Days in AR Benchmark: Why Your Number Isn't Comparable to Anyone Else's

Days in AR is the most-cited number in practice financial reporting and the least standardized. The formula is stable — total AR divided by average daily charges — but every input in it is a choice, and no two practices make the same set of choices. Change the denominator from gross to net charges and the same practice reports a number that is 80% higher. Pull patient balances out of the numerator and it drops again. Nothing about the practice changed.

That matters because days in AR is usually the number that triggers action. It goes to the physician-owners, it sets the collections agenda, and sometimes it sets someone's bonus. If the number isn't comparable to the benchmark you're measuring it against, the action it triggers is aimed at the wrong thing.

This article takes one synthetic practice, computes its days in AR five defensible ways, and shows the spread. Then it covers the one interval that no version of the metric isolates, and what to report alongside it so the number becomes diagnostic rather than decorative.

The formula is stable. The inputs are not.

Days in AR = Total AR ÷ (Charges over N days ÷ N)

Three inputs, three definitional forks:

1. Which charges go in the denominator — gross or net? Gross charges are what you billed. Net (adjusted) charges are gross minus expected contractual adjustments. Because gross is the larger number, a gross-charge denominator produces a larger average daily charge and therefore a smaller days-in-AR figure. Practices with high gross-to-collected ratios look dramatically better on gross. Benchmark sources differ on which they publish; check your source's stated methodology before you compare [verify].

2. What's in the numerator? Candidates that practices include or exclude inconsistently: patient responsibility balances, credit balances and unapplied cash, claims held in a scrubber but not yet transmitted, small balances below your write-off threshold, and contractual adjustments that posted late and are still sitting on the account. Each inclusion decision moves the number.

3. How long is the charge window? A 90-day trailing window is common; some sources use 365 days [verify]. If your charge volume is seasonal — surgical practices with a Q4 deductible-met surge, pediatrics with a back-to-school and flu-season pattern — the two windows produce materially different daily charge rates, and the shorter window produces a number that moves for reasons unrelated to collections.

Worked example: one practice, five benchmark numbers

A synthetic multi-provider practice. All figures constructed for illustration; no real data.

90-day trailing gross charges      $1,800,000  →  $20,000/day
Contractual adjustment rate               45%
90-day trailing net charges          $990,000  →  $11,000/day
365-day trailing gross charges     $6,351,000  →  $17,400/day   (Q4-weighted volume)

Total AR on the books                $520,000
  Contractual adjustments not yet cleared      $ 38,000
  Payer payments in transit (835 issued)       $ 46,000
  Patient responsibility                       $104,000
  Denied lines awaiting appeal                 $ 61,000
  Claims awaiting adjudication                 $271,000

That $38,000 of uncleared contractual is worth naming precisely, because it's the most common numerator defect. It comes from lines like this one:

SVC*HC:99213*160.00*74.20**1~
CAS*CO*45*61.80~
CAS*PR*2*24.00~

SVC02 submitted $160.00 = SVC03 paid $74.20 + CO-45 contractual $61.80 + PR-2 coinsurance $24.00. The $61.80 is a contractually agreed reduction — nobody owes it to you. If it's still on the account thirty days after the 835 posted, it's inflating your AR total and your days in AR, and it isn't collectible by anyone.

Now compute the metric five ways on identical operations:

20.8 to 47.3 days. A 26.5-day spread, and every row is defensible.

Suppose your benchmark source publishes a 30-day target for your specialty [verify your source's current figure — targets vary by specialty and by publisher]. This practice beats it comfortably, misses it badly, or lands close, depending entirely on a reporting convention that was probably set once by whoever configured the PM system and never revisited.

Two operational consequences. First, comparing your number to a published benchmark is only meaningful if you've matched the benchmark's methodology — otherwise you're comparing measurement conventions, not performance. Second, comparing your number to your own number from last quarter is only meaningful if nothing about the convention, the charge master, or the payer mix changed in between.

The interval no version of the metric isolates

Every row in that table treats AR as one undifferentiated pool aging at one rate. It isn't. In the composition above, $46,000 is money the payer has already adjudicated and committed to — the 835 exists, the amount is final, the decision is made — and it's aging purely because of how the funds move.

On the standard healthcare EFT path, the payer originates an ACH credit and the money settles on a banking-day cycle. Same Day ACH exists but has submission windows and a per-payment cap [verify current NACHA limit], and a payment that misses the last window waits. Weekends and Federal Reserve holidays are dead time. The 835 and the deposit arrive on separate channels, so reassociation by TRN02 becomes a manual matching exercise when the descriptors don't line up.

None of that is a collections problem, a payer-relationship problem, or a staffing problem. It's transport, and it lands in your days-in-AR number indistinguishable from a claim nobody has worked in six weeks.

Instant rails change the transport, not the adjudication. FedNow and RTP are credit-push, settle in seconds, run 24/7/365 with no cutoff windows and no weekend or holiday dead time, and are final on receipt — funds return only by request, not by unilateral return within a NACHA window. Both carry ISO 20022 structured remittance capability, which turns reassociation into a lookup. Per-transaction caps differ by network and have been raised more than once; [verify] current limits and confirm your account's receive-side enablement with your bank, because receiving is not on by default.

Be precise about what this does and doesn't do: it does not make a payer adjudicate faster. Clean-claim rate, payer edits, and prior-auth discipline are what move that interval. What real-time settlement removes is the gap between "the payer decided" and "we can spend it." How many days that is for you is measurable — take twenty claims per major payer, record the BPR16 effective entry date and the date funds were actually available, and take the median. That interval is your upside stated in days rather than assumed. Payer-side origination adoption is uneven, so most practices will run mixed rails for a while. That's the normal migration path, not a failure state.

Building a benchmark you can act on

Days in AR is a fine trend line and a poor diagnostic. Report it with the context that makes it interpretable:

• [ ] Write down your methodology and freeze it. Gross or net, window length, and an explicit inclusion list for the numerator. Put it in the footer of the report. Every future comparison depends on it.
• [ ] Report aged AR over 90 days as a percentage of total AR. This is harder to game than the headline number and it moves when real problems appear. Days in AR can hold flat while the aged tail grows.
• [ ] Report net collection rate alongside it. Payments ÷ (charges − contractual adjustments) tells you what you actually collected of what you were entitled to. Days in AR measures speed; net collection rate measures completeness. A practice can be fast and still leave money behind.
• [ ] Split the number by payer and by financial class. A blended figure hides the two payers driving the problem. Patient AR and payer AR should never share a denominator without also being reported separately.
• [ ] Quantify your contractual residue. Every aged dollar tracing to a CO-45 is a posting defect inflating the metric. Fix it before you optimize anything else.
• [ ] Measure `BPR16`-to-funds-available per payer. Twenty claims each, majors only. This is the component of days in AR that no amount of follow-up work will touch.
• [ ] Recompute last year under this year's methodology before you claim an improvement.

FAQ

What's the right days-in-AR target for my practice? There isn't a single one, and any source quoting a universal number is describing an average across specialties, payer mixes, and reporting conventions that probably don't match yours. Use your specialty society's or MGMA's current published figure [verify], confirm the methodology it was computed under, and match it. Then treat your own trailing twelve months as the more useful comparison.

Should patient AR sit in the same number as payer AR? Report them separately. They behave differently and respond to completely different actions — a payer balance closes through follow-up or appeal, a patient balance closes only when someone sends a request the patient can act on. Blending them produces a number that can't tell you which one moved. If your headline metric must include both, publish the split next to it.

Our days in AR dropped four days last quarter. Did we improve? Check the denominator first. A charge volume increase raises average daily charges and lowers days in AR with no change in collections at all — the metric improves because you billed more, not because you collected faster. Then check whether a bad-debt or contractual write-off batch cleared during the period, which reduces the numerator the same way. If both are flat and the aged-over-90 percentage also fell, the improvement is real.

Run your own numbers through it

The shteg.ai sandbox takes a de-identified 835 and returns the inputs this article depends on: per-line contractual, payer-paid, patient-responsibility, and denial splits; SVC/CAS balancing validation to catch posting defects; allowed-amount variance against a contract table you supply; and a settlement timeline comparing ACH against instant rails on that same remittance.

One file is enough to see how much of your AR is transport rather than collections.

[Open the shteg.ai sandbox →](/sandbox) [verify link target before publish]

Educational content for practice administration and revenue-cycle staff. Not legal, billing, coding, accounting, or compliance advice. All claim segments, dollar amounts, and practice figures in this article are synthetic illustrations and contain no patient data. Benchmark targets and methodologies vary by specialty and by publishing source — verify current figures and definitions with your source before comparing. Payment network capabilities, transaction limits, cutoff windows, and operating rules change — verify current specifications with your financial institution and clearinghouse before making operational decisions. Patient balance collection is governed by your payer contracts and by applicable state and federal law; consult counsel for your jurisdiction.

Spec check: ~1,520 words · H1 + 6 H2 · worked example with SVC02 = SVC03 + CAS arithmetic and five-convention computation · one comparison table · 3-question FAQ · sandbox CTA · disclaimer. All dollar figures synthetic; every unverifiable claim (benchmark targets, NACHA/FedNow/RTP caps, window conventions) is marked [verify].

<!-- generated_by: claude-draft-1 -->