Skip to main content

TriNetX EHR query tool

Image
TriNetX Logo

TriNetX is a helpful external web-based tool to assess feasibility and plan for clinical trial and other research studies, using electronic health record (EHR) data. In TriNetX, you can create a cohort of patients, then explore the cohort's summary data, and even perform analyses. Although TriNetX only provides de-identified summary data; if you share your cohort(s) with (Banner) honest (data) brokers, they can extract more detailed data.

The Center for Biomedical Informatics and Biostatistics (CB2) provides an overview of how to use TriNetX, and links to the proper authorities and experts who provide access and training, in the sections below.  

Access to TriNetX network members’ data

Currently, only Banner Health’s Cerner EHR data is accessible in TriNetX to U of A and Banner researchers.  Please note this means data from other TriNetX network members is currently not available to U of A and Banner researchers.

Log in to TriNetX

Request access to Banner's TriNetX

Watch TriNetX demonstration and training

Explore how TriNetX can help you

Use TriNetX to find patient cohort(s)

  1. Log in to TriNetX with your U of A (arizona.edu) email (and password you created for TriNetX)
  2. Click Create new Study button (top right)
    1. specify required info.:
      1. Study Name
      2. >= 1 Research Purpose
      3. Primary Therapeutic Area
    2. Click Create button (bottom right)
  3. Create and revise as many queries you need, to find and define patient cohort(s)
    1. Search Terms for “MUST…” and “CANNOT…” “…HAVE” events/criteria and Add To Query
    2. Create Groups, Add related groups to Set Relationships between groups and more
Verify cohort definition(s)

Contact CB2 to verify you found the actual patients you need, and as needed, help you refine your queries.

Assess feasibility (perform a reality check):

  1. Confirm expected number of patients meet your criteria: check the number of patients appearing (at the top) after you execute your queries

Then, from TriNetX menu on the left:

  1. Confirm you selected the patients you need: Explore your cohort’s demographics; + top diagnoses, procedures and medications

Also, as needed:

  1. Analyze your cohorts to see additional summary statistics, graphs and other visualizations
Verify appropriate number of patients

Contact CB2 to help you find the patients needed, especially if found more or less than expected.

Share your TriNetX study with others

  • So other team members can create queries for your study
  • To confirm cohort(s) are defined as expected, with team members and CB2
  • If need, once obtain approval, to request individual patient data from honest brokers (HBs)
  1. Click Study Management (from menu on the left), then click Team 
  2. Click Share Study button (top right)
  3. Search (top right) for name(s), "CB2" or "HB" then click box(es) next to resulting name(s), then click Save Team button (bottom right)
Request patient level data if needed

Contact CB2 to help request approval and share your cohorts with honest (data) brokers.

Questions?

Review answers to frequently asked questions (FAQs) about TriNetX below.

FAQs:

Anyone, including faculty, staff and students, at the University of Arizona has access to the same data and can request access to Banner's TriNetX, which only has data from Banner Health. Thus, U of A and Banner researchers currently cannot query data from other TriNetX network members. However, the pharmaceutical industry, which supports the platform, can query the network of TriNetX members to find cohorts required to meet enrollment targets.  If they find cohorts at Banner, they can send clinical trial opportunities to determine if there is an investigator interested in pursuing the project.

Only summary statistics, not individual patient level data, is available from TriNetX. If you need patient level data, CB2 can share your patient cohort(s) with Banner honest brokers, once you have approval from Banner and U of A

  1. Go to the Filters section at the top of the Query Builder to filter the entire cohort on:
    1. Demographics, inc. current* Age, and Sex, of the cohort
    2. Facility: Banner clinics: Contact CB2 for detailed lists and instructions to filter by facility and/or state
  2. Add Search Terms to the MUST HAVE and CANNOT HAVE sections for a Group in the Query Builder to currently** search for:
    1. Diagnoses
    2. Procedures, inc. Labs 
    3. Medications
    4. Visits
  3. Click Explore Cohort in a study on left side in TriNetX to see available summary data on a cohort, which currently** includes: 
    1. Demographics inc. current* age of the cohort
    2. Number and percentage of patients in the cohort by the following, but only the where >= 10 patients have the given term (to protect patient privacy)
      1. Diagnoses
      2. Procedures, inc. Labs 
      3. Medications
Please note:
*Age has restrictions and filtering is different depending on which age you need to filter.  Read answer to next FAQ for details
**Currently, only the data listed above is available for Banner patients; despite other categories of data being listed in TriNetX

To protect patient privacy: the numbers of patients with ages >89 years are grouped into a single category "Patients 90 and Older..." (seen at the bottom right of the graph of age in the Demographics section of Explore Cohort)
Filtering on age has restrictions and is different depending on which age you need to filter:
  • Ages used/shown in the following sections are the current ages of patients in the cohort (i.e. age as of the day of your search):
    • Filters at the top of the Query Builder
    • Demographics of Explore Cohort 
  • If you need to filter for age at a specific event (such as diagnosed as an adult), then:
    • Add term filters to a search term by hovering over the search term (such as a diagnosis code you chose for which you want to filter by age diagnosed)
    • Click the funnel icon which appears, then:
      • Set the age range (such as to >=18 years) 
      • Click the Save button appearing at the top of the term filter box

  1. Go to Design Assistance on bottom left of menu when in a study in TriNetX 
  2. Click Help button (question mark top right) when logged into TriNetX
  3. Read general questions about TriNetX (NOT specific to Banner's TriNetX)
  4. Sign up for or review training from TriNetX, on not just the different functions of TriNetX, but also "Best Practices" for various therapeutic areas and regions (you will see some types of data not available from Banner in TriNetX)

Ways to reuse TriNetX criteria:

  1. Create a study from a template: on the Studies page when you log in, there are templates with pre-built criteria which can be edited as needed: click on the template, then click the Create study from template button
  2. Create a duplicate of a study: on your Studies page:
    1. hover over the name of study to duplicate 
    2. click the right arrow which pops up, then click the Duplicate button 
  3. Copy/paste a group within your study, or to another study: 
    1. in your study to copy from, to the right of a group title are several icons, one looks like 2 boxes on top of each other (to the left of the trash can delete icon), click this copy icon to copy the group  
    2. paste by clicking the Paste Logic button to the right of the Create a New Group button (in existing study, or open & then paste in another study)
  4. Import a query from an existing study: in a study:
    1. click the + New Query button at top right of study into which you want to import a query
    2. click Import Query button which will be in the Build a new query box in the middle
    3. select the study, then the query to import
    4. click the Import button at bottom right
    5. if needed, copy any group you want to copy from, then click Paste Logic button to another query in study

When searching for terms, either using known diagnosis, procedure or other codes; or keywords; sometimes nothing will be found.  First make sure you are not making a typo in your search by making sure the "Show Terms with Zero Patients" box is checked at bottom left of the search results to see if the term shows up, but just with zero patients. Also make sure if you are searching using diagnosis, procedure, or other codes, include decimal points if the vocabulary uses those, such as ICD codes do, or try searching by a higher level code, i.e. just the part of the code before the decimal point. If you are seeing the term, but with zero patients, then Contact CB2 so they can check with Banner to see if those terms are just missing from TriNetX, or if there really are no Banner patients with those terms.

To keep data de-identified in TriNetX:

  • numbers shown are close estimates of the actual number of patients
  • criteria where < 10 patients have the given criteria show “10”

Time range of available data is the same as Cerner (source of the data): 2017-09-30 - present. Some facilities go back as far as 2014. If needed, Contact CB2 to find out the first year a facility has data and the visit count from that year so that you get an idea of whether it is a full year of data or just partial. 

Data in TriNetX is updated weekly:  to check when last updated, log into TriNetX, go to the Browse Network tab on the left menu and select the Banner network

Unlike most other instances of TriNetX outside of Banner, it does NOT contain data from any other sources other than Banner itself:  

Data is from Banner Health's OMOP, a common data model (CDM) database, which is derived from Banner Health's Cerner EHR:

  • What is a CDM? Data, usually from multiple data sources such as admin./billing and EHRs, stored in a standard common format, allowing queries to be executed across multiple databases, sources, and sites; and use of standard vocabularies 
  • What is the OMOP CDM? 
    • By OHDSI.org (Observational Health Data Sciences and Informatics)  – an international organization for standardizing “observational” health data (Note: OHDSI is pronounced “Odyssey,” as in Homer’s Odyssey) 
    • OHDSI created OMOP (Observational Medical Outcomes Partnership), which created the OMOP CDM, a single relational database 
    • it is “open-community” so can get answers more easily, and give feedback too

Contact CB2 if you still have questions

About the information on this page

Updated: Spring/Summer 2026

Thank you to the following for the above information and answers to FAQs: the Banner TriNetX and honest broker teams.

Disclaimer

TriNetX was not created by U of A. In order to guide researchers, CB2 at U of A reviewed TriNetX and the process to obtain EHR data from owners of EHR data outside of U of A.  The information on this page does not necessarily reflect all of the data Banner Health, or any other HCO, would have in TriNetX; nor the process to obtain access; and updates to TriNetX can occur which might make some of the information out of date. Please Contact CB2 if any instructions on this page do not work for you so CB2 can direct you to the appropriate person(s) with proper authority and verify the information.

Design and graphics by Manuel Snyder