Purpose of the Course
NAACCR is reaching critical mass in terms of having population-based survival data from numerous registries. The purpose of this course is to provide the training and tools needed to properly carry out survival analysis using central registry data.
The course will address the principles, methods and application of statistical methods in population-based cancer survival analysis and cover central concepts, such as how to estimate and model relative survival, as well as recent methodological developments including cure models, flexible parametric models, and estimation of crude probabilities of death. Comparison of alternative methodological approaches (e.g., to estimating relative survival and to modeling relative survival) will be a focus of the course and participants will get the opportunity to apply and contrast a range of methods to real data. The course will consist primarily of lectures and hands-on computing sessions.
Class participants must supply their own laptop computer.
Data analysts, epidemiologists, biostatisticians, and others interested in learning about population-based survival analysis.
The primary focus of the course will be on statistical methods. A degree in statistics or mathematical statistics is not required, but we expect participants to possess basic knowledge of the fundamentals of epidemiology and biostatistics and be comfortable fitting statistical models in epidemiology (e.g., logistic regression, Poisson regression, or Cox regression). We expect participants to have a wide range of backgrounds, and some of the content will be directed at those with formal training in statistics but the main emphasis of the course will be on concepts and application with a minimum of complex mathematical detail. Participants will gain most if they have some previous knowledge of basic concepts in survival analysis such as survival functions, Kaplan-Meier curves, and Cox regression.
The faculty use Stata for most of their analysis and have written Stata commands to implement some of the more advanced methods that will be taught during the course. A time limited trial version of Stata will be provided. We will provide extensive exercises with printed solutions as well as Stata .do files that participants can use as templates for analyzing their own data. The course is designed to be accessible to participants without previous experience of using Stata. There will also be a demonstration of SEER*Stat and CanSurv with the aim of contrasting the approaches with what is available in Stata. Participants should load SEER*Stat and CanSurv on their laptops prior to the course. As we will be using SEER data for most of the exercises, participants will need to sign a data use agreement for the SEER Research Data File at the beginning of the course.