Building a data aggregation engine using PostGreSQL to identify opportunities to improve clinical documentation at an academic medical center
Dr. Janos G. Hajagos is currently research assistant professor and Chief of Data Anayltics in the newly formed department of Biomedical Infomatics at Stony Brook University. Previously, he was the lead data analyst for a unique partnership between SUNY and the New York State Department of Health. He has a Ph.D. in Ecology and Evolutionary Biology and has published widely from risk analysis to applications of the semantic web to healthcare.
No video of the event yet, sorry!
The process of assigning diagnosis codes to an inpatient hospitalization requires a coordinated effort across a medical center’s clinical and administrative domains. Coded data is increasingly being used to evaluate quality metrics, such as, length of stay and mortality. The Clinical Documentation Improvement (CDI) team supports the medical review of cases to improve coding quality.
An integrated data engine was built to more accurately identify cases for CDI team review. In this talk, I will demonstrate how the capabilities of PostGreSQL including JSONB support was used to develop a system which integrates both administrative and EHR (Electronic Health Records) data sources. Part of the pipeline includes applying trained machine learning models to the processed data. I will discuss practical issues including how to build a system to meet HIPAA security requirements in a time and cost sensitive environment.
- 50 min
- PostgresConf US 2018
- Regulated Industry Summit