Modeling Clinical Trial Attrition Using Machine Intelligence: A driver analytics case study using 1,325 trials representing one million patients

This file is the readme.txt file for the code folder that contains the R code that was used in the model

TABLE OF CONTENTS

• Full Author List

• Introduction

• Requirements

• Folder Contents

• Contact

Full Author List

Emmette Hutchison, Youyi Zhang, Sreenath Nampally, Imran Khan Neelufer, Vlad Malkov, Jim Weatherall, Faisal Khan and Khader Shameer

Introduction

Patient attrition, also referred to as dropout or patient withdrawal, occurs when patients enrolled in a clinical trial either withdraw or are lost to follow-up by the clinical site and trial sponsor.

Requirements

This was performed using RStudio and R. The versions of RStudio and R are listed below: RStudio 1.0.44, which can be downloaded from https://rstudio.com/products/rstudio/download/ R 3.5.2 (2018-12-20). RStudio can be installed from https://cran.r-project.org/mirrors.html

The project requires the following R packages. The version numbers indicate the version of the packages that were used in the analysis. Please install the following packages using the command below in your R Environment

SuperLearner==2.0-26
MASS==7.3-51.5
ranger==0.12.1
ipred==0.9-9
kernlab==0.9-29
arm==1.10-1
dplyr==1.4.2
caret==6.0-84
parallel==3.5.2

Folder Contents

This folder contains the data files that was used in the analysis. The file descriptions are listed below:

|---- readme.md :  readme file for the data folder
|
|---- code : This folder contains the code required for the pre-processing of raw data and running the predictive model results,
|			 further explanation is available in a readme file within this folder
|
|---- data:  This folder contains two sub-folders, another readme file in the folder will explain the overview on each folder content.
		|---- analysis_ready
|---- requirments.txt : is a file which lists the libraries used in the model building exercise, this file can be used to install the packages needed.
|		
|		Using command "Install.Packages("package name") one can easily install the required packages

Contact

shameer.khader@astrazeneca.com