I have found that in the epitab section of Stata ,when calculating
confidence intervals for risk ratio or odds ratio using the exact (or
Cornefield) method,
if one of the cells in the 2x2 table is 1 or 2 ,the upper limit is
infinite or cannot be calculated. Is this a numerical problem or there
is another explanation?
Basilio de Braganca Pereira
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|