Question 2: Draw a survival curve for time to readmission to hospital.
We have time to event data and so we use the st commands. The time variable is "Follow-up time (days)" and the Status variable is "Readmitted". We must first set up the survival time (st) functions usy=ing stset:
. stset time, failure(readmit)
failure event: readmit != 0 & readmit < .
obs. time interval: (0, time]
exit on or before: failure
------------------------------------------------------------------------------
1034 total obs.
3 event time missing (time>=.) PROBABLE ERROR
------------------------------------------------------------------------------
1031 obs. remaining, representing
540 failures in single record/single failure data
526790 total analysis time at risk, at risk from t = 0
earliest observed entry t = 0
last observed exit t = 990
We want a survival curve, so we use sts graph. This gives us:
This would be OK for information while doing the analysis, but for publications we would improve it. I would like to give a different title to the time axis, change the vertical axis labels, make the labels horizontal, remove the grid lines, give the vertical axis a title, remove the "Kaplan-Meier survival esimate" title, and scale up all the text:
. sts graph, xtitle("Time from discharge (days)") ylabel(0 (.2) 1, angle(horiz) nogrid)
ytitle(Proportion not yet readmitted) title("") scale(1.5)
Back to Exercise: Readmission to hospital for asthmatic children.
To Biostatistics in Research Practice index.
To Martin Bland's M.Sc. index.
This page maintained by Martin Bland.
Last updated: 6 February, 2008.