Venous thromboembolism: prophylaxis: antiplatelet drugs reduce the risk of DVT and PE in high risk medical and surgical patients.

Clinical bottom line (level 1a)

  1. In high risk surgical patients, antiplatelet drugs given for about two weeks reduce the risk of deep vein thrombosis (NNT = 11 at unknown) and pulmonary embolism (NNT = 58 at unknown) .
  2. The benefits are greater the more risky the surgery (traumatic orthopaedic>elective orthopaedic>general surgery).
  3. Antiplatelet drugs given for about two weeks reduce the risk of DVT in high risk medical patients (NNT = 13 at unknown) . The effect on PE is unclear.
  4. Bleeds requiring transfusion (NNH = 291 at unknown) and bleeds requiring re-operation or causing a wound haematoma or infection due to bleed (NNH = 45 at unknown) are more common.
Antiplatelet Trialists' Collaboration : British Medical Journal 1994; 308 (6923): 235-246
Expires September 2003

The study

Systematic review of unconfounded randomised controlled trials of
  • Patients: having general, traumatic orthopaedic or elective orthopaedic surgery and considered 'high risk' for thromboembolism
  • Intervention: antiplatelet prophylaxis compared with no antiplatelet therapy
  • Outcome: deep vein thrombosis or pulmonary embolism


    • Articles found in all using MEDLINE and Current Contents, up to March 1990 (search terms: not stated ) and manual search of selected journals, lists of conference abstracts and meeting reports, bibliographies of relevant studies and review articles, collaboration with the trial register of the International Committee on Thrombosis and Haemostais and correspondence with colleagues, manufacturers of antiplatelet drugs and collaborating trialists

    Selection criteria: as above
    Appraisal criteria: set criteria detailed in text.
    Articles excluded if: none stated

    53 trials (53 involving PE and 45 involving DVT) involving 8891 patients were included
    • Patients received antiplatelet drugs or placebo for an average of two weeks. Antiplatelets used were: aspirin, dypiridamole, hydroxychlorquine, ticlodipine

    The evidence

    Outcome Time to outcome CEREERRRR
    (95% CI)    		ARR
    (95% CI)    		NNT
    (95% CI)
    DVT-all surgery unknown 814
    (34.8%)    		 601
    (26.0%)    		 25%
    (19% to 32%)    		 8.84%
    (6.21% to 11.5%)    		 11
    (9 to 16)
    DVT-general surgery unknown 396
    (27.1%)    		 278
    (19.4%)    		 29%
    (18% to 38%)    		 7.76%
    (4.69% to 10.8%)    		 13
    (9 to 21)
    DVT-traumatic orthopaedic unknown 186
    (41.9%)    		 163
    (35.9%)    		 14%
    (-1% to 27%)    		 5.99%
    (-0.38% to 12.4%)    		 17
    (NNT = 265 to infinity;
    NNH = 8 to infinity)
    DVT-elective orthopaedic unknown 232
    (53.2%)    		 160
    (37.5%)    		 30%
    (18% to 39%)    		 15.7%
    (9.18% to 22.3%)    		 6
    (4 to 11)
    DVT-high risk medical unknown 61
    (22.9%)    		 39
    (14.9%)    		 35%
    (6% to 55%)    		 7.99%
    (1.34% to 14.6%)    		 13
    (7 to 75)
    PE-all surgery unknown 121
    (2.72%)    		 44
    (0.99%)    		 64%
    (49% to 74%)    		 1.73%
    (1.17% to 2.29%)    		 58
    (44 to 86)
    PE-general surgery unknown 58
    (1.70%)    		 16
    (0.47%)    		 72%
    (52% to 84%)    		 1.23%
    (0.74% to 1.72%)    		 82
    (58 to 136)
    PE-traumatic orthopaedic unknown 34
    (6.88%)    		 14
    (2.78%)    		 60%
    (26% to 78%)    		 4.10%
    (1.45% to 6.76%)    		 24
    (15 to 69)
    PE-elective orthopaedic unknown 29
    (5.40%)    		 14
    (2.65%)    		 51%
    (8% to 74%)    		 2.75%
    (0.40% to 5.10%)    		 36
    (20 to 248)
    PE-high risk medical unknown 8
    (2.86%)    		 3
    (1.09%)    		 62%
    (-42% to 90%)    		 1.77%
    (-0.54% to 4.07%)    		 57
    (NNT = 185 to infinity;
    NNH = 25 to infinity)
    fatal bleed unknown 0
    (0.00%)    		 2
    (0.05%)    		 %
    (% to %)    		 -0.05%
    (-0.11% to 0.02%)    		 -2200
    (NNT = 931 to infinity;
    NNH = 5750 to infinity)
    non-fatal bleed requiring transfusion unknown 15
    (0.39%)    		 28
    (0.74%)    		 -87%
    (-250% to 0%)    		 -0.34%
    (-0.68% to -0.01%)    		 -291
    (-16000 to -150)
    bleed requiring re-operation, wound haematoma, infection due to bleed unknown 129
    (5.59%)    		 177
    (7.80%)    		 -39%
    (-74% to -12%)    		 -2.21%
    (-3.65% to -0.76%)    		 -45
    (-130 to -27)
    DVT- heparin vs heparin + aspirin unknown 45
    (18.1%)    		 41
    (16.3%)    		 10%
    (-32% to 39%)    		 1.80%
    (-4.80% to 8.41%)    		 55
    (NNT = 21 to infinity;
    NNH = 12 to infinity)
    PE- heparin vs heparin + aspirin unknown 11
    (1.68%)    		 4
    (0.61%)    		 64%
    (-13% to 88%)    		 1.07%
    (-0.08% to 2.23%)    		 93
    (NNT = 1200 to infinity;
    NNH = 45 to infinity)

    Comments

    1. Adding aspirin to heparin may provide a small additional benefit - but results are not clear cut.
    2. Dosing regimen of antiplatelet drug used appears unimportant.
    3. Definition of high risk medical patietns was not given. Looking at titles of papers selected, they include spinal cord injury, recurrent DVT/PE, post stroke and ?TIA, decompensated heart failure, post MI and ?unstable angina.
    4. Control rates from summed data taken to be the patient's expected event rate for occlusion- this was used to calculate NNT.
    5. The analysis recommended that prophylaxis is continued so long as the risk is substantial.

    Citation

    1. Antiplatelet Trialists' Collaboration , : Collaborative overview of randomised trials of antiplatelet therapy-III: reduction of venous thrombosis and pulmonary embolism by antiplatelet prophylaxis among surgical and medical patients. British Medical Journal 1994; 308 (6923): 235-246
    Contributor: Chris Ball and Clare Wotton, May 2000
    Reviewer: Alex Gallus

    Clinical Question.
    Patient undergoing surgery and 'high-risk' of VTE
    Intervention or Exposure antiplatelet therapy
    Comparison no antiplatelet therapy
    Outcome DVT or PE