What can DEWARS be used for?

USC research teams can use the research data warehouse for study planning (feasibility assessments, power analyses) through the I2b2 portal. The self-service website allows them to find the number of patients at CHLA and Keck that meet your study inclusion and exclusion criteria. They can also extract identifiable patient information for screening or recruitment purposes if they have IRB approval to do so. Once the warehouse data has been validated for quality, research teams will be able to perform large-scale analyses on the data already existing in the warehouse (e.g., epidemiology studies).

How can it be used for study planning?

Users can search health system data for de-identified counts of patients who meet certain inclusion and exclusion criteria (i.e., cohort discovery). The search tool is accessible via a web browser and is self-service. Instructions are available below.

Can research teams receive a list of eligible patients and their contact information, or is only generic information available (i.e., names and MRNs)?

Depending on the protocol’s IRB approval, study teams can get a specific patient list with contact information and specific EHR data elements.

How do I access DEWARS?

The data warehouse can be accessed via web browser through the i2b2 web tool: https://calv-i2b2appdev.med.usc.edu/webclient/

To request an i2b2 account, contact the ITS HelpDesk (323-442-4444; HelpDesk at med.usc.edu). Before creating an account, our staff will verify that you are a USC or CHLA employee and have Human Subjects and HIPAA certifications. (To obtain certifications, see instructions here).

Are instructions available?

Basic instructions:

Once you log in and select a project, start a query by selecting a search term under the "Navigate Terms" tab in the upper left of the screen. Drop your desired term in the Group 1 box under Query Tool. You can drop two terms in one box to search for patients who meet either of those conditions (an OR operation), or you can drop a second term in Group 2 to search for patients who meet both conditions (an AND operation). Once you've selected your search term(s), hit Run Query and choose "Patient Count." Your results will appear below in the Query Status area. To see a breakdown of that count by demographics, check other boxes after hitting Run Query, and your results will once again appear below.

To submit a request for a dataset, run your query and select "Patient Set" as the result type. Select Analysis Tools from the bar on the top of the page. Select Project Request from the Plugins area on the bottom right of the screen. Fill in the form that appears. Under “Enter Title,” type in the title of your study AND the IRB number. Under “Cases,” you will be asked to drop patient set(s), which are accessible from the Previous Queries tab on the bottom left. Be sure to expand the folders until you see "Patient Set," and drag and drop the file into the first box. You can provide more details by following the prompts, or you can simply hit “Submit Query” and a staff member will contact you about next steps.

More detailed instructions are available from external sites:

Our CRI support desk is also available for help on your project. Please submit a ticket through the ITS HelpDesk to request support.

What is the project timeline?

We are still implementing the technology, workflows, and data cleansing procedures that will support the fully functional data warehouse (target completion December 2016), but some of the services are available currently.

Who can access it?

All investigators and research staff that currently use USC or CHLA IRB and have current certifications in Human Subjects and HIPAA can request access. Study roles in approved IRBs will determine the granularity of access. We currently have a waiver to re-distribute safe-harbor de-identified clinical data from the health system to USC and CHLA researchers.

How long does it take to get information out of it?

Queries using the i2b2 web tool are very quick. Results are typically returned in under one minute, depending on complexity. The turn-around time for dataset requests depends on many factors, including approval status of the study and type of data being requested. We are refining the process to shorten the wait time. It will be automated in the long term.

Can research teams receive a list of eligible patients and their contact information, or is only generic information available (i.e., names and MRNs)?

Depending on the protocol’s IRB approval, study teams can get a specific patient list with contact information and specific EHR data elements.

Can research teams request information regarding who the physicians are that are seeing the patients?

Yes.

How can I learn more about the i2b2 tool?

The tool is described in more detail at the NIH-funded i2b2 Center website.

Browser not supported