Supplementary material for:

Cost-effectiveness analysis in R using a multi-state modelling survival analysis framework: A tutorial

All of the files detailed below can be downloaded from the zip.file tutorial files  *UPDATED - March 2019*

The R commands.R  file should be opened and run in R. It will read into R the example dataset and the R functions and other R programs detailed below used to implement the multi-state modelling approach.

File Description

Algorithm step (if applicable)
FILES DEMONSTRATING THE APPROACH
  
data.txt‌ example dataset used in the tutorial  
R commands.R‌

syntax used for the full cost-effectiveness analysis shown in the tutorial

  
SEPARATE R PROGRAMS RUN FROM WITHIN R COMMANDS.R
  
assessment of fit for prog to death.R‌ assessment of fit for the progression to death transition  4
building all combinations of msms.R

builds all combinations of models using each distribution for each transition, for each treatment arm separately

  4
 visual assessment of fit for progression-free to progression.R  plots to assess the fit for progression-free to progression  4
 visual assessment of fit for progression-free to death.R  plots to assess the fit for progression-free to death  4
 one way sens.R one-way sensitivity analysis

 7
sensitivity analysis - alternative fits ICERs table.R one-way sensitivity analysis considering alternative distributions  7
BASE CASE ANALYSIS
  
function modelparam.R‌

R function that displays the coefficients after fitting a parametric survival regression

  2
function Markov.R‌ R function to fit a Markov parametric multi-state model 3,7
function semiMarkov.R  R function to fit a semi-Markov parametric multi-state model 3,7
README file.txt‌

description of how to overcome a warning with the Markov or  semiMarkov functions

 3,7
function MarkovHaz.R‌ R function to help overcome a warning with the Markov function 3,7
function semiMarkovHaz.R‌

R function to help overcome a warning with the semiMarkov function

 3,7
function visualMarkov.R‌ R function to visually assess the fits of a Markov parametric multi-state model 4
function visualsemiMarkov.R R function to visually assess the fits of a semiMarkov parametric multi-state model 4
function proportion.R‌ R function to create the observed proportion in the intermediate state 4
function meanLY.R‌ R function to create the mean Life Years in a given state for a given treatment arm 5
ONE-WAY SENSITIVITY ANALYSIS
  
function Markov_notrteffect.R‌ R function to remove treatment effect in extrapolation period  7
function Markov_varyHR.R R function to vary hazard ratio in extrapolation period 7
function semiMarkov_notrteffect.R‌ R function to remove treatment effect in extrapolation period 7
function semiMarkov_varyHR.R‌ R function to vary hazard ratio in extrapolation period 7
PROBABILISTIC SENSITIVITY ANALYSIS
  
 function PSAprob.R R function that returns the state occupancy probabilities for each draw 8
 function PSAmeanLY.R R function that returns the mean Life Years in a given state for a given treatment arm, by calculating the mean across each draw in the PSA 9
function PSAQALY.R‌ R function that returns the QALYs in a given state for a given treatment arm 9
function CEplane.R‌ R function to plot a cost-effectiveness plane 11
function CEAC.R R function to plot a cost-effectiveness acceptability curve 12
FUNCTIONS THAT CAN BE USED WHEN INDIVIDUAL PATIENT DATA (IPD) IS NOT AVAILABLE
  
function Markov_noipd.R R function that fits a Markov multi-state model when IPD is not available  
function semiMarkov_noipd.R R function that fits a semiMarkov multi-state model when IPD is not available  
function Haz_noipd.R R function to help overcome a warning with the Markov_noipd.R or semiMarkov_noipd.R functions