Objective: To review the published clinical data on prophylaxis for thrombo
embolism in order to develop general guidelines to encourage the establishm
ent of local protocols for management.
Data sources: Published papers on thromboembolism over the period 1991-1997
were identified by Medline search and/or from the authors' personal litera
ture collections and reviewed.
Study selection: A total of 981 studies were identified. Only those papers
reporting randomized studies with clearly defined diagnostic methods and cl
ear end-points were included in this review.
Data extraction: The available evidence for each specialty was summarized a
nd reviewed by the authors responsible for each specialty, prior to present
ation and discussion of their findings within the group. Where a consensus
opinion was achieved in a speciality, general guidelines for thromboprophyl
axis were summarized. Where a consensus could not be agreed, recommendation
s for further work were made.
Data synthesis: There is evidence to support the preferred use of low-molec
ular-weight heparins (LMWHs) over unfractionated heparin (UFH) in orthopaed
ic surgery, major trauma and general surgery. However, the ideal duration o
f thromboprophylaxis has yet to be defined. The use of once daily subcutane
ous administration of LMWH offers major practical advantages and may have s
ignificant cost saving implications. Further work is required to investigat
e the use of thromboprophylaxis in minimal access surgery, trauma, elective
lower limb surgery, hip fracture and pregnancy; to compare the efficacy of
LMWH and mechanical prophylaxis; and to investigate extended prophylaxis a
fter discharge.
Conclusions: There is overwhelming evidence that thromboembolic prophylaxis
reduces the incidence of postoperative deep vein thrombosis and pulmonary
embolism. Recommendations concerning the manage ment of these patients when
stratified into low, moderate and high risk are made with the suggestion t
hat hospitals develop their own guidelines for the treatment of these patie
nts.