Accounts receivable recovery: working down old claims without writing off money you could still collect.
Every practice accumulates some unpaid claims. The question that actually matters is not how large the balance looks on a report, but how much of it is realistically recoverable, and that answer changes completely depending on why each claim is still open.
Key takeaways
- Not all old A/R is equally dead. A claim still inside its appeal window is far more recoverable than one past it, regardless of age.
- Segment before you work. A blended pile of claims wastes effort on the wrong ones; sort by age, reason and dollar value first.
- A backlog is a symptom, not just a problem. Fix the upstream cause or the same claims pile up again next quarter.
- A specific write-off reason is worth more than a clean balance sheet. It is the data that tells you what to fix upstream.
Start with aging, not age alone
Standard aging buckets, 0–30, 31–60, 61–90, and 90-plus days, exist for a reason: they roughly track when a claim moves from "still in normal process" to "something has gone wrong" to "running out of time to fix." But age by itself is a blunt instrument. A 120-day claim that is still inside its appeal window is a very different problem from a 45-day claim past its appeal deadline. The number that should drive priority is not how old a claim is, it is how much runway remains before the deadline that determines whether it can still be recovered at all.
| Bucket | What it usually means | Typical action |
|---|---|---|
| 0–30 days | Normal processing time for most payers | Monitor; no action needed unless payer's typical turnaround has passed |
| 31–60 days | Slower than expected, or an unworked denial | Check claim status, resolve denials as they surface |
| 61–90 days | A genuine problem: denial, missing information, or payer delay | Escalate; confirm the appeal or resubmission window is still open |
| 90+ days | High risk of missed deadlines and reduced recovery odds | Triage: recover what is inside a window, document and write off what is not |
Segment the backlog before working it
Working a backlog in date order, oldest first, feels intuitive but is usually the wrong approach. A better first pass sorts the whole balance three ways: by dollar value, since a handful of large claims often account for a disproportionate share of the total; by denial reason, since claims denied for the same fixable cause can often be corrected and resubmitted as a batch rather than individually; and by remaining window, since a claim about to lose its appeal eligibility needs to move ahead of a claim with more time left, regardless of which one is older. This sorting takes a few hours up front and routinely saves far more than that in wasted effort on claims that were never going to be collectible.
When to keep chasing, and when to write it off
Not every dollar is worth the staff time to pursue. A claim is a reasonable candidate to write off once the appeal window has genuinely closed, once a patient balance has gone through a documented, good-faith collection attempt without result, or once the remaining balance is small enough that continued pursuit costs more in staff time than the balance is worth. What separates a healthy write-off process from a sloppy one is specificity: recording the actual reason, timely filing expired, patient uncollectible after three attempts, provider write-off per contract, rather than lumping everything into one generic adjustment code. That reason data is what eventually tells you whether the real problem is a slow eligibility process, an undertrained coder, or a genuinely uncollectible patient population, and each of those has a different fix.
Before writing off any batch of old claims, run them through one more check: group by denial reason and see if a single fixable issue, a missing modifier, an outdated payer ID, an eligibility mismatch, explains a large share of them. Practices are frequently surprised to find that what looked like an unworkable backlog was actually one repeated, correctable mistake spread across dozens of claims.
Stopping the backlog from rebuilding
Clearing an existing backlog without addressing why it built up in the first place buys a few months of clean numbers before the same pile reappears. The recurring causes are usually upstream of the claim itself: eligibility not verified before the visit, a scrubber that is not catching payer-specific rejection patterns, or a denial queue that is not worked on a fixed schedule, so claims sit for weeks before anyone looks at them. A practice that reviews its top denial reasons monthly and assigns denial follow-up as a specific, tracked responsibility, not a task that happens whenever someone has spare time, is the practice whose A/R balance stays low rather than the one that needs a periodic backlog cleanup project.
Frequently asked questions
How old does a claim have to be before it's basically uncollectible?
There is no universal number, but recovery odds drop sharply after 90 days and drop again after 120, mainly because timely filing and appeal windows close during that period, not because the money itself becomes harder to justify. A claim still inside its appeal window at 150 days is more recoverable than a claim past its window at 60 days. Track the deadline, not just the age.
Should we write off old balances or keep trying to collect them?
Write off a balance once you have confirmed there is no further legitimate recovery path: the appeal window is closed, the patient is confirmed uncollectible after a documented attempt, or the amount is below what it would cost in staff time to keep pursuing. Keep the reason code specific rather than lumping everything into one generic write-off category, since that reason data is what tells you which upstream process to fix.
What is a realistic days-in-A/R target?
Under 40 days on average is a reasonable target for most specialties, though the achievable number depends on payer mix. What matters more than hitting an exact number is the trend: a rising days-in-A/R average, even if still under 40, is an early warning that something upstream, eligibility, coding or submission timing, is slipping before it shows up as a cash problem.
Can old A/R actually be recovered, or is a backlog mostly a lost cause?
A meaningful share of an old backlog is usually recoverable, particularly claims denied for a fixable reason like a missing modifier or an eligibility mismatch rather than a true coverage exclusion. The recovery rate on a backlog is rarely zero and rarely 100%; the value of a dedicated review is finding out which claims in your specific backlog are which, rather than assuming the whole balance is dead.
Do you take over A/R that's already old, or only new claims going forward?
Both. When we start an engagement we inventory the existing backlog, separate what is realistically recoverable from what should be written off, and work the recoverable portion in parallel with taking over new claims. You are not choosing between a clean start and a resolved backlog.
Sitting on an A/R backlog you haven't had time to triage?
We inventory what's actually recoverable, work it alongside your new claims, and give you a monthly report on where the balance stands.