Since 1999 the routine cflpois.ado has been available from SSC archives
(Stata Source code libraries) for making tables like:
cflpois inj pyrs , by(hrs_wk)
hrs_wk N inj pyrs rate l a h
1-13 234 47 54433 86.3 62.9 - 115.7
14-35 136 38 151370 25.1 17.9 - 34.1
36+ 298 178 729396 24.4 21.0 - 28.3
t N inj pyrs rate l a h
Total 668 263 935199 28.1 25.0 - 31.9
That is a stratified (conditional) table of rates which shows confidence
limits, personyears and number of observations.
In the very first version of that routine I used the values for expected N
shown in reference:
Haenszel W, Loveland D, Sirken MG. Lung cancer mortality as related to
residence
and smoking histories. J Natl Cancer Inst 1962; 28: 947-1001
reprinted in: Breslow NE, Day NE: Statistical Methods in cancer Research.
Lyon,
IARC, 1987. P. 70.
But I forgot to change the help file and the !* lines of the routine to
tell that the Stata routine cii actually made the confidence intervals.
I hope this has not caused problems to anyone.
The reference for cflpois is cii routine not the mentioned one above. An
update of the routine has been sent to the SSC-library, where the correct
reference is stated. You can access the routine with "findit cflpois" from
any Stata running with connection to internet. Wait a few days if you want
the corrected help file. Estimates are still the same computed with cii.
(I have not checked if one can now do the stratification with a direct
command in Stata7/8)
Apologies and regards
Jens M.Lauritsen, Odense Denmark
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
