Implementing an AI NLP Tool to Address SDOH Needs

Tuesday, April 18 at 1:30 PM - 2:30 PM CT
South Building, Level 1 | S105 C

NorthShore, in partnership with Linguamatics, is developing a natural language processing (NLP) system to extract important insights hidden in our vast amounts of unstructured data. The initial use case is to identify and extract social determinants of health (SDOH) data. Currently we are in the nascent stages of screening for SDOH needs in a standardized format so our SDOH data is frequently buried in the clinical note text. This data collection format makes it challenging to bring the relevant information to the forefront for care teams to act on. By leveraging NLP to store SDOH data in a structured format, we can feed this information back into our EMR for care team member review and intervention. The pilot program will focus on providing our social workers a list of patients who are in the Emergency Department (ED) with an NLP-identified SDOH need. The social workers will verify the SDOH need with the patient and provide community resources or enroll the patient in care management where applicable. Where previously the social workers may not have interacted with the patient, they will now be able to intervene on any unmet social needs.

Learning Objectives

  • Describe a process for validating and improving vendor-offered Named Entity Recognition (NER) tools
  • Identify use-cases for integration of SDOH data in the care continuum
  • Design user interface in the EMR for prospectively monitoring and improving queries based on end-user feedback
Clinical Informaticist, Data Scientist, Population Health Management Professional


Danielle Jungst
Implementation Manager
Northshore University Health System
Emily Kwan
Senior Programmer Analyst
NorthShore University Health System
Urmila Ravichandran
Data Scientist
NorthShore University HealthSystem