I have 2-piece model in a Partitioned Survival Model, where first part is KM curve and second part is extrapolated (exponential here). How can I conduct the uncertainty in the KM curve in the Excel model, specifically on how to generate the 95% CI with bootstrapping and use this in the PSA ? Thank you.
To be honest, I doubt that Excel is a good choice of software. I would investigate R packages which likely would have something you can use. R is free and very powerful. Attached is an introductory R book. Another very useful book is Jared Lander's, R for everyone which contains a lot of useful code including survival analysis. Best wishes, David Booth
I'm not really a fan of bootstrapping, or resampling, which is really just 'fake replication' imo. You could do AUC or even simple statistics
@Andrew, I am not sure that you could do AUC or simple stats here. I would suggest as a starting reference:
https://books.google.com/books?id=Sj1yDAAAQBAJ&printsec=frontcover&dq=editions:8ccNMVZCXawC&hl=en&sa=X&ved=0ahUKEwjvl8jZs-LgAhXM64MKHSg_AVAQ6AEIJjAA#v=onepage&q&f=false
Best wishes, David Booth
- Badges
- Science topic
- Similar topics
- Statistics
- Time Series