Dear All,
My colleagues and I were wondering if anyone could shed some light on
to the differences between the trend tests available in Stata and their
advantages and disadvantages. We understand that in Stata there are
several commands e.g. nptrend, ptrend, vwls, mhodds and tabodds that
provide tests for trend, but we are unsure as to which one we should be
using as they all seem to give different answers. We are interested in
using these tests for investigating whether there is a trend in
prevalence of toxicity grades 0-4 between treatment groups.
Many thanks
Lucy
Lucy Porter, MSc
Statistician
Institute of Cancer Research
Clinical Trials & Statistics Unit
Brookes Lawley Building
15 Cotswold Road
Sutton, Surrey
SM2 5NG
020 8643 8901 x 4814
Fax: 020 8722 4368
|