How do you recover old or aged accounts receivable at a mental health practice?

What counts as “aged” A/R at an outpatient practice?

A balance is aged once it sits unpaid past the point it should have resolved. It is tracked in an aging report that sorts every open balance into buckets by how long it has been outstanding — typically 0–30, 31–60, 61–90, 91–120, and 120-plus days — measured from the date you billed. Days in A/R is the related summary figure: roughly how long, on average, a dollar waits before you collect it.

The number that tells you whether you have a problem is not a single benchmark. It is how much of your balance is stuck in the older buckets, and why. A/R that is young and moving is normal; A/R piling up past 90 and 120 days is where claims quietly cross filing and appeal deadlines and turn into write-offs. The practical health signal most practices can read off their own report is the share of A/R sitting past 90 days, watched as a trend rather than compared to a borrowed number.

Be skeptical of generic targets, because the credible ones are not specific to this specialty — that data gap is the whole subject of our guide to normal denial, days-in-A/R, and clean-claim numbers. A practice billing a few hundred small session claims a month does not have the same aging shape as one carrying a handful of large balances, and a Medicaid-heavy panel does not look like a commercial one. Judge your own trend and the shape of your buckets.

Is old A/R still recoverable, or should you just write it off?

Some of it is recoverable, and writing a balance off on age alone is a decision made without information — but “recoverable” is not “collectible,” and this is a section where a billing company has every incentive to overstate. What decides recoverability is why a claim aged and how much filing or appeal window is left. Some old A/R genuinely is gone. Some of it was simply never worked.

The economics of abandonment are the reason the pile exists. The same AHIMA Journal article that reports the industry average for unresubmitted returned claims also puts the average cost to rework a denied claim at $25 per claim for practices— and at $181 for hospitals, which is a different segment, not the top of a range. Read both as what they are: industry averages a professional journal repeats from 2017–2019 trade reporting, not measurements anyone took. Neither is a behavioral-health number, and neither is a number about your practice. They explain a pattern; they do not size your pile.

Now the number that cuts the other way, because you should have it before you decide. KFF’s analysis of 2024 HealthCare.gov data found insurers denied about 19% of in-network claims, that consumers appealed fewer than 1% of those denials, and that where consumers did appeal, 66% of appeals were upheld — that is, the original denial stood. Two thirds of those appeals lost. That is a real check on any story about old claims being easy money.

Be precise about what those numbers cover, though, because it is not a measure of the work a biller does. The appeals are consumer-initiated — filed by patients, not by practices or their billing companies — on qualified health plans sold through HealthCare.gov: not marketplaces a state runs on its own platform, and not employer group plans. KFF reports 165,863 denials upheld and notes it counts at least 262,982 appeals, because CMS suppresses counts under ten; it does not publish the denominator behind the 66%, so do not compute one from those two numbers. The 19% has its own definition worth knowing in an article about resubmission: KFF states that “claims that were initially denied, then subsequently resubmitted and paid, are not included as denied claims in the denial rate” — so 19% is denials that stuck. Denial information was reported for 75% of returning insurers’ claims, and none of this is mental-health-specific.

No comparable public figure exists for provider-initiated appeals, so we are not going to claim ours run better. What the KFF data supports is narrower than it first looks: consumers almost never appeal, and the appeals that do get filed mostly lose. It says nothing about how often a practice or its biller challenges a denial. The lesson for an aging report is still the right one, though — working old A/R is a search for the winnable claims and the documentation to win them, not a presumption that an old balance pays.

Why does A/R age at a mental health practice specifically?

Because the claims are small, numerous, and unusually easy to get administratively wrong — so the same friction that produces denials also produces claims that stall without ever being denied. The recurring drivers are predictable once you have watched them repeat, and most of them start before the claim is ever sent.

The carve-out is the classic one: when a separate managed behavioral health organization administers the benefit under its own payer ID, filing address, and clock, claims sent to the plan on the card sit at the wrong entity and age in silence. Eligibility and benefit checks at intake fail quietly — a plan change between the first call and the third session does not surface until the remit does. Visit authorizations run out mid-episode, and the sessions after the last authorized unit are billed and then denied. Telehealth place-of-service and modifier rules have changed repeatedly and by payer, and a wrong POS is a rejection that is cheap to fix and easy to never notice. Underneath all of it, the per-claim dollar is small, so a single balance rarely feels urgent enough to chase — which is precisely how the aggregate gets large.

Then there is the triage dynamic, which is our own explanation rather than a sourced finding, though it is the one billers recognise fastest. When the day is full, submitting this week’s claims always beats working last quarter’s denials, because this week’s claims are how this month gets paid. Old claims lose that contest every single day, and they lose it quietly. That is why aged A/R and denials are largely the same problem seen at different stages, and why catching the problem before the claim goes out is worth more than any recovery process — recovery is what you do about the claims where prevention already failed.

How do you recover aged A/R, step by step?

It is a triage process, not a mass resubmission. Pull the full aging report; segment it by payer, age, dollar, and status; check what filing and appeal window each claim has left; separate the no-response claims from the denials and the underpayments; then work each on its own path, escalating through the payer’s appeal levels when the first pass fails.

The distinction that decides the path is status, and getting it wrong wastes the window. A no-response (pending) claim is not a denial: the payer never adjudicated it, there is no remittance, and the fix is to follow up or correct and resubmit — usually while the timely-filing clock is still running. A denial was adjudicated: it carries a remittance with CARC and RARC codes, and it needs an appeal, not a resubmission. Resubmitting a true denial just earns the same outcome and burns days you may not have. An underpayment looks closed and is not: a claim paid below your contracted rate shows as resolved on the aging report, so it never enters the recovery queue at all unless someone reconciles the remit against the fee schedule.

Protect the window at every step, and know that the filing clock and the appeal clock are two different clocks. Medicare sets a one-calendar-year filing limit under 42 CFR 424.44 — the CMS Medicare Claims Processing Manual, Chapter 1 states that “in general” such claims must be filed “no later than 12 months, or 1 calendar year, after the date the services were furnished,” and §70.1 measures that from the date of service, using the line-item “From” date on professional claims. Its exceptions (§70.7) are narrow and specific — administrative error or misrepresentation by a Medicare employee, contractor, or agent of the Department acting within the scope of its authority; retroactive Medicare entitlement; retroactive entitlement involving a State Medicaid agency; and retroactive disenrollment from a Medicare Advantage or PACE plan. None of them is “nobody got to it.” Commercial and Medicaid deadlines are shorter and set by contract and by state, and a denial starts its own separate appeal clock — which, contrary to the usual advice, is not reliably shorter than the filing window: Texas Medicaid allows 95 days to file and 120 to appeal, and UnitedHealthcare’s 2026 guide gives 12 months unless law or your agreement says otherwise. There is a separate guide on which claims survive a filing deadline and which are simply gone.

Which aged claims should you work first?

Work by remaining window, winnability, and dollar value — not oldest-first. A 45-day claim with a filing deadline next week outranks a 200-day claim with months of appeal window left, every time. Chasing the report in date order spends the time the winnable claims needed.

Sort by what the claim turns on. Administrative denials — timely filing, eligibility, coordination of benefits, registration and demographic errors, wrong payer entity — are won by proving a fact with a document, so they clear fastest and should go first. Authorization and medical-necessity denials need clinical documentation and more calendar time, so they need the window most and should be started earliest even if they finish last. No-response claims still inside the filing window are, in our experience rather than by any published measure, often the cheapest money on the report, because many of them need a status check and a resubmission rather than an argument. And group the report by payer before you start: one payer portal, one enrollment quirk, one filing address usually explains a whole cluster of aged claims at once, and fixing the cause clears them in a batch instead of one at a time.

What happens to your old A/R when you change billing companies?

This is where aged A/R most often becomes permanent, and it is worth settling in writing before you sign anything. When a practice switches billers, the old balances sit in a gap: the outgoing company has little reason to keep working them, and the incoming one is usually being paid a percentage on new collections, which makes old, hard claims the least profitable work on its desk.

Four things need an answer. Who works the claims already in flight during the run-out — the ones submitted but not yet adjudicated when the handover happens. Who owns the denials that come back afterwards on claims the previous company submitted, since the remittance may land weeks after they have gone. Whether the old A/R is priced separately at all — if it is folded into an ongoing percentage, nobody is paid enough to chase it, and it will quietly age out while everyone stays polite about it. And what the snapshot is: the agreed list of balances, as of a specific date, that the new company is responsible for, so that six months later there is no argument about which recoveries were whose.

Ask what has to change in your own systems too — the answer is usually less than practices expect, though the electronic-remittance enrollment genuinely does move, and it moves one payer at a time. Old claims paid after that switch can land in a different place than you are watching, which is its own way to lose track of a recovery.

Should you work aged A/R in-house or hand it to someone?

It depends on whether anyone on your team has uninterrupted time and the payer-specific knowledge to work old claims before their windows close. In-house keeps control and costs nothing extra in cash, but it competes with daily billing — and, as above, daily billing wins. Outside help adds capacity and pattern knowledge but adds cost, unless it is priced on what it actually recovers.

Contingency pricing is the structure that fits this work, for a reason worth stating plainly: it is paid out of balances that would otherwise likely have been written off, so the downside is bounded — which is why what the snapshot excludes matters as much as the rate. Money that was already going to arrive should not carry a recovery fee, and the exclusions are where you check that. It is also why old balances are normally priced separately from an ongoing billing percentage rather than folded into it — whoever works them carries the risk that a given claim never pays, and a rate set for routine ongoing work does not cover that risk. What a particular company charges for old A/R is a question to ask it directly; ours is below.

The practical test is a specific question, not a rate: can the people working your old claims name how your payers behave on carve-out routing, authorization exhaustion, and telehealth place-of-service — and tell you what they will do when a claim comes back with a reason code they have not seen? If your aging report is where they learn your payers, the recoverable claims will keep aging while they do.

Key takeaways

• Aged does not mean lost, but it does not mean collectible either: recoverability turns on why a claim aged and how much window is left, not on the number of days.
• No-response, denied, and underpaid are three different jobs. The remittance tells you which — and underpayments look closed on the report, so nothing surfaces them but reconciliation against your fee schedule.
• Filing clocks and appeal clocks are separate. Medicare allows one calendar year from the date of service to file, with four narrow exceptions; commercial and Medicaid windows are shorter and set by contract and state.
• Prioritise by remaining window, winnability, and dollars — not oldest-first — and group by payer, because one routing or enrollment problem usually explains a whole cluster.
• Old A/R is most often lost in the handover between billing companies. Settle the run-out, the late denials, the snapshot, and how the old pile is priced before you sign.
• Appeals are not free money: in KFF’s HealthCare.gov data, insurers upheld 66% of the consumer appeals that were filed. Look for the winnable claims and document them.