Development and external validation of prediction models for adverse health outcomes in rheumatoid arthritis: a multinational real-world cohort analysis (EHDEN Study-a-thon Barcelona 2020)

• Analytics use case(s): Patient Level Prediction
• Study type: Clinical Application
• Tags: EHDEN
• Study lead: Cynthia Yang, Ross Williams
• Study lead forums tag: cynthiayang, RossW
• Study start date: January 16, 2020
• Study end date: -
• Protocol: Word file
• Publications: -
• Results explorer: Shiny app

Prediction of adverse health outcomes after methotrexate initiation.

NOTE: The following information below requires review from the Study Coordinator before using the package.

Introduction

This is a package to train models to predict for Patients who are: [EHDEN RA] New users of methoxtrexate monotherapy used for PLP [EHDEN RA] Female new users of methoxtrexate monotherapy used for PLP (we are predicting breast and uterus cancer and so are limiting for these to female patients)

Who will develop: [EHDEN RA] Stroke (ischemic or hemorrhagic) events (any visit) [EHDEN RA] Acute myocardial infarction events (in any visit) [EHDEN RA] Pancytopenia events using diagnoses and measurements [EHDEN RA] Opportunistic Infections [EHDEN RA] Serious Infection events [EHDEN RA] Persons with a Malignant neoplasm of breast 1 dx [EHDEN RA] Persons with a Malignant neoplasm of uterus 1 dx [EHDEN RA] Persons with a Malignant neoplasm of colon and rectum 1 dx [EHDEN RA] Serious Infection, opportunistic infections and other infections of interest event [EHDEN RA] Leukopenia events using diagnoses and measurements [EHDEN RA] Pancytopenia or leukopenia events using diagnoses and measurements in 90-days, 2 years and 5 years time at risk.

To run validation scroll down to "Install Validation Package"

Instructions To Build Package

• Build the package by clicking the R studio 'Install and Restart' button in the built tab

Instructions To Run Package

• Share the package by adding it to the OHDSI/StudyProtocolSandbox github repo and get people to install by running but replace 'EHDENRAPrediction' with your study name if not using atlas:

  # get the latest PatientLevelPrediction
  install.packages("devtools")
  devtools::install_github("OHDSI/PatientLevelPrediction")
  # check the package
  PatientLevelPrediction::checkPlpInstallation()
  
  # install the network package
  devtools::install_github("https://github.com/ohdsi-studies/EhdenRaPrediction")

• To execute the study by running the code in (extras/CodeToRun.R)

  library(EHDENRAPrediction)
  # USER INPUTS
#=======================
# The folder where the study intermediate and result files will be written:
outputFolder <- "./EHDENRAPredictionResults"

# Specify where the temporary files (used by the ff package) will be created:
options(fftempdir = "location with space to save big data")

# Details for connecting to the server:
dbms <- "you dbms"
user <- 'your username'
pw <- 'your password'
server <- 'your server'
port <- 'your port'

connectionDetails <- DatabaseConnector::createConnectionDetails(dbms = dbms,
                                                                server = server,
                                                                user = user,
                                                                password = pw,
                                                                port = port)

# Add the database containing the OMOP CDM data
cdmDatabaseSchema <- 'cdm database schema'
# Add a database with read/write access as this is where the cohorts will be generated
cohortDatabaseSchema <- 'work database schema'

oracleTempSchema <- NULL

# table name where the cohorts will be generated
cohortTable <- 'EHDENRAPredictionCohort'
#=======================

execute(connectionDetails = connectionDetails,
        cdmDatabaseSchema = cdmDatabaseSchema,
        cohortDatabaseSchema = cohortDatabaseSchema,
        cohortTable = cohortTable,
        oracleTempSchema = oracleTempSchema,
        outputFolder = outputFolder,
        createProtocol = F,
        createCohorts = T,
        runAnalyses = T,
        createResultsDoc = F,
        packageResults = T,
        createValidationPackage = F,
        minCellCount= 5)

• You can then easily transport the trained models into a network validation study package by running:

  
  execute(connectionDetails = connectionDetails,
        cdmDatabaseSchema = cdmDatabaseSchema,
        cohortDatabaseSchema = cohortDatabaseSchema,
        cohortTable = cohortTable,
        outputFolder = outputFolder,
        createProtocol = F,
        createCohorts = F,
        runAnalyses = F,
        createResultsDoc = F,
        packageResults = F,
        createValidationPackage = T,
        minCellCount= 5)
  

• To create the shiny app and view run:

  
populateShinyApp(resultDirectory = outputFolder,
                 minCellCount = 10, 
                 databaseName = 'friendly name'
                 ) 
        
viewShiny('EHDENRAPrediction')
  

Install Validation Package

install.packages("devtools")
library(devtools)
install_github("ohdsi-studies/EhdenRaPrediction/validationPackage/EHDENRAPredictionValidation")

To run the vaidation package:

# If not building locally uncomment and run:

#install.packages("devtools")
#devtools::install_github("OHDSI/StudyProtocolSandbox/EHDENRAPredictionValidation")

library(EHDENRAPredictionValidation)

# add details of your database setting:
databaseName <- 'add a shareable name for the database you are currently validating on'

# add the cdm database schema with the data
cdmDatabaseSchema <- 'your cdm database schema for the validation'

# add the work database schema this requires read/write privileges 
cohortDatabaseSchema <- 'your work database schema'

# if using oracle please set the location of your temp schema
oracleTempSchema <- NULL

# the name of the table that will be created in cohortDatabaseSchema to hold the cohorts
cohortTable <- 'EHDENRAPredictionValidationCohortTable'

# the location to save the prediction models results to:
outputFolder <- getwd()

# add connection details:
options(fftempdir = 'T:/fftemp')
dbms <- "pdw"
user <- NULL
pw <- NULL
server <- Sys.getenv('server')
port <- Sys.getenv('port')
connectionDetails <- DatabaseConnector::createConnectionDetails(dbms = dbms,
                                                                server = server,
                                                                user = user,
                                                                password = pw,
                                                                port = port)

# Now run the study
EHDENRAPredictionValidation::execute(connectionDetails = connectionDetails,
                 databaseName = databaseName,
                 cdmDatabaseSchema = cdmDatabaseSchema,
                 cohortDatabaseSchema = cohortDatabaseSchema,
                 oracleTempSchema = oracleTempSchema,
                 cohortTable = cohortTable,
                 outputFolder = outputFolder,
                 createCohorts = T,
                 runValidation = T,
                 packageResults = T,
                 minCellCount = 5,
                 sampleSize = NULL)
                 

Once you have successfully executed the study run you will find a compressed folder in the location specified by '[outputFolder]/[databaseName]' named 'resultsToShare.zip'. The study should remove sensitive data but we encourage researchers to also check the contents of this folder (it will contain an rds file with the results which can be loaded via readRDS('[file location]'). Please send the compressed folder results to Cynthia Yang c.yang AT erasmusmc.nl.

Development status

Under development. Do not use