The EMERSE Research Network allows participating sites to securely search the free text repositories at all other participating sites to support cross-institutional cohort identification. This capability is extremely important for some types of research including studies involving rare diseases for which no specific ICD-10 or other code exists. For those familiar with the i2b2/ACT network, the EMERSE Research Network works in a similar manner except that with EMERSE the source data are unstructured free text notes, not structured data. No identifiable information ever leaves a site — only obfuscated counts travel across the network. And thresholds can be set to ensure that small cell counts cannot be obtained.

Screenshot of the Results page for the EMERSE Research Network. Also shown is the threshold feature (< 10) for small counts


Joining the network is easy: a simple Terms of Use document must be signed by each participating institution. Not even a Data Use Agreement is needed since the information moving across the network is not considered to be human subjects data. From a technical perspective joining is also easy: once EMERSE is installed for local use, no additional software has to be installed, and no firewall changes are needed; only a few settings need to be configured and turned on.

This network was built with security in mind, and multiple security mechanisms are in place. The system is HIPAA-compliant. All network queries are logged. For additional technical details, see our Networking Guide. For user-focused details, see our User Guide.

While the network is currently small, we hope to have more sites join us in the near future. Interested in EMERSE or in joining the network? Contact Us.

Participating SiteDate JoinedDetails
University of Michigan 2022-03 All patients at the University of Michigan
University of Kentucky 2022-03 Only patients from the Markey Cancer Center