Value-Based Care

Population Health Analytics for ACOs: Risk Stratification That Actually Drives Care Management

By the Vizier Editorial Team  ·  May 19, 2026  ·  10 min read

Risk stratification is only valuable if it routes a real care manager to a real patient this week. The models, the workflow, and the analytics that matter.

Risk stratification is one of the most-discussed and least-operationalized capabilities in ACO analytics. The score is easy to compute; routing the right care manager to the right patient at the right time is the hard part. The patterns that work share a small number of properties — none of them about the model.

What good risk stratification produces

Three concrete outputs that change behavior:

  • A weekly “new high-risk” list — patients who weren't high-risk last week and are now.
  • A persistent “chronically high-risk” list with care management assignment and last-contact date.
  • An “emerging risk” list — patients in the middle quintile with concerning trajectory features.

Everything else (PMPM cost projections, risk-adjusted utilization patterns, etc.) is supplementary. The three lists drive the workflow.

The model is the easy part

Whether the underlying model is HCC-weighted, claims-derived, EHR-derived, or a combination, the choice matters less than commonly assumed. The well-validated public models (LACE, HCC, ACG, CCI) all perform within 5-10% of each other on health-system populations. Picking the “best” model is less important than building the workflow that acts on it.

What does matter operationally

  1. Frequency. Weekly refresh of the risk stratification list. Monthly is too slow; daily is overkill and trains the care management team to ignore noise.
  2. Care manager assignment. Each high-risk patient has a named owner. Without assignment, the list is a report; with assignment, it's a workflow.
  3. Touch tracking. When was the last care management contact? Patients with no contact in >30 days are either over-stratified (don't need contact) or being missed.
  4. Outcome feedback. Did the intervention work? Patients de-risked off the list because their condition stabilized vs. patients who escalated despite contact tell you different things.

The cohort views ACOs actually use

  • High-risk + no recent primary care visit → outreach for AWV / problem-list refresh.
  • High-risk + recent ED visit → care manager follow-up within 72 hours.
  • High-risk + recent inpatient discharge → TCM workflow within 7 days.
  • High-risk + chronic disease + non-adherent medication → pharmacy outreach.
  • High-risk + multiple specialist visits → care coordination check.

Where Vizier fits

Vizier's ACO module ships with weekly risk stratification, care manager assignment, and the five cohort views above. Pair with attribution analytics and quality measure tracking for the full reconciliation picture.

See Vizier vs Arcadia for ACO analytics for the head-to-head if you're evaluating dedicated population health platforms.

Related on Vizier

See Vizier with your data.

Direct EHR connectors. Plain-English queries. BAA in 1 business day. Bring an export or wire up a connector — answer in 60 seconds.

Request a Demo →See EHR Connectors