Skip navigation

Thromboprophylaxis and death after total hip replacement


Bandolier has visited the issue of thromboprophylaxis or compression stockings after surgery before ( Bandolier 17 , 18 ). A new meta-analysis [1] asks and answers the question of what is the risk of fatal pulmonary embolism and total mortality after total hip replacement (THR) and to what extent do different methods of thromboprophylaxis affect mortality?

Background

Studies conducted in the 1960s found an incidence of fatal pulmonary embolism after THR of approximately 1%, and authoritative sources currently recommend thromboprophyl-axis for patients undergoing THR. Many deep venous thromboses (DVTs) detected in trials are clinically asymptomatic, and DVTs are used as surrogate end-points because they can be measured and because they occur commonly (40% to 80% of cases). Many studies have shown a venographic reduction in the risk of DVT with various prophylactic agents (heparin, warfarin, aspirin, etc.). The assumption inherent in these investigations is that DVT sometimes causes pulmonary embolism which, at times, results in death.

Meta-analysis

A MEDLINE search (from 1966-95) was conducted to find all the relevant literature that discussed fatal pulmonary embolism and total mortality in patients receiving THR. It was not restricted to prospective randomised controlled trials, but included all studies of THR in which the number of patients being studied and total number of deaths or fatal pulmonary embolisms were reported.

Types of thromboprophylaxis were broken down into six categories:

  1. None (including patients who received placebo, compression stockings, or no prophylaxis)
  2. Heparin (including low molecular weight and unfractionated),
  3. Warfarin,
  4. Aspirin,
  5. Dextran,
  6. Other (usually a combination of medications)

Patients receiving active prophylaxis may have also worn compression stockings. The results of this heterogeneous mix of studies were combined with appropriate meta-analytic techniques.

Outcomes

The primary outcomes were the rates of fatal pulmonary embolism and total mortality after total hip replacement.

Results

About 93,000 patients in 181 papers were analysed. Over the last 15 years, the rate of fatal pulmonary embolism was 0.11% (95%CI 0.07% to 0.16%) and total mortality was 0.38% (95%CI 0.29% to 0.47%). These rates had declined dramatically from 0.64% for fatal pulmonary embolism and 1.1% for overall mortality in the 1960s, probably because of better techniques.

No differences emerged in the efficacy of various thromboprophylaxis regimens preventing fatal pulmonary embolism. This was primarily due to the very low incidence of this outcome. For example, only 4 fatal pulmonary embolisms were reported in 3432 patients (0.12%) who did not receive prophylaxis and only 8 fatal pulmonary embolisms were reported in 10,356 patients (0.08%) who received heparin. Overall mortality did not differ between any of the prophylaxis regimens or when they were compared with no prophylaxis.

Comment


This meta-analysis challenges the current dogma regarding thromboprophylaxis in patients receiving THR. The summary suggests that because there is not enough evidence in the literature to conclude that any form of prophylaxis decreases the death rate after THR, "guidelines which recommend their routine use to prevent death after hip replacement are not justified".

This is strong stuff. The paper indicates that thromboprophylaxis is unlikely to reduce the rate of fatal PE after THR by more than 0.05% (that is, 5 in 10,000). If the death rate from complications of thromboprophylaxis (bleeding, for instance) is greater than 5 in 10,000, then thromboprophylaxis may cause more harm than benefit. The risk of clinically significant thrombotic events caused by heparin-induced thrombocytopenia is about 3% in patients receiving unfractionated heparin [2].

It would be helpful to quantify the harm resulting from thromboprophylaxis (probably though new systematic reviews looking specifically at harm), because randomised trials of sufficient size to measure the low rates of death and benefit would be huge, expensive, and take a long time.

References:


  1. DW Murray, AR Britton, CJK Bulstrode. Thromboprophylaxis and death after total hip replacement. Journal of Bone Joint Surgery (Br)1996 76-B: 863-70.
  2. TE Warkentin, MN Levine, J Hirsh, et al. Heparin-induced thrombocytopenia in patients treated with low-molecular weight heparin or unfractionated heparin. New England Journal of Medicine 1995 332: 1330-5.



previous or next story in this issue