Patient Stratification

My Folder

Product Overview

Patient Stratification integrates current and cost trends, chronic conditions,  and social determinants risk models and disparate sources to identify the individuals most likely to benefit from proactive care management programs. Users can build and analyze different stratification algorithms based on proven risk models and patient utilization to rank and ultimately determine the most important candidates for intervention through complex care management, chronic condition management, readmission prevention or other programs. (This Care Management application combines and replaces Patient Identification & Patient Risk Stratification)


  • Dynamic stratification algorithm creation and saving for targeting individuals for a variety of care programs
  • Analyze cohort or patient-specific attributes
  • Supports complex attribution to the individual’s PCP via integration with Attribution Modeler or incorporation of client-provided logic.
  • Complex patient filtering capabilities to create precise cohort registries
  • Automates hand off to care management staff for intake processing


Benefits include:

  • Expand ability to identify the most critical candidates for Care Management programs
  • Ability to tailor criteria to meet unique program requests
  • Reduce resource demands to identify individuals for Care Management programs
  • Increased timeliness of identifying individuals for Care Management programs
  • Increased closed loop capabilities to determine the most impactful algorithms for outcomes improvement


Included stratification variables:

  • Utilization (ED visits, admits, ICU stays, costs)
  • Risk (Charlson-Deyo, Readmission (LACE), HHS-HCC, Predicted, and Rising Risk)
  • Conditions (High, Moderate, Low Acuity)
  • Medications (Current number and High Risk Medications as defined by CMS)
  • Social determinants

See Sample Screenshots of Patient Stratification

Data Sources

  • Acute & Ambulatory EMR – billing, longitudinal active medication list and ADT data
  • Inpatient & Outpatient Claims – Commercial, Medicare, Medicaid
  • Pharmacy claims
  • Catalyst EDW component dependencies: CAP 3.1 or above, Attribution modeler OR client mechanism to attribute patients to physicians, Claims common if have claims sources, Risk SAM

Patient Stratification: A Deeper View


Successful health organizations understand that often around 5% of patients can drive more than 50% of spending. Controlling future costs is much more than just looking at historical costs.  Multiple factors impact the health and future cost of an individual – historical costs, chronic conditions, rising risk patterns, social and environmental factors, and behavior. Identifying the right individuals to allocate care management resources to can help avoid the higher likelihood of hospitalization, ED encounters or other high cost care.

What types of problems does Patient Stratification address?

Health organizations too often lack visibility into complete, longitudinal patient data that will help them manage their populations. Those that do attempt to identify high risk individuals spend enormous time and valuable resources, often looking at fragmented source data that doesn’t provide a holistic view of individual risk.  Some of the key but often missing elements include: identification of patients with the highest risk of requiring high cost care using multiple models, comparing in and out-of-network claims, grouping or analyzing by: care programs, encounter details, diagnosis codes, providers, age cohorts, and zip codes.

Use Case

Care Manager wants to identify:

  • The top 5% of patients who are high risk, high cost, on multiple medications and with high acuity conditions to refer to a complex care management program
  • Individuals for chronic care management programs
  • Individuals to refer to readmission reduction programs

Success Measures Examples

Opportunity Identification:

  • Identify the most critical candidates (“high risk, high cost”, etc) for Care Management