While I agree with Dr. Sigurdsson’s
findings, I just want to comment on one point he makes – and
yes I am not a statistician, so acknowledge in advance the
potential for being wrong. Dr. Sigurdsson states “I guess
anybody with some statistical knowledge will see that the
Number Needed to Treat (NNT) for one year should be a higher
number than the NNT for 3.5 years”. Most of the time
this may be correct, but I question if it can be a rule of
thumb.
NNTs are derived from absolute risk
difference (ARDs). ARD is a function of control event rate and
relative risk. If the control event rate varies substantially
by time, then so also would NNTs. And if sicker patients are
liklier to experience adverse events earlier, then NNTs would
be lower at the start than later in follow-up. To add to the
complexity, even relative risk may not be constant if the
composition of the population changes dramatically overtime
because sicker patients are dying earlier and less susceptible
ones are surviving.
However, the way NNTs were actually
calculated in the JACC paper, Dr. Sigurdusson is right even
when his principle may not be.
-interesting practical problem with
statistics
Confidentiality
Statement - The contents of this e-mail, including its
attachment, are intended for the exclusive use of the
recipient and may contain confidential or privileged
information. If
you are not the intended recipient, you are strictly
prohibited from reading, using, disclosing, copying, or
distributing this e-mail or any of its contents. If you received this
e-mail in error, please notify the sender by reply e-mail
immediately or the Privacy Office
([log in to unmask]) and permanently delete this
e-mail and its attachments, along with any copies thereof. Thank you.
Avis
de confidentialit� � Ce courriel, y compris ses pi�ces
jointes, s�adresse au destinataire uniquement et
pourrait contenir des renseignements confidentiels. Si
vous n��tes pas le bon destinataire, il est strictement
interdit de lire, d�utiliser, de divulguer, de copier ou
de diffuser ce courriel ou son contenu, en partie ou en
entier. Si vous avez re�u ce courriel par erreur,
veuillez en informer imm�diatement l�exp�diteur ou le
bureau de la Protection des renseignements personnels
([log in to unmask]), puis effacez le
courriel ainsi que les pi�ces jointes et toute autre
copie. Merci.
