Start Submission Become a Reviewer

Reading: Transforming the Premier Perspective® hospital database to the OMOP Common Data Model

Download

A- A+
dyslexia friendly

Empirical research

Transforming the Premier Perspective® hospital database to the OMOP Common Data Model

Authors:

Rupa Makadia ,

Janssen Research & Development
X close

Patrick B. Ryan

Janssen Research & Development
X close

Abstract

Background: The Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) has been implemented on various claims and electronic health record (EHR) databases, but has not been applied to a hospital transactional database. This study addresses the implementation of the OMOP CDM on the U.S. Premier Hospital database.

Methods: We designed and implemented an extract, transform, load (ETL) process to convert the Premier hospital database into the OMOP CDM. Standard charge codes in Premier were mapped between the OMOP version 4.0 Vocabulary and standard charge descriptions. Visit logic was added to impute the visit dates. We tested the conversion by replicating a published study using the raw and transformed databases. The Premier hospital database was compared to a claims database, in regard to prevalence of disease.

Findings: The data transformed into the CDM resulted in 1% of the data being discarded due to data errors in the raw data. A total of 91.4% of Premier standard charge codes were mapped successfully to a standard vocabulary. The results of the replication study resulted in a similar distribution of patient characteristics. The comparison to the claims data yields notable similarities and differences amongst conditions represented in both databases.

Discussion: The transformation of the Premier database into the OMOP CDM version 4.0 adds value in conducting analyses due to successful mapping of the drugs and procedures. The addition of visit logic gives ordinality to drugs and procedures that wasn’t present prior to the transformation. Comparing conditions in Premier against a claims database can provide an understanding about Premier’s potential use in pharmacoepidemiology studies that are traditionally conducted via claims databases.

Conclusion/Next steps: The conversion of the Premier database into the OMOP CDM 4.0 was completed successfully. The next steps include refinement of vocabularies and mappings and continual maintenance of the transformed CDM.

How to Cite: Makadia R, Ryan PB. Transforming the Premier Perspective® hospital database to the OMOP Common Data Model. eGEMs (Generating Evidence & Methods to improve patient outcomes). 2014;2(1):15. DOI: http://doi.org/10.13063/2327-9214.1110
Published on 11 Nov 2014.
Peer Reviewed

Downloads

  • PDF (EN)