Session Time: 12:30pm-2:00pm
Location: Research Hub - Kiosk 6
Disclosures: Monique Diaz, MD: Nothing to disclose
Objective: Our objective is to measure and represent outcomes in stroke care. This project uses SQL to query an electronic data warehouse (EDW) at UI Health (Chicago) and strategically create an enterprise dashboard to tell the story of patient reported outcomes and functional independence post-stroke as they relate to acute interventions.
Design: We created an interdisciplinary team of a clinicians and biovisualization experts. After querying the EDW we built data dictionaries and ontologies specific to stroke rehabilitation. From this we were able to build a common data model (CDM) for both operational and future research use. The CDM guided us as we wire-framed an entity relationship diagram for a stroke database. The resulting tables fed into Tableau data-visualization software to create the enterprise dashboard.
Setting: Urban academic medical center
Participants: The operational definition of stroke patients for this project is: Acute stroke (both hemorrhagic and ischemic) presenting to UI Health emergency department (ED) or transfer from outside ED with financial quarter year time points.
Interventions: Deployment of enterprise stroke dashboard, including illustration of average Glasgow coma scales on admit and time from symptoms onset to ED arrival in the context of Chicago(land) neighborhoods.
Main Outcome Measures: In addition to measuring dashboard views we will track improvements in stroke metrics as this data becomes widely available to hospital staff.
Results: Early stroke intervention translates to improved average cerebral artery patency (mTICI 2+) and coordinated tPA administration 20 minutes on average prior to UIH ED arrival. Dashboard creation encourages requests from physicians to view their individual performance metrics.
Conclusions: Functional outcomes after stroke such as Modified Rankin Scores should, in our opinion be visualized beyond number representation with simple, but compelling graphics. Clusters of late-show stroke presentations by neighborhood motivates providers to target stroke education in those communities. Greater physician engagement in value-based care can be encouraged by enterprise dashboards.
Level of Evidence: Level V
To cite this abstract in AMA style:Diaz M, Krasna D. Data Architecting for Value-based Stroke Care [abstract]. PM R. 2019; 11(S2)(suppl 2). https://pmrjabstracts.org/abstract/data-architecting-for-value-based-stroke-care/. Accessed December 3, 2023.
« Back to AAPM&R Annual Assembly 2019
PM&R Meeting Abstracts - https://pmrjabstracts.org/abstract/data-architecting-for-value-based-stroke-care/