Domain

Clinical Informatics

Type

Conceptual or Process Model/Framework

Theme

effectiveness; population

Start Date

7-6-2014 10:25 AM

End Date

7-6-2014 11:45 AM

Structured Abstract

Introduction

Recent developments in establishing observational health data networks have focused on the use of a common data model and the foundational infrastructure to enable observational research. While this work constitutes an important means, it does not in and of itself represent an end, as the ultimate objective is to generate reliable evidence about disease natural history, healthcare delivery, and the effects of medical interventions. To advance the field of observational health data sciences, we introduce a new open source analytics framework designed to leverage advances in health information standardization and allow organizations to realize the benefits of this standardization in the form of reporting and visualization of summary statistics across various types of health information.

Background

No single observational data source is likely to be sufficient to meet all expected outcome analysis needs, so there is interest in assessing and analyzing multiple data sources concurrently. In attempting to analyze multiple data sources their disparate formats made impossible to use standardized methods or generate reproducible results. Analyzing these disparate formats would require de novo programming each time a new potential data source was identified which introduced the opportunity for inconsistency across analyses. The Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) enabled organizations to standardize multiple observational health information data sources from disparate formats to a standard data model. The CDM provided a layer of abstraction upon which the OMOP community was able to successfully develop and execute large-scale statistical analyses capable of enabling active medical product safety surveillance and comparative effectiveness research across medical treatments.

Innovation

We will demonstrate how building upon the foundation of the OMOP CDM allows for the development of an open source analytics framework capable of providing standard reports to help describe cohorts identified from the populations contained across networks of observational health information data sources. This analytics framework includes access to open source programs that provide the ability to characterize each cohort with respect to demographics, comorbidities, concomitant medications, outcome rate summaries, procedures, risk factor analysis, and risk identification. We will also demonstrate a frontend web interface that provides access to these results through interactive visualization.

Lessons Learned

Standardized analytics can be designed and implemented to enable real-time exploration of population-level summary statistics across an international network of patient-level observational health datasets. The open source analytics framework can be applied across disparate computing environments once data are standardized to the OMOP Common Data Model and is scalable to support an array of different reports to facilitate investigations into large collections of diseases, drugs, and procedures.

Next Steps

Observational Health Data Sciences and Informatics (OHDSI, http://ohdsi.org) is releasing this open source analytics framework to the community, enabling all those groups that have standardized their data assets on the OMOP CDM to have immediate access to the insights provided by the various capabilities of the framework. We are looking to engage the open source community and leverage the broad expertise across the industry to further enhance the framework capabilities.

Acknowledgements

N/A

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License.

Share

COinS
 
Jun 7th, 10:25 AM Jun 7th, 11:45 AM

Establishing an Open Source Informatics Framework to enable Observational Health Data Sciences

Introduction

Recent developments in establishing observational health data networks have focused on the use of a common data model and the foundational infrastructure to enable observational research. While this work constitutes an important means, it does not in and of itself represent an end, as the ultimate objective is to generate reliable evidence about disease natural history, healthcare delivery, and the effects of medical interventions. To advance the field of observational health data sciences, we introduce a new open source analytics framework designed to leverage advances in health information standardization and allow organizations to realize the benefits of this standardization in the form of reporting and visualization of summary statistics across various types of health information.

Background

No single observational data source is likely to be sufficient to meet all expected outcome analysis needs, so there is interest in assessing and analyzing multiple data sources concurrently. In attempting to analyze multiple data sources their disparate formats made impossible to use standardized methods or generate reproducible results. Analyzing these disparate formats would require de novo programming each time a new potential data source was identified which introduced the opportunity for inconsistency across analyses. The Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) enabled organizations to standardize multiple observational health information data sources from disparate formats to a standard data model. The CDM provided a layer of abstraction upon which the OMOP community was able to successfully develop and execute large-scale statistical analyses capable of enabling active medical product safety surveillance and comparative effectiveness research across medical treatments.

Innovation

We will demonstrate how building upon the foundation of the OMOP CDM allows for the development of an open source analytics framework capable of providing standard reports to help describe cohorts identified from the populations contained across networks of observational health information data sources. This analytics framework includes access to open source programs that provide the ability to characterize each cohort with respect to demographics, comorbidities, concomitant medications, outcome rate summaries, procedures, risk factor analysis, and risk identification. We will also demonstrate a frontend web interface that provides access to these results through interactive visualization.

Lessons Learned

Standardized analytics can be designed and implemented to enable real-time exploration of population-level summary statistics across an international network of patient-level observational health datasets. The open source analytics framework can be applied across disparate computing environments once data are standardized to the OMOP Common Data Model and is scalable to support an array of different reports to facilitate investigations into large collections of diseases, drugs, and procedures.

Next Steps

Observational Health Data Sciences and Informatics (OHDSI, http://ohdsi.org) is releasing this open source analytics framework to the community, enabling all those groups that have standardized their data assets on the OMOP CDM to have immediate access to the insights provided by the various capabilities of the framework. We are looking to engage the open source community and leverage the broad expertise across the industry to further enhance the framework capabilities.