Intended for healthcare professionals


Prophylaxis for venous thromboembolism during treatment for cancer: questionnaire survey

BMJ 2003; 327 doi: (Published 11 September 2003) Cite this as: BMJ 2003;327:597
  1. C C Kirwan, surgical research fellow1,
  2. E Nath, research assistant1,
  3. G J Byrne, senior lecturer in surgical oncology (gedbyrne{at},
  4. C N McCollum, professor of surgery1
  1. 1 Education and Research Building, Wythenshawe Hospital, Manchester M23 9LT
  1. Correspondence to: G J Byrne
  • Accepted 17 July 2003


Venous thromboembolism is common in patients with cancer and is often the cause of death.1 Patients receiving treatment for cancer are at even greater risk of thromboembolism. Thromboembolism occurs in 5% of patients receiving chemotherapy for early breast carcinoma,2 and up to 17.6% of patients receiving chemotherapy for metastatic breast disease are affected.3 Patients with node-negative breast cancer taking tamoxifen were six times more likely to develop venous thromboembolism.4

Adjuvant use of tamoxifen carries a relative risk of 1.22 compared with no treatment. Combining methods of treatment further increases the risk of thromboembolism. Chemotherapy with tamoxifen increases risk by 3.5 times compared with chemotherapy alone,5 and preoperative radiotherapy for rectal carcinoma doubles the postoperative risk of venous thrombosis.5 Low doses (1 mg) of warfarin throughout chemotherapy for metastatic breast cancer are associated with a relative risk reduction of 85% with no increase in serious bleeding complications.1

Participants, methods, and results

We sent a postal questionnaire to all oncologists in northern England, identified by internet search and in the Medical Directory 2002. We used a scoring system to establish specialty, main type of cancer treated, main method of treatment (chemotherapy, hormone therapy, or radiotherapy), and current prophylaxis practice and estimate of risk of venous thrombembolism.

Of the 123 responses to the 166 questionnaires we sent, 106 (64%) were acceptably completed. Half the oncologists (56) specialised in clinical oncology, 31 in medical oncology, seven in surgery, five in gynaecological oncology, five in paediatrics, one in urology, and one in radiology. We have no information about the specialties of oncologists who did not respond. The most common treatment was chemotherapy, used by 41 (39%) oncologists; 10 (9%) used hormone therapy and 44 (42%) used radiotherapy. The oncologists treated many types of tumour.

A total of 29 (27.4%; 95% confidence interval 19.8% to 36.5%) oncologists thought their patients were not at risk of venous thromboembolism. This response was independent of the type of tumour treated. Considering different methods of treatment individually, 71 oncologists believed that hormone therapy posed little or no increased risk to patients; 83 thought the same for chemotherapy and 96 for radiotherapy.

Of the 106 oncologists, 84, 79, and 86 reported not using prophylaxis routinely in chemotherapy, hormone therapy, and radiotherapy. As risk factors, nine of the 10 oncologists who used prophylaxis in chemotherapy mentioned central venous lines and six of the 11 who used prophylaxis in hormone therapy mentioned stilboestrel. Preferred prophylaxis was aspirin (13), warfarin (12), subcutaneous heparin (10), stockings (9), and randomising patients with central venous lines to different warfarin regimens (3). Many factors cause oncologists to prescribe prophylaxis (table). A total of 19 (17.9%; 11.8 to 26.3%) oncologists never used prophylaxis for venous thromboembolism.

Factors that cause 106 oncologists in northern England to initiate use of prophylaxis*

View this table:

Of those who answered, 37% of oncologists estimated that less than 1% of their patients werecurrently using prophylaxis for venous thromboembolism and 62% estimated less than 5%; 16% did not answer this question.


More than a quarter of oncologists do not recognise the thrombogenic effects of treatments for cancer, and thromboprophylaxis is rarely used in patients undergoing treatment for cancer. Oncologists estimated that a surprisingly low percentage of patients were using prophylaxis bearing in mind that half of respondents mentioned previous venous thromboembolism and immobility as indications for routine prophylaxis. The response of a third ofoncologists that venous thromboembolism does not pose a risk is not biased by lower risk specialties such as paediatrics. The good response rate to our questionnaire demonstratesa reliable representation of current practice in the north of England. National guidelines on prophylaxis for venous thromboembolism during cancer treatment are needed.


  • Contributors CCK designed the questionnaire and analysed the responses. EN collected data and identified oncologists. GJB and NMMcC helped to design the questionnaire and write the paper. CCK is guarantor.

  • Funding No additional funding.

  • Competing interests None declared.

  • Ethical approval Not needed.


View Abstract