st: Problem with confidence intervals in stptime when using weights

["Dr. Till Ittermann" <Itterman@uni-greifswald.de>]

Dear statalist,

I want to calculate age- and sex-specific incidence rates using 
pweights. In some of the age-and sex-specific subgroups incidence rates 
are very low. For these subgroups confidence intervals are not correctly 
calculated. This only happens when analysis are weighted. If no weights 
are used confidence intervals are correctly specified.

Example:

stset ftime [pweight = fweights], failure(diab_1=1) id(zz_nr)
bysort sex_0: stptime if diab_0 == 0, by(ag_kat) per(1000),

which will result in an output:

    ag_kat         |  person-time   failures        rate            
[95% Conf. Interval]
-----------+---------------------------------------------------------------------------------------------
 20 - 29 years |    461611.79      528.7   1.1453347   .4140403    .4362638

I don't understand what is going wrong here. I tried several options of 
stptime incl. the jacknife option or enlarging calculations to 10000 
person years, but nothing helped. Has anyone an idea how to solve this 
problem? I am using stata 12.1 with the latest updates.

Best,

Till





*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/