Online Program

Return to main conference page
Thursday, May 30
Data Science Techologies
Practice and Applications
Data Science Applications E-Posters, II
Thu, May 30, 5:30 PM - 6:30 PM
Grand Ballroom Foyer

Multiple uses for chronic condition data mart (306375)

Norris Kamo, Virginia Mason 
*John Massman, Virginia Mason 
Nicholas Moy, Virginia Mason 
Sonali Tamhankar, Virginia Mason 

Keywords: healthcare, data mart, multiple stakeholders

At Virginia Mason, we built a data mart tracking four long-term conditions (diabetes, hypertension, colorectal cancer screening, statin) for quality improvement purposes. Patient-specific data is available to care providers and outreach staff through condition/disease specific registries. Summary statistics and employee performance comparison is available to physicians and management. Summaries by insurance are available to our contracting team. We worked with providers and managers to (a) identify patients in each category directly from our electronic health record (EHR) system, (b) set patient-level metrics for “out of care patient”, and (c) set physician-level goals for “high achieving physician”. We worked with physicians and outreach staff to determine the data needed to prioritize patients for immediate contact. We created a training video (narrated by a physician) for the dashboards. Data is extracted from our EHR and stored on a SQL Server data mart. Power BI dashboards are used for all reports. R and T-SQL are used for data munging. All data is refreshed nightly.