Building a model

In order to build a model, we will incorporate all regular predictors as well as predictors that have a lesser value of p (0.3-0.35). As methodux consists of three steps, the preceding predictor will be incorporated along with a dummy variable as well as methodux1 as a reference. A dummy variable can be formed through the xi command with Toxiq:

Toxiq age metfordrug j.sets 1.metho, Lsssdzc

Fail_c: patient response
Analysis_tm: time

Iteration 0: log Likelihood = - 38645.678, Iteration 1: log Likelihood = -38546.876
Iteration 2: log Likelihood = -38546.956, Iteration 3: log Likelihood = -38546.567

Refined estimate: Iteration 0: Log Likelihood  = -38456,567

Cox regression -- Breslow method
Total subjects- 520
Total observations – 520
Failures – 294
Risk time – 13259
Log likelihood - -38456.567        LRRRchiiirl(3) – 36.90
                                   Prob> chiiiir(3) = 0

t_time	Coefficient	Standard Error	y	P> !y!	[ 93% Conff interval]
age	-0.12456 ...