Skip to Main Content



Instructions to UW Medicine and AMC account holders on using the Leaf cohort discovery tool
warning image

Due to sensitive data in Leaf, access is restricted to users who meet certain criteria:

  1. UW Medicine AMC account
  2. Requested access through UW Medicine Analytics, with supervisor confirmation
  3. Using the Pulse Secure/Ivanti VPN if not on a hospital computer

For questions about access, please contact


Leaf logo


Leaf is a powerful, flexible, and user-friendly self-service web application that allows UW researchers to query the electronic health records within the UW Medicine data warehouse in real-time. This data can help investigators support cohort discovery answer a variety of questions about patient populations across medical and demographic information, clinical procedures, diagnoses, and department statistics.

With just a few clicks, Leaf's HIPAA-compliant drag-and-drop interface can produce de-identified data sets for researchers seeking a high-level view to gauge cohort size potential and support QI investigations, as well as identifiable datasets for those with IRB approval from UW Research's Human Subjects Division.


Accessing Leaf

Requesting Access

Due to its direct access to the university health data system, Leaf is currently only available to investigators and research teams affiliated with UW Medicine who have or are eligible to request an AMC account. New users must have completed UW Medicine's HIPAA Compliance Training and the Privacy, Confidentiality, and Information Security Agreement before requesting access. 

  1. Navigate to the UW Medicine Analytics site and click “Request Access” on the right-hand side.
  2. Complete the Requestor, Supervisor (must be affiliated with UW Medicine), and Affiliation fields.
  3. Select “Leaf” in the Tool Needed field and check "General Access" as the Data Requested.
  4. Provide a short description explaining why you are requesting access to Leaf.


Training & Orientation

Successful completion of the UW Medicine's Data Access Policy training, offered via the online learning management system, is required for all Leaf users. It can be found by searching “data access” in the search box in the upper left.

You can also access the training modules by logging through your AMC account to the UW Medicine Analytics site.


Logging onto Leaf

Once you have received confirmation and completed your training, you can access Leaf by going to


Pulse Secure VPN

warning imageUsers must have UW Medicine's Pulse Secure/Invati VPN running if accessing Leaf from off-campus. If you attempt to access Leaf without it, your browser will not go anywhere and eventually say “this site can’t be reached.”

Instructions for setting up Pulse Secure/Ivanti can be found on the UW Medicine intranet:

Identifiable/De-identified Data

Log in screen for Leaf where users choose identified or deidentified data

By default, data accessible through Leaf has been de-identified to comply with HIPAA regulations. De-identification methods include removing PHI (Protected Health Information) and date shifting.

Researchers who have IRB (Institutional Review Board) approval from a regulatory institution or board can access identifiable information by:

  1. Selecting "Research"
  2. Selecting "Yes"
  3. Providing an official IRB Number/ID, the approving institution or body, and the IRB expiration date.

Leaf Interface Overview

Overview of the Leaf search interface

Leaf uses a simple drag-and-drop interface to let users query the UW Medicine data warehouse across a wide variety of avenues.

The Concept Tree (1) serves as the basic building blocks for the search, letting users select criteria such as demographics, diagnoses, imaging, family and social history, and many more. The hierarchical concept tree is built on the Unified Medical Language System and uses internationally-recognized coding systems, such as ICD-9 and ICD-10, Healthcare Common Procedure Coding System, and Logical Observation Identifiers Names and Codes.

The concept tree can be accessed through a Search Bar (2) that can be used to search across all concepts and offers auto-complete suggestions.

To create a search, individual concepts can be dragged to one of the Panels (3) that use Boolean search operators (AND, OR, NOT). Once a concept has been dropped, users can select additional parameters such as date range and frequency (e.g. find patients who have had three or more MRIs). 

Once one or more concepts are in place, users can Run their Query (4) and see the number of Patient Results (5) within seconds. Users can then move to the Visualization and Patient List tabs (6) to dive deeper into the results.