This query comes from a colleague, please send replies to
her directly at [log in to unmask]
I have just been forwarded a paper dealing with diagnostic
tests. In the results section they have calculated the
positive and negative predicted values (ppv and npv) for a
range of prevalences other than observed. However they have
also put CIs around the ppv and npv.
It is not clear to me what method should be used. The paper
refers to Fleiss's rates and proportions book but this does
not have any mention of this particular CI.
----------------------
Jim Jeffs
[log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|