Is there a difference between Chi-square (from coxph -> anova) and Wald Chi-square (from cph -> anova)?

And how do I have to interpret these Chi-square values? What does a P<0.05 mean in this case? Why does the sum of Chi-square values of each variable not equal TOTAL? My idea was that each Chi-square indicates the predictive information of each variable and TOTAL that of the entire model.

> library(survival) > library(rms) > > data(colon) > d <- colon > m1 <- cph(Surv(time, status) ~ age + sex + nodes, data=d) > anova(m1) Wald Statistics Response: Surv(time, status) Factor Chi-Square d.f. P age 0.03 1 0.8612 sex 0.93 1 0.3349 nodes 189.79 1 <.0001 TOTAL 192.01 3 <.0001 > 0.03+0.93+189.79 # = 190.75 [1] 190.75 > m2 <- coxph(Surv(time, status) ~ age + sex + nodes, data=d) > anova(m2) Analysis of Deviance Table Cox model: response is Surv(time, status) Terms added sequentially (first to last) loglik Chisq Df Pr(>|Chi|) NULL -6424.0 age -6423.6 0.7147 1 0.3979 sex -6423.4 0.5019 1 0.4787 nodes -6356.9 132.8685 1 <2e-16 *** --- Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 

Is there a difference between chi-squared (from coxph -> anova) and Wald chi-squared (from cph -> anova)?

And how do I have to interpret these chi-squared values? What does a P<0.05 mean in this case? Why does the sum of chi-squared values of each variable not equal TOTAL? My idea was that each chi-squared indicates the predictive information of each variable and TOTAL that of the entire model.

> library(survival) > library(rms) > > data(colon) > d <- colon > m1 <- cph(Surv(time, status) ~ age + sex + nodes, data=d) > anova(m1) Wald Statistics Response: Surv(time, status) Factor Chi-Square d.f. P age 0.03 1 0.8612 sex 0.93 1 0.3349 nodes 189.79 1 <.0001 TOTAL 192.01 3 <.0001 > 0.03+0.93+189.79 # = 190.75 [1] 190.75 > m2 <- coxph(Surv(time, status) ~ age + sex + nodes, data=d) > anova(m2) Analysis of Deviance Table Cox model: response is Surv(time, status) Terms added sequentially (first to last) loglik Chisq Df Pr(>|Chi|) NULL -6424.0 age -6423.6 0.7147 1 0.3979 sex -6423.4 0.5019 1 0.4787 nodes -6356.9 132.8685 1 <2e-16 *** --- Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 

Is there a difference between Chi-square (from coxph -> anova) and Wald Chi-square (from cph -> anova)?

And how do I have to interpret these Chi-square values? What does a P<0.05 mean in this case? Why does the sum of Chi-square values of each variable not equal TOTAL? My idea was that each Chi-square indicates the predictive information of each variable and TOTAL that of the entire model.

> data(colon) > d <- colon > m1 <- cph(Surv(time, status) ~ age + sex + nodes, data=d) > anova(m1) Wald Statistics Response: Surv(time, status) Factor Chi-Square d.f. P age 0.03 1 0.8612 sex 0.93 1 0.3349 nodes 189.79 1 <.0001 TOTAL 192.01 3 <.0001 > 0.03+0.93+189.79 # = 190.75 [1] 190.75 > m2 <- coxph(Surv(time, status) ~ age + sex + nodes, data=d) > anova(m2) Analysis of Deviance Table Cox model: response is Surv(time, status) Terms added sequentially (first to last) loglik Chisq Df Pr(>|Chi|) NULL -6424.0 age -6423.6 0.7147 1 0.3979 sex -6423.4 0.5019 1 0.4787 nodes -6356.9 132.8685 1 <2e-16 *** --- Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 

Is there a difference between Chi-square (from coxph -> anova) and Wald Chi-square (from cph -> anova)?

And how do I have to interpret these Chi-square values? What does a P<0.05 mean in this case? Why does the sum of Chi-square values of each variable not equal TOTAL? My idea was that each Chi-square indicates the predictive information of each variable and TOTAL that of the entire model.

> library(survival) > library(rms) > > data(colon) > d <- colon > m1 <- cph(Surv(time, status) ~ age + sex + nodes, data=d) > anova(m1) Wald Statistics Response: Surv(time, status) Factor Chi-Square d.f. P age 0.03 1 0.8612 sex 0.93 1 0.3349 nodes 189.79 1 <.0001 TOTAL 192.01 3 <.0001 > 0.03+0.93+189.79 # = 190.75 [1] 190.75 > m2 <- coxph(Surv(time, status) ~ age + sex + nodes, data=d) > anova(m2) Analysis of Deviance Table Cox model: response is Surv(time, status) Terms added sequentially (first to last) loglik Chisq Df Pr(>|Chi|) NULL -6424.0 age -6423.6 0.7147 1 0.3979 sex -6423.4 0.5019 1 0.4787 nodes -6356.9 132.8685 1 <2e-16 *** --- Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1