Hi all, I need to compare the median survival for a couple of patient groups in Stata for a project and see if median survival is statistically significantly different among these groups. After using stset to set up my data, I am able to obtain median overall survival for groups using the stci commands, which also gives me confidence intervals.
But I can't find a way to compare the survival of groups using the survival analysis commands in Stata, to obtain a p-value. I've looked online several places and on several Stata guides on survival analysis. Just hoping to see if anyone may know of a command to do this.
As you can see below, if I look at medina survival by mrd_flow, median survival is 21.9 for Positive and not reached for Negative, but I want to be able to say these are (or are not) statistically significantly different. Similarly, when I look at therapyintense, median survival is 36for HAM and 61 for HI--I want to be able to directly compare and see if they are statistically significantly different from each other (i.e., obtain a P-value).
Appreciate the help,
Mike
Stata commands:
stset os_months, failure(status==1) //ST set the data with status=1 meaning death.
Survival-time data settings
Failure event: status==1
Observed time interval: (0, os_months]
Exit on or before: failure
--------------------------------------------------------------------------
388 total observations
0 exclusions
--------------------------------------------------------------------------
388 observations remaining, representing
226 failures in single-record/single-failure data
10,118.786 total analysis time at risk and under observation
At risk from t = 0
Earliest observed entry t = 0
Last observed exit t = 96.03571
. stci if mrdtx_grp!=.
Failure _d: status==1
Analysis time _t: os_months
| Number of
| subjects 50% Std. err. [95% conf. interval]
-------------+-------------------------------------------------------------
Total | 214 58.5 . 40.4643 .
.
. stci if mrdtx_grp!=., by(mrd_flow)
Failure _d: status==1
Analysis time _t: os_months
| Number of
mrd_flow | subjects 50% Std. err. [95% conf. interval]
-------------+-------------------------------------------------------------
Negative | 178 . . 46.4286 .
Positive | 36 21.92857 2.940085 18.4286 40.4643
-------------+-------------------------------------------------------------
Total | 214 58.5 . 40.4643 .
. stci if mrdtx_grp!=., by(therapyintens)
Failure _d: status==1
Analysis time _t: os_months
| Number of
therapyint~s | subjects 50% Std. err. [95% conf. interval]
-------------+-------------------------------------------------------------
HAM | 55 36.10714 . 20.25 .
HI | 159 61.17857 . 42.5 .
-------------+-------------------------------------------------------------
Total | 214 58.5 . 40.4643 .
. stci if mrdtx_grp!=., by(mrdtx_grp)
Failure _d: status==1
Analysis time _t: os_months
| Number of
mrdtx_grp | subjects 50% Std. err. [95% conf. interval]
-------------+-------------------------------------------------------------
MRD-, Hi | 134 . . 46.4286 .
MRD+, Hi | 25 24.89286 3.955319 18.7143 .
MRD-, HM | 44 . . 25.4643 .
MRD+, HM | 11 18.42857 5.262849 6.17857 .
-------------+-------------------------------------------------------------
Total | 214 58.5 . 40.4643 .
But I can't find a way to compare the survival of groups using the survival analysis commands in Stata, to obtain a p-value. I've looked online several places and on several Stata guides on survival analysis. Just hoping to see if anyone may know of a command to do this.
As you can see below, if I look at medina survival by mrd_flow, median survival is 21.9 for Positive and not reached for Negative, but I want to be able to say these are (or are not) statistically significantly different. Similarly, when I look at therapyintense, median survival is 36for HAM and 61 for HI--I want to be able to directly compare and see if they are statistically significantly different from each other (i.e., obtain a P-value).
Appreciate the help,
Mike
Stata commands:
stset os_months, failure(status==1) //ST set the data with status=1 meaning death.
Survival-time data settings
Failure event: status==1
Observed time interval: (0, os_months]
Exit on or before: failure
--------------------------------------------------------------------------
388 total observations
0 exclusions
--------------------------------------------------------------------------
388 observations remaining, representing
226 failures in single-record/single-failure data
10,118.786 total analysis time at risk and under observation
At risk from t = 0
Earliest observed entry t = 0
Last observed exit t = 96.03571
. stci if mrdtx_grp!=.
Failure _d: status==1
Analysis time _t: os_months
| Number of
| subjects 50% Std. err. [95% conf. interval]
-------------+-------------------------------------------------------------
Total | 214 58.5 . 40.4643 .
.
. stci if mrdtx_grp!=., by(mrd_flow)
Failure _d: status==1
Analysis time _t: os_months
| Number of
mrd_flow | subjects 50% Std. err. [95% conf. interval]
-------------+-------------------------------------------------------------
Negative | 178 . . 46.4286 .
Positive | 36 21.92857 2.940085 18.4286 40.4643
-------------+-------------------------------------------------------------
Total | 214 58.5 . 40.4643 .
. stci if mrdtx_grp!=., by(therapyintens)
Failure _d: status==1
Analysis time _t: os_months
| Number of
therapyint~s | subjects 50% Std. err. [95% conf. interval]
-------------+-------------------------------------------------------------
HAM | 55 36.10714 . 20.25 .
HI | 159 61.17857 . 42.5 .
-------------+-------------------------------------------------------------
Total | 214 58.5 . 40.4643 .
. stci if mrdtx_grp!=., by(mrdtx_grp)
Failure _d: status==1
Analysis time _t: os_months
| Number of
mrdtx_grp | subjects 50% Std. err. [95% conf. interval]
-------------+-------------------------------------------------------------
MRD-, Hi | 134 . . 46.4286 .
MRD+, Hi | 25 24.89286 3.955319 18.7143 .
MRD-, HM | 44 . . 25.4643 .
MRD+, HM | 11 18.42857 5.262849 6.17857 .
-------------+-------------------------------------------------------------
Total | 214 58.5 . 40.4643 .
