Cohort Builder


Product Overview

The Cohort Builder application is designed for clinicians, clinical analysts, and leaders of outcome improvement projects—people who aren’t necessarily data analysts or database experts but need to conduct complex queries of data stored in the EDW. The app allows users to identify specific populations of patients based on demographic and many clinical criteria (diagnosis, medication, lab, and orders details; and download information about these populations. Users can specify the level of detail (“grain”) they need for a cohort (patient-, episode- or encounter-centric data) and may deploy the tool in a patient de-identified configuration to facilitate study design and pre-IRB analysis.


  • Population data available at three fundamental levels of granularity: patient-centric data such as demographics, problem lists, medication, labs, and orders (e.g., asthma, beta 2 agonists, and PFT orders); episode-centric data gathered from facility billing and specific to a facility account (e.g., pregnancy); and encounter-centric data derived from a patient encounter (e.g., appendectomy).
  • Populations stratified by diagnosis, place in the clinical hierarchy (clinical program, care process family, and care process), orders, medications, labs, and demographics.
  • Exportable patient lists for all levels of granularity
  • Support for de-identified version to support research needs.
  • Future (2016) support for sharing of defined registries across Health Catalyst Analytics Platform for your health system.

See Sample Screenshots of Cohort Builder

Data Sources

  • EMR
  • Patient problems
  • Medications administered
  • Laboratory orders and results
  • Orders
  • APRDRGs (billing data)
  • ICD diagnoses and procedures

Cohort Builder: A Deeper View


Clinical quality improvement initiatives and clinical studies require precisely defined patient populations. Having a clinician-friendly way to dynamically and simply interact with the data in the EDW can make this patient cohort definition process more efficient.

What types of problems does Cohort Builder address?

The task of defining patient cohorts for quality improvement projects or clinical studies is often shared by several people working in various clinical and technical roles. Typically, the complex questions of the clinician experts must be transmitted to technical team members, who then conduct the queries and produce a custom-built report—one that will likely require several iterative reviews before it meets the needs of the clinicians. This process of iteratively defining patient cohorts can be time-consuming—and it delays progress on meaningful clinical improvement and research.

Use Cases

  • A team leading the hospital’s pediatric improvement team needs to validate a hunch that head CT scans are being overused for pediatric patients with head injuries.
  • A case manager wants to generate a list of asthma patients who smoke to invite them to participate in a tobacco cessation program.
  • An improvement team needs to identify the patients that have received naloxone and are also mechanically ventilated to screen for potential in-hospital iatrogenic respiratory failure.
  • A researcher is planning a study to investigate the use of ACE inhibitors in elderly patients with diabetes to determine the need for potassium supplementation. The researcher would like to know how many patients in the last year have diabetes, are older than age 65, have at least one order for an ACE inhibitor, and at least one serum potassium measurement.

Anticipated Improvements

  1. Decreased time and resources required to create more precise patient registries for improvement projects and clinical studies.
  2. Increased physician and leadership engagement with EDW data due to the ease and speed of “self-service” analytics.
  3. Greater ease in planning resource needs and duration for studies, due to the ability to quickly determine number of members likely to qualify for research study within a given time period.

Success Measure Examples

  • Registry identification: Identify potential $ saved from more efficient definition of patient cohorts.
  • Accelerated improvement: Lower time and resources spent on iterative population registry definitions for improvement work teams.
  • Better Information: Deliver rapid, clear insight into study recruitment potential, inclusion criteria, associated resources, and potential study timelines.