Pulmonary embolism: prophylaxis with heparin during surgery reduces fatal PE but increases excessive bleeding.

Clinical bottom line (level 1a)

  1. Patients undergoing surgery who receive prophylactic heparin are less likely to develop a deep vein thrombosis (NNT = 15 at unknown) or a pulmonary embolism (NNT = 83 at unknown) .
  2. Patients are less likely to die from a pulmonary embolism (NNT = 340 at unknown) .
  3. Patients are more likely to have an episode of excessive bleeding or require a transfusion of blood (NNH = 43 at unknown) .
  4. Death from haemorrahge is very rate (< 0.1%) - there is no clear increase when using heparin.
Collins et al: New England Journal of Medicine 1988; 318 (18): 1162-1173
Expires October 2003

The study

Systematic review of all randomised controlled trials of
  • Patients: undergoing general, orthopaedic or urologic surgery
  • Intervention: subcutaneous heparin compared with placebo or no therapy
  • Outcome: deep vein thrombosis (detected by fibrinogen scan), pulmonary embolism, bleeding, death


    • Articles found in ?all languages using Medline, ?1966 to 1988 (search terms: not given ) and scanning reference lists of retrieved articles and reviews, hand-searching journals and abstracts, and contactin colleagues, investigators and manufacturers to identify missed studies.

    Selection criteria: see above
    Appraisal criteria: not given. Authors were contacted for missing information.
    Articles excluded if:
    • quasi-randomised
    • compared different regimens of heparin


    74 studies found involving 15598 patients
    Tests for heterogeneity were not performed.

    The evidence

    Outcome Time to outcome CER OR
    (95% CI)    		 NNT
    (95% CI)
    deep vein thrombosis unknown 1107/4144
    (26.7%)    		 0.68
    (0.65 to 0.71)     		15
    (13 to 16)
    non-fatal pulmonary embolism unknown 147/7239
    (2.03%)    		 0.40
    (0.29 to 0.51)     		83
    (70 to 100)
    fatal pulmonary embolism unknown 56/6777
    (0.826%)    		 0.64
    (0.49 to 0.80)     		340
    (240 to 610)
    episodes of excessive bleeding or need for transfusion unknown 244/6504
    (3.75%)    		 1.66
    (1.55 to 1.77)     		-43
    (-51 to -37)

    Comments

    1. The use of I-125 scanning especially in the hip surgery trials (with limited senistivity) may underestimate the true benefit of perioperative heparins.
    2. Using heparin had no clear effect on the rate of fatal haemorrhage. 0.08% of patients died from bleeding.
    3. Patients undergoing each of the different types of surgery had similar degrees of benefit.

    Citation

    1. Collins R, Scrimgeour A, Yusuf S, et al: reduction in fatal pulmonary embolism and venous thrombosis in perioperative admininstration of subcutaneous heparin: overview of results of randomized trials in general, orthopedic and urologic surgery. New England Journal of Medicine 1988; 318 (18): 1162-1173
    Search Terms: ?
    Contributor: Chris Ball and Clare Wotton, October 1999
    Reviewer: Alex Gallus

    Clinical Question.
    Patient orthopaedic, urologic, general surgery
    Intervention or Exposure heparin
    Comparison no heparin
    Outcome deep vein thrombosis (DVT), pulmonary embolism (PE), death from PE, bleeding, death from bleeding