By Nicolaides, A N; Fareed, J; Kakkar, A K; Breddin, H K; Et al
Disclaimer
Due to the evolving field of medicine, new research may, in due course, modify the recommendations presented in this document. At the time of publication, every attempt has been made to ensure that the information provided is up to date and accurate. It is the responsibility of the treating physician to determine best treatment for the patient. The authors, committee members, editors, and publishers cannot be held responsible for any legal issues that may arise from citation of this statement.
Evidence and grades of recommendation
We have used the following consistent method lor grades of recommendations in keeping with most other guidelines in this area,4 and first adopted by our group in 1999.
Grade A recommendations are based on Level 1 evidence from randomized controlled trials with consistent results (e.g., in systematic reviews), which are directly applicable to the target population. Single randomized controlled trials have not been accepted as Level 1 even when they were of a high quality and methodologically sound, and have been classified as Grade B.5-7
Grade B recommendations are based on Level 1 evidence from randomized controlled trials with less consistent results, limited power, or other methodological problems, which are directly applicable to the target population. Grade B recommendations are also based on Level I evidence from randomized controlled trials extrapolated from a different group of patients to the target population.
Grade C recommendations are based on Level 2 evidence from well- conducted observational studies with consistent results, directly applicable to the target population.
Recent proposals that a Grade C+ category be used instead of B for extrapolation from randomized trials and that A, B or C recommendations be prefaced by the Grade 1 or 2 according to the balance of benefits, risks and costs made by the “Seventh ACCP Conference of Anti-thrombotic and Thrombolytic Therapy: Evidence- Based Guidelines” 8 have not been used. Developers of national or local guidelines, which include all stakeholders and all relevant healthcare professionals, the public, patients, and healthcare funders, make such judgments more appropriately.
Only fully published, peer-reviewed papers of directly randomized comparisons for each prophylactic method have been used to determine risk reduction. Non-randomized comparisons of outcome in different trials such as those reported by Colditz et al.,9 Mohr el al.,10 and Imperiale et al.11 have not been included as they are potentially biased.
The relationship between the incidence of asymptomatic and the incidence of symptomatic venous thromboembolism (VTE) including pulmonary embolism (PE) has been known for some time.12-14 Reduction in the incidence of asymptomatic deep vein thrombosis (DVT) has recently been shown to be accompanied by a corresponding reduction for symptomatic DVT.15, 16 Demonstration that asymptomatic below knee DVT is associated with subsequent development of the post- thrombotic syndrome (PTS) 17 also validates adoption of surrogate endpoints for efficacy evaluation. Thus, evidence is presented for surrogate outcomes such as the incidence of asymptomatic DVT at screening as well as clinical outcomes (symptomatic DVT or PE) depending on availability of data.
This document presents the evidence in a concise format and attempts to indicate not only the magnitude of the effect of different prophylactic regimens but also the quality of the studies. Information on safety (clinically relevant bleeding and other adverse effects) is also provided. When randomized controlled studies are not available, the lack of data is stated and recommendations for the design of appropriate studies are made.
Regulatory bodies in Europe and North America now consider the various low molecular weight heparins (LMWHs) to be distinct drug products. They require clinical validation for specific indications for each drug and that each LMWH must be dosed according to the manufacturer’s label and recommendations. Therapeutic interchange among these products is not appropriate. The choice of LMWH should reflect the level of clinical evidence and the approval of the regulatory authorities for each indication. This is emphasized throughout the document.
Finally, evidence has been provided for and reference has been made to methods of prevention that are rarely or no longer used or the drug has been withdrawn (dextran, antiplatelet therapy, dihydroergotamine, melagatran/ximelagatran) in order to provide a complete picture to the clinicians and researchers who are new in the field. The reasons for no longer recommending these drugs have been stated.
The problem and the need for prevention
DVT and PE are major health problems with potential serious outcomes. Acutely, PE may be fatal. In the long term, pulmonary hypertension can develop from recurrent PE. Often overlooked is post- thrombotic chronic venous insufficiency (CVI) occurring as a result of DVT causing deep venous reflux or obstruction with skin changes and ulceration with adverse impact on quality of life and escalation of health care costs. In North America and Europe, the annual incidence is approximately 160 per 100 000 for DVT, 20 per 100 000 for symptomatic non-fatal PE and 50 per 100 000 for fatal autopsy- detected PE.18-23 The prevalence of venous ulceration is at least 300 per 100 000 and approximately 25% are due to DVT.24, 25 Estimates of the overall annual costs of CVI vary from 600-900 million euro (US$720 million-1 billion) in Western European countries,26, 27 representing 1-2% of the total health care budget, to 2.5 billion euro (US$3 billion) in the USA.28
Virchow’s triad of factors that predispose to VTE are venous stasis, alterations in blood constituents, and changes in the endothelium; these are as true today as when postulated in the 19th century. It is often necessary for at least 2 factors to coexist for VTE to occur. Principal clinical predisposing influences are immobilization, trauma, surgery, infection and the postpartum period.29 Other predisposing influences are age, obesity, malignancy, previous history of venous thrombosis, varicose veins, dehydration and hormone therapy.23, 30-39 In the background for all of these is predisposition due to thrombophilia.40 The type of risk factor, whether acute or persistent, determines the type and duration of therapy.
Patients admitted to hospital are particularly at risk for VTE and the problem continues after discharge.16, 41-44 Without prophylaxis, the incidence of DVT is high.
Although VTE is an appealing target for maximally effective prophylaxis, this goal has been difficult to achieve. Renewed efforts to educate should maximize utilization, coupled with a system of incentives, warnings, and, where feasible, computerized electronic alerts.44
General and vascular surgical patients
The risk
Patients who undergo non-orthopedic operative procedures are at risk of developing VTE 9, 45-49 (Tables I-IV).
The risk is increased by age, obesity, malignancy, prior history of venous thrombosis, varicose veins, and thrombophilic states. It is also affected by the nature and duration of the operation, type of anesthesia, immobility, dehydration and sepsis.30-34, 177, 178 Known clinical risk factors allow for classification of patients into high, medium or low risk of developing VTE (Tables VI and VII).
Studies in patients having abdominal or pelvic surgery 41, 180, 181 demonstrate that the risk continues after discharge from hospital. Further studies are needed before recommendations can be made on the optimal duration of prophylaxis.
Despite the use of intraoperative heparin or other perioperative antithrombotic agents, vascular surgical patients are at moderate risk. The incidence of postoperative asymptomatic DVT is of the order of 18% in patients having abdominal vascular surgery and 15% for those having peripheral vascular reconstruction (Table I). The reported incidence of proximal DVT (DVT in popliteal or more proximal veins) in patients having abdominal vascular reconstruction is 4-6%.139, 141 The incidence of symptomatic VTE within 90 days of major elective or urgent vascular procedures has been found to be 1.7% to 2.8% in one study.182 A large prospective European registry of vascular surgical procedures showed that the incidence of clinical DVT was 0.9% following aortic procedures and 0.7% following femoro-distal bypass operations.183
The risk of VTE in patients undergoing laparoscopic surgery appeal’s to he low. Three small prospective studies in which no prophylaxis was used showed an incidence of DVT detected by duplex ultrasound or venography in the range of 0-2%.184, 185 Other prospective studies in which some form of prophylaxis was used confirmed the low incidence 186-190 with the exception of one in which 11 of 20 patients developed DVT.191 Large series from surveys,192-194 registries,195-198 a literature review 199 and a population study 182 indicate that the risk for clinical postoperative VTE after laparoscopic procedures is less than 1%. The use of prophylaxis in these studies is not reported in detail, but there appears to be a wide variation from none to LMWH in 80% of patients in some hospitals.
Obesity is an independent risk factor for sudden postoperative fat\al PE.200,201 Bariatric surgery is associated with clinical VTE in 1.2% and with fatal PE in 0.3% of cases.202
Prophylactic methods and recommendations
GENERAL CONSIDERATIONS
In the 1970s, low dose unfractionated heparin (LDUH) (5 000 IU 8 or 12 hourly subcutaneousIy) was found to reduce both DVT (Level I evidence) and fatal PE.203-105 During the late 1980s, 2 published meta-analyses concerning prophylaxis with LDUH compared with no prophylaxis or placebo 46, 47 showed that the incidence of asymptomatic DVT was reduced from 22% to 9% (relative risk-RR-0.41; 95% confidence interval-CI-0.35 to 0.47) and fatal PE from 0.8% to 0.3% (RR 0.39; 95% CI 0.17 to 0.87), The price was a small increase in bleeding complications from 3.8% to 5.9% (RR 1.56 95% CI 1.21 to 1.99).
A multicenter study has found that LMWH not only reduces the incidence of fatal PE but also the overall surgical mortality as compared to controls without prophylaxis.206 Two small randomized placebo-controlled trials in patients having major oncological abdominal surgery207 and emergency abdominal surgery 208 demonstrated the effect of LMWH in reducing asymptomatic DVT.
Sixteen studies have compared LMWH with LDUH,209-224 and 6 studies compared different doses of LDUH or LMWH.214, 225-229
Nine meta-analyses and systematic reviews have compared LMWH with LDUH.230-238 There are some differences between the studies regarding selection of publications. Four of the meta-analyses reported that there was no difference in total mortality comparing LMWH with LDUH.231, 233-235 Two meta-analyses reported reduced incidence of symptomatic PE with LMWH from 0.7% to 0.31% (RR 0.43; 95% CI 0.33 to 0.54) 231, 233 and One showed a decrease of symptomatic VTE.235 The overall conclusion is that there is no large difference between LMWH and LDUH, but the latter has to be given 2- 3 times daily whereas LMWH is administered once daily.
LMWHs have a lower risk of heparin-induced thrombocytopenia (HIT) than LDUH.239, 240 High dose LMWH is more effective but is associated with a higher incidence of hemorrhagic complications than LDUH, whereas a low dose of LMWH has a similar efficacy but with less bleeding.233
Regulatory bodies in Europe and North America now consider the various LMWHs to be distinct drug products. They require clinical validation for specific indications for each drug. Therapeutic interchange among these products is not appropriate.
In a recent double-blind double-placebo randomized study in 2 927 (2 048 valuable) patients having high risk major abdominal surgery, fondaparinux 2.5 mg daily was at least as effective as perioperative LMWH (dalteparin 5 000 U daily) in preventing venographically detected DVT without any increase in bleeding.241
In the meta-analysis by Clagett et al.,46 dextran was also analysed and was found to reduce the incidence of fatal PE (RR 0.22; 95% CI 0.11 to 0.44) although the effect of dextran on DVT was relatively small (RR 0.76; 95% CT 0.64 to 0.91). The preventive effect of dextran on fatal PE has been updated and verified.242 It appears that fibrin formed in the presence of dextran is not cross- linked so that it is easily lysed by the body’s natural fibrinolytic activity.241, 244 However, dextran has inherent risks of fluid overload and anaphylactoid reactions 46 and its routine daily use has been abandoned. Currently, the administration of one perioperative infusion is considered effective in short-stay surgery and is used by some centers in patients where prophylaxis is considered indicated.
Graduated elastic compression (GEC) stockings reduce the incidence of asymptomatic DVT by approximately 50-60% as shown by several studies (Table VIII) and 3 systematic reviews,252-254 but the number of patients studied has been too small to assess the effects on PE.
Intermittent pneumatic compression (IPC) (Table IX) reduces the incidence of asymptomatic DVT by approximately 69% (95% CI 58% to 77%) but the number of patients studied has been too small to assess the effects on PE.
Aspirin reduces DVT by 30% (Table X) and PE by 50% (Table XI).
Combined modalities: evidence from randomized controlled studies shows that combinations of prophylactic methods are more effective than using each method singly. These include LDUH with GEC (Tables XII and XIII), LDUH and antiplatelet agents,170 LDUH and IPC,246 263 dextran and GEC,264 GEC and IPC,270, 271 and LDUH and dihydroergotamine which is no longer manufactured or used because of the risk of vasospasm.272-279
However, the number of studies is relatively small for some combinations and more are needed, particularly in high-risk patients. A randomized study involving 2 551 patients undergoing cardiac surgery has demonstrated a reduction in the incidence of PE from 4% in the LDUH group to 1.5% in the group receiving LDUH combined with IPC (RR 0.37 95% CI 0.22 to 0.63).280
In the majority of studies, the duration of prophylaxis has been for 5-7 days. However, several studies suggest that the risk continues after discharge from hospital.41, 180, 281, 282 Extended prophylaxis to 1 month reduces asymptomatic DVT further by approximately 50-70%.229, 283, 284 However, further studies are needed to determine the optimum duration of prophylaxis beyond 1 week in different groups of patients.
In a review of members of the American Society for Bariatric Surgery, 95% of the surgeons routinely used some form of thromboprophylaxis.285 In one study, a higher dose of LMWH in combination with GEC and IPC was associated with fewer thrombotic events compared to a lower dose group alone (0.6% vs 5.7%). Bleeding was rare occurring in 2/481 patients.286
RECOMMENDATIONS
Low-risk patients are those without risk factors undergoing minor surgery. The data are insufficient to make any recommendations. On the basis of risk/benefit ratio and extrapolation from studies in moderate-risk patients, it is the practice in some countries to use GEC stockings in addition to early ambulation and adequate hydration (Grade C).
Moderate-risk patients are those over the age of 40 years undergoing major surgery for benign disease. The use of LDUH 5 000 IU commenced preoperatively and continued twice or 3 times daily, or LMWH initiated and dosed according to manufacturers recommendations for moderaterisk patients are recommended (Grade A). An alternative method, especially in patients at risk for or with active bleeding, is IPC with GEC compression used continuously until the patient is ambulant (Grade A).
High-risk patients are those aged over 60 years with additional risk factors. LDUH (5 000 IU commenced 2 h before operation and continued postoperatively 3 times a day) (Grade A) or LMWH initiated and dosed according to the manufacturers recommendations are recommended (Grade A). Both may be combined with mechanical methods (GEC or IPC) (Grade B). Fondaparinux (one study) is a Grade B recommendation.
These recommendations are extrapolated to patients undergoing vascular or bariatric surgical procedures in the absence of evidence from prospective clinical trials (Grade C).
Laparoscopic surgery for major prolonged procedures will reduce venous flow in the legs and activate blood coagulation.186 199 Prophylactic subcutaneous LDUH, LMWH, or IPC with GEC are recommended only in patients with additional risk factors (Grade C).
Urologic surgery
The risk
In the 1970s, the incidence of DVT in the absence of prophylaxis was found to be 32% in patients having retropubic prostatectomy and 9% in patients having transurethral resection (Table I). The incidence of symptomatic VTE is currently in the range of 1-5% and PE is the most common cause of postoperative death.182,287,288
Prophylactic methods and recommendations
GENERAL CONSIDERATIONS
One small randomized study demonstrated that IPC was effective in preventing silent DVT when compared with UFH or no prophylaxis 107 (RR 0.27 95% CI 0.14 to 0.52). LDUH was effective in reducing asymptomatic DVT in 3 randomiz.ed studies in which the control groups did not have prophylaxis (RR 0.22 95% CI 0.11 to 0.47).108, 112, 289 A large study of 579 patients having radical prostatectomy did not find any difference in the number of pelvic lymphoceles or blood loss between those receiving LDUH and those not having prophylaxis.290 LMWH has not been studied by randomized controlled studies and there are no studies for patients having transurethral resection.
RECOMMENDATIONS
IPC with GEC is recommended based on one randomized study and by extrapolation from trials in patients having general surgery (Grade B). LDUH (Grade A) giving 5 000 U commenced 2 h before operation and continued 3 times a day in the postoperative period is recommended. An alternative is LMWH initiated and dosed according to the manufacturer’s recommendations by extrapolation from general surgery (Grade C).
Gynecology
The risk
Thromboembolic complications after gynecological surgery occur with approximately the same frequency as for general surgeiy. Patients undergoing major gynecological surgery (e.g. over 30 min duration) aged 40 years or over have a significant risk of postoperative VTE (Table I). This risk will increase with additional risk factors such as obesity, previous VTE, malignancy or immobility. However, the incidence appears to be much lower for benign gynecological surgery and vaginal procedures (Table I). PE is a leading cause of death following gynecological cancer surgew 291 and accounts for approximately 20% of perioperative hysterectomy deaths.292 Risk factors for DVT include those listed for general surgery.293, 294
An additional risk for VTE is the use of estrogen containing combined oral contraceptives (COC),295 which are used by 18% of women in a UK studv.296 COC increase the risk of VTE.295 However, the absolute risk is small and represents an increase from 5 to 15- 30 per 100000 women years.297 The latter is lower than the risk of pregnancy, which is estimated at 100 cases per 100 00\0 maternities. The risk of postoperative VTE showed an increase from 0.5% to 1% for pill users versus non-users in early studies.298 The absolute excess risk in COC users has to be balanced against the risk of stopping the pill 4 to 6 weeks before surgery which includes unwanted pregnancy, the effects of surgery and anesthesia on a pregnancy, and the risks of subsequent termination. Each case should be assessed in relation to additional risk factors. Before major surgery, COC should be discontinued for at least 4 weeks and alternative contraception advised. If it is elected not to discontinue COC then the patient should receive prophylaxis as if for at least a moderate- risk patient. Other estrogen-containing preparations should be considered to carry the same risk as COC at least until studies become available. In emergency surgery or when COC have not been discontinued, VTE prophylaxis should be given at least as moderate- risk category. COC do not need to be discontinued before minor surgery without immobilization. Progestogen-only oral contraceptives need not be discontinued even when immobilization is expected.299
Hormone replacement therapy (HRT) should be included as a risk factor for VTE when assessing patients for elective or emergency surgery.300 HRT does not need to be stopped routinely prior to surgery provided that appropriate thromboprophylaxis is used such as LDUH or LMWH.301 Transdermal HRT has less effect on blood coagulation and appears to have a lower VTE risk than oral HRT.302
Prophylactic methods and recommendations
GENERAL CONSIDERATIONS
Low-risk patients: a Level I study 25 demonstrated a lower DVT rate with the use of GEC (0% vs 4%; P
Moderate-risk patients: subcutaneous LDUH (5 000 IU, 12 hourly),130, 150, 152, 303 or LMWH (initiated and dosed according to the manufacturer’s recommendations) 304, 305 are effective for preventing DVT.
IPC has been shown to be as effective as LDUH or LMWH for preventing DVT when used continuously for 5 days,129, 257, 306 with no bleeding complications.306 Thus, in patients with a high risk of bleeding, IPC can be used as an alternative to heparin prophylaxis until the patient is ambulatory.
Laparoscopic surgery for major prolonged procedures will reduce venous flow in the legs and activate blood coagulation.186, 199 Prophylactic subcutaneous LDUH, LMWH, or IPC combined with GEC are recommended only in patients with additional risk factors (Grade C).
High-risk patients: LMWH 130, 210 305, 306 initiated and dosed according to the manufacturers recommendations), or IPC (throughout hospital stay) 129 are equally effective. Randomized controlled studies in patients having gynecologic oncology surgery have shown no difference in efficacy between LMWH and LDUH given 3 times a day for thromboprophylaxis against DVT or PE and no difference in the risk of bleeding.210, 307-309 The risk of wound hematomas appears to be reduced by avoiding subcutaneous injection near the wound. LMWH has the advantage of once daily injection and is less likely to cause HIT.
RECOMMENDATIONS
Low-risk patients: they should receive GEC (Grade B) in addition to early ambulation and adequate hydration.
Moderate-risk patients: LDUH (5 000 IU, 12 hourly), LMWH (initiated and dosed according to the manufacturer’s recommendations) or IPC are Grade A recommendations. LMWH is the preferred method because it has the advantage of once daily injection and is less likely to cause HIT. IPC is the method of choice in patients with a high risk of bleeding.
High-risk patients: LMWH (initiated and dosed according to the manulacturer’s recommendations) (Grade A), LDUH (5 000 IU 8 hourly) (Grade A) or IPC (throughout hospital stay) (Grade B) are recommended. LMWH or LDUH combined with IPC or GEC stockings provides optimal prophylaxis (Grade B).
Obstetrics
The risk
Pregnancy is a risk factor for VTE with a 10-fold increase compared with the risk for non-pregnant women. The time of greatest risk is the puerperium. PE remains the leading direct cause of maternal deaths in the UK.310 Additional risk factors tor VTE in pregnancy and the puerperium include obesity, age over 35 years, thrombophilic states, Cesarean section and surgical procedures during pregnancy and the puerperium.311 Risk assessment for VTE is recommended for all women in early pregnancy.310
Prophylactic methods and recommendations
GENERAL CONSIDERATIONS
The Cochrane Review of VTE prophylaxis in pregnancy and the puerperium examined 8 trials involving 649 women. It was not possible to assess the effects of interventions because of the limited number of trials and the small sample sizes.312 Large scale randomized trials of currently used interventions are required.
Table XIV summarizes management strategies for various clinical situations.
Women at high risk of VTE including those with previous confirmed VTE should be offered prepregnancy counseling to agree to a management plan. The thrombotic risk exists from the beginning of pregnancy.
Women with previous VTE or a strong family history of VTE, particularly where VTE occurs at a young age (
Systematic reviews and retrospective studies have concluded that LMWH is now the treatment of choice in pregnancy compared to UH in view of efficacy and safety 313-317 (Grade B). The risks of HIT and osteoporosis during pregnancy are reduced with LMWH as compared with LDUH.318, 319
The overall risk of recurrence of DVT during pregnancy has been reported as 2-3% 320 and is higher in those with an underlying thrombophilia or idiopathic thrombosis (6%). A further study repotted a recurrence rate per 100 patient-years 10.9% during and 3.7% outside pregnancy.321 Thus, women in whom a previous VTE occurred in association with a temporary risk factor that is no longer present and no known thrombophilia or additional risk factors should be offered postpartum thromboprophylaxis with LMWH (Grade C). GEC stockings during pregnancy should be considered in addition to postpartum prophylaxis (Grade C). Women in whom a previous VTE was estrogen-related (pregnancy or the combined contraceptive pill), or additional risk factors are present such as obesity should be started with throiriboprophylaxis with LMWH as early as possible in pregnancy (Grade C).
Women with thrombophilias have an increased risk of VTE in pregnancy and the risk varies with the specific thrombophilia. Women with previous VTE and thrombophilia should be offered thromboprophylaxis with LMWH antenatally and throughout the 6 weeks postpartum (Grade B).
Women who are on long-term anticoagulant thromboprophylaxis for VTE and women with anti-thrombiri deficiency are at very high risk (30%) during pregnancy. Those on vitamin K antagonists (VKA) should be advised to switch to LMWH as soon as pregnancy is confirmed because of the risk of embryopathy from warfarin between the 6th and 12th week of pregnancy. In both situations, LMWH dosage should be similar to that used for the treatment of VTE (Grade B).
Table XV shows the most recent Royal College of Obstetricians and Gynecologists (RCOG) guidelines301 for recommended LMWH dosage in pregnancy. Further reports since this guideline was published suggest that a once-daily dosage of tinzaparin provides adequate 24- h cover 322, 323 but more studies are needed concerning its safety and especially the risk of osteoporosis.
Women with a previous VTE and a lhrombophilia such as protein C deficiency, Factor V Leiden, Prothrombin 20210A or protein S deficiency who are at moderately increased risk of VTE should receive LMWH (e.g. enoxaparin 40 mg daily, dalteparin 5000 U daily or tinzaparin 4500 U daily in women of normal body weight) from early pregnancy (Grade C).
Women with no personal history of venous thrombosis but who have a thrombophilic defect identified may require thromboprophylaxis. This will depend on the type of thrombophilia, the family history, and the presence of additional risk factors (e.g. obesity, immobilization and hyperemesis). All should be offered anticoagulant prophylaxis following delivery. The risk of thrombosis should be discussed with the patient antenatally and GEC stockings should be considered (Grade C).
Women with antiphospholipid antibody syndrome (lupus anticoagulant or anticardiolipin antibodies) and recurrent miscarriages should receive thromboprophylaxis with unfractionated heparin (UFH) and low dose aspirin (75 mg/day) 324, 325 from the time of diagnosis of pregnancy (Grade A). Although such therapy is aimed at preventing pregnancy loss that is related to placental thrombosis, these women are also at risk for VTE and prophylaxis with LMWH should continue for at least 2 to 5 days alter delivery.301 In women with antiphospholipid syndrome and previous VTH, postpartum prophylaxis should he continued for 6 weeks (Grade C).
Delivery and the pueperium.-Management of delivery: patients on LMWH antenatally and who wish epidural anesthesia should have heparin prophylaxis discontinued with the onset of labor. An epidural or spinal block is not advised for at least 12 h after prophylactic LMWH administration and 24 h after therapeutic doses have been discontinued.126 LMWH should not be given for at least 4 h after the epidural catheter has been inserted or removed and the catheter should not be removed within 10 to 12 h of the most recent injection.327 For delivery by elective Caesarea\n section, the woman should receive a thromboprophylactic dose of LMWH on the day before delivery. On the day of delivery the thromboprophylactic dose of LMWH should be given 3 h postoperatively or 4 h after removal of the epidural catheter (Grade C).
Management of the puerperium: in addition to previous VTE and thrombophilias, other risk factors should be considered for postpartum prophylaxis: age over 35 years, obesity, Caesarean section (particularly an emergency procedure during labor), gross varicose veins, pre-eclampsia and immobilization (Grade C).
Postpartum thromboprophylaxis is recommended in women with previous VTE, known thrombophilias and other thrombotic risk factors. The first postpartum daily dose of subcutaneous LMWH (enoxaparin 40 mg, dalteparin 5 000 U daily or tinzaparin 50 U/kg) should be given 3 to 4 h after delivery. Postpartum anticoagulation should be continued for a minimum of 6 weeks in patients with previous VTE or thrombophilia. In other patients, prophylaxis should continue until discharge from hospital, and the need for prophylaxis should be reviewed if the hospital stay continues beyond 5 days (Grade B).
If a patient does not wish to continue on selfinjections of LMWH, VKA can be commenced on the 1st or 2nd postpartum day. LMWH can be discontinued when the INR has been within the target range of 2-3 for 2 consecutive days. GEC stockings can be added to LMWH in high- risk patients and should be used where LMWH is contraindicated. Where anticoagulants are contraindicated, GEC stockings should be worn for at least 6 weeks following delivery and can be combined with 75 mg of Aspirin daily (Grade C).
Patients who develop VTE during pregnancy or the puerperium should be referred for hematological screening to determine if they have underlying thrombophilia and counseled about the increased risk of COC pills. Progestogen-only contraception is suitable for these women. They should also be counseled about the need for prophylactic treatment in any future pregnancy.
Breast feeding is not contraindicated with either LMWH, UFH or warfarin (Grade C).328, 329
Orthopedic surgery and trauma
Elective hip replacement
The risk
In the absence of prophylaxis, patients undergoing elective major joint replacement and those with hip fracture have a DVT risk of approximately 50% (Table I).330-332 The frequencies of proximal DVT (Table II) and PE (Tables III and IV) are also high. Symptomatic events range from 2-5%.333 Studies on clinical DVT and PE indicate a postoperative risk period of approximately 3 months 331,333.334 (Table V). Recent mortality studies have confirmed a reduced survival for 2-3 months following elective total hip replacement (THR) surgery with the highest death rate initially early after operation.335,336
There is a high incidence of proximal DVT (18-36%) in patients having THR 64, 67, 68, 70, 73, 337-340 in contrast to patients having total knee replacement (TKR) in whom the preponderance of thrombosis is distal.83-85, 341, 342
Modern THR surgery is performed with a continuing reduction in hospital stay (3-6 days) so that patients are discharged while still at risk. Thus, the majority of clinical events appear after hospital discharge, giving a false impression of a decreasing problem.182, 334
Prophylactic methods and recommendations
GENERAL CONSIDERATIONS
The majority of studies in elective orthopedic surgery have been carried out on elective THR patients. Prophylactic methods that have been investigated include aspirin, clextran, fixed LDUH, adjusted dose UFH, fixed LDUH with dihydroergotamine (no longer manufactured), LMWH, heparinoid, recombinant hirudin, the direct thrombin inhibitor melagatran (withdrawn due to hepatotoxicity) fixed mini-dose and adjusted doses of VKA, GEC stockings, IPC and foot impulse technology (FIT). To determine the risk reduction lor each prophylactic method, only randomized studies with systematic screening tests for DVT are used for this analysis (Tables X, XI, XVI-XVIII).
LDUH (5 000 IU 8 or 12 hourly) was found to be effective for reducing DVT from 46.8% to 23.3% (RR 0.5 95% CI 0.43 to 0.58) (meta- analysis of 20 randomized controlled studies(Level I evidence)47 in patients having elective THR and was the method of choice in the 1980s.
It has since been demonstrated that LMWH is superior to UFH for elective THR surgery reducing DVT from 21.2% to 13.8% (RR 0.66 95% CI 0.52 to 0.84) and PE from 4.1% to 1.7% (RR 0.4; 95% CI 0.19 to 0.84) 231, 232, 234, 236, 340, 359-363 (Level I evidence). Thus, LDUH is no longer recommended.
As indicated in the section on “General and vascular surgical patients”, regulatory bodies in Europe and North America now consider the various LMWHs to be distinct drug products. They require clinical validation for specific indications for each drug. Therapeutic interchange among these products is not appropriate.
Randomized controlled studies have shown that recombinant hirudin is more effective than LDUH 364-366 or LMWH.365 Of 2 079 patients studied, 1 587 were included in the primary efficacy analysis. Overall, DVT was reduced with hirudin 15 mg twice daily compared with 40 mg enoxaparin from 25.5% to 18.45% (P=0.001; RRR 28%). The safety profile was the same in both groups.365
Several randomized controlled trials have compared VKA with LMWH. LMWH was found to be more effective 355, 356, 367, 368 or at least as effective 358 in the prevention of asymptomatic DVT. However, this was at the expense of a slight increase in hemorrhagic complications. When LMWH was started before or immediately after surgery, there was a marked reduction of proximal DVT from 3% to 0.8% (RR 0.28; 95% CI 0.1 to 0.74).369 Symptomatic DVT was also reduced from 4.4% in the warfarin group to 1.5% in the LMWH group (RR 0.32; 95% CI 0.12 to 0.88). A recent meta-analysis of VKA in orthopedic surgery 37 showed a RR of 0.56 (95% CI 0.37 to 0.84) for DVT and 0.23 for PE (95% CI 0.09 to 0.59) compared with placebo. VKA were less effective than LMWH in preventing total DVT (RR 1.51, CI 1.27 to 1.79) and proximal DVT (RR 1.51 95% CT 1.04 to 2.17) although the risk of wound hematoma was increased from 3.3% in the VKA recipients to 5.3% in LMWH recipients (RR 2.29; CI 1.09 to 7.75).
In a recent clinical trial on THR patients,371 1 279 patients were randomized on the 3rd postoperative day to LMWH or to warfarin for the subsequent 6 weeks. The primary endpoint was the overall clinical failure rate, i.e. symptomatic VTE (radiologically confirmed), major hemorrhage or deaths. The failure rate was 3.7% in the LMWH group and 8.3% in the warfarin group (P=0.01). Major bleeding occurred in 1.4% in the LMWH group and in 5.5% in the warfarin group. It appears that reduced bleeding seen initially after surgery due to the slow onset of action for warfarin is off- set by long-term increased bleeding. Furthermore, national drug registries have shown warfarin to be a major cause of readmission and fatal bleeding.372, 373 With these data, and because of the need for monitoring, the small therapeutic window and the risk of drug interactions, some surgeons find it difficult to see an advantage for VKA over LMWH.
In contrast to LMWH, the pentasaccharide fondaparinux, is a pure synthetic chemical compound. It is a potent indirect inhibitor of factor Xa and acts by a catalytic effect facilitating antithrombin binding to activated factor X and represents one of many attributes of heparins. The drug is administered by subcutaneous injection once daily. It has been registered internationally for major orthopedic surgery. Two large randomized controlled trials compared fondaparinux to enoxaparin.374, 375 Reduction of asymptomatic DVT was 26% (RR 0.74 95% CI 0.47 to 0.89) and symptomatic PE was 56% (RR 0.44 95% CI 0.27 to 0.66) with fondaparinux. For the 2 studies combined, the incidence of major bleeding was 3% in the fondaparinux and 2.1% in the enoxaparin patients (P>0.05). Fondaparinux may accumulate and increase bleeding in patients with impaired renal function.
A meta-analysis in the early 1990s 170 demonstrated that antiplatelet therapy in elective hip surgery is only moderately effective for protection against DVT (RR 0.7; 95% CI 0.61 to 0.82) (Table X) but the observed reduction in the risk of PE was substantial (RR 0.49; 95% CI 0.26 to 0.92) (Table XI). However, the recent PEP study 262,376 showed that aspirin is not as valuable as the metaanalysis suggested. Over 13 000 hip fracture patients were randomized to have either aspirin or placebo. The overall death rate was identical in each group. Risk reduction for symptomatic VTE was from 2.5% to 1.6% and this was only 1/2 of that expected from LMWH and 1/3 from pentasaccharide. The reduced risk of VTE was matched by an increased risk of blood transfusion, gastrointestinal bleeding and wound bleeding. In a supplementary group of 4 000 elective hip and knee replacement patients, there was an insignificant difference in symptomatic VTE.376 The relative weak thrombophylactic effect of aspirin therefore carries an alternative complication rate and its use might deprive patients of safer or more effective prophylaxis.
The Cochrane database 254 and an earlier metaanalysis 252 shows that GEC is effective in reducing DVT in hospitalized patients, but there are few robust studies specific to orthopedic surgery.69, 377 In addition, there are disadvantages to graduated elastic compression in trauma cases in which the limb has to be regularly inspected. Because other methods of prevention are more effective, GEC stockings on their own are not recommended.
IPC is effective in patients having THR 64, 68, 344 (Table XVI, hip replacement) reducing DVT from 43.6% in the control groups to 21% in the compression groups (RR 0.48; 95% CI 0.36 to 0.64) (Level I evidence). It offers an alternative for surgeons with concerns or patients with contraindications to chemical prophylaxis. It can also be used as an additi\onal method for those at particularly high risk. In a recent randomized study 378 in 131 patients having THR and TKR, the combination of LMWH plus IPC was more effective than LMWH plus GEC stockings (DVT incidence 0% versus 28%).
Recent data demonstrate that FIT combined with GEC is effective in reducing the incidence of proximal DVT in patients having hip or knee replacement (Table XVII) with reduced bleeding and swelling. Direct comparisons with chemical prophylaxis are sparse; there is probably superiority to UFH 348 and equivalence with LMWH for THR 350, 379 but not for TKR.351
Mechanical methods are intuitively attractive to orthopedic surgeons because of the lack of risk for bleeding. The small number of randomized studies available sets mechanical methods at a disadvantage compared with the considerable number of large studies of chemical prophylaxis. Mechanical methods are generally cumbersome and need supervision from health personnel. Nevertheless, they offer an alternative during the early postoperative period for surgeons with concerns about bleeding or for patients with contraindications to chemical prophylaxis. They may also be used as an additional method for those at particularly high risk 378 although this aspect requires confirmation with further studies.
Modern technology has made IPC devices light, silent, more portable and more effective in preventing stasis by sensing venous volume so that the compression period follows immediately alter venous refilling. In addition, different sleeve designs and materials have been used to improve patient compliance.380
Melagatran is a direct thrombin inhibitor that had been developed for subcutaneous administration and as the prodrug ximelagatran for oral administration. It had been approved for short-term use in orthopedic surgery in some countries. Four randomi/.ed controlled studies have compared melagatran/ximelagatran with enoxaparin.381- 384 when melagatran was administered immediately before surgery and on the evening of surgery and then orally twice a day, the incidence of DVT was lower than with enoxaparin that was started the evening before surgery and continued each evening thereafter. Jn the EXPRESS trial,381 total VTE was reduced from 26.6% to 20.3% (P=0.0003). Fatal bleeding, critical site bleeding and bleeding requiring re- operation did not differ between the 2 groups. “Excessive” bleeding, i.e. bleeding as subjectively judged by the local investigator was 1.2% with enoxaparin vs 3.1% with melagatran. For clarification bleeding events in the EXPRESS trial 381 were evaluated and classified (i.e. “adjudicated”) by a group of physicians external to the sponsor, not subjectively by the local investigator. However, one category of bleeding called excessive bleeding, was not adjudicated but instead was a subjective assessment by the local investigators of bleeding observed which they thought was in excess of what they usually observed. When the melagatran/ximelagatran regimen was started after operation, the incidence of DVT was lower in the enoxaparin group (METHRO III) 384 and bleeding rates were the same in both groups. In a North American study where both oral ximelagatran 24 mg bid and enoxaparin 30 mg bid were commenced the day after THR surgery and continued for 7-12 days, venographic DVT plus symptomatic VTE was detected in 4.6% of the enoxaparin group and 7.9% of the ximelagatran recipients (P=0.03). Major bleeding was less than 1% in both groups.
The melagatran/ximelagatran regimen had been approved by 18 countries across Europe, South America and Asia for short-term prophylaxis in major hip and knee surgery. Because of concerns about an increase in myocardial infarctions, elevation of liver enzymes and a case of fatal liver failure in the long-term studies, the FDA had not given approval. Because of a second case of severe liver damage a week after short-term therapy, melagatran/ximelagatran has been withdrawn from the market and its development terminated (http:/ / www.astrazeneca.com/pressrelease/ 5217.aspx).
Meta-analysis shows that spinal and epidural anesthesia reduce both thromboembolism and perhaps mortality in hip fractures,385, 386 and TKR.387-389 This method does not reduce risk sufficiently on its own, but should be regarded as a useful adjunct. Initial European experience suggested that neuraxial anesthesia could be safely used in the presence of LMWH.390 However, more recently there have been concerns that a spinal hematoma may develop on very rare occasions.391, 392 Guidelines have been suggested.327,393 LMWH (or pentasaccharide) can be given safely 4 h after removal of the epidural catheter (see pregnancy). However, LMWH or pentasaccharide should be avoided whilst a continuous postoperative neuraxial block is in place. The catheter should not be inserted until serum levels of the chemical agent used are at their lowest. This means that postoperative administration of the agent is, generally, safer and more predictable than pre-operative administration when epidural analgesia is needed.
Timing of prophylaxis.-VTE prophylaxis involves a balance of risks and benefits. Chemical prophylaxis poses a dilemma: for a given dose the closer it is administrated to surgery, the better the thromboprophylaxis but the greater is the risk of bleeding complications.355 In Europe, LMWH is given at a lower dose prior to operation, providing an anticoagulant effect to counteract the intraoperative activation of coagulation factors and venous stasis. However, if a given dose of the drug is administered too long before surgery then, intraoperative blood levels would be inadequate for effective prophylaxis, whereas if given too close to surgery then surgical bleeding is a threat. In North America, LMWH is given after surgery at a higher dose and more frequently. This should reduce the risk of surgical bleeding, yet intraoperative thrombogenesis is not prevented and thrombi may have already begun forming. The drug is now expected to be therapeutic as well as prophylactic. Therefore, prophylaxis needs to be given close but not too close to surgery.369,394
IPC and FIT sleeves are available in sterile packages that allow their intraoperative use, reducing the risk of bleeding and the time the patient is not under prophylaxis.346,395,396
Duration of proplivlaxis in elective orthopedic surgery.-Studies in patients having THR 174, 181, 331,334,355,397-401 demonstrate that there is prolonged risk with 45-80% of all symptomatic events occurring after discharge from hospital.203,331,402,403
Randomized controlled studies in patients having THR indicate that prolonged thromboprophylaxis with LMWH for up to 35 days is safe and effective irrespective of whether in-hospital prophylaxis was with LMWH or warfarin. It decreases the frequency of venographically detected total DVT, proximal DVT and symptomatic VTE after the 7th day by more than 50%.359,401,404-407 Further studies are needed before recommendations can be made for prophylaxis beyond 35 days. The optimal duration of prophylaxis is unknown. Epidemiological data on postoperative death rates indicate a much longer duration of risk in subgroups such as emergency patients (e.g. hip fracture) and patients with co-morbidity (e.g. rheumatoid arthritis) in which vascular deaths dominate.336,408
One randomized controlled trial compared warfarin prophylaxis (INR 2-3) for 9 days with warfarin extended for 1 month after hospital discharge. VTE occurred in 5.1% of in-hospital prophylaxis patients and 0.5% in those having extended prophylaxis (RR 9.4; 95% CI 1.2 to 73.5).409 This study was prematurely terminated because of the superiority of prolonged prophylaxis. As indicated above, it has been subsequently demonstrated that extended prophylaxis with warfarin is associated with more hemorrhagic complications than with LMWH.371
RECOMMENDATIONS
LMWH initiated and dosed according to the manufacturer’s recommendations, fondaparinux, oral anticoagulant therapy, IPC or FIT combined with GEC are Grade A recommendations. The preferred methods are LMWH or fondaparinux for in-hospital prevention. IPC or FIT combined with GEC stockings are an equivalent alternative to LMWH for those surgeons or anesthetists concerned about bleeding either in all or in certain patients. These devices can be used as long as tolerated and then replaced with chemical prophylaxis for the rest of the 5-week period of risk. Recombinant hirudin (Revasc) is approved for short-term prophylaxis in approximately 20 European countries and can be used in patients with HIT (Grade A).
Prophylaxis with LMWH should be initialed either before or after operation depending on the adopted regimen (Grade A). Fondaparinux should be started at least 6 to 8 h after surgeiy Prophylaxis should be continued for 4-6 weeks with LMWH (Grade A) or fondaparinux (Grade C, extrapolation from hip fracture trial).
Elective knee joint replacement
The risk
Data from THR should not be extrapolated to TKR. The incidence of asymptomatic DVT detected by venography is higher in patients having TKR than THR. However, the incidence of above knee DVT is lower than in patients having THR (see section on THR above).
Prophylactic methods and recommendations
GENERAL CONSIDERATIONS
IPC is effective in patients having knee replacement (RR 0.27; 95% CI 0.14 to 0.49) (Table XVI, knee replacement). One small study demonstrated that IPC reduced the incidence of asymptomatic DVT from 65% to 6%.80 A subsequent study found IPC to be more effective than aspirin.345 IPC was found to be less effective than coumadin for preventing venographically detected DVT (32% vs 19%).410 FIT was also effective in 2 studies 86, 342 but showed inferiority when compared to LMWH in 2 other studies 351, 352 (Table XVII).411
A randomized controlled study demonstrated that LMWH was more effective than placebo. It reduced venographically detected DVT from 65% in the placebo group to \19% in the LMWH group (RR 0.3; 95% CI0.16 to 0.58).41-1 Subsequent studies demonstrated that LMWH was more effective than LDUH (RR 0.75; 95% CI 0.58 to 0.92) 413, 414 or warfarin (RR 0.68; 95% CI 0.62 to 0.76) (Table XVIII).
Fondaparinux (2.5 mgs.c. once daily stalling 6 h alter surgery) was more effective than enoxaparin (30 mg s.c. bid starting 12-24 h after surgery) in one study.415 VTE (defined as venographically detected DVT, symptomatic DVT or symptomatic. PE) was reduced from 27.8% in the enoxaparin group to 12.5% in the fondaparinux group (RR 0.45; 95% CI 0.32 to 0.62). However, major bleeding was more common with fondaparinux (2.1% vs 0.2%, P=0.006). This increased rate of bleeding when fondaparinux was given within 6 h of surgery was confirmed in a meta-analysis 416 including this study and 3 other randomi/.ed controlled trials comparing fondaparinux with enoxaparin in patients having orthopedic surgery other than TKR.
Subcutaneous melagatran followed by oral ximelagatran had the same efficacy and safety as LMWH.382, 417 Three randomized control studies compared oral ximelagatran 24 mg bid or 36 mg bid starting the morning after surgeiy with adjusted dose warfarin (INR range 1.8 to 3) starting on the evening after surgeiy418-420 The incidence of DVT was the same in both groups for the 24 mg bid dose but the overall VTE or death rate was reduced with the 36 mg bid dose. There was no difference in hemorrhagic complications between either dose of ximelagatran and warfarin.
The melagatran/ximelagatran regimen had been approved by 18 countries across Europe, South America and Asia for short-term prophylaxis in major hip and knee surgery. Because of concerns about an increase in mvocardial infarctions, elevation of liver enzymes and a case of fatal liver failure in the long-term studies, the FDA had not given approval. Because of a second case of severe liver damage a week after short-term therapy melagatran/ximelagatran has been withdrawn from the market and its development terminated (http:/ /www.astrazeneca.com/pressrelease/5217. aspx).
DURATION PROPHYLAXIS
The effect of extending prophylaxis using LMWH to 30-42 days beyond hospitali/.ation on svmptomatic DVT in patients having TKR is less (OR 0.74; 95% CI 0.26 to 2.15; P>0.05) than in patients having THR (OR 0.33; 95% CI 0.19 to 0.56; P
RECOMMENDATIONS
LMWH (initiated and dosed according to the manulacturer’s recommendations) or warfarin (although less effective) are Grade A recommendations. Fondaparinux is a Grade B recommendation (one study). IPC or FIT plus GEC stockings are alternative options but more studies are needed (Grade B).
Hip fracture surgery
The risk
Patients having hip fracture surgery have the highest rates of DVT (46-60%) (Tables I, III, IV) 98, 421, 422 and fatal PE (2.5- 7.5%).338, 422, 433 Jhe VTE risk period lasts for 2-3 months after hip fracture surgery in spite of common short-term prophylaxis 333,334 and the 90-day risk of overall death is 13%.424 After hip fracture, the risk is greater than the standardized mortality, the majority dying of vascular events despite the fact that most patients receive some form of short-term prophylaxis.355, 336
Prophylactic methods and recommendations
GENERAL CONSIDERATIONS
Because the risks of DVT and PE, including fatal PE, are high in patients with hip fracture (Tables I, III, IV), prophylaxis should start as soon as possible after diagnosis and should be the same as that recommended for elective hip surgeiy
Reduction in asymptomatic DVT has been demonstrated by IPC (RR 0.2; 95% CI 0.07 to 0.55) 344 (Table XVI) and FIT in combination with GEC 353 (RR 0.32; 95% CI 0.32 to 0.67) (Table XVII). In the most recent study 343 the combined endpoint of PE and proximal DVT using duplex ultrasound was reduced from 12% in the group without prophylaxis to 4% in the IPC group. More studies are needed.
A meta-analvsis 170 demonstrated that antiplatelet therapy in traumatic orthopedic surgery is only slightly effective for protection against DVT (RR 6.86; 95% CI 0.73 to 1) (Table X) but the observed reduction in the risk of PE is substantial (RR 0.4; 95% CI 0.22 to 0.71) (Table XI). In the randomized, placebo-controlled trial of patients undergoing surgery for hip fracture (13 356 patients) or for elective hip or knee arthroplasty (4 088 patients), aspirin in a dose of 160 mg daily started preoperativelv was used as the primary prophylactic agent for 35 days. In the patients with hip fracture, aspirin reduced the incidence of symptomatic DVT by 29% (95%. CI 3% to 48%; P=0.03) and PE by 43% (95% CI 18% to 60%; P=0.002). PE or DVT was confirmed in 105 (1.6%) of 6 679 patients assigned aspirin compared with 165 (2.5%) of 6 677 patients assigned placebo, which represents an absolute reduction of 9 per 1 000 and a proportional reduction of 36% (95% CI 19% to 50%; P=0.0003). However, the complication rate (transfusion requirements and bleeding) offset much of the reduction in symptomatic VTE. The death rate was equivalent in the placebo and aspirin groups.262 Since other methods are more effective, aspirin on its own is not recommended for routine thrombophylaxis.
In the 1970s, LDUH had been found to be effective in reducing asymptomatic DVT (RR 0.51 ; 95% CI 0.42 to 0.62) (Table XVII) and although an overview of trials had not demonstrated a significant reduction in total PE, there was a significant reduction in fatal PE.47
LMWH has been assessed against placebo,70, 229 LDUH,425 danaparoicl,426 high dose (40 mgenoxaparin) LMWH 427 and fondaparinux.428 LMWH has been found to be equally effective as LDUH without increase in hemoirhagic complications.429
Three randomized controlled trials have demonstrated that VKA are effective in preventing asymptomatic DVT with a 61% RR reduction for DVT and 66% for proximal DVT, compared with no prophylaxis.98, 430, 431 The increase in hemoirhagic complications reported varied from 0% to 47% without any increased bleeding in the most recent trial.98
Foridaparinux given for 11 days was more effective compared with LMWH in reducing VTE from 19.1% to 8.3% (RR 0.46; 95% CI 0.32 to 0.59) and proximal DVT from 4.3% to 0.9% (RR 0.22 95% CI 0.09 to ().53).428 There was no difference in major bleeding but minor bleeding was increased from 2.1% in theenoxaparin group to 4.1% in the fondaparinux group (P=0.02). In a second study, patients who received fondaparinux for 7 days were randoinized to continual ion with fondaparinux or placebo for a further 3 weeks.432 The incidence ol venographic DVT was 1.4% in the extended prophylaxis group and 35% in the placebo group (RR 0.04; 95% CI 0.01 to 0.13). Symptomatic VTE was 0.3% and 2.7%, respectively (RR 0.11; 95% CI 0.01 to 0.88). There was no difference in hemorrhagic complications.
Delayed admission to hospital or delayed surgery following hip fractures is associated with a high incidence of DVT developing prior to surgeiy433-436 The incidence of preoperative DVT as shown by venography can be as high as 62% for all DVT and 14% for proximal DVT when the delay is 48 h or more.436 Thus, it is strongly recommended that if surgical delay is anticipated, prophylaxis with LDUH or LMWH is commenced as close to the fracture as possible. Prophylaxis should be restarted once postoperative hemostasis has been achieved.
RECOMMENDATIONS
LMWH (initiated and dosed according to the manufacturer’s recommendations), fondaparinux, adjusted dose VKA (INR: range 2-3), or LDUH are Grade A recommendations. IPC or FIT combined with GEC should be used when there are contraindications for pharmacological prophylaxis (Grade B). If surgery is likely to be delayed, prophylaxis should be initiated with LMWH or IPC or FIT plus GEC as close to the fracture as possible (Grade C).
Knee arthroscopy
The risk
Knee arthroscopy is a very common procedure varying from a simple diagnostic technique to an extensive repair of injured soft tissues. A tourniquet is usually used.
The incidence of DVT in patients undergoing arthroscopic procedures in the absence of prophylaxis as demonstrated by routine venography or duplex ultrasound is approximately 7% for all thrombi and 1.4% for proximal DVT (Table I).84, 163-168, 437 The risk is minimal for diagnostic arthroscopy,165, 334 but the risk is increased if the tourniquet is applied for more than 1 h or if therapeutic arthroscopy is performed.163, 165
Symptomatic VTE occurs after arthroscopy without prophylaxis but it is very rare. In a 10-year prospective study, clinical and radiologically confirmed symptomatic DVT occuired in 0.6%.165, 334 One prospective study lound symptomatic PE within 5 weeks after surgery in 1/101 patients who had received LMWH for about 2 days.438
Prophylactic methods and recommendations
GENERAL CONSIDERATIONS
LMWH was effective compared with no prophylaxis in reducing the incidence of ultrasound detected DVT in 2 blind randomized studies of patients scheduled for arthroscopy, with 239 patients in one 167 and 130 patients in the other.168 The incidence of DVT was reduced from 8% in the control groups to 1% in the LMWH groups (RR 0.12 95% CI 0.03 to 0.53). Bleeding complications were not increased.
Thus, although clinical VTE is uncommon and fatalities are rare, the huge number of patients undergoing knee arthroscopy surgery makes VTE complications potentially relatively frequent. There is a clear correlation between age and degree of trauma with VTE.84 This justifies prophylaxis in patients with additional risk factors along with prolonged tourniquet application or when extensive surgery beyond a simple diagnostic procedure is performed.
RECOMMENDATIONS
Recommendation for simple diagnostic arthroscopy: routine prophylaxis is not recommended unless other risk factors are present (Grade C).
Recommendation for arthroscopic surgery (e.g. ligament reconstmctions): LMWH starting before or alter surgery (Grade B) or IPC in the presence of contr\aindications to LMWH are recommended (Grade C) until full ambulation.
Isolated below knee injuries
The risk
Patients with below knee injuries and immobilization have a DVT incidence in the range of 10-35% depending on the type and severity of injury (Table I) 144-149 and carry a risk of clinical PE in the range of 0.4-2.1%.424 The frequency of symptomatic events is unknown.
Prophvlactic methods and recommendations
GENERAL CONSIDERATIONS
This group is so heterogeneous that studies and recommendations are difficult to devise.
In one study of 253 patients with plaster casts of which the majority had soft tissue injuries, the ultrasonic incidence of DVT at cast removal was reduced from 17% in the control group to 5% in a LMWH group.146 It was reduced from 4% in the control group to 0% in the LMWH group 146 in another study of 339 patients.147 Considering both studies the RR was 0.21 (95% CI 0.09 to 0.49).
In patients with lower leg fractures, the 5-week incidence of venographic DVT was reduced from 18% in the control group to 10% in the LMWH group in one study (n=293) 148 and from 13% to 11% in another (n=150).149 Considering both studies, the effect of LMWH on DVT was not significant (P>0.05) (RR 0.64 95% CI 0.39 to 1.05). More studies are needed in well-defined groups of patients.
RECOMMENDATIONS
Currently available data based on small studies of a mixture of different types of injury do not allow routine prophylaxis to be recommended for isolated limb trauma. However, a thorough risk assessment and an approach using LMWH standardized by an institution yet individualized for each patient is recommended.
Multiple trauma
The risk
The incidence of DVT in patients who have sustained major trauma is in excess of 50% 77, 78, 439-442 (Table I) and PE is the 3rd leading cause of death in those who survive beyond the 1st dav.77, 443-445 The risk is particularly high in patients with spinal cord injury, pelvic fracture and those needing surgery.77, 78, 446-448
Prophylactic methods and recommendations
GENERAL CONSIDERATIONS
Patients with polytrauma have a particularly high risk for VTE. The tissue factor released by multiple injuries is potentiated by the likely surgical interventions and the subsequent prolonged immobility 446 produces marked venous stasis. Routine venography has shown a DVT frequency of 58% in these patients.77
Well-designed studies in this area are few and thromboprophylaxis has to be assessed according to the risk for bleeding. However, in the absence of intracranial bleeding and when bleeding is under control, LMWH (enoxaparin 30 mg bid) started within 36 h of injury has been shown to be more effective than LDUH (5 000 IU bid).439 It reduced the incidence of venographic DVT from 44% in the LDUH to 31% in the LMWH group (RR 0.7; 95% CI 0.51 to 0.97). The superiority of LMWH to LDUH has been confirmed by a subsequent study and a meta- analvsis.441, 449 A study comparing nadroparin fixed daily dose versus a weight-adjusted dose did not demonstrate any significant difference (0% vs 3% DVT).450
Three randomized controlled trials have tested the efficacy of IPC. The first was in patients with pelvic fractures but the study was small and underpowered so that the DVT reduction from 11% in the control group to 6% in the IPC group was not significant (P>0.05).343 IPC or FIT was compared with enoxaparin 30 mg bid in the second with an incidence of DVT of 2% and 1%, respectively.451 In the third, IPC was compared with FIT with an incidence of DVT of 6% and 21%, respectively (P
Mechanical methods are attractive if chemical prophylaxis is contraindicated. However, more studies are needed to confirm the efficacy of IPC as this would be the method of choice in patients in whom LMWH is contraindicated because of increased or continuing risk of bleeding.
RECOMMENDATIONS
LMWH starting as soon as bleeding risk is acceptable (Grade A) or IPC in the presence of contraindications to LMWH (Grade B) and continued until lull ambulation.
Elective spine surgery
The risk
The incidence of DVT detected by routine venography in the absence of prophylaxis has been found to be 18% (Table I), 154, 453 A review of studies on complications in patients having spinal fusion reported a 3.7% incidence for symptomatic DVT and 2.2% for PE.454
Prophylactic methods and recommendations
GENERAL CONSIDERATIONS
Two small randomized controlled studies comparing no prophylaxis with LDUH 455 and with enoxaparin 768 demonstrated that prophylaxis reduces the incidence of asymptomatic DVT from 20% and 10% respectively to 0%. In a prospective non-randomized study of 306 patients769 venographically detected DVT was found in 6% of those having IPC and in 21% of those without prophylaxis.
RECOMMENDATIONS
Mechanical method: IPC (Grade B); drug: LMWH (Grade B); initiation: before operation for IPC or after operation for LMWH; duration: during hospilalization (Grade C).
Spinal cord injury
The risk
In the absence of prophylaxis the incidence of silent DVT is of the order of 35% (Table I). In this group of patients, PE is the 3rd leading cause of death.456, 457 In a series of 1 649 patients undergoing rehabilitation, symptomatic DVT occurred in 10% and PE in 3%.458
Prophylactic methods and recommendations
GENERAL CONSIDERATIONS
LDUH and a combination of LDUH with GEC were compared with no prophylaxis in a randomized controlled study.108 The incidence of venographic DVT was 47% and 50% in the control and LDUH groups and 7% in the group that received combined prophylaxis. Subsequently, 2 small randomi/ed controlled
Comments