Hi everyone, I have data on cancer deaths in the cohort of 316 oncological patients, that are under the influence of diagnostic radiation during the period of time since 1st diagnostic procedure to death or 10-year follow-up end.
The hypothesis to test is if the dose of radiation will impact on frequency of deaths (crd) during follow up.
Is it appropriate to use poisson regression with offset time variable, or binomial regression with the same condition?
If not, what regression model is of better choise?
Code:
. sum crd, d
death from cr
-------------------------------------------------------------
Percentiles Smallest
1% 0 0
5% 0 0
10% 0 0 Obs 347
25% 0 0 Sum of Wgt. 347
50% 1 Mean .6397695
Largest Std. Dev. .4807604
75% 1 1
90% 1 1 Variance .2311306
95% 1 1 Skewness -.5822912
99% 1 1 Kurtosis 1.339063
Is it appropriate to use poisson regression with offset time variable, or binomial regression with the same condition?
If not, what regression model is of better choise?
Comment