Data Object Exchange (DOEx) as a Method to Facilitate Intraorganizational Collaboration by Managed Data Sharing: Viewpoint

Abstract

Background To help reduce expenses, shorten timelines, and improve the quality of final deliverables, the Veterans Health Administration (VA) and other health care systems promote sharing of expertise among informatics user groups. Traditional barriers to time-efficient sharing of expertise include difficulties in finding potential collaborators and availability of a mechanism to share expertise. Objective We aim to describe how the VA shares expertise among its informatics groups by describing a custom-built tool, the Data Object Exchange (DOEx), along with statistics on its usage. Methods A centrally managed web application was developed in the VA to share informatics expertise using database objects. Visitors to the site can view a catalog of objects published by other informatics user groups. Requests for subscription and publication made through the site are routed to database administrators, who then actualize the resource requests through modifications of database object permissions. Results As of April 2019, the DOEx enabled the publication of 707 database objects to 1202 VA subscribers from 758 workgroups. Overall, over 10,000 requests are made each year regarding permissions on these shared database objects, involving diverse information. Common “flavors” of shared data include disease-specific study populations (eg, patients with asthma), common data definitions (eg, hemoglobin laboratory results), and results of complex analyses (eg, models of anticipated resource utilization). Shared database objects also enable construction of community-built data pipelines. Conclusions To increase the efficiency of informatics user groups, a method was developed to facilitate intraorganizational collaboration by managed data sharing. The advantages of this system include (1) reduced duplication of work (thereby reducing expenses and shortening timelines) and (2) higher quality of work based on simplifying the adoption of specialized knowledge among groups.