DNFB to cash: the 13-week compression method.
Cut Discharged Not Final Billed from roughly 11 days to roughly 4 in one fiscal quarter. No new contracts, no rate concessions, no volume change. Just days pulled out of the unbilled bucket, one week at a time.
Executive summarySeven days of held revenue, released in a quarter.
DNFB is the single fastest cash lever a hospital revenue cycle owns. It needs no contract renegotiation and no rate change. This method moves the archetype hospital from roughly 11 unbilled days to roughly 4 across one fiscal quarter, on a visible weekly curve the CFO can take to the board.
Seven released days on a $1M-per-day book is roughly $7M of one-time cash.
Take an archetype 280-bed hospital running about $1M of net patient revenue per day. Every DNFB day held is about $1M sitting unbilled. Compress from 11 days to 4 and you release roughly 7 days, a one-time cash pull of about $7M, followed by a permanently faster billing clock. The figure is illustrative and scales linearly with the hospital's daily net revenue. It is not tied to any named client.
The metricWhat DNFB actually measures.
Discharged Not Final Billed counts encounters that have left the building but have not dropped a final bill. Expressed in days, it is the count of net-revenue days sitting in the unbilled bucket before the cash clock even starts.
Illustrative decomposition of an 11-day DNFB into the five root causes the method attacks. A small number of causes hold most of the days.
| HFMA KPI | What it measures | Archetype start | Target |
|---|---|---|---|
| DNFB days | Days of net revenue discharged but not final billed | 11.0d | ≤ 4d |
| Discharged not submitted | Final-coded charts not yet transmitted to the payer | 3.5d | ≤ 1.5d |
| Clean claim rate | Claims accepted first pass with no edit or rejection | 82% | ≥ 95% |
| Cost to collect | Total RCM cost as a share of net revenue collected | 3.6% | ≤ 3.0% |
| Initial denial rate | Claims denied on first submission by dollars | 9.5% | ≤ 5% |
These five KPIs are the HFMA MAP Key family that frames the entire method. DNFB days is the headline; discharged-not-submitted, clean claim rate, cost to collect, and initial denial rate are the supporting instruments that tell you whether the days are actually leaving the bucket or just moving around inside it. Every number above is an illustrative archetype figure, not a named-client result.
DiagnosisThe five root causes of a held chart.
DNFB does not rise for one reason. It rises because five distinct failure points each add days, and they compound. The method assigns a workstream to each.
A chart held for three causes is not held three times. It is held for the longest one, then the next.
Held days do not simply add. A chart waiting on a query and stuck behind a coding backlog clears the query first, then joins the backlog. Attacking causes in isolation moves days from one bucket to another and the headline DNFB barely moves. The method runs all five workstreams in parallel so a chart cleared of one blocker does not immediately hit the next.
The methodThree phases. Thirteen weeks. One curve.
The quarter breaks into three phases: stabilize the inflow, compress the backlog, and lock the gain with a standing cadence. Each phase owns a slice of the compression curve.
The curve, in numbersWeek by week, 11 down to 4.
The same curve that leads this paper, expressed as a weekly plan. The steepest drop lands in the compress phase, weeks 5 through 9, when the backlog burns down and the query SLA takes hold.
| Week | Phase | Primary move | DNFB days | Clean claim % |
|---|---|---|---|---|
| 1–2 | Stabilize | Baseline, freeze queue, interim CDI | 11.0 | 82% |
| 3–4 | Stabilize | Triage aged charts, kill stale edits | 9.6 | 84% |
| 5–6 | Compress | Queue rework, 24h query SLA on | 8.7 | 87% |
| 7 | Compress | Charge-capture gaps closed at discharge | 6.8 | 90% |
| 8–9 | Compress | Aged backlog burn-down, cohort by cohort | 5.3 | 93% |
| 10–11 | Lock | Permanent CDI model, cash-forecast cadence | 4.4 | 95% |
| 12–13 | Lock | Hold ≤ 4d as a standing control | 4.0 | 96% |
Roughly 60 percent of the total day reduction lands in the compress phase. That is by design. Stabilize buys the visibility and the capacity; compress spends it; lock keeps the gain from decaying. Numbers are illustrative and modeled on the 280-bed archetype.
The payoutWhere the released cash comes from.
Each compressed day is a slice of net revenue pulled forward. This waterfall walks the illustrative 7-day release on a $1M-per-day book from the 11-day baseline down to the 4-day target.
A one-time cash pull of roughly $7M as DNFB compresses from 11 to 4 days, attributed to the workstream that released each slice.
How it runsA cadence, not a takeover.
The method is a working cadence run alongside the hospital's own CDI and coding teams. ASP-RCM supplies the instruments and the discipline; the hospital owns the charts.
One standing meeting turns DNFB from a lagging report into a forward forecast.
Every week the cadence reads three numbers: DNFB days this week, the projected bill-drop for next week, and the resulting cash forecast. Because the forecast is built from the actual unbilled inventory rather than a trailing average, finance can see the cash coming before it lands, and any drift in DNFB shows up as a forecast miss the same week rather than a surprise a month later. This cadence is what separates a one-time backlog cleanup from a permanent control. Proof of the pattern lives in our case studies.
Common questionsFrequently asked: DNFB compression.
What is DNFB and why does it matter?
What is a healthy DNFB target?
Why 13 weeks?
What actually causes high DNFB?
Does this replace the hospital's coding or CDI team?
Are the numbers in this paper from a real client?
Want the 13-week curve run against your DNFB?
Send a DNFB-by-cause export and one quarter of unbilled inventory. Inside 30 days you get a written DNFB baseline, a root-cause decomposition with day impact, and a phased 13-week compression plan. Yours to keep.