Cohort2Trajectory

Introduction

R-package Cohort2Trajectory creates patient orientated treatment trajectories from cohorts defined in OHDSI ATLAS. The package can be used with access to OMOP CDM database. The package creates discrete and continuous time trajectories (and outputs them as .csv file) which describe patients’ treatment through time.

Usage

We have created wrappers to use the CohortContrast output for creating trajectories in Cohort2Trajectory. NOTE! It will use only the concepts which are inside the data_feature dataframe and reported in trajectoryDataList object. The same features are shown on plots if you load the CohortContrast object in the graphical user interface.

Firstly, we have to initiate the database connection and load (or create) the CohortContrast object.

# Initiate database connection & load data
pathToResults <- "./inst/example"
db <- DBI::dbConnect(duckdb::duckdb(), dbdir = CDMConnector::eunomia_dir("GiBleed"))
cdm <- CDMConnector::cdm_from_con(db , cdm_name = "eunomia", cdm_schema = "main", write_schema = "main")
data <- readRDS("./inst/example/example.rds")

Secondly, we have to import the data for the trajectories from the database.

# Import raw data for trajectory creation
data <- CohortContrast::createC2TInput(data, cdm)

Thirdly, we will call Cohort2Trajectory execution function. It has many parameters, but an easy example is given here, for more info check references.

# Create trajectories
trajectories <- CohortContrast::C2TCaller(
  data,
  pathToResults
)

This will output a trajectory dataframe as well as save this a a csv file to the pathToResults path.