ACO Mental Health Population Size Analysis Case Study

Client question

A client needed to estimate the size of the mental health patient population associated with each ACO. The client wanted ACO-level counts that could support market sizing, outreach planning, product prioritization, and comparison across organizations.

Data foundation

FastHSR used Medicare claims, beneficiary enrollment information, ACO reference data, provider identifiers, diagnosis codes, procedure codes, place-of-service information, and pharmacy data where medication-based definitions were needed. The output was built at the beneficiary level first, then summarized to the ACO level.

• ACO universe: identify active ACOs and relevant contract/entity identifiers for the measurement year.
• Beneficiary attribution: assign beneficiaries to ACOs using available ACO assignment data or claims-based attribution logic.
• Enrollment checks: confirm eligibility, coverage periods, and observation time before counting beneficiaries.
• Claims sources: use inpatient, outpatient, carrier, post-acute, and pharmacy data as needed for cohort definitions.

Mental health cohort definitions

The analysis created multiple mental health cohort definitions so the client could compare a strict diagnosed population with broader treated or service-based populations. This avoided relying on a single definition that might undercount or overcount the market.

• Diagnosed mental health population: beneficiaries with qualifying ICD-10 diagnosis codes on claims.
• High-need cohort: beneficiaries meeting configurable diagnosis or utilization logic for more intensive behavioral health needs.
• Common mental health cohort: beneficiaries meeting configurable diagnosis patterns for broader mental health needs.
• Treated behavioral health population: beneficiaries with qualifying behavioral health visits, evaluations, or other relevant services.
• Medication-supported definitions: optional flags based on Part D fills for relevant therapeutic classes, when useful for sensitivity analysis.

ACO-level counting logic

FastHSR counted unique beneficiaries rather than claim lines. Each beneficiary was assigned mental health flags, service-use flags, and ACO attribution fields before aggregation. This made it possible to calculate both population size and prevalence rates for each ACO.

• Unique beneficiary counts by ACO.
• Total attributed population denominator for each ACO.
• Mental health population count and rate per ACO.
• Configurable cohort counts for selected mental health categories.
• Treated-patient counts based on behavioral health service utilization.
• Optional geographic cuts by beneficiary county, state, or market.

Quality checks

Because ACO-level population sizing can be sensitive to attribution and clinical definitions, FastHSR built quality checks into the workflow before delivering final tables.

• Compare ACO attributed-population totals with expected ACO reference data.
• Review diagnosis logic across claim types and care settings.
• Check for double counting when beneficiaries have multiple diagnoses or service types.
• Apply cell suppression and small-count handling where needed.
• Produce sensitivity cuts under narrow, broad, and treated-population definitions.

Findings

The analysis showed meaningful variation across ACOs in mental health population size and prevalence, depending on the cohort definition and denominator used. The final outputs helped the client identify where the largest attributed populations were located and where additional segmentation was needed before outreach or investment decisions.

The report emphasized that ACO-level counts should be interpreted with attention to attribution method, measurement year, encounter completeness, diagnosis coding patterns, and whether the client wanted a diagnosed, treated, or broader behavioral health population.

Deliverables

• ACO-level mental health population size table.
• ACO-level denominator and prevalence rate table.
• Configurable cohort counts for selected mental health categories.
• Treated behavioral health population counts based on service utilization.
• Optional geography cuts by state, county, or market.
• Suppression-ready output for small cells and external reporting.
• Documentation of cohort definitions, attribution logic, and sensitivity analyses.

Use cases

• Behavioral health market sizing by ACO.
• ACO outreach and partnership prioritization.
• Mental health population segmentation.
• Product planning for behavioral health vendors and care management organizations.
• ACO-level benchmarking of diagnosed and treated mental health populations.

Frequently asked questions

Why count beneficiaries instead of claims?

Population sizing requires unique patient counts. Claims are used to identify diagnoses, services, and medications, but the final ACO-level output counts beneficiaries.

Why create multiple mental health definitions?

A strict diagnosis-only definition may miss treated patients, while a broad service-based definition may capture a larger behavioral health population. Multiple definitions let clients understand the range.

Can this analysis be repeated over time?

Yes. The same cohort and attribution logic can be applied by year or quarter to track changes in ACO-level mental health population size and prevalence.

For ACO mental health population sizing, behavioral health market analysis, or Medicare claims cohort development, please email us.

See more case studies