A primary care group is offered a contract: a per-member-per-month quality bonus if it hits screening and control measures, a share of the savings if its attributed patients cost the payer less than a benchmark, and, in the version the payer prefers, a capitated payment per patient in place of fee-for-service. The group's revenue cycle is built for visits; its data shows what it billed, not which patients are attributed to it or whether they got their screenings; and the contract's arithmetic turns on a benchmark the payer computes. Value-based contracts can raise margin and deepen payer relationships; they can also transfer risk to a group that has no way to manage it. Readiness is the work that decides which.
What is attribution, and why is it the first problem?
The group is paid for the patients the payer says are its. Attribution rules assign each member to a group based on where they received the plurality of primary care visits over a look-back period, or on a patient's selection, or on an algorithm the payer runs; the group's attributed panel is the denominator for every measure and every dollar — and the group usually does not know who is on it until the payer sends a list, often quarterly, often late.
Panel management is the operating discipline. The attributed list reconciled to the group's active patients; patients attributed but not seen in twelve months identified and outreached, because they count in the denominator and are the ones most likely to be hospitalized; patients seen but not attributed flagged, because the group is doing the work without the credit; and the panel size by provider against the capacity the contract's care model requires.
Which quality measures must be reported before they can be earned?
Measures are defined by the payer and reported from the group's data. Screening rates — cancer, diabetes-related, depression; control measures — blood pressure, blood-sugar markers; immunizations; care-transition follow-up; patient experience; each with a specification, a numerator and denominator drawn from the attributed panel, and a reporting mechanism — claims, supplemental data files, or registry submissions.
The infrastructure. The electronic health record configured to capture each measure's data elements as structured fields rather than free text; a registry or population-health report that shows the gap list — attributed patients missing a screening or out of control — by provider; a workflow that closes gaps at visits and through outreach; and supplemental data submission to the payer so that care delivered is counted, because claims alone under-report most measures.
The measurement baseline. The group's current performance on each measure, from its own data, before signing — because a bonus threshold set at a level the group cannot see itself reaching is not a bonus, and a group that cannot compute its baseline cannot compute the contract's value.
How is capitation modeled?
Capitation replaces fee-for-service with a fixed per-member-per-month payment. The group receives the payment for each attributed member whether they are seen or not; it bears the cost of the primary care those members use; margin depends on utilization — visits per member per year — and on the group's cost per visit.
The model at three utilization scenarios. Attributed members times the per-member-per-month rate gives revenue; low, expected, and high utilization — say 2.4, 3.2, and 4.3 visits per member per year — times the group's fully loaded cost per visit gives cost; contribution at each; the expected scenario should leave a margin the group would accept, and the high scenario should not be a loss the group cannot carry; if the capitation rate works only at low utilization, it is a bet on the panel's health the group should decline.
Carve-outs and what is inside the payment. Which services the capitation covers — office visits, preventive care, chronic-care management — and which remain fee-for-service — procedures, ancillaries, after-hours; the model follows the contract's definitions, because a capitation that covers ancillaries the group had been earning margin on is a different deal.
Cash timing improves; risk timing does not. Capitation pays monthly in advance, which shortens the cash cycle; the utilization risk arrives over the year, and the group's monthly report must show visits per member against the model so a drift is seen in month three rather than at reconciliation.
How do shared savings work, and what does the group control?
The group shares in savings against a benchmark. The payer computes the total cost of care for the attributed panel — including hospital, specialist, pharmacy, and post-acute costs the group does not deliver — against a benchmark derived from the panel's history and risk adjustment; if actual cost is below the benchmark by more than a minimum savings rate, the group receives a share, often conditioned on quality performance; the benchmark, the risk adjustment, and the attribution are the payer's calculations.
What the group controls. Primary care access and continuity that keep patients out of emergency departments; chronic-condition management that avoids admissions; referral patterns to lower-cost specialists and sites; transitions of care after discharge; and documentation that captures risk accurately, because risk adjustment moves the benchmark.
What the group must insist on. Transparent benchmark methodology; timely claims and attribution data from the payer — monthly, not annually — so the group can see cost trends; a minimum savings rate it can plausibly beat; a reconciliation timeline; and, in the first years, upside-only terms.
Downside risk and the reserve. A contract with downside — the group pays the payer if cost exceeds the benchmark — transfers insurance risk; the group's exposure is modeled at an adverse scenario, a reserve is held against it, and stop-loss protection is negotiated or purchased; a group that signs downside risk without a reserve has made a bet with its working capital.
What is the panel size that makes fixed payment work?
Small panels are volatile. A panel of 800 attributed members has a cost of care that one or two catastrophic cases can move by a fifth; a panel of 5,000 averages out; capitation and shared-savings arithmetic depend on the law of large numbers, and the group should know the panel size at which its expected result is stable — often several thousand members per contract — before accepting risk on a smaller one.
What is the twelve-month build?
Months one to three. Attribution reconciliation and panel management workflow; electronic health record configuration for measure capture; the measurement baseline computed.
Months four to six. Gap-closure workflow in visits and outreach; supplemental data submission tested with the payer; referral-pattern and site-of-care analysis.
Months seven to nine. Capitation and shared-savings models built at three scenarios on the group's own cost per visit and the panel's cost history; contract terms drafted — attribution, benchmark transparency, data timeliness, minimum savings rate, upside-only, carve-outs, reserve and stop-loss for any downside.
Months ten to twelve. Quality performance tracked monthly against thresholds; the first pay-for-reporting or upside-only agreement signed; the monthly value-based report — attributed panel, gap closure, measure performance, visits per member, cost-of-care trend from payer data, projected bonus and savings — in the management pack.
Why is this CFO work?
Attribution reconciliation is a data discipline; the measurement baseline is a reporting build; capitation and shared savings are contract models at multiple scenarios; the reserve is a balance-sheet position; the panel-size analysis is a volatility calculation; and the monthly value-based report is the finance function's product. The medical director owns the care model and the gap closure; the CFO owns the economics and the contract terms.
Worked example
A primary care group with nine providers and about 11,000 active patients, offered by its largest commercial payer a contract with a US$3.50 per-member-per-month quality bonus, 40 percent shared savings against a benchmark with a 3 percent minimum savings rate, and an option to move to US$42 per-member-per-month capitation for primary care.
| Readiness item | Found at offer | After the build |
|---|---|---|
| Attributed panel reconciled | Payer list of 6,800; 1,340 attributed patients not seen in 12 months; 900 active patients not attributed | Outreach program; attribution corrected to 7,400 |
| Measurement baseline (six measures) | Could not compute from the record; estimated 48–71% | Computed: 61–84%; three measures above the bonus threshold, three below |
| Supplemental data submission | None | Tested and accepted; measured rates rose 9–14 points on care already delivered |
| Capitation at US$42 — contribution at 2.4 / 3.2 / 4.3 visits per member | Not modeled | +US$410,000 / +US$190,000 / −US$120,000 (fully loaded cost per visit US$118) |
| Shared savings — benchmark transparency and data timeliness | Annual reconciliation only; methodology not disclosed | Monthly claims feed and methodology disclosed, negotiated as conditions |
| Panel-size volatility | — | 7,400 members: one-in-ten-year adverse scenario moves cost of care 7%; downside risk declined |
| Contract signed | — | Year one: quality bonus and upside-only shared savings; capitation deferred to year two pending utilization data |
Illustrative figures for a hypothetical organization; not a client's data.
The group would have signed the full offer on the payer's terms; the readiness build showed it could not compute its own measures, that a fifth of its attributed panel had not been seen, and that the capitation rate was a loss at the high-utilization scenario. The year-one contract — quality bonus plus upside-only shared savings, with monthly data and benchmark transparency as conditions — earned the group about US$260,000 on care it was largely already delivering and produced the utilization data that priced the year-two capitation conversation.