Statalist The Stata Listserver

[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: time varying variable in Cox regression

From   "Svend Juul" <SJ@SOCI.AU.DK>
To   <>
Subject   Re: st: time varying variable in Cox regression
Date   Wed, 28 Mar 2007 14:58:20 +0200

Raoul continued ...

Could I also set the period between the first birth and the third to
missing? So I would know that up to the first birth (for which the dob
is known) the subject didn't have any children. Then set the period
between the first and the third birth to missing because I don't know
whether they had one or two children in that period
and then when the third child is born set it to 3 (so I would stsplit

period up to first birth: child_num=0
period between first birth and third child_num=.
period after third birth: child_num=3

By doing this the subject will at least contribute some person-years
up to the first birth and after the third birth. I guess I could also
add 9 months to the first birth, because I'm pretty sure that the
subject didn't give birth in that period. However, I 'm not sure
whether the above method is ok (though It would make it more complex).


If there is uncertainty not only about the timing of births, but also
about the number of children, I think the problem is much worse. The
latency period for many cancers is counted in decades, and an 
imprecision of a couple of years concerning the dates of birth is
hardly very important in this context. But I don't have any idea about
how to handle an uncertainty about the number of children.



Svend Juul
Institut for Folkesundhed, Afdeling for Epidemiologi
(Institute of Public Health, Department of Epidemiology)
Vennelyst Boulevard 6
DK-8000  Aarhus C, Denmark
Phone: +45 8942 6090
Home:  +45 8693 7796

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index