What is i2b2?

i2b2 is an acronym for the Informatics for Integrating Biology & the Bedside (i2b2) Cohort Discovery System. UCLA i2b2 is a web-based application that enables UCLA investigators to identify potential research study cohorts using clinical data resources obtained through the UCLA Health System. Users can conduct interactive searches of data derived from patient care activities at Ronald Reagan UCLA Medical Center, Santa Monica UCLA Medical Center and other UCLA-affiliated clinics and departments.


What data is available through this tool?

Cohort counts rather than data sets are available through UCLA i2b2. These are based on de-identified data extracted from UCLA’s clinical data warehouse, transformed into a common data representation, and stored in a separate, dedicated data repository. Currently, the following search terms are available for querying: Demographics, Diagnoses, Procedures, Laboratory Tests, Area Deprivation Index (ADI), Medications, Visit Details, Vital Signs and Vital Status.

What can UCLA i2b2 not do?

UCLA i2b2 is a cohort discovery tool, meaning only patient counts can be retrieved. Individual patient-level data cannot be accessed through this tool, but it is possible to gather this information outside of i2b2. Please contact us directly at i2b2@mednet.ucla.edu if you would like individual patient-level data. These requests will be reviewed on a case by case basis.

Who can use the tool?

To access UCLA i2b2, researchers must be affiliated with a UCLA principal investigator and complete a training session.

If you don't have an account for UCLA i2b2, please email i2b2@mednet.ucla.edu to schedule a training session and receive additional information regarding the application.

Users must also agree to the terms of use.


Watch the i2b2 tutorial


Data is obtained from administrative records and electronic health records of patients that received care across the UCLA Health System (this does not include any of the LA County hospitals).

The time range in which data is available varies and depends on the data type. UCLA i2b2 currently has data on 4.5 million unique patients and is continually being updated. We will display a chart with data counts in the near future.

There is currently no option for patients to request that his/her information be excluded from UCLA i2b2, but all of the fields for every patient specified by HIPAA rules for safe harbor de-identification will be excluded from the data repository except for dates and zip codes.

ICD codes are based on billing data from both inpatient and outpatient encounters.

The Area Deprivation Index (ADI) score is a measure of the socioeconomic deprivation experienced in a geographic area. It was developed by Gopal Singh, PhD, MS, MSc using 17 different markers of socioeconomic status based on the 1990 Census Data. The current index score in UCLA i2b2 is based on the 2000 Census data. Some examples of the markers used to derive the ADI score include: income disparity, median home value, percent of families below federal poverty level and median family income. Higher index values represent higher levels of deprivation.

To learn more about the ADI score and the complete list of socioeconomic status markers, please review the Area Deprivation Index homepage

UCLA i2b2 currently has the top 150 ordered laboratory tests from each UCLA institution with the earliest lab data going back to 1992. We are working on integrating all laboratory tests using the Logical Observation Identifiers Names and Codes (LOINC) system and hope to have this available in the near future.

If you are interested in a specific test that is not in UCLA i2b2, please contact us as we are interested in learning what researchers need.

We have integrated both inpatient and outpatient medications using the Anatomical Therapeutic Chemical (ATC) Classification System hierarchy.

Yes, each patient reported by UCLA i2b2 is unique. Whether a query consists of multiple ICD-9 codes, demographics, date restrictions, etc., each patient is only counted once.

UCLA i2b2 is a cohort discovery tool and, thus, will only provide patient counts for a specific query. To go further and receive identified patient data, you will need to follow the steps below:

Obtain patient counts using UCLA i2b2. Obtain UCLA IRB approval. Obtain the patient recruitment list from the UCLA CTSI's Informatics Program.

To get more information on how to access patient-level data from UCLA i2b2, please click on the link below:

Obtaining Medical Record Data to Conduct Your Research

We are performing a couple of initial comparisons between UCLA i2b2 data, data from UHC (university hospital consortium), and the state and the results reported from UCLA i2b2 were identical to the two other data sources.

Our next step in UCLA i2b2 is to perform more in-depth data quality checks. However, please report any query results that appear inaccurate so we can investigate right away.

Release Notes

Version Release: May 2016


A web browser is required to use UCLA i2b2. The following four browsers are currently supported:

  • Mozilla Firefox
  • Microsoft Internet Explorer
  • Apple Safari
  • Google Chrome

Please note that JavaScript must be enabled on your web browser.

New Features:

Temporal Queries

The Temporal Query Tool extends the query functionalities provided by the classic Query Tool. Users can now define a sequence of events using time-based relationships to further add to their query criteria. For more details regarding the Temporal Query Tool, you can review the LADR Tutorial Video at the 6:35 mark or read instructions on using the new tool at the following link: Temporal Query Tool

Search Concepts

New search concepts are available under the Navigate Terms pane in the Terminology Subject Areas tree. The new search concepts include:

  • Visit Details: Users can include Age at Visit, Length of Stay, Visit Type and Visit Location.
  • Demographics
    • Age: Users can include Age by Month for subjects under the age of 3.
    • Marital Status: Users can include Life Partner or Other.
    • Race: Users can include Multiple Races or No Information.