At 08.55 03/09/98 -0500 Rosanne Leipzig wrote:
> Can anyone point me to a way to calculate NNTs, and especially
> confidence intervals, for pooled data? It seems to me the NNT is
> straight forward, but the CIs may need something special. Thanks
The new version of CATmaker (July 1998) - the resourse specifically
developed for critical appraisal from Centre for Evidence-based Medicine -
has a module for systematic review that calculates NNTs and CIs from OR of
metanalyses.
You can require a free copy of CATmaker to [log in to unmask]
I hope you find it useful
Bests
------------------------------------------------------------------
Gruppo Italiano per la Medicina Basata sulle Evidenze
------------------------------------------------------------------
Antonino Cartabellotta, MD Phone:+39 91 407153
Head of GIMBE Fax: +39 91 407153
Passaggio L. da Vinci, 16 e-mail:[log in to unmask]
90145 - Palermo (Italy) http://www.gimbe.org
------------------------------------------------------------------
G I M B E
------------------------------------------------------------------
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|