A small fraction of your Medicaid panel is quietly driving most of its cost. That isn’t a moral judgment or a data glitch—it is the defining arithmetic of the program. Medicaid cost concentration is steep, and it is stable year over year: a thin slice of high utilizers accounts for close to half of everything Medicaid spends. For a value-based practice, understanding why that slice exists, and who sits inside it, is the fastest lever you have on both quality scores and total cost of care.
The 5/50 rule isn’t a metaphor
Health-policy analysts have a shorthand for this pattern: the “5/50 rule.” Roughly the most expensive 5 percent of enrollees account for roughly half of spending. In Medicaid specifically, KFF found that fewer than 5 percent of enrollees—each exceeding $25,000 in annual costs—account for almost half of all Medicaid spending [1].
The concentration shows up just as clearly when you cut the data by who Medicaid covers. In 2023, adults 65 and older and people eligible on the basis of a disability made up about 19 percent of full-benefit enrollment but 51 percent of spending; children were 35 percent of enrollment and only 15 percent of spending [2]. Per person, that works out to roughly $20,950 a year for enrollees with disabilities and $20,194 for older adults, against $3,321 for a child [2]—close to a six-to-one gap. The label “frequent flier” describes the utilization. The economics underneath it are concentration.
Why the high-utilizer cohort exists
None of this is random, and it isn’t a billing artifact. Medicaid covers the populations with the most medically complex needs—people managing several uncontrolled chronic conditions at once, often layered with behavioral health needs and long-term services and supports. Cost follows complexity, and complexity clusters in a small number of people. Coverage churn compounds it: a member who loses Medicaid and defers care often re-enters the system through the emergency department, sicker and costlier than continuous coverage would have left them.
That reframes the “frequent flier” from a problem patient into a predictable population. The same handful of members will drive a disproportionate share of admissions, readmissions, and emergency department visits next year, too—because the underlying complexity doesn’t resolve on its own. The strategic question isn’t whether your panel has a high-cost tail. It’s whether you can see that tail before it shows up in next year’s claims run.
Finding the top 5 percent before the claims catch up
This is where risk stratification earns its keep. Waiting for a member to become a high utilizer—reading it off last quarter’s claims—means you are always managing the cohort a year late. Prospective risk models flip that: they estimate who is most likely to be high-cost going forward, so proactive care can start before the expensive event, not after it.
What the risk model actually flags
Equality Health’s stratification inside CareEmpower® runs on the Johns Hopkins ACG Case-Mix System, a population-health risk model. To land in the high-risk category, a member has to clear three bars: predicted to be in the top 10–25 percent of highest-risk members per the ACG model, greater than a 10 percent probability of hospitalization, and one or more uncontrolled or rising-risk chronic conditions in the registry—diabetes, heart disease, chronic kidney disease, asthma, COPD, or depression [3]. Because risk distribution is uneven from practice to practice, the activity-based payment targets the top 5 percent highest-risk members for every payer contract, though any given panel may run higher or lower [3].
What to do with the list once you have it
A stratified list is only useful if it changes the visit schedule. The point of identifying the top 5 percent is to move that cohort from reactive, ED-driven care to proactive, planned management—the High Risk Member Management model builds around a minimum of quarterly visits with those members [3]. Each visit is a standing chance to reconcile medications, order the labs a complex chronic patient actually needs, and close the preventive screenings that otherwise slip through [3].
None of that requires guessing which patients matter most. The stratification has already answered that question. It lets a practice spend its scarce proactive-outreach capacity exactly where the total-cost-of-care math concentrates, instead of spreading it thin across a panel where most members are, in any given year, comparatively inexpensive.
The concentration is the opportunity
Medicaid cost concentration reads like a burden—a few high utilizers absorbing half the budget. Turned around, it is the most tractable fact in value-based care. A cost problem spread evenly across thousands of members would be nearly impossible to manage. A cost problem that concentrates in a definable, findable 5 percent is a work list. Find that cohort early, build the schedule around them, and the same arithmetic that makes Medicaid expensive becomes the clearest path your practice has to better outcomes and a lower total cost of care.
References
[1] KFF. “Medicaid’s High-Cost Enrollees: How Much Do They Drive Program Spending?” https://www.kff.org/medicaid/medicaids-high-cost-enrollees-how-much-do/
[2] KFF. “A Look at Variation in Medicaid Spending Per Enrollee by Group and Across States” (2023 data). https://www.kff.org/medicaid/a-look-at-variation-in-medicaid-spending-per-enrollee-by-group-and-across-states/
[3] Equality Health. “Pathways to Better: High Risk Member Management (HRM).” Provider guide, June 2025.
