It counts conditions, not people
One member with four chronic conditions contributes four to the denominator. That is why the gap count runs several times the member count, and why the rate cannot be compared with anything counted per member.
Guide · Risk adjustment
The HCC recapture rate: what it measures, how it is calculated, and what moves it.
A working definition for coders, analysts, and practice administrators: what the rate counts, what it leaves out, why it resets every January, what a good number looks like at each point in the year, and the six things that raise it.
By Jonan Portal Alvarez, Chief Executive Officer, Alara Healthcare (management services) · Updated September 10, 2026 · About 9 minutes
The definition
What the recapture rate measures
Under the CMS-HCC model that pays Medicare Advantage plans, a member's risk score is rebuilt from scratch every calendar year. A chronic condition documented in a qualifying face-to-face encounter last year counts for last year. This year it counts only if it is documented again. Risk adjustment has no memory, and the recapture rate is the measure of how much of last year's picture you have re-established so far this year.
The definition: of the conditions that scored for your members last year and are expected to persist, the share that has been documented and submitted again this year. It is a completion figure for the recapture work, and its mirror image is the open HCC gap count. As one rises, the other falls.
It is a progress measure, so where you are in the year matters more than the number. Forty percent in February is on pace. Forty percent in November is a finding.
It resets every January 1
The denominator is rebuilt from last year's documented conditions, and the numerator starts at zero. A comparison across years compares two different denominators.
It is continuity, not quality
The rate says whether last year's documented conditions were documented again. It says nothing about whether the care was good, and it is not a coding target. A condition that has resolved should not be recaptured, and a code the note cannot support is worse than a gap.
The formula
How it is calculated
| Term | What it means | What to watch |
|---|---|---|
| An HCC gap | A condition that scored for a member in a prior year and has not yet been documented this year, in the payment categories of the model in force for the year. | One member can carry several. The unit is the condition, never the person. |
| Closed: revalidated | The same condition documented this year in a qualifying encounter and submitted. | The normal route, and the one your visit schedule produces. |
| Closed: higher specificity or equivalent | A more severe or more specific code in the same hierarchy, or an equivalent code that maps to the same category. Within a hierarchy only the most severe condition scores. | Counts as closed for the rate. It may change what the condition is worth. |
| Closed: outside provider | A provider outside your panel documented the condition, and the plan's data shows it. | Counts toward the rate but leaves your actionable list. You did not do the work, and you cannot rely on it happening again next year. |
| Excluded from the denominator | Members who left the plan or died, and conditions that can no longer be submitted for the payment year. | This is why the denominator moves during the year. A shrinking denominator can raise the rate with no work done. |
Across a period the rate is recomputed from totals, not averaged from monthly rates, so a quarter with few gaps does not weigh the same as a quarter with many. On Alara Population Health the same set of closed statuses feeds the Dashboard tile and the RAF Gap Analysis tile, so the two screens agree. Your role and location scope decide whose gaps are counted, which is why two people on the same screen can correctly see different rates.
The target
What a good rate looks like
Two numbers get quoted. Industry writing commonly cites 90 percent or higher for chronic conditions as the mark of a mature program. Alara Population Health ships with 85 percent as the default target, with a warning band ten points below it, and the target is configurable, because the right number depends on the mix of conditions, the share of members who leave mid-year, and how the denominator is defined.
More useful than the year-end target is the pace. A rate is on track when it is ahead of the share of the year's visits that have already happened, because the members carrying the most conditions are the ones who most need a visit on the books.
January to February
Near zero and rising slowly. The denominator has just been rebuilt from last year's conditions, and January visits have not produced claims yet. A rate that is already high in February usually means last year's late closures are landing, not that this year's work is done.
March to June
Climbing steadily with the visit schedule. This is where the year is won: the members carrying the most open conditions should have a visit booked by now, and pre-visit review should be running on every one.
July to September
The halfway checkpoint. Members with open conditions and no visit are the list to work, provider by provider. A rate stuck below the warning band here will not recover in the fourth quarter without weekend clinics.
October to December
The last visits that count for the service year. The rate keeps rising for months afterward as claims and submissions land, so December's number is not the year's number.
The following year
Submissions for the service year continue into the next year on the payer's schedule, and the rate climbs on its own as they are accepted. Then it resets, and this year's closed conditions become next year's denominator.
Two effects make the rate look better than the work. It climbs on its own as claims arrive, because documentation that happened weeks ago is only counted when the claim lands. And the denominator shrinks when members leave. Read the rate beside two other figures: the open gap count, which is the workload, and RAF Impact, which is what the still-actionable gaps are worth. A high rate on a small number of gaps can be worth less than a lower rate on a large one.
The work
What actually moves it
Six steps, in the order they pay off. None of them is a report.
- 1
Start from the right list
The denominator comes from last year's documented conditions for the members attributed to you now. If the panel is stale, the list is wrong before anyone opens a chart. Reconcile the roster first.
- 2
Rank by value, not by count
Gaps are not worth the same. A queue worked by length sends coders to twenty light gaps while three heavy ones expire. Rank by what each gap adds and by how much evidence supports it.
- 3
Book the visit early
A gap closes in a qualifying face-to-face encounter, so the recall list is the recapture program. Members with the most open conditions and no visit scheduled are the first calls, in the first quarter.
- 4
Put the list in the visit
The provider sees the member's open conditions with the patient in front of them, with the evidence behind each suspect, and documents what is clinically true. A condition that is not monitored, evaluated, assessed, or treated in the note has not been recaptured, whatever the problem list says.
- 5
Review, submit, confirm
A coder checks that the note supports the code, the diagnosis goes to the payer in a format it accepts, and the closure is confirmed against the plan's data rather than assumed from the chart. A rejected submission is a gap that looks closed and is not.
- 6
Measure per provider and per location, monthly
A panel-level rate hides the one clinic where nothing is scheduled. The rate by provider, read beside the visit schedule, is the management tool.
Every one of those steps is a task with an owner, a due date, and a status, and the last one is a figure that only means something beside a quality signal. How the Work List carries them
The boundaries
What the rate is, and what it is not
The rate is
- A completion figure for this year's re-documentation work.
- A pace measure, read against the calendar.
- A ratio of conditions, recomputed from totals.
- A lagging figure that keeps rising after the work stops.
The rate is not
- A clinical quality measure. It measures documentation continuity, not care.
- A coding target. A condition that resolved should not be recaptured, and a code the note cannot support is an audit finding, not a closure.
- A revenue forecast. RAF Impact is the value view, and even that is a ceiling.
- Comparable across years, or across two screens with different scopes.
The model
What V28 changed about recapture
The CMS-HCC model was revised in version 28, phased in over payment years 2024 through 2026 and now fully in force. It has more payment categories than the previous version but maps fewer diagnosis codes into them, and several condition groups now carry one coefficient regardless of severity. Some of last year's documented conditions therefore map to nothing this year, and some that used to differ in value no longer do.
For the recapture rate this means two things. The denominator is built from the model in force for the year you are looking at, so a condition that no longer maps is not a gap and not a target. And ranking by value matters more than it did, because the spread between a heavy gap and a light one changed. The V28 transition has its own guide in this series.
What moved, what still counts, and what to recapture first under the new model: The CMS-HCC V28 transition
Questions people ask
Questions about the recapture rate
Why did our recapture rate go up when nobody worked a gap this week?
Two reasons, and both are normal. Claims from visits weeks ago are landing, so documentation that already happened is only now being counted. And members who left the plan drop out of the denominator. Neither is work done this week. To tell movement from data apart from movement from work, read the rate beside RAF Impact, which is measured in value rather than count.
What is a good recapture rate?
Industry writing commonly cites 90 percent or higher for chronic conditions. Alara Population Health ships with 85 percent as a configurable default target and a warning band ten points below it. The honest answer is that pace matters more than the year-end number: a rate ahead of the share of visits already completed is on track, whatever month it is.
Does a condition documented by a specialist outside our practice count?
For the rate, yes. If the plan's data shows the condition documented this year by any provider, the gap is closed. It leaves your actionable list, because there is nothing left for you to do, but it should not be relied on: an outside closure this year says nothing about next year.
Why do two people see different rates on the same screen?
The payment year selected and each person's role and location scope decide whose gaps are counted. A provider sees the rate for their own attributed panel; a location manager sees their locations; an organization role sees everything. The closed-status set is the same for all of them, so the difference is scope, not arithmetic.
Is the recapture rate the same as the RAF?
No. The rate is a ratio of conditions, written as a percentage. The RAF is an index to three decimals that scores a whole person, and it never carries a percent sign. RAF Impact is the value of the still-actionable gaps, which is the figure to read beside the rate when deciding what to work first.
When is a gap actually closed?
When the condition has been documented this year in a qualifying encounter, submitted to the payer in a format it accepts, and accepted. A note that mentions the condition without monitoring, evaluating, assessing, or treating it does not close the gap. A submission the payer rejected does not either, and it is the most common gap that looks closed and is not.
See your recapture rate on your own file.
Bring last year's risk file to the demo and we will show the denominator it builds, the gaps it ranks, and who would own each one.