46. Combining Survival/Hazard Estimates - Compound Curves
Survival analysis examines clinical data to generate inputs for health economic models. This can take the form of Kaplan-Meier tables, fitted survival distributions, hazard tables and other forms. These inputs are then typically used to drive disease progression in Partitioned Survival models and in Markov models.
This chapter focuses on combining these survival or hazard estimates into a single curve to model disease progression. Here are a few examples of why you need this.
-
You might want to start with clinical Kaplan-Meier survival curve data for the observation period then transition to a distribution for the extrapolated period.
-
You might want to use two different fitted distributions to represent high/low bounds on survival for uncertainty/sensitivity analysis.