Skip to main content

Frequently asked questions about the Carolina Data Warehouse for Health including: What systems feed into the CDW-H? What data are in the CDW-H? What are the characteristics of the population of patients in the UNC CDW? How much history is available in the CDW-H? What institutions are represented in the CDW-H? and more...

CDW-H Service Line Guide for Researchers

The NC TraCS IDSci program is excited to introduce a new model for the Carolina Data Warehouse for Health (CDW-H) data request service.

We are organizing our work into distinct service lines: Self-Service, Self-Service Plus, Standard Data Extraction, and Custom Data Extraction.

Our aim with Service Lines is to continue to offer the services you are familiar with – providing lists of patients potentially eligible for your trial and datasets for secondary analyses – in a way that is more sustainable and easier to understand. This document provides detail about the offerings in each service line.

[Updated: 11/18/25]

1. Self Service

Cost: None

Options:

2. TriNetX Re-ID Service

With this service, you can build your own query in TriNetX, and TraCS can provide you with an identified list of patients in that cohort. Information on cost, requirements, and data available can be found on the TriNetX Re-ID page.

3. Standard Data Extraction

With Standard, we offer a set "menu" of commonly requested data domains and variables, available in one of three formats: OMOP, PCORnet, or Epic Clarity. (All of these are sourced from UNC Health data.)

Requirements: approved IRB, approved data request, moderate funding (covering, on average, 20-40 hours of effort)

Cost: Effort is billed at the going hourly rate.

Available Add On: Recurring queries - Through our autoreporter tool, we can provide automated, recurring reports at standardized intervals (daily, weekly, etc.). Note that we are unable to make revisions as part of the recurring report process.

Constraints and limitations: Standard does not support customizations to the data output or additions to the options available in our defined menu. Standard requests are designed to be supported with low to moderate effort. As such, only projects with a planned hours estimate of < 40 hours will be considered Standard. (Requests with higher estimates will be supported by the Custom service line.)

Options:

OMOP Recommended option for most projects that do not require near real-time data.

  • Benefits: Includes data from both Epic and legacy; well-documented data model; format standardized across other institutions that use OMOP
  • Limitations: 1 month latency
  • Data available: visit learn-idsci.unc.edu
  • Sample data available by request
  • Available add-on: Standard Sidecars -- Sidecars refer to the practice of enhancing OMOP data extract with additional data outputs from Epic Clarity. We offer a few commonly needed sidecars as part of the Standard service line: identifiers to support chart review or recruitment, death cause, appointment no-shows, future appointments, ADT transactions, visit-based payer information, care site details, and care provider details.

PCORnet Primarily used for PCORnet studies.

  • Benefits: Includes data from both Epic and legacy; documented data model; format standardized across other institutions that use PCORnet
  • Limitations: 1 month latency
  • Data available: visit learn-idsci.unc.edu
  • Available add-on: Standard Sidecars — Sidecars refer to the practice of enhancing PCORnet data extract with additional data outputs from Epic Clarity. We offer a few commonly needed side cars as part of the Standard service line: identifiers to support chart review or recruitment, appointment no-shows, future appointments, facility details, provider details.

Epic Clarity

  • Benefits: Near-real time data (1 day latency)
  • Limitations: limited documentation, format less standardized across institutions, no legacy data (i.e., no data available before Epic Go Live on 4/4/2014)
  • Data available: Available upon request
  • Available add-on: Recruitment via MyChart Foundation and Message (extra approvals required)

4. Custom

Requirements: Approved IRB, approved data request, substantial funding to cover at least 40-80 hours of effort. Effort for Custom requests varies widely depending on your needs. Very high effort requests (100+) may require review by TraCS leadership and justification for the custom requirements.

Cost: Effort is billed at the going hourly rate.

Available Add On: Recurring queries - Through our autoreporter tool, we can provide automated, recurring reports at standardized intervals (daily, weekly, etc.). Note that we are unable to make revisions as part of the recurring report process.

Constraints and limitations: Even within the Custom service line, we are not able to support all customizations to data requests. For example, we cannot rename variables per your specifications, "roll up" variables, or otherwise custom format your dataset.

Example Services

Due to the nature of the Custom service line, we do not have a standard list of options. Below are examples of the type of services that would fall into the Custom service line.

  • Requests for data domains or variables not supported in Standard (e.g., notes, cancer registry, surgical, obstetrics)
  • Data linkages
  • Advanced phenotyping/high complexity cohort definitions (think high-specificity)
  • Custom calculated fields or outputs
  • Projects with multiple rounds of iteration

Services Not Supported

TraCS is responsible for providing CDW-H data extracts for hundreds of researchers a year. In order to meet this demand, we are unable to support all data and informatics needs. Below is a non-exhaustive list of services we are unable to support.

  • Building of study-specific infrastructure
  • "Rolling-up" of variables (e.g., patient has diabetes diagnosis y/n)
  • Custom formatting or custom naming of variables
  • Customized free text parsing/natural language processes
SUBMIT A REQUEST

related services

more information