Aged Claims at Timely-Filing Risk
Every follow-up queue is sorted by age, and age is a measure of the past. The quantity that decides what is worth working today is the time remaining before a claim stops being collectible at all. Those two orderings disagree, and not marginally: filing periods differ by payer and commonly run from the date of service rather than from the date anyone noticed, so a long-aged claim under a generous window can be safer than a younger one under a short one. Sorting by what is left is the change.
Updated 7 min read
On this page
Key takeaways
- Rank by remaining time, not elapsed time. Age measures the past; only the remainder measures what can still be lost.
- Filing periods differ by payer and usually run from the date of service, so identical ages carry very different risk.
- A claim that has never been adjudicated and one awaiting a decision are at different risk even at the same age.
- Compute the remainder per payer from the contract or program rule, and treat it as a field on the claim rather than a fact someone recalls.
- Escalate on the remainder crossing a threshold, not on a bucket boundary. Bucket edges are an accounting convention.
- Medicare enumerates exceptions to its filing limit at 42 CFR 424.44(b) — including a contractor's own error and retroactive entitlement.
- Those exceptions turn two situations practices read as bad luck into arguments, if the evidence was kept when it appeared.
- A claim past the window is not automatically worthless, and a claim inside it is not automatically safe.
Why the aging report sorts the wrong way
An aging report is built from one date and today. It answers how long a balance has been open, which is a useful accounting fact and a poor work-ordering signal, because nothing about the risk of losing a claim is a function of its age alone.
- The window is not shared. Each payer's filing period is a term of its own agreement or program rule, so two claims of identical age sit at different distances from their deadlines. A single bucket boundary cannot represent that.
- The clock usually started earlier than the report thinks. Where the period runs from the date of service, time spent unbilled has already been spent — which is the fact unbilled and held claims is built on, and it means a claim can arrive in the queue with much of its window gone.
- Age does not distinguish the states. A claim the payer has never seen and one awaiting a decision carry different risk at the same age. No response from the payer covers separating them.
- Bucket edges are a convention. Nothing happens to a claim when it crosses one. Something does happen when its remaining time reaches zero.
Making the remainder a field rather than a memory
Record the window per payer, from its source
The contract term or the program rule, held against the payer rather than in somebody's head. This is one of the questions a payer contract inventory exists to answer quickly.Record what starts it
Date of service, discharge, or something else. This single field changes every remainder on the report, and it is a term rather than a convention.Compute remaining time per claim, and store it
Not derived on demand in a report nobody runs. A stored remainder can be sorted, filtered, alerted on and audited, and it survives the person who understood the rule.Set the escalation threshold on the remainder
Far enough out that there is time to do something — obtain the status, produce the evidence, resubmit — rather than at the point where the only remaining action is to write it off.Re-derive after every event that moves the claim
A resubmission, a corrected claim or a rejection can change what applies. A remainder computed once at intake and never revisited is a number that becomes wrong quietly.
Two queues, not one sorted list
A claim past the window is not automatically finished
The triage above is prevention. It is worth knowing what remains when it fails, because the common reaction — writing the claim off as out of time — is sometimes wrong, and the enumerated cases are ones practices meet often.
Medicare's exceptions are a list, not a discretion
- These are Medicare's. A commercial agreement may recognize similar situations, more, or none — that is a term to find rather than a rule to assume, and it is worth establishing before it is needed.
- Evidence has to exist at the time. A contractor's error and a retroactive entitlement notice are both documented moments. Keeping the document when it appears is the difference between an argument and a recollection.
- The denial route is separate. Once a filing denial has been issued, the timely filing denial covers what can be argued and how the balance must be treated in the meantime.
- Patterns beat claims here too. A cluster of at-risk claims under one payer is usually a submission, enrollment or configuration problem — a question for A/R by payer analysis rather than a run of individually unlucky accounts.
Common questions
Why not just work the oldest claims first?
Because age measures how long a balance has been open and not how much of its life is left, and the two diverge sharply once several payers are involved. Filing periods are terms of each agreement or program, and they commonly run from the date of service rather than from the date the claim entered a queue — so a long-aged claim under a generous window can carry less risk than a younger one under a short one. Ranking by remaining time puts the claims that can actually be lost at the top, which is a different list from the oldest ones and usually a much shorter one.
What should the escalation threshold be?
Far enough from the limit that something can still be done, which makes it a function of the practice's own turnaround rather than a general number. The actions a claim near its window needs — establishing status, obtaining evidence of the original submission, correcting and resubmitting — all take time, and a threshold set at the point where only a write-off is possible has recorded the loss rather than prevented it. Deriving it from how long those steps actually take in that practice, per payer, is the honest way to set it.
Is a claim past the filing limit always lost?
No, and treating it as automatically lost is where recoverable money goes. For Medicare, 42 CFR 424.44(b) enumerates exceptions: an error or misrepresentation by an employee, Medicare contractor or agent of HHS acting within the scope of Medicare functions; a beneficiary who was not entitled to Medicare at the time of service and was later notified of entitlement effective retroactively; that situation combined with a State Medicaid agency recovering its payment from the provider; and a beneficiary enrolled in a Medicare Advantage plan or PACE organization who was later disenrolled retroactively, where that organization recovered its payment. Two of those describe things practices meet regularly and read as bad luck. Commercial agreements may recognize similar situations or none — that is a term to find rather than a rule to assume.
Does resubmitting restart the clock?
Generally not, and assuming it does is a common and expensive error. The period is usually measured from the date of service or another fixed event, so resubmission produces a new claim inside the same window rather than a new window. That is why evidence of the original, timely submission matters more than the resubmission itself: where a window has closed, the argument available is that the claim was filed on time and something happened to it, which is an argument only the acknowledgment trail can support.
Where should the remaining-time figure live?
On the claim, as a stored field, rather than in a report that has to be run and interpreted. A stored remainder can be sorted on, filtered, alerted against and audited, and it does not depend on anybody remembering which payer has which window. It also has to be re-derived after events that change the claim, since a value computed once at intake and never revisited goes wrong silently — which is the failure mode that produces a queue everyone trusts and nobody checks.
Key terms in this article
Defined once, on their own pages.
Continue learning
The report this re-sorts, and what happens if the window closes.
A/R Aging Buckets
Why bucket edges are a convention, and what a threshold that means something looks like.
Unbilled and Held Claims
The claims already spending their window before anyone has billed them.
No Response From the Payer
Separating the never-adjudicated from the pending, which changes the risk at the same age.
The Timely Filing Denial
What can be argued once the window has closed and a denial has been issued.
Timely Filing Calculator
Work out what a filing window leaves, from your own dates.
Authoritative sources
- 42 CFR § 424.44(b) — Time limits for filing claims: exceptions (opens in a new tab)
Enumerates the exceptions under which the time for filing a Medicare claim is extended where CMS or one of its contractors so determines: a failure caused by error or misrepresentation of an employee, Medicare contractor (including a Medicare Administrative Contractor, intermediary, or carrier), or agent of HHS that was performing Medicare functions and acting within the scope of its authority; a beneficiary who was not entitled to Medicare at the time the service was furnished and who subsequently received notification of Medicare entitlement effective retroactively to or before the date of the furnished service; that situation together with a State Medicaid agency recovering the Medicaid payment for the furnished service from the provider or supplier a period after the service was furnished; and a beneficiary who was enrolled in a Medicare Advantage plan or Program of All-inclusive Care for the Elderly organization at the time of service, was subsequently disenrolled effective retroactively to or before the date of the furnished service, and where that organization recovered its payment from the provider or supplier a period after the service was furnished.