses are now used, typically 81 mg once daily [135] or, less commonly, 160 mg once daily [98].

●Direct oral anticoagulants – Suggested dosing for rivaroxaban (10 mg once daily started 6 to ≥10 hours after surgery), dabigatran (initially 110 mg given one to four hours after surgery and thereafter 220 mg once daily), and apixaban (2.5 mg twice daily starting ≥12 hours after surgery), are discussed in detail separately. (See "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects".)

LOWER EXTREMITY INJURY REQUIRING IMMOBILIZATION — This is a heterogeneous group of patients, often young, that includes fractures below the knee, tendon ruptures, cartilage injuries of the knee and ankle, foot and ankle surgery. In addition to typical procedure- and patient-related issues, factors that can contribute to the increased risk of VTE in this population include degree of immobilization and casting, proximity to the knee (surgery closer to the knee is at higher risk), and type of surgery (eg, Achilles tendon rupture has a higher risk) [157]. Thromboprophylaxis in patients with severe lower extremity injury (multiple fractures, concomitant vascular injury) are discussed separately. (See "Venous thromboembolism risk and prevention in the severely injured trauma patient", section on 'Extremity injury' and "Surgical management of severe lower extremity injury", section on 'VTE prophylaxis' and "Lower extremity amputation", section on 'Thromboprophylaxis'.)

For most patients with isolated lower extremity injury requiring immobilization, we suggest no thromboprophylaxis, other than early ambulation when feasible, rather than pharmacologic or mechanical methods. This approach is based upon the reported low risk of VTE for most patients in this population and evidence that suggests, on balance, no benefit from low molecular weight (LMW) heparin. However, given the wide range of risk among this population (up to 40 percent), there should be a low threshold for administering pharmacologic prophylaxis particularly in those with additional risk factors for VTE (eg, previous history of VTE [158], patient not fully mobile) or undergoing surgeries known to be high risk (eg, Achilles tendon rupture, femoral fracture, tibial plateau fracture). When prophylaxis is chosen, although we have a preference for LMW heparin, the optimal method is unknown and should be individualized. For example, aspirin is frequently administered in this population, particularly in individuals with low thrombotic risk who are not yet weight bearing. Clinicians should recognize that mechanical methods may only be applied to the unoperated leg (thereby limiting their efficacy). The duration of thromboprophylaxis is generally for the period of immobilization.

In patients with lower extremity injury requiring immobilization, several randomized trials and meta-analyses of patients are conflicting regarding benefit from LMW heparin [159-164]. However, although fraught with several methodologic limitations, these studies did consistently demonstrate that the risk of VTE remains wide, suggesting that some patients, particularly those deemed at high risk (eg, prior history of VTE, Achilles tendon rupture [162]), are likely to benefit from pharmacologic prophylaxis.

●In an open label randomized trial (POT-CAST) of 1519 patients undergoing casting of the lower leg (requiring surgical or conservative management), the rate of VTE was similar among those receiving LMW heparin for the full period of immobilization compared with those not receiving anticoagulant (1.4 versus 1.8 percent) [163]. No major bleeding events occurred in either group. However, patients with a higher risk of VTE (eg, previous VTE) were excluded from this study. Similarly, several randomized trials and a meta-analysis of 22 studies (that did not include POT-CAST) report no differences in the rate of radiologic or clinically significant DVT among patients who received pharmacologic prophylaxis compared with untreated patients [157,162,164].

●In contrast, a meta-analysis of eight randomized trials (almost 3680 patients; that included patients from POT-CAST) reported a lower incidence of VTE using LMW heparin compared with placebo in patients with lower leg immobilization (odds ratio, 0.45) [165]. However, the rates of VTE were wide-ranging (0 to 37 percent [LMW heparin)] versus 4 to 40 percent [placebo]). Results were consistent across subcategories: operated patients, conservatively treated patients, patients with fractures, patients with soft-tissue injuries, patients with proximal thrombosis, patients with distal thrombosis, and patients with below-knee casts. Major bleeding events were rare (0.3 percent). A subsequent meta-analysis of six studies found similar results [166].

●In another randomized trial of 3604 patients who were immobilized after non-major lower limb orthopedic surgery (eg, knee ligament repair, ankle/Achilles tendon and foot surgery), rivaroxaban (10 mg orally daily) was associated with a reduction in symptomatic and asymptomatic VTE compared with patients treated with enoxaparin (40 mg subcutaneously once daily; 0.2 versus 1.1 percent; 0.2 versus 0.6 for symptomatic VTE) [167]. There was no difference in the rates of major bleeding between the groups (approximately 1 percent). However, interpretation of these results should consider that more than two-thirds of VTE events were asymptomatic proximal DVT or symptomatic distal DVT, which may not be clinically important. In addition, the study was stopped early due to poor enrollment, and the overall event rate was low.

KNEE ARTHROSCOPY — Thromboprophylaxis is controversial in patients undergoing arthroscopy and arthroscopic-assisted procedures of the knee, many of which are outpatient-based and performed in young healthy individuals [87,163,168-173]. On balance, data do not support routine anticoagulant prophylaxis with low molecular weight (LMW) heparin in patients undergoing arthroscopy. However, individualization of the approach is needed, particularly in those with concomitant individual risk factors for VTE (eg, previous history VTE). For example, aspirin is frequently administered in this population, particularly in individuals with low thrombotic risk who are not yet weight bearing.

Best supporting this practice is one randomized trial of 1543 patients (POT-KAST) undergoing arthroscopy, which reported the rate of VTE was unchanged among those treated with LMW heparin (for eight days) compared with those not receiving prophylactic anticoagulation (0.7 versus 0.4 percent) [163]. There were no major bleeding events in either group (0.1 percent each).

The Efficacy of Rivaroxaban for thromboprophylaxis after Knee Arthroscopy (ERIKA) trial randomly assigned 241 patients to rivaroxaban or placebo for seven days after knee arthroscopy [169]. The primary outcome (composite of all-cause death, symptomatic thromboembolism, asymptomatic proximal DVT at three months) occurred in significantly fewer patients in the rivaroxaban group compared with placebo (1 of 120 [0.8 percent] versus 7 of 114 [6.1 percent]; absolute risk difference, -5.3 percent, 95% CI, -11.4 to -0.8). No major bleeding was observed. The small sample size, high exclusion rate, and low number of reconstruction procedures limited this study.

Other trials reported possible benefit from pharmacologic thromboprophylaxis but at the expense of an increased risk of minor bleeding [87,168] . However, the interpretation of these data is limited since the calculated risk was based upon a low number of events and most of the events that were prevented were clinically insignificant (eg, asymptomatic distal deep vein thrombosis).

OTHERS — The administration of thromboprophylaxis in patients undergoing orthopedic procedures other than arthroplasty, arthroscopy, or lower extremity injuries, should be individualized given the wide range of possible risk. For example patients assessed as having a high level of risk should probably receive pharmacologic prophylaxis, preferably with low molecular weight (LMW) heparin, when feasible. Examples of patients in this category include those with pelvic fractures and those with multiple sites of orthopedic trauma, or spinal fractures. However, these patients often have a high risk of bleeding, and alternative strategies are needed. (See "Venous thromboembolism risk and prevention in the severely injured trauma patient".)

In contrast, those undergoing fracture repair of the upper extremity are at lower risk, are often ambulatory, and may be candidates for no prophylaxis or for aspirin.

Efficacy of thromboprophylaxis in orthopedic surgeries other than major orthopedic surgery, lower extremity injury, or arthroscopy are poorly studied. One study of patients undergoing tibial osteotomy, reported no benefit from the direct oral anticoagulant, edoxaban, compared with mechanical methods [174].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Superficial vein thrombosis, deep vein thrombosis, and pulmonary embolism" and "Society guideline links: Anticoagulation".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●The Basics topics (See "Patient education: Deep vein thrombosis (blood clots in the legs) (The Basics)" and "Patient education: Pulmonary embolism (blood clot in the lungs) (The Basics)" and "Patient education: Choosing a medicine for blood clots (The Basics)" and "Patient education: Taking medicines for blood clots (The Basics)".)

●Beyond the Basics topics (see "Patient education: Deep vein thrombosis (DVT) (Beyond the Basics)" and "Patient education: Pulmonary embolism (Beyond the Basics)" and "Patient education: Warfarin (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●The risk of postoperative venous thromboembolism (VTE; deep venous thrombosis and pulmonary embolism) in orthopedic patients, particularly those undergoing major hip and knee surgery, is among the highest of all surgical specialties. (See 'Introduction' above.)

●The risk of postoperative VTE and bleeding in orthopedic patients depends upon both procedure- and patient-related risk factors, all of which need to be assessed prior to surgery. Most patients undergoing major orthopedic surgery (hip and knee arthroplasty and hip fracture repair) are considered high risk for VTE while those undergoing minor orthopedic surgery (eg, arthroscopy) are considered low risk. In general, major shoulder, hip, and knee surgeries, and some foot and hand surgeries are considered at high risk of bleeding while minor procedures like arthroscopy are considered as having a lower bleeding risk. (See 'Risk of thrombosis and bleeding' above.)

●For initial VTE prophylaxis in patients undergoing total hip arthroplasty (THA) and total knee arthroplasty (TKA), in whom the risk of bleeding is low, our approach is the following (see 'Low bleeding risk: Pharmacologic prophylaxis (up to 10 to 14 days)' above):

•We recommend pharmacologic prophylaxis with or without intermittent pneumatic compression (IPC) devices rather than no prophylaxis (Grade 1B). While many experts combine mechanical and pharmacologic methods in high risk orthopedic patients, few studies have examined the benefits associated with this strategy.  

•For most patients with THA and TKA we suggest low molecular weight (LMW) heparin or a direct oral anticoagulant (DOAC), rather than warfarin (Grade 2B). When a DOAC is selected, we prefer rivaroxaban or apixaban (rather than edoxaban or dabigatran) since experience is greater with these agents and evidence, on balance, suggests that, at minimum, they have equal efficacy with LMW heparin without any significant increase in the risk of bleeding.

•For those in whom LMW heparin or DOACs cannot be administered, options include fondaparinux, low dose unfractionated heparin (UFH]), and warfarin. Since LMW heparin, fondaparinux, and many of the DOACs are generally avoided in those with severe renal insufficiency (eg, creatinine clearance <20 to 30 mL/min), we prefer UFH in this population; alternatively, warfarin may be used if heparin injections are undesirable.

•Aspirin should not be used as a standalone agent for initial VTE prophylaxis but may be part of a hybrid strategy that follows a short course of anticoagulation in select low risk populations.

•In patients with hip fracture surgery (HFS), we use similar principles for agent selection but prefer to avoid DOACs until their safety and efficacy are proven in this population.

●For patients undergoing THA, TKA, or HFS with contraindications to pharmacologic prophylaxis (table 4) or patients at high risk of bleeding (eg, severe trauma, coagulopathy), we prefer mechanical methods rather than no prophylaxis. IPC devices are preferred since more robust data are available for these devices in this population. Inferior vena cava filters should not be routinely used. Switching to or adding a pharmacologic agent, should be done as soon as hemostasis is assessed as adequate, bleeding risk becomes acceptably low, and/or the bleeding diathesis has been reversed. (See 'High bleeding risk: Mechanical methods' above.)

●For patients undergoing THA, TKA, or HFS in whom LMW heparin is indicated, we prefer administration at least 12 hours or more preoperatively and/or 12 hours or more postoperatively. This strategy is based upon evidence that suggests an unacceptable increase in the bleeding risk when pharmacologic agents are given perioperatively (ie, within four hours of surgery). Fondaparinux is typically started six or more hours after skin closure and oral agents are generally administered 4 to 12 hours after surgery but vary with individual agents. Mechanical devices are typically placed on the patient just prior to the start of surgery. (See 'Timing of initiation' above.)

●For patients undergoing THA, TKA, or HFS, we recommend that pharmacologic prophylaxis is administered for a minimum of 10 to 14 days (Grade 1B). (See 'Duration' above.)

•For those undergoing THA, we suggest that pharmacologic prophylaxis is continued for up to 35 days after surgery (Grade 2B).

•For those undergoing TKA, shorter courses at the 10 to 14 day end of the spectrum may be preferred.

•For low risk patients receiving prophylaxis with rivaroxaban, we suggest a hybrid strategy that involves switching to daily aspirin (81 mg) at day 5 (Grade 2B) and continuing aspirin for the remaining duration of prophylaxis. This strategy appears to have similar efficacy and bleeding risk compared with continuation of rivaroxaban for the full duration or prophylaxis. Low risk patients are considered patients with any of the following: elective unilateral THA/TKA who ambulate within 24 hours after surgery and do not have additional risk factors for VTE (eg, prior VTE, active cancer, known hereditary or acquired thrombophilia), indications for long term anticoagulation, lower limb or hip fracture in the previous three months, or expected major surgery in the oncoming three months. Other patients who are not considered low risk should be continued on the initial agent of choice.  

•In patients with hip fracture surgery (HFS), we use similar principles for duration of prophylaxis, although few data are available to draw similar conclusions regarding efficacy.

●For most patients with isolated lower extremity orthopedic injury requiring immobilization, or undergoing arthroscopy or arthroscopic-assisted procedures of the knee, early and frequent ambulation is typically preferred rather than pharmacologic or mechanical methods. However, given the wide range of risk among this population, clinicians should have a low threshold for administering pharmacologic prophylaxis in those assessed as having a higher than usual risk (eg, patients previous history of VTE or undergoing Achilles tendon repair, femoral fracture, tibial plateau fracture); in such cases, when prophylaxis is chosen, although we have a preference for LMW heparin or rivaroxaban, the optimal method is unknown and should be individualized. For example, aspirin is frequently administered in this population, particularly in individuals with low thrombotic risk who are not yet weight bearing. Prophylaxis is generally administered for the period of immobilization. (See 'Lower extremity injury requiring immobilization' above and 'Knee arthroscopy' above.)

●For patients undergoing arthroscopy and arthroscopic-assisted procedures of the knee, data do not support routine prophylactic anticoagulation with LMW heparin; however, individualization of the approach is needed, particularly in those with concomitant individual risk factors for VTE (eg, previous history VTE). (See 'Knee arthroscopy' above.)