Now that we have an NNT in terms of preventing asymptomatic DVT, and an NNH in terms of causing serious bleeds, we can start to consider whether prophylactic anti-coagulants are worthwhile in this group of people.
The first thing we have to establish is how important the benefit is, and how serious the harms are.
Consider the NNT and NNH we have calculated in the light of the following information from the paper:
On the benefits
We are given very little information in the paper about the clinical importance of asymptomatic DVT. We do know that it does not cause much discomfort, as it is asymptomatic. But what are the serious consequences, and how common are they? This is the only useful bit of information we are given on this subject:
"...in the outpatient setting proximal [asymptomatic] DVT was found in only 3% of patients with a low suspicion for DVT."
This does not tell us much, but it does suggest that DVT is fairly common, affecting as many as 1 in 30 outpatients. But personal experience suggests that relatively few people are dying suddenly of problems relating to an undiagnosed DVT. Serious consequences may be quite rare, with little to no consequence otherwise. The NNT should perhaps be adjusted to account for the rarity of serious side effects. By what factor? Without more information, your guess is as good as ours.
On the harms
There was no difference in overall mortality, so the only major harm to consider is major bleeding. This is defined by the authors as:
"major bleeding defined as bleeding requiring transfusion of two or more units of packed red blood cells, involving a critical site (eg retroperotineal), or that was fatal"
This definition excludes more "minor" bleeds, but if you had required a blood transfusion following, say, a nose-bleed you might be entitled to consider it a pretty major harm from treatment. This suggests that there might be some additional harms from treatment that are not reflected in the NNH. How large these are is another guess, in the absence of more information.
Given this information, do you think prophylaxis is worthwhile?
THE ANSWER...
The answer to this question depends a great deal on how heavily you weight the harms compared to the benefits. It is clear that a simple NNT substantially overstates the true benefit and that a simple NNH understates the true harms. It would take some additional research to establish the appropriate weight to give each number.
Click the "up" link below to return to the index of the Clinical Significance section, or click here to move on to Interpreting meta-analyses...