If I understood your question correctly, you are wondering if you
can calculate NNT from survival curves that are crossing each other. I don't
believe that you should, but then again there are more competent people at
this discussion group to answer your question.
ben
Benjamin Djulbegovic, MD
Associate Professor of Medicine
H. Lee Moffitt Cancer Center & Research Institute
at the University of South Florida
Division of Blood and Bone Marrow Transplant
12902 Magnolia Drive
Tampa, FL 33612
e-mail:[log in to unmask]
http://www.hsc.usf.edu/~bdjulbeg/
phone:(813)979-7202
fax:(813)979-3071
> -----Original Message-----
> From: Jeong H. Oh, MD MPH [SMTP:[log in to unmask]]
> Sent: Thursday, January 28, 1999 5:57 PM
> To: Djulbegovic, Benjamin
> Cc: 'Dave Sackett'; 'Evidence-based discussion group'; 'Iztok Hozo'
> Subject: Re: NNT or NNH
>
> I have a question:
>
> For the past years we have started to use in our studies statistical
> methods that try to adjust for differences between experimental/control
> groups, etc.
>
> Most of those studies are reported in RR or OR from logistic regression.
>
> Can you use the formula for NNT based on OR for systematic reviews in
> these therapy studies?
>
> I find it very distressing to use NNT based on "spot" outcomes. I have
> seen several studies where there is a crossover of outcomes just 1 year
> before end of the study according to the Kaplan-Meyer graphs, where
> despite NNT I would much prefer to give the "other drug" because of
> early survival and late crossover.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|