Dear Jason,
at a very first blush your procedure seems reasonable, provided that your
choice is to adopt a non parametric approach (such as KM) in survival
analysis (SA).
By the way, I have two remarks about your research approach:
What if a subject die during the observation time for whatever reason
different from the injury itself? Do you consider it a right-censored
observation?
Besides, is there any way to reduce the number of injuried workers as a
result of a "on the job safety training" programme (ie, is a turning point
detectable)? Should the answer be positive, you would be probably more
comfortable with a semiparamentric Cox model (please see for a detailed
example and many other interesting stuff about SA topics, Cleves MA, Gould
WW, Gutierrez RG: An Introduction to Survival Analysis using Stata. Revised
edition. College Station: Stata Press, 2004: 160-162).
Kind Regards and Alkl the Best for the just started New Year,
Carlo
-----Messaggio originale-----
Da: [email protected]
[mailto:[email protected]] Per conto di [email protected]
Inviato: venerdì 2 gennaio 2009 5.50
A: [email protected]
Oggetto: st: Survival Analysis Comparing Younger and Older
I want to compare time to first injury for younger people vs. older
people in a specific occupation.
So far I have set up my data as follows:
stset smonth, failure(injury) exit(event==2 3) enter(event==1) id(empid)
smonth: the number of months worked
injury: 0-not injured, 1-injured
event: 1-first day on job, 2-injured, 3-last day on job
I have used Kaplan-Meier survival plots with the by() option to
compare older and younger worker's time to first injury.
Does this sound reasonable? Are there any other methods I could try?
*
* 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/
*
* 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/
