**************** **************** ** stset ** define a survial data outcome **************** **************** Overview: To use survival analysis methods such as the Kaplan-Meier or Cox regression we need to identify the follow-up time, T, and the censoring indicator, delta. Usage: "stset ftime, failure(dvar)" 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) Summaries: STATA will return a brief summary of the number of records, the number of observed events, and the total person-time under study. Options: (1) "stset ftime, failure(status) enter(stime)" -- where the variable stime records the time when the subject entered the study. **************** **************** ** sts graph ** Plotting estimated survival curves **************** **************** Overview: The "sts graph" command calculates and plots Kaplan-Meier survival curves. Usage: "sts graph" "sts graph, by(Svar)" Summaries: The "sts graph" function generates a Kaplan-Meier curve -- with a curve for each group given with the "by" option if selected. Options: (1) "sts graph, by(Svar)" The "by" option creates a separate line for each level of the argument Svar. (2) "sts graph, lost" The "lost" option requests that the number lost (or censored) be shown on the plot. (3) "sts graph, gwood" The "gwood" option requests that point-wise confidence bands be displayed. (4) "sts graph, adjustfor(xvar1 xvar2)" This option displays the fitted survival function for xvar1=0, xvar2=0 after using Cox regression. If you also specify "by(Svar)" then separate Cox regressions are used to model each level of Svar separately, while if you use "strata(Svar)" then a single stratified Cox model is estimated and the baseline survival is returned. **************** **************** ** sts list ** Listing the estimated survival curves **************** **************** Overview: The "sts list" command will print a tablular version of the survival estimates and standard errors. Usage: "sts list" "sts list, by(Svar)" Summaries: The "sts list" command returns the summaries that are used to estimate the Kaplan-Meier curves. See the course notes on page 258-2 for illustration. Options: (1) "sts list, by(Svar)" -- separate summaries for each level of Svar. **************** **************** ** sts test ** Testing survival curves **************** **************** Overview: The "sts test" command Usage: "sts test Svar, logrank" "sts test Svar, wilcoxon" Summaries: The "sts test" function is used to test the null hypothesis that survival curves are equivalent. It will return a small table that shows the number of observed events for each level of Svar and the corresponding expected number of events under the null hypothesis. Finally, the test statistic and p-value are given. Options: (1) "sts test, logrank" -- calculates the log-rank test as described in the course notes pp 270-276. (2) "sts test, wilcoxon" -- calculates the wilcoxon-gehan-breslow test as described in the course notes pp 277-281.