Hi Sas analysts!
This post announces the availability of SAS macros for performing flexible parametric regression analysis of relative survival. These macros represent an implementation of Stata's stpm2 user-supplied program, originally developed by Paul Lambert of University of Leister, UK. Aside from their use in regression analysis of relative survival, many other measures useful in the examination of population-based cancer survival can be examined: crude probabilities of death (cancer/other) in a competing risks framework, conditional survival, hazard ratios and differences, cure proportion, age-standardised survival, future life years lost. A separate macro that performs life table analysis of relative survival is also present in the repository, developed from an original program by Paul Dickman of the Karolinska Institutet, Stockholm, which includes the Pohar-Perme estimator as output.
The GitHub repository below contains material that was presented in the survival workshop hosted at the Albuquerque NAACCR meeting, including example datasets and annotated exercises. Full macro documentation is included, along with some of the key academic papers.
If you are a SAS user and have wanted to perform these kinds of analyses using your registry's data, this post is for you. As a teaser, the attached SAS plot shows life table (Pohar-Perme) estimates of relative survival using the included melanoma dataset, along with the equivalent modelled relative survival estimates (see exercise A100 in the repository for full code).
The GitHub repository is available at GitHub.com/FlexSurv/repo
------------------------------
Ron Dewar
[CompanyName]
------------------------------