****************
****************
** stcox ** Cox proportional hazards regression
****************
****************
Overview: To model survival as a function of covariates the Cox model
may be used. This model focuses on estimates of hazard
ratios.
Usage: "stset ftime, failure(status)"
"stcox tx"
"stcox tx wbc"
Where - ftime is the follow-up time
dvar is the censoring/death indicator (dvar=1 if
the event time was observed and 0 if censored)
Where - "tx wbc" are the covariates
Summaries: STATA will return estimates of the hazard ratios (exponentiated
coefficients), standard errors, and confidence intervals.
Options: (1) "stcox tx wbc, nohr" -- the "nohr" option means that
regression coefficients (log hazard ratios) will be
reported rather than hazard ratios.
(2) "stcox tx wbc, basesurv( s0 )" -- this option saves the
estimate of the baseline survival as the variable "s0".
NOTE: if you are saving s0, then this is for X=0 (all of
them). So, you might want to center your predictors
so that the value X=0 is meaningful.
(3) "stcox tx wbc"
"lrtest, saving(1)" -- this option is used similar to
logistic regression for computing likelihood ratio tests.
(4) "stcox tx, strata(Svar)" -- this means a separate baseline
hazard is assumed for every level of "Svar" but hazard
ratios within these strata are assumed to be equal
(common). A single "tx" HR is estimated.
****************
****************
** stcoxkm ** Cox proportional hazards model compared to Kaplan-Meier
****************
****************
Overview: This compares Kaplan-Meier curves to the estimated survival
curves based on the Cox regression model.
Usage: "stset ftime, failure(status)"
"stcoxkm, by(tx)"
Where - ftime is the follow-up time
dvar is the censoring/death indicator (dvar=1 if
the event time was observed and 0 if censored)
Where - "tx " is the covariates
Summaries: STATA will create a KM plot and overlay the fitted Cox
model estimates of survival curves.
Options: (1) The "by(Svar)" is required.
OTHER COMMENTS:
(*) Some graphical output after regression:
This is based on the command "sts graph" Example:
sts graph, strata(tx) adjustfor( wbc )
This will create survival function estimates for each level of tx
after adjusting for wbc. What this means is that we use a Cox
regression with
log[ h(t,X) ] = log[h0(t)] * tx + b2 wbc
and then looks at the fitted survival curves for different
levels of tx (ie. tx=1 and tx=0) for a fixed value of the
adjustment variables (wbc=0 in this case, and in general
X_j=0 for all of the adjustment variables).
This is not exactly what I had hoped for... In the end we obtain
separate baseline survival estimates for each level of "tx".
options:
by(tx) -- a separate Cox regression for each level of tx. This
means that any adjustment variable is included as a
main effect *and* interaction with tx.
i.e. log[h(t,X)] = log[h0(t)]*tx + newlwbc*tx
strata(tx) -- now adjustment is calculated by using a stratified
Cox regression, so that a common regression estimate
is used for the adjustment variable.
i.e. log[h(t,X)] = log[h0(t)]*tx + newlwbc