I dont have the original dat in front of me, but could it be the duration of
followup, e.g was 10 years the mean follow up? The NNT might have been based
on a survival curve analysis that would be different than using the total
number of events and the mean followup, since events occur at different
periods over the followup.
-----Original Message-----
From: [log in to unmask] [mailto:[log in to unmask]]
Sent: Monday, October 05, 1998 5:26 PM
To: Henry Barry; 'Evidence List'
Subject: RE: UKPDS
I'm having five of the residents in our program critically appraise the
articles recently published from this study, and so far, we found a similar
problem in the article on atenolol/ACE's in control of hypertension (in BMJ,
i don't have the reference handy)...the resident was doing the calculations
right and I even ran the numbers through CATmaker and got different NNT's
than the authors reported...curious...
> -----Original Message-----
> From: [log in to unmask]
> [mailto:[log in to unmask]]On Behalf Of Henry
> Barry
> Sent: Monday, October 05, 1998 10:37 AM
> To: 'Evidence List'
> Subject: UKPDS
>
>
> I have been muddling through the UKPDS 33 (Lancet 1998; 352:837-53). I
> am impressed with the magnitude of this undertaking. I suspect there
> will be much discussion about its findings and how they are applied.
> The main results are summarized on page 843 (Figure 4).
>
> When I calculated NNT (based upon aggregate of 963 endpoints/2729
> subjects in the intensive control group and 438 endpoints/1138 in the
> conventional therapy group, and 10 years of follow up), I come up with
> an absolute risk reduction of .032 and NNT = 31.2 for 10 years.
> However, in the text on page 847, the authors report NNT of 19.6. Any
> insights into why my NNT is different from theirs?
>
> Henry
>
>
> Henry C. Barry, M.D., M.S.
> Associate Professor
> Senior Associate Chair
> Department of Family Practice
> B-104 Clinical Center
> Michigan State University
> East Lansing, Michigan 48824-1315
> Phone: 517-353-0851 x 456
> Fax: 517-355-7700
> E-mail: [log in to unmask]
> [log in to unmask]
> <mailto:[log in to unmask]>
>
>
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|