Power for interrupted time-series analysis

[Bruce Weaver]

I've been asked for advice about power analysis for an interrupted time-series analysis (looking at counts of a medical condition in one location over several years). My searches are turning up no sub-categories of -power- that are relevant to time-series analysis. Google searches are suggesting that simulation is the way to go. Before I advise my colleagues to roll their own simulations (so to speak), I thought I should check here to see if anyone knows of any user-written programs that might be helpful. Pointers to good general tutorials on interrupted time-series analysis would also be helpful.

Cheers,
Bruce

[Bruce Weaver]

Someone alerted me to this via private e-mail.

ssc describe itspower