MENU

Click here to

×

Are you sure ?

Yes, do it No, cancel

From Data Extraction to Abstraction: Lessons Learned in Acquiring and Interpreting High Quality Clinical Data

M Schmidt*, D Caruthers , F Reynoso , W Bosch , C Robinson , J Kavanaugh , J Hilliard , N Knutson , S Mutic , G Hugo , Washington University School of Medicine in St. Louis, Saint Louis, MO

Presentations

(Sunday, 7/14/2019) 4:00 PM - 4:30 PM

Room: Exhibit Hall | Forum 5

Purpose: Enterprise level software solutions for gathering clinical and research data require an understanding of disparate data sources, the stakeholders and end-users of the application, the nature of retrospective data querying, and specifications for output format. We sought to characterize a project backlog that is defined by clinical and research use cases, the data sources for this backlog and create a minimally viable software application to test these cases.

Methods: A request for project proposals throughout the department yielded a compelling number of requests centered on gathering data. 14 of 21 total requests for projects submitted would require some programmatic data extraction with 10 of the 14 being considered categorically as “Planning Consistency and Automation�. A multi-disciplinary task-group was formed to develop a systematic approach characterizing the inputs, outputs, and quality of relevant data sources that address this category.

Results: A multi-disciplinary approach to such effort is necessary in order to reduce duplicative efforts and identify potential synergies. Representatives from IT, informatics, medical physics and radiation oncology often request data through parallel channels and from varying sources. Each source of clinical information, its data integrity and consistency across multiple clinical systems was well documented and accounted for within a custom application build. The user interface was specifically designed for customization and templating of requested information, and a configurable output schema was introduced for the varying types of data requested.

Conclusion: Clinical networks increasingly need to utilize the potential knowledge to be gained from their clinical experience. An inclusive approach to data collection is possible for most clinics, but there are some obstacles to consider before embarking on a task of this magnitude.

Keywords

Data Acquisition

Taxonomy

Education: Knowledge of methodology

Contact Email