At our critical appraisal seminars for GPs (held in Suffolk) which looked at
the UKPDS 38 and HOT trials, we became aware that there is a source of
confusion when trials report events in patient-years. Two participants were
debating at cross purposes because they used different methods of
calculating NNTs with divergent results.
The event rates in such trials could be worked out in at least two different
ways and they give two different answers. You could calculate the event rate
from the events per patient-years as stated in the reportıs tables (method
1) OR you could calculate it from the number of events divided by the number
in the grou over the median or mean follow up (method 2). Method 1, I would
say, is the more accurate of the two. This confusion is not confined to our
group! Examples can be seen elsewhere, for example, Atle Klovningıs
calculations for the HOT trial use method 2 (which can be viewed at
http://www.uib.no/isf/people/atle/toohot.htm).
A fuller explanation of why these methods differ can be viewed at:
http://www.suffolk-maag.ac.uk/ebm/pt-yrs&NNTs.html
but please - before you shoot me down, remember I am not a mathematician,
just a doctor trying to make sense of it all!
Contrasts of the two methods when applied to the UKPDS 38 and HOT trials can
be viewed at:
http://www.suffolk-maag.ac.uk/ebm/cpukpds.html
http://www.suffolk-maag.ac.uk/ebm/cphot.html
Kev Hopayian, GP, Suffolk England
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|