|
|
Jon
NNT are only one piece of the therapeutic puzzle. They are an either/or
end point. Most therapeutic agents improve significant numbers of people
but not necessarily by the NNT end point amount ie walking 100 yards
might be the NNT threshold but many people might increase their walking
by 50 yards on the new treatment. I think hard endpoints conceal a lot
of therapeutic effect but are much easier for the statistical
interpretation of the study.
So there is possibly an equation of NNT *(1/proportion of people with
any benefit) (see excel).
It doesn't help define the uncertainty as we don't know who will get
benefit and by how much until we try the drug. That's why n of 1 are in
theory so important but in reality nowhere to be seen
Martin
|
|
|
|