Introduction to the week

This week, we will start our journey into the world of survival analysis and failure time. We will look at the structure of survival data, some models for survival data, and statistics that are often used to summarize and evaluate survival experiences. We will move from estimating survival to statistically comparing the survival experiences of groups. We will also learn to evaluate whether some standard parametric models are good fits to the data.

Along the way, we will be introduced to a fundamental concept in survival analysis, the proporational hazards assumption. We will see how it is useful to help evaluate survival, and approaches we can use if it is not met.

Course notes

Link

Readings

• “Tutorial in Biostatistics: Survival Analysis in Observational Studies” by Kate Bull and David J. Spiegelhalter PDF link
• “Statistical Inference Enables Bad Science; Statistical Thinking Enables Good Science” by Christopher Tong PDF link
• “Clinical Utility of Tumor-Naïve Presurgical Circulating Tumor DNA Detection in Early-Stage NSCLC” by Hong et al PDF link. This is a recent paper I contributed to which uses survival analysis to identify a strong prognostic biomarker for lung cancer. Yes I do use survival analysis a lot.

Laboratory work

Github Classroom link

Assignment

The weekly assignment is available on GH Classroom here. A rubric is available on Canvas. The assignment is due on Thursday, Sept 26 by 11:59pm

Solution

Resources