Intended for healthcare professionals


Prophylaxis rates for venous thromboembolism and gastrointestinal bleeding in general medical patients: too low or too high?

BMJ 2012; 344 doi: (Published 18 May 2012) Cite this as: BMJ 2012;344:e3248
  1. Shoshana J Herzig, instructor in medicine12,
  2. Michael B Rothberg, associate professor of medicine34
  1. 1Division of General Medicine and Primary Care, Beth Israel Deaconess Medical Center, Boston, MA, USA
  2. 2Harvard Medical School, Boston
  3. 3Division of General Medicine, Baystate Medical Center, Springfield, MA, USA
  4. 4Tufts University School of Medicine, Boston
  1. Correspondence to: S J Herzig sherzig{at}
  • Accepted 28 February 2012

Recommendations and general conceptions surrounding venous thromboembolism and gastrointestinal bleeding in medical patients are divergent, despite similar efficacy and risks. Shoshana Herzig and Michael Rothberg examine how industry influence and research bias have contributed to the difference

Imagine a hypothetical condition that affects about 1 in 200 hospitalised general medical patients and that can occasionally be fatal. Imagine that the drug used to treat this condition could also be used for prophylaxis, and that prophylaxis is inexpensive but does have side effects. If prophylaxis halves the incidence of the condition but does not reduce mortality, and 30% of medical patients receive such prophylaxis, would that rate be too low or too high?

Curiously, the answer seems to depend on the type of prophylaxis. The estimates of incidence, severity, and efficacy of prophylaxis for venous thromboembolism and gastrointestinal bleeding are almost identical. Yet, despite similar rates of use, in the case of venous thromboembolism, the predominant message derived from published studies and guidelines is that doctors don’t do enough prophylaxis, while for gastrointestinal bleeding they do too much. Why this dichotomy?

Similar numbers, divergent recommendations

Clinically serious venous thromboembolism and gastrointestinal bleeding both occur in less than 0.5% of hospitalised general medical patients.1 2 Preventive treatment of these two conditions—with heparin or acid suppressive drugs—has similar efficacy, with a 40-60% reduction in incidence for both.2 3 4 5 Notably, neither treatment has been shown to reduce mortality.3 4 5 Studies investigating rates of use find that 30-50% of medical patients receive venous thromboembolism prophylaxis and 20-30% receive acid suppressive drugs to prevent stress ulcer.1 6 7 8

Given the similarity of the figures for these two types of prophylaxis, it is difficult to understand why in the case of venous thromboembolism, the focus of published studies has been on increasing use …

View Full Text

Log in

Log in through your institution


* For online subscription