Optimizing the migration of a data warehouse to the cloud using network analysis

Presentation Time: 10:45 AM - 11:00 AM

Abstract Keywords: Administrative Systems, Data Mining, Change Management
Primary Track: Applications
Programmatic Theme: Clinical Informatics

Passage of the HITECH Act in 2009 led to an increase in the adoption of electronic health record (EHR) and health information technology (HIT) systems by hospitals. These systems require hospitals to maintain complex data warehouses supporting clinical and operational activities. Periodically modernizing this data infrastructure is a critical task for IT departments, but this is often managed heuristically. Our team at Children’s Hospital of Philadelphia (CHOP) sought an efficient, data-driven method to assist in this modernization activity by prioritizing data warehouse assets for migration into a cloud environment. We created a network graph to capture the dependencies between assets and used graph theoretic methods to score the relative influence of each asset within the overall network. The influence score, based on centrality measures, is proportional to the number of downstream dependencies of an asset. Using this score, we proposed a data-driven and rational strategy for efficiently migrating assets.

Speaker(s):
Ashley Oliver, MPH


Author(s):
Abdul Tariq; Jake Riley, BA - Children's Hospital of Philadelphia; Hojjat Salmasian, MD, MPH, PhD, FAMIA - Children's Hospital of Philadelphia;