stnet is a Stata command that implements the Pohar Perme estimator of net survival using a life table approach. This work is a collaboration between Enzo Coviello, myself, Karri Seppä, and Arun Pokhrel. Enzo Coviello did essentially all of the Stata coding and played a major role in adapting the Pohar Perme estimator to discrete time.
The Pohar Perme estimator estimator was developed for continuous survival times, yet cancer registries often have only discrete survival times (for example, survival time in completed months). We proposed an approach to estimation that is also appropriate when survival times are discrete (ties are common). When ties are rare (e.g., when survival times are measured in days), there is little practical difference between our modified estimator and the continuous time estimator. Our estimator, however, makes more reasonable assumptions in the presence of ties and is more appropriate when exact times are not available. See Table 1 in our Stata Journal article for a comparison of the estimators and Seppä et al for another comparison and additional discussion.
stnet can be installed from the Boston College Statistical Software Components (SSC) archive using the following command (from the Stata command line)
ssc install stnet
If you use this package, please cite the associated paper in The Stata Journal.
Enzo Coviello, Karri Seppä, Paul W. Dickman, Arun Pokhrel, 2015. Estimating net survival using a life-table approach, The Stata Journal, StataCorp LP, vol. 15(1), pages 173-185.