If you lead an ABA practice, there’s a good chance you’ve done this at least once: blocked out a week, pulled the team off everything else, worked through the aging report line by line, and felt the relief of a clean AR column on a Friday afternoon.
And then, six months later, you’re staring at the same report. Different claims. Same shape.
Most articles on ABA billing backlogs explain how to recover from one. That’s useful the first time. But the more interesting question for any ABA leader on their second or third cleanup is different:
Why doesn’t the fix hold?
The honest answer is that the cleanup usually addressed the old claims sitting in AR without touching the conditions that produced them. The backlog cleared because you threw hours at it. The conditions stayed, and the next backlog started forming the same week the last one was resolved.
This piece walks through where those conditions actually live: the specific points in your session-to-claim pipeline where a day quietly turns into three, three into a week, and a week into next quarter’s AR aging problem.
A Backlog Is a Pipeline Problem, Not a Billing Problem
In most ABA organizations, the path from “session delivered” to “claim paid” passes through six or seven hands and at least three systems. Each handoff is a chance for a session to wait. None of those waits feels urgent on the day they happen. Stacked across a month of caseload, they’re the backlog.
If your billing team is running clean, submitting what reaches them, working denials, posting payments, and you’re still ending the quarter behind, the work isn’t failing at the billing step. It’s failing upstream. The billing team is just where it becomes visible.
That’s why pulling them off forward work to clean up history doesn’t solve anything structural. You’ve moved the bottleneck temporarily. The pipeline keeps producing.
The Six Places ABA Practices Lose Days Before a Claim Is Ever Submitted
Here’s where the days actually accumulate. None of these is dramatic. That’s the point.
Session note completion sliding into “later”
When RBTs can’t complete session notes during the session because the documentation tool is clumsy, because data collection is done on paper to be transcribed later, or because the workflow simply doesn’t make it possible, notes are finished that evening, the next morning, or two days out. A 24-hour drift across 40 sessions per week, per BCBA caseload, compounds quickly.
What it looks like in your data: the median time between session end and note completion exceeds a few hours, and the p90 exceeds 48 hours.
Supervisor co-signs queued in weekly batches
BCBAs review and co-sign RBT documentation. If they’re carrying clinical hours plus supervision plus parent training, co-signs are the thing that gets done “when there’s time”, which usually means Friday afternoon or Sunday night. That’s a five-to-seven-day delay engineered into the workflow before billing has any say.
What it looks like in your data: a stable lag between note completion and co-sign, measured in days rather than hours.
Authorization usage tracked outside the system
When units used vs. units remaining live in a spreadsheet, a shared doc, or someone’s head, two things happen. Sessions get scheduled past the authorized cap and become unbillable retroactively. And authorizations expire mid-week without anyone catching it, which forces a renewal scramble that holds an entire week of sessions in limbo until the new authorization lands.
This is one of the most expensive failure points because the loss is invisible at the time. The sessions were delivered. The clinical work happened. The revenue just won’t follow.
What it looks like in your data: a meaningful percentage of sessions submitted past auth, or a recurring spike of held claims around month-end renewals.
Payor changes that the practice learns about last
A family’s insurance shifts because a parent changed employers. The change happens silently from your end. The first signal is usually a denied claim three weeks later, sometimes longer if the new payor takes its time. By then, several weeks of sessions have been delivered under credentialing that doesn’t apply, and the renewal/re-credentialing process can stretch the recovery out for another month.
This isn’t avoidable in every case. But practices that catch the change early through systematic eligibility re-verification rather than discovering it at denial lose days, not months.
The documentation-to-billing handoff where re-keys live
A session note arrives at billing. A modifier is missing. A measurable goal data field is blank. The supervisor’s signature is technically there, but on the wrong line. The biller bounces it back to the RBT, who is now mid-session with a new client. The note sits in a queue for two more days while the original session ages another two days in AR.
Multiply this across the percentage of session notes that need any revision at all, and you have a structural delay that no amount of billing-team effort can fix from the billing side.
What it looks like in your data: revision rate on session notes is north of 5–10%, with no visibility into which RBT, BCBA, or note type is producing the corrections.
Multi-location and multi-team coordination overhead
Two locations are using slightly different note templates. Three BCBAs with three different completion rates. A new RBT cohort whose first month of documentation needs more review. None of this is wrong; it’s normal growth. But without a single operational view, leadership only finds out at month-end which team is creating its own backlog, by which point the backlog already exists.
What “Fixed” Actually Looks Like
If you audit only your billing department, you’ll see only billing-side metrics. To know whether the pipeline is healthy, measure upstream too. A short, honest dashboard:
- Median and p90 time from session end to note completion
- Median time from note completion to supervisor co-sign
- Percentage of sessions submitted within 48 hours of completion
- Session-note revision rate, segmented by team or BCBA
- Authorization utilization with proactive alerts at 80% and 90%
- Percentage of payor coverage changes caught before the first denial
You don’t need all of these on day one. You need a few that your current system can produce, honestly. If your system can’t produce any of them, that’s the answer to a different question.
These numbers stop billing from being something you explain after the fact. They make it something you can forecast.
Book a Caretap walkthrough and we’ll show you the session-to-claim audit applied to your real pipeline. No pressure pitch.
Why a Connected ABA Workflow Changes the Math?
Most of the failure points above are coordination problems wearing a billing costume. A session note isn’t slow because RBTs don’t care. It’s slow because the documentation tool is one system, the schedule is another, the authorization tracker is a spreadsheet, and the billing platform is a fourth thing that only billers log into.
When session documentation, scheduling, authorization tracking, and billing prep run on the same platform, the handoffs that produce delays stop being handoffs. A co-sign queue is visible to the BCBA in the same place they review progress. Authorization caps show up at scheduling, not at billing. Notes that aren’t billing-ready get flagged when they’re saved, not three days later when a biller pulls them.
Caretap was built around this, one connected workflow for scheduling, clinical documentation, session tracking, authorization management, and billing preparation, with operational dashboards that surface the upstream numbers above. Practices using the platform report up to 40% faster documentation completion and up to 25% faster billing preparation, which is mostly the compounding effect of removing the handoffs rather than any single feature.
The point isn’t the platform. The point is that until the upstream conditions change, the next backlog has already begun to form. The cleanup that lasts is the one that addresses where the days are actually getting added, not the one that resolves what’s already aged.
What to Do This Week?
Three things, in order:
- Pull the upstream numbers at a minimum, the median time from session end to note completion and the percentage submitted within 48 hours. If you can’t pull them, note which system would need to change for them to become visible.
- Identify your single biggest pipeline drag from the six above. Most practices have one that accounts for the majority of their delay. Fixing it disproportionately matters.
- Resist the urge to pause forward work to chase the aging report. Stabilize the pipeline first. History will shrink as the present stops producing new accumulation, and the team won’t be exhausted on the other side.
The backlog is real. But it’s the symptom, not the diagnosis.
Conclusion
A billing backlog rarely starts in the billing department. It starts days or even weeks earlier, in delayed documentation, pending co-signatures, authorization gaps, disconnected systems, and small operational bottlenecks that quietly compound over time.
That’s why clearing aged claims can feel like a victory in the moment, yet leave practices facing the same challenge again a few months later. The backlog disappears, but the process that created it remains unchanged.
The practices that consistently stay ahead of AR don’t just work harder on billing. They create visibility across the entire session-to-claim journey, identify where delays are introduced, and eliminate those friction points before they impact revenue. When documentation, authorizations, scheduling, and billing operate as one connected workflow, claims move faster, staff spend less time chasing issues, and revenue becomes more predictable.
The next time a backlog appears, resist the temptation to view it as a billing problem. Treat it as an operational signal. The goal isn’t simply to clean up old claims; it’s to build a process that prevents the next backlog from forming in the first place. That’s the difference between temporary relief and lasting financial stability.
