Skip navigation

Graduated compression stockings to prevent postoperative venous thromboembolism

Postoperative venous thromboembolism can be prevented by methods that reduce the coagulability of blood or by methods that reduce blood stasis in the deep veins of leg. Graduated compression stockings reduce stasis, are a simple prophylactic approach, are inexpensive, easy to use and free of side effects.

Graduated compression stockings may be used alone or in combination with other methods of prophylaxis. They may be used in circumstances where the risk of venous thromboembolism is relatively low (as with abdominal surgery) or where it is high (after hip replacement).

The question is whether they are effective. A recent meta-analysis has gone a long way towards proving the point.


A group from Hamilton, Ontario, has performed a first class meta-analysis on published papers. They had a prior set of criteria for inclusion or exclusion based on methodology and diagnosis of thromboembolism. The inclusion criteria included:-
  • Only randomised trials with proper randomisation
  • Reliable objective tests of proven accuracy for the diagnosis of postoperative deep vein thrombosis
  • Independent and blind interpretation of venography (when this was the outcome measure) by observers without knowledge of the patients' assignment or symptoms, and predefined criteria for an abnormal result.


The searching strategy yielded 122 articles, of which only 35 were randomised trials. After further exclusions because of methodological inadequacy, the result was 12 trials for analysis, one in orthopaedic surgery and the others in moderate risk non-orthopaedic (mainly abdominal) surgery.

In the 11 moderate risk studies there were 1752 patients - some of whom were their own controls in studies where legs rather than patients were randomised.

In the control group without compression stockings 930 legs suffered 164 episodes of venous thromboembolism. In the group treated with graduated compression stockings, 932 legs had 58 episodes of venous thromboembolism. The pooled odds ratio was 0.28 (95% CI 0.23 - 0.42, p<0.001), with a risk reduction of 68% (95% CI 53 - 73).


The NNT was 9 (95% CI 7 - 13). If 100 postoperative patients are treated prophylactically with graduated compression stockings, 82% will not have a venous thrombosis anyway, 6 will have a thrombosis despite the treatment, but 12 will have had an episode of venous thrombosis prevented.


Surveys have shown that the use of graduated compression stockings varies greatly from less than 3% to over 70%. There is confusion over whether they are effective. Clearly, in moderate risk patients graduated compression stockings are highly effective, and given that they are cheap, easy to use, and have negligible adverse effects their use should be promoted widely in these patients.

In patients at high risk of venous thromboembolism, those undergoing hip replacement, for instance, there is insufficient evidence of their effectiveness either alone or in combination with anticoagulant prophylaxis. The single RCT identified in this paper where graduated compression stockings were used alone in orthopaedic patients did not show a significant reduction in episodes of venous thromboembolism.


  1. PS Wells, AWA Lensing, J Hirsh. Graduated compression stockings in the prevention of postoperative venous thromboembolism. Archives of Internal Medicine 1994 154: 67-72.

previous or next story in this issue