Hi, all
Using relsurv package - rsadd and rsmul I got the following coefficients for a colorectal cancer study:
Addictive - Estève Multiplicative - Andersen
Coef (SE) Exp(coef) p-value Coef (SE) Exp(coef) p-value
sex female -0.06957 (0.0646) 0.9328 0.28145 0.51641 (0.0559) 1.6760 <2e-16
Age 45-54 0.15426 (0.1668) 1.1668 0.35513 -0.67519 (0.1635) 0.5091 3.63e-05
Age 55-64 0.27521 (0.1535) 1.3168 0.07300 -1.28001 (0.1497) 0.2780 <2e-16
Age 65-74 0.39229 (0.1515) 1.4804 0.00963 -2.05659 (0.1279) 0.1279 <2e-16
Age 75+ 0.84512 (0.1569) 2.3283 7.22e-08 -2.63199 (0.1488) 0.0719 <2e-16
I can see that the variables that are used in lifetables, like sex and age changed its sign.
I noticed, too, that for other covariates, like cancer stage, coefficients are bigger in addictive model than in multiplicative, but keeping the same sign.
I'm not being able to interpret these differences. Can someone help?
Thanks,
Carlos Falcao
<br/><hr align="left" width="300" />
View this message in context: <a href="http://r.789695.n4.nabble.com/RELSURV-difficulty-interpreting-RSADD-and-RSMUL-coefficients-tp4719386.html">RELSURV - difficulty interpreting RSADD and RSMUL coefficients</a><br/>
Sent from the <a href="http://r.789695.n4.nabble.com/datatable-help-f2315188.html">datatable-help mailing list archive</a> at Nabble.com.<br/>