3. What is i2b2?
Developed at Harvard, Integrating Biology and the Bedside (i2b2) is a suite of applications
that organizes and mines clinical data in a repository and allows researchers to ask
questions about data stored in the repository https:/www.i2b2.org.
4. What is SHRINE?
The Shared Health Research Information Network (SHRINE), also developed at Harvard, is a
web-based distributed query tool that allows investigators to determine the aggregate number of
patients seen at participating hospitals who meet clinical criteria of interest. SHRINE technology
links the five UC i2b2 instances to allow researchers to conduct queries across the 12 million
UC patient records.
5. Who is authorized to use Data Explorer?
UC researchers and faculty who are authorized to access clinical informatics tools can use
1. What data is currently available in Data Explorer?
Data Explorer is currently populated with the following de-identified data:
|Laboratory Orders||150+ top UC ordered labs (does not include lab values).|
|Medications||A proof of concept of (4) medications are included.|
2. What is the source of the Data Explorer data?
Data are derived from administrative records and electronic health records of unique patients
that received services at participating UC biomedical centers.
3. How is the data transformed?
The sources of information for the Data Explorer are de-identified data-sets that are extracted
from each UC biomedical institution's clinical data warehouse, transformed into a common
data representation, and stored in a separate, dedicated data repository at each institution.
The patient information from each site is de-identified using recognized best practices.
4. How many patient records are included in UC Data Explorer?
Data Explorer currently has 12 million± patient records representing the combined patients of
the five UC medical centers.
5. How many years of data are included in the UC ReX Virtual Clinical Repository?
Years of Data Loaded by campus:
Davis: 10 years
Irvine: 20 years
Los Angeles: 7 years
San Diego: 8 years
San Francisco: 20 years
6. Why do I receive a slightly different count as a result of running the same query?
A random number is intentionally added or subtracted to obfuscate the results especially
for diagnoses or procedures with dates. This obfuscation is necessary to ensure patient
7. How is patient information protected?
SHRINE architecture has security requirements that ensure queries are only broadcast
to other authorized databases in the Data Explorer network.
a. Only network sites can communicate with other sites in the network via a secure public
b. SHRINE is a secure network independent of each institutions production security. If a
machine housing the SHRINE application is somehow compromised, the primary firewall
of the institution could provide additional security.
c. SHRINE is a peer network based on secure IP addresses. Traffic that does not originate
from one of the authorized peer IP addresses is not granted access.
d. A password secure account is required to access the de-identified data in the Data
Explorer and permissible users are only provided counts as a result of the query.
Each medical center's de-identified repository is deployed on a dedicated i2b2 instance
that is physically and logically separate from the institution's Clinical Data Warehouse (which includes PHI and HIPAA identifiers). All procedures used for the extraction, transformation
and loading (ETL) of data are performed on the Clinical Data Warehouse server, under
supervision of medical center personnel. All patient identifiers are removed before loading
data into the i2b2 repository. Users access the i2b2 repositories at the 5 UCs only via the
SHRINE instance at their own institution.