EMERSE: Electronic Medical Record Search Engine

EMERSE is a search engine that facilitates the use of free-text documents in medical records (clinical, radiology, pharmacy, and pathology).

Developed at the University of Michigan, EMERSE is an open-source search engine that facilitates the use of free-text documents in medical records (vs billing/diagnostic codes). The CICB/University Hospitals collaboration participates, together with an expanding group of consortium members, including the University of Cincinnati, University of North Carolina, Columbia University, the University of Kentucky, Dana Farber Cancer Institute, MD Anderson, City of Hope, and University of California San Diego. This scale facilitates the implementation of cohort building that can result in thousands of study subjects.

Estimates are that at least half of the information in clinical records is in an unstructured format. While Natural Language Processing (NLP) has frequently been used as a tool, EMERSE makes it easier to identify cases, and put together longitudinal trajectories of patient records via simple searches, rather than extended via programming.

Implementing EMERSE in the greater Cleveland area

Currently in a pilot phase at University Hospitals, EMERSE will be launched system-wide at UH in the fall of 2021.  In early 2022, we will announce how EMERSE can be used by CWRU researchers in conjunction with our University Hospitals partners.

This implementation covers more than 1.5 million patients, with more than 33 million searchable clinical notes that are updated via nightly feeds from the EHR.  The search engine allows for patient searches using specified parameters (or term bundles), including advanced searches based on Boolean Logic. The inclusion of a clinically curated set of  >1.5 million terms and phrases allows investigators to search across the entire patient population in one pass rather than having to perform and then combine the results of multiple searches. The search engine also allows for exclusion of criteria and the inclusion of negation.

Mark Beno, MSM, and Paola Saroufim, PharmD, MPH, presented on using EMERSE at the 2022 AMIA conference. View their presentation below.

Learn more about EMERSE here.

 

Core Contact:

Mark Beno, MSM, Senior Director of Strategic Operations

cicb_info@case.edu

Back to Computational Phenotyping Core

Bioinformatics Core

Data Management Core