This repository is intended to hold code to produce Kaplan–Meier plots in the style recommended following the KMunicate study (TP Morris et al. Proposals on Kaplan–Meier plots in medical research and a survey of stakeholder views: KMunicate. BMJ Open, 2019, 9:e030215). An example is given at the bottom of this readme.
Since publication of the paper, many people have said 'I want to do those plots. Where is the code? The embarrassing answer is: I did the table as a one-off assuming nobody would like it – who likes tables!
I have written a Stata program kmunicate (in the main directory of this repo because I could not work out how to make net from detect stata.toc from a subdirectory). To install it, use
net from https://raw.githubusercontent.com/tpmorris/kmunicate/master/
Alessandro Gasparini (@ellessenne) wrote an R package kmunicate; repo here.
# install.packages("devtools")
devtools::install_github("ellessenne/KMunicate-package")
Back in 2020 I launched a challenge on twitter for people to produce a script or program that implements these plots. The winning prizes are wooden spoons I carved (see the twitter thread).
Ok, time for some criteria. Here is what a winner would need to produce:
- A figure similar to the one below that can be reproduced with data.
- The numbers in the table must align with the time points on the figure.
- The tables must be structured for each group (e.g. trial arm) with a row for 'at-risk', 'censored' and 'events' (see example below).
- Confidence intervals must be translucent so that two or more can be seen when they sit on top of each other.
- This must be showcased with a publicly-available dataset.
- The result must be reproduced.
- One winner for each of Stata, R and SAS.
- The Stata prize is harder: I have already produced a script. You have to produce a general program as an ado file. (I would also prefer this for R and SAS but beggars can't be choosers.)
The fastest person gets to choose their spoon and second and third choose from the remainder (no pressure to take one). The competition closed on 1st September 2020.
The first winners came in fast:
- Ruth Keogh (LSHTM) produced an R script and picked the coffee scoop (pictured centre)
- Alessandro Gasparini (Karolinska Institutet) wrote the KMunicate R package (see Alessandro’s KMunicate repo). (He suggested I sell the spoon on Etsy then donate proceeds, but it was so impressive that I sent him the one withtand donated from the next one sold on Etsy.)
- Amritendu Bhattacharya (George Institute India) wrote a SAS script and won the toasted sycamore pocket spoon.
- The final spoons were awarded to Sarwar Mozumder and Austin Nichols for their Stata scripts.
