Stop the write-off before it becomes one.

Every health system has a bucket it stopped looking at.

Denials that timed out before anyone appealed them. Accounts that aged past internal attention. Balances stranded by a system conversion. Patient balances worked with tools built for insurance follow-up.

None of it was uncollectable when it arrived. It became uncollectable waiting.

How the money gets stranded

Denials that run out of clock. A denied claim has a fixed appeal window. Denial volume scales with payer automation; rework capacity scales with headcount. The difference falls off the end of the window — not adjudicated on its merits, simply timed out.

Aging triage. Every revenue cycle team prioritizes rationally: fresh claims first, because recovery probability declines with age. That's correct — and it means that as new volume arrives daily, the aging bucket is never reached. Not through negligence. Through arithmetic. Accounts cross the write-off threshold not because they were uncollectable, but because nobody had capacity to look before the clock ran out.

System conversions. An EHR or billing system migration strands whatever was in flight. Legacy AR sits in a system nobody logs into, worked by nobody, owned by nobody, because the team that understood it has been retrained on the new platform. This bucket is often large and almost always older than anyone wants to admit.

Self-pay. Patient responsibility has grown into a substantial share of net revenue, and self-pay balances behave differently from insurance AR: they respond to timing, communication design, and payment flexibility rather than to follow-up persistence. Most internal teams are staffed and trained for insurance follow-up, so self-pay gets worked with the wrong tools.

The measurement failure underneath. Most organizations track a final denial rate — what stuck after appeals — and a write-off total. Neither says how much of what was written off was ever genuinely worked. The gap between what was denied and what was pursued is the real size of this lever, and most organizations cannot produce it on demand.

Why the obvious fixes don't work

"We'll get to it next quarter." You won't. The arithmetic that prevented it this quarter is unchanged.

"We'll hire more people to work denials." Recovery capacity grows one hire at a time; denial volume does not. It adds fixed cost to a variable problem, and the long tail still goes unworked.

"We'll appeal everything." Appeal economics are unforgiving. A meaningful share of denials cost more to appeal than they return, and blanket appeal policies consume the capacity that should be going to the high-value ones.

A generalist collection agency. They will collect some of it and cost you something more valuable. In a community where you're also the emergency department, an aggressive collections experience is a reputational event — and reputational events in healthcare are asymmetric. Nobody remembers the account you resolved gracefully.

Write it off and move on. Defensible for genuinely uncollectable balances. The problem is that "uncollectable" is being defined by your capacity rather than by the account's actual merit.

What actually moves the needle

Work the denial while the clock is still running. Prioritize by appeal deadline and likelihood of recovery, not by arrival order — and put agents on the determinate follow-up, so the long tail is reached before it expires. The agents themselves live under Scale Without Headcount.

Specialists, not generalists. Early-out self-pay, aged insurance follow-up, and legacy AR liquidation are different disciplines, and each needs people trained for it.

Protect the relationship while you collect. Representatives trained in both medical billing and healthcare collections, working within FDCPA, HIPAA, and ACA constraints, resolve balances without generating the experience that ends up on local news. This isn't a soft consideration — it's the constraint that makes the whole category safe to pursue.

Close the loop back to the front end. Every recovered denial has a cause. Route it back to the registration, scheduling, or clinical team that generated it, at a cadence short enough to change behavior. Preventing the next one is the work of Get Paid Right the First Time.

Where to start

RevDNA measures initial denial rate, final denial rate, and the gap between them; appeal economics by denial type; aged AR by vintage and payer; legacy balances stranded by conversions; and self-pay recovery by balance age. The output is how much of what you write off was defined as uncollectable by your capacity rather than by the account's merit.