Attribution & Eligibility Reconciliation in Risk Contracts
Most performance disputes in risk arrangements are attribution disputes. See the eight failure modes — mid-month effective dates, retroactive enrollment, level-of-care mismatch — the reconciliation grain that makes a figure defensible, and a monthly cadence you can run now.
Key points
- Before you argue about an outcome, prove the population. Most performance disputes in risk arrangements are attribution disputes wearing a clinical costume.
- Coverage starts or ends inside a billing month while census is counted in whole days. The plan’s member-month count and your attributed-day count disagree by construction unless both are computed from dated events.
- A resident becomes attributable for a period already closed and reported. Without version history, last month’s reported figures silently change and nobody can explain the movement.
- Coverage is withdrawn after care was delivered under the assumption it applied. The clinical work is real; the attribution is not.
- Assignment methodology differs by program and participation option. Operators reporting against a prospective list while the counterparty measures on a retrospective one produce two defensible, incompatible answers.
- Institutional or institutional-equivalent status is a clinical determination with a date. If the record cannot show the assessment that established it, the eligibility is asserted rather than evidenced.
- A resident who moves between units, buildings, or levels of care mid-period is easy to double-count and easy to lose entirely.
- A resident can be inside a plan and inside a network relationship at once. Without contract-level attribution on the event, the same transfer is claimed and blamed twice.
Frequently asked questions
- Is this an eligibility verification or clearinghouse product?
- No. Eligibility verification and clearinghouse transactions are payer-facing plumbing. This is the operator-side reconciliation layer: your census, level of care, dated events, and coverage held in one record so a counterparty file can be checked against something rather than accepted as the population of record.
- Can SeniorCRE pull our plan roster automatically today?
- No. No integration to any plan, TPA, EHR, payroll, or clearinghouse is live in operator production today, and no clinical or enrollment data is transmitted to any external party today. Roster ingestion, matching, and variance recording are architected and sequenced in the implementation plan and rehearsed in validation environments before any external transmission is enabled.
- Why keep both versions instead of correcting ours to match the plan’s?
- Because the difference is the finding. A mid-month effective date, a retroactive disenrollment, and an identifier mismatch have different operational responses, and all three disappear the moment you overwrite your figure with theirs. Holding both with the variance reason-coded is what makes a disagreement negotiable.
- Does correcting attribution improve financial performance?
- We do not publish a dollar figure or percentage for that and will not until an operator’s own data supports it under a stated methodology. What we will say structurally: rates computed on the wrong denominator are wrong in an unknown direction, so settling population before arguing outcomes is a prerequisite for any performance conversation, not a savings claim.
- What is disclosed to the plan during reconciliation?
- Minimum-necessary fields for the stated purpose, scoped per disclosure, with role boundaries and an operator-readable audit trail — consistent with HIPAA minimum-necessary guidance cited in the sources. Reconciling attribution does not require handing over the clinical record.
https://seniorcre.com/value-based-care/attribution-and-eligibility