help for stcband
-------------------------------------------------------------------------------

Title

stcband -- Hall-Wellner and equal precision confidence band for the survival and cumulative hazard function

Syntax

stcband [if] [in] [, nair tlower(#) tupper(#) na transform(string) genlow(newvar) genhigh(newvar) level(#) nograph twoway_options ]

options Description ------------------------------------------------------------------------- Options nair specify equal precision (Nair) confidence band be computed. Default is Hall-Wellner confidence band tlower specify the lower time boundary for confidence band. Default is the smallest time event tupper specify the upper time boundary for confidence band. Default is the largest time event na compute confidence band for the cumulative hazard function transform produce log, linear or arcsine transformed confidence band. Default is log genlow create a variable saving lower confidence band genhigw create a variable saving upper confidence band level set confidence level. Values allowed are 90, 95 and 99. Default is 95 nograph suppress graph

Y-Axis, X-Axis, Legend twoway_options some of the options documented in [G] twoway_options

Add plot plot(plot) addplot(plot) add other plots to generated graph -------------------------------------------------------------------------

You must stset your data before using stcband; see stset.

Description

stcband graphs and optionally saves the confidence bands for the survival function.

The usual pointwise confidence intervals for the survival function are valid for a single fixed time. Often limits which contain S(t) for all t in an interval [tlow,tup] are of interest. Two important types of confidence bands are the Hall-Wellner bands (Hall and Wellner, 1980) and the equal precision bands (Nair,1984). To construct these confidence bands critical values are taken from NairTables.dta or HallWellnerTables.dta (see reference). There are three possible forms for the confidence bands: linear, log-minus-log and arcsine-square root transformed.

Examples

use "C:\Data\BMT", clear stset Time2,failure(DeathRelapse)

Graph equal precision 95% confidence band in time interval 100 - 600 in patients with acute lymphoblastic leukemia

stcband if Dis==1, nair tlower(100) tupper(600) xla(0(100)600) yla(0(.2)1, gstyle(minor))

Generate and graph Hall-Wellner 95% confidence band in time interval 100 - 600

stcband if Dis==1, tlower(100) tupper(600) genlo(lohw_band) genhi(uphw_band) xla(0(100)600) yla(0(.2)1, gstyle(minor))

Graph simultaneously pointwise log-log confidence interval and Hall-Wellner arcsine transformed 95% confidence band

sts gen pw_lo=lb(s) pw_hi= ub(s) if Dis==1 stcband if Dis==1, tlower(100) tupper(600) xla(0(100)600) transform(arcsine) yla(0(.2)1, gstyle(minor)) legend(ring(0) pos(7) size(*.7)) plot(line pw_hi pw_lo _t if _t<600, sort c(J J) lc(red red))

Downloading ancillary files in one of your `"`c(adopath)'"' directory you can run this example.

(click to run)

Also see

[ST] sts gen

References

Klein JP, and Moeschberger ML: Survival Analysis: techniques for Censored and Truncated Data - 2nd ed., pp. 109 - 117. New York: Springer-Verlag, 2003

Authors

Enzo Coviello (enzo.coviello@alice.it)