Skip to main content

Archived Comments for: Methods for stratification of person-time and events – a prerequisite for Poisson regression and SIR estimation

Back to article

  1. The Stratify macro has been updated and enhanced

    Klaus Rostgaard, Statens Serum Institut, Denmark

    19 November 2010

    Dear reader,

    In the fall of 2010 I updated and enhanced the stratify macro substantially.

    Most of the things that appear as post-processing in this paper is now performed by default internally in the macro. Also the BRTF statement has been replaced by the SRTF statement which has exactly the same syntax, but is more general. Most examples in the paper have to be modified slightly to work in the same way (often just the addition of COMPLETE=no as an option). These changes have been made in the textfile containing the macro and sample code. The enhancements made have made the macro (relatively) more useful for Cox regression, which can now be performed directly on output data from the stratify macro. Also see my new suggestions for making a “period” time scale in the manual.

    The updated macro is available in the same place as before on

    Klaus Rostgaard
    The author

    Competing interests