Enrichment Analysis of Clinically Relevant Concepts in Common Data Model Cohort Data

Identifies clinically relevant concepts in Observational Medical Outcomes Partnership Common Data Model cohorts using an enrichment-based workflow. Defines target and control cohorts and extracts medical interventions that are over-represented in the target cohort during the observation period. Users can tune filtering and selection thresholds. The workflow includes chi-squared tests for two proportions with Yates continuity correction, logistic tests, and hierarchy and correlation mappings for relevant concepts. The results can be optionally explored using the bundled graphical user interface. For workflow details and examples, see < https://healthinformaticsut.github.io/CohortContrast/>.

CohortContrast

The goal of CohortContrast is to facilitate the comparison between cohorts in specified domains across all OMOP CDM datasets. It enables users to analyze and visualize the contrast between target and control cohorts effectively.

Installation

The development version of the package from GitHub:

# install.packages("devtools")
devtools::install_github("HealthInformaticsUT/CohortContrast")

Usage

For complete setup and workflow code, use:

1. Project wiki (setup + workflow): https://github.com/HealthInformaticsUT/CohortContrast/wiki
2. Package site (function reference + vignettes): https://healthinformaticsut.github.io/CohortContrast/

If you are running CohortContrast on an air-gapped server, see the wiki article here: https://github.com/HealthInformaticsUT/CohortContrast/wiki

Outputs

The CohortContrast package generates the following outputs:

1. Running CohortContrast returns a list of tables (patient level summarised data for target and control) and saves a study folder with parquet files that can be analysed in the GUI directly.
2. Using viewer helpers with runCohortContrastViewer generates plots from parquet-formatted results.
3. Example studies are available under inst/example/st.
4. Public study outputs can be explored at http://omop-apps.cloud.ut.ee/CohortContrast/.
5. A larger example with 1080 diagnosis-level outputs is available in the CohortContrast Atlas at http://omop-apps.cloud.ut.ee/CohortContrastAtlas/.

More information

CohortContrast provides much more insight generation possibilities. See the package site for details: https://healthinformaticsut.github.io/CohortContrast/

For feature requests create issues on Github or contact Markus Haug ([email protected]) personally. Bug reports can be opened at https://github.com/HealthInformaticsUT/CohortContrast/issues.