Hello all:
I note a slight difference between months elapsed using -datediff_frac- function and the _t in stset with the scale option using date of censor (docens) and date of diagnosis (dodx). What would the appropriate way to set the scale to show in months ? Is 30.455 a reasonable approximation in scale?
gen tos = datediff_frac(dodx,docens,"month")
format tos %9.1f
label variable tos "Time from Dx to study exit(months)"
stset tos, fail(dead)
vs.
stset docens, origin(time dodx) fail(dead) scale(30.455)
I note a slight difference between months elapsed using -datediff_frac- function and the _t in stset with the scale option using date of censor (docens) and date of diagnosis (dodx). What would the appropriate way to set the scale to show in months ? Is 30.455 a reasonable approximation in scale?
gen tos = datediff_frac(dodx,docens,"month")
format tos %9.1f
label variable tos "Time from Dx to study exit(months)"
stset tos, fail(dead)
vs.
stset docens, origin(time dodx) fail(dead) scale(30.455)
