Skip navigation

Low molecular weight heparin and knee replacement

Deep venous thrombosis (DVT) is a common postoperative complication of knee replacement, and is a common cause of re-admission. It can occur in the lower leg, or proximally in the popliteal vein. Proximal vein thrombus formation is associated with a small but finite risk of pulmonary embolism, and a small proportion of those affected will die.

Trying to prevent thrombus formation is often done prophylactically, and agents used may include warfarin or heparin. Low molecular weight heparin (LMW heparin) has been around for more than a decade now, and has theoretical advantages because of longer half-life and because it can be used at fixed doses subcutaneously without laboratory monitoring.

The question of what additional benefit, if any, obtains from the use of LMW heparin is answered by a meta-analysis of randomised trials [1].


The search was comprehensive, and sought randomised trials in knee replacement of thromboprophylaxis with LMW heparin compared with placebo or active control in which routine screening for DVT was done.


Ten studies were found; eight used an active pharmacological control, either standard heparin or warfarin. The other two used placebo. Four different makes of LMW heparin were used and the doses are given in the paper; for enoxaparin in five trials it was 30 mg twice daily or 40 mg once daily. The duration of the studies was 8 to 14 days or until discharge, and venograms were used to screen for DVTs in all cases.

  • A proximal DVT occurred in 93/2039 (5%) patients given LMW heparin and in 166/1884 (9%) of patients given control (Figure). The relative risk was 0.52 (0.41 to 0.67) and the number needed to treat to prevent one proximal DVT was 24 (17 to 37).
  • In nine trials for which there was information, a pulmonary embolus occurred in 4/1974 patients given LMW heparin and 9/1823 given control.
  • In nine trials for which there was information, a bleeding complication occurred in 58/1863 (3.1%) of patients given LMW heparin and 45/1708 (2.6%) of those given control.
  • There were no deaths with LMW heparin in six trials, and three with control.


Argument might still rage about the use and possibly abuse of LMW heparin, but this paper will be useful in helping to resolve the issues. Be careful while reading it, as Bandolier found a few minor errors in some of the tables in the calculation of relative risk for individual trials. We used the total of patients randomised to give an intention to treat outcome for relative risk and NNT. The paper appears to use evaluated patients rather than all patients randomised.

The paper also shows us in sensitivity analyses that the type of control is not important, and that papers of poor reporting quality may give higher estimates of treatment effect (though it doesn't actually give the quality scores).

That aside, this paper could easily be used to do some 'back-of-stamp' health economics to evaluate whether the additional costs of LMW heparin repay themselves in other ways.


  1. AW Howard, SD Aaron. Low molecular weight heparin decreases proximal and distal deep venous thrombosis following total knee arthroplasty. Thromb Haemostat 1998 79: 902-6.

previous or next story in this issue