Risk Stratification and Finding Opportunity

An ACO’s data is only useful if it tells the organization where to act. Turning raw feeds into a ranked list of opportunities is the analytic core of running an ACO, and it draws directly on the population-health methods from the intro course, applied with an operator’s eye.

The stratification pyramid, revisited

The intro course introduced risk stratification; here is how an ACO uses it operationally. The attributed population is sorted into tiers:

• The high-cost, high-need apex: a few percent of patients driving a large share of spending, who get intensive care management.
• The rising-risk middle: patients whose trajectory is worsening but not yet acute.
• The stable base: generally healthy patients who need reliable prevention and access.

Worth remembering: the evaluation course’s warning applies directly to where an ACO spends its effort. Last year’s highest-cost patients partly regress to the mean on their own, so a program aimed only at them can look effective while doing little. The rising-risk middle is often the higher-value target, because that is where proactive intervention can still bend a trajectory. A sophisticated ACO targets impactability, not just cost.

The specific opportunities to hunt

Beyond risk tiers, an ACO’s analytics look for concrete, actionable patterns:

• Care gaps: overdue screenings, uncontrolled chronic conditions, missing follow-up, the difference between guideline care and actual care.
• Avoidable utilization: emergency visits and admissions that better outpatient management could prevent.
• Leakage: attributed patients getting care outside the ACO’s network, where the ACO has less influence over cost and quality (Module 4).
• Post-acute patterns: where discharged patients go, and whether they land in efficient, high-quality settings.

From list to action

The output of all this analysis is not a report; it is a work queue. The Medicaid and intro courses made the point and it is worth repeating for ACO operators: a registry or risk list changes nothing until someone owns it and works it daily. The analytic sophistication of an ACO is measured not by its dashboards but by whether the right patient name reaches the right care manager while intervention can still help.

Key takeaways

• An ACO stratifies its population into high-need, rising-risk, and stable tiers to focus effort.
• Rising-risk patients are often the highest-value target because their deterioration is still preventable and less subject to regression to the mean.
• Analytics must produce daily work queues owned by real people, not dashboards; action, not insight, generates savings.