--- title: "Biostatistics III in R" author: - Code by Johan Zetterqvist and Mark Clements output: prettydoc::html_pretty: theme: cayman highlight: github --- # Exercise 13. Modelling the diet data using Cox regression # ----------- Now fit a model to the localised melanoma data where the outcome is observed survival (i.e. all deaths are considered to be events). ```{r setup, cache=FALSE, message=FALSE, echo=FALSE} library('knitr') read_chunk('../q13.R') opts_chunk$set(cache=FALSE) ``` You may have to install the required packages the first time you use them. You can install a package by `install.packages("package_of_interest")` for each package you require. ```{r loadDependencies, message=FALSE} ``` Load the melanoma data and explore it. ```{r loadPreprocess, results='hide'} ``` ## (a) ## ```{r 13.a, warning=FALSE, message=FALSE} ``` These two models are conceptually different since the Cox model adjusts for ‘time’ even though this is not explicit in the `coxph` function. In this example, ‘time’ refers to ‘time on study’ (time since entry) which we do not expect to be a strong confounder. That is, we would expect the estimates of the effect of high energy to be similar for the two models, which they are. ## (b) ## If we use a different timescale then this amounts to adjusting for a different factor. As such, we would not expect the estimates to be identical. Attained age, unlike time since entry, is expected to be a confounder but we see that it is not a strong confounder. ```{r 13.b, warning=FALSE, message=FALSE} ```