I recently had to produce a detailed report of incidence rates for many
variables and by groups. Stata's -ci. had a number of problems regarding
incomplete displays and poor formatting. So taking Nick Cox's usual advice,
I created a new program - really a wrapper for Stata's ci.
It displays the number of cases (Stata's ci does not), produces output in
the desired number of person years, allow well-formatted output for each
level of a by group, allows replacement of dependent variable and
grouping variable names with descriptions, allows formatting of numerical
output and allows for fine tuning of the display. Here's an example.
