How Do Therapists Protect Themselves From 90837 Audits and Extrapolated Recoupment Demands?
What Actually Stops a 90837 Sample From Becoming a Five-Figure Demand
The goal is simple: every 90837 note proves the time on its own, the utilization pattern does not invite a review, and if a records request comes, the sample holds. Here is what does that, move by move.
1. Put Real Start and Stop Times in Every Note
The single most common 90837 audit finding is time that the note does not support. Writing 60 minute session or a round hour is not proof; the reviewer wants an actual start and an actual stop time in the record for every claim. Before a 90837 goes out, the note has to show the session began and ended at specific clock times that add up to 53 minutes or more. That one habit takes the most common downcode off the table, because there is nothing left for the auditor to challenge when the clock is written down.
2. Kill Cloned Content So Each Note Stands Alone
Auditors flag notes that read like a template dropped into every session: the same interventions, the same progress language, the same phrasing client after client. Even when the care was real and distinct, cloned content reads as documentation that was not actually written for that visit, and it weakens the whole sample. Each note needs to reflect what genuinely happened in that session, in language specific to that client, so a reviewer sees individualized care rather than a copy-paste pattern that invites a deeper look.
3. Watch Your Own 90837 Utilization Before a Payer Does
Payers profile billing patterns, and a clinician billing 90837 on a very high share of sessions stands out against peers who use the shorter codes when the time is shorter. That does not mean you cannot bill 90837 when the session earns it; it means you should know your own ratio and be able to explain it before an auditor asks. When your utilization is visible to you and every one of those claims has the time documented, a high 90837 share is a defensible clinical reality instead of a red flag with no support behind it.
4. Respond to a Records Request as a Defense, Not a Formality
A records request is the start of the audit, not paperwork to clear off your desk. Every note pulled should be reviewed against the code before it goes back: does the time support 53 minutes, is the content individualized, is the medical necessity clear. If the sample holds, the extrapolation has nothing to stand on. If a note is weak, you want to know before the auditor scores it, so the response is prepared and the strongest version of the record is what gets reviewed, not whatever was easiest to pull.
5. Hand Documentation Review to a Dedicated Team
Practices that stop losing five figures to a five-note sample do it by handing pre-claim documentation review and audit response to a dedicated team: remote specialists who check every 90837 for time, individualized content, and necessity before the claim goes out, and who own the records-request response when one lands, live in 1 to 2 weeks. The clinicians go back to seeing clients, a trained backup covers every gap, and the audit exposure stops being the thing nobody has time to watch. Below is what it sounds like when nobody owns it yet, in providers’ own words.
Key Pain Points and Discussions by Providers
real reports from practice staff, lightly edited
“They pulled five notes, decided the documented time did not support 90837, downcoded the sample to 90834, and then applied that failure rate to years of claims. A handful of notes turned into a demand for around fourteen thousand dollars. The sessions all happened; I just wrote sixty-minute session instead of the actual clock.” – practice owner, outpatient therapy group
“Nobody told me that billing 90837 on almost every session was the thing that put me on their radar. I bill it because my sessions run long, but the pattern flagged me before anyone even read a note, and once they were reading, every round-number time in my chart was a problem.” – licensed therapist, behavioral health practice
“The auditor said my notes looked cloned. Same interventions, same wording, session after session. The care was real and each client was different, but on paper it read like a template, and that alone made them dig deeper into the whole sample.” – clinical director, group practice
“I got the records request and treated it like routine paperwork. I pulled whatever notes were easiest and sent them back. That was the mistake. Half of them had the weak time documentation, and those were the ones that got scored and multiplied.” – billing lead, mental health practice
“I have learned to write actual start and stop times in every single note now. The day the clock is on the record, there is nothing for them to downcode, and a small sample stays a small sample instead of a five-figure clawback.” – office manager, outpatient therapy practice
Our Answer
Here is what we actually do. A dedicated remote specialist reviews every 90837 before the claim goes out: they confirm the note shows real start and stop times supporting 53 minutes or more, that the content is individualized rather than cloned, and that the medical necessity is clear. They track your 90837 utilization so a high share is documented and defensible instead of a silent red flag, and when a records request lands, they own the response, reviewing every pulled note against the code so the sample that gets scored is the strongest version of your record. Our specialists are credentialed professionals, including US-licensed nurses and clinicians trained in behavioral health billing and audit workflows, working inside the EHR and billing tools you already use, with AI drafting the first-pass review and a human verifying every note. This is our medical billing support paired with an AI-first workflow, in one paragraph.
Why This Keeps Happening
If the sessions really happened, why does a small sample turn into a five-figure demand? Because of how the audit math works. The reviewer does not read every claim; they pull a small sample, score the failure rate, and extrapolate that rate across your entire claim history for the lookback period. A 90837 downcoded to 90834 is a modest dollar difference on one claim, but multiplied across years of sessions the recoupment climbs into the tens of thousands fast. The sample is the lever, and a few weak notes move the whole thing.
The reason those notes fail is almost never the care; it is the record. CPT 90837 requires a documented 53 minutes or more of psychotherapy, and the American Medical Association’s CPT definitions are explicit about the time thresholds separating 90832, 90834, and 90837. When a note reads sixty-minute session instead of an actual start and stop time, the reviewer has no way to confirm the clock, and the safest call for them is to downcode. Add cloned content that reads as templated, and the sample looks weak before the medical necessity is even in question. Closing that gap before the claim goes out is exactly what a documentation-review workflow with human oversight is built to do, and it is the same discipline behind our AI medical coding support.
And the trigger for the whole review is often the pattern, not any single claim. Payers profile billing, and 90837 is one of the most audited psychotherapy codes precisely because it is billed heavily. A clinician using it on a very high share of sessions stands out against peers, and that ratio can start the records request before anyone reads a note. When you cannot see your own utilization and your notes carry round-number times, you are exposed on two fronts at once: the pattern that draws the audit and the documentation that fails it.
Most groups have already tried the obvious fixes before they talk to anyone. Each one fails the same way: the work lands back on the practice. The pattern, in one table:
| What you tried | What actually happened | Who ended up doing the work |
|---|---|---|
| Wrote 60 minute session in the note | Round numbers do not prove the clock; the reviewer downcoded to 90834 because nothing showed 53 real minutes | Whoever wrote the note that day |
| Reused a template for progress notes | Cloned content read as documentation not written for the visit and weakened the whole sample | A copy-paste habit nobody flagged |
| Sent back whatever notes were easiest for the records request | The weak-time notes got scored and extrapolated across years of claims | Whoever pulled the charts in a hurry |
| Gave documentation review to a dedicated remote specialist | Every 90837 checked for real times, individualized content, and necessity before the claim went out, and the records request owned | Someone whose whole job it is |
The Solution
So what does “someone whose whole job it is” look like on a 90837 claim? The specialist starts before the claim ever goes out. They read the note against the code: does it show a real start and stop time supporting 53 minutes, is the content specific to that client rather than cloned, is the medical necessity clear. The claims that are clean go out clean, and the ones that are weak get flagged back to the clinician while the session is still fresh enough to document correctly. Most recoupment exposure is a documentation problem caught too late, and catching it early is exactly what dedicated medical billing support is built to do.
Then there is the pattern nobody watches until a payer does. The specialist tracks your 90837 utilization so you know your own ratio and can explain it, and so every high-share claim has the time documented behind it. When a records request lands, they own the response: pulling the right notes, reviewing each one against the code, and preparing the strongest defensible version of the record so the sample that gets scored holds up instead of getting downcoded and multiplied. That is the difference between a five-note review that ends there and one that becomes a five-figure demand.
Behind all of it, AI drafts the first-pass review and a credentialed human verifies. The workflow reads each note, checks it against the time and content rules, and flags the weak ones; a person confirms the clinical documentation is right and owns the audit response. Every security control that protects the chart data moving through that process is documented and auditable, and the whole approach is described on our HIPAA and security page, because moving behavioral health records through a billing workflow is only safe when the controls are real.
Who Actually Does This Work
Fair question: why would an outsourced team defend your 90837 documentation better than your own staff? Because reading notes against the code and building an audit-ready record is their entire day, not the thing a clinician squeezes in after a full caseload. The people working your documentation are credentialed medical professionals: overseas-trained physicians, US-licensed nurses and pharmacists, and PharmDs, all trained in US behavioral health billing and audit workflows. They know what a psychotherapy auditor scores, how time documentation has to read to hold, and what a cloned-note flag looks like before a reviewer ever sees it. That is not a generalist task handed to whoever is free; it is a specialty.
We are not a call center. We are a clinical operations partner, a healthcare BPO built on dedicated virtual staff: 500+ credentialed professionals, 24/7 coverage, and the AI-first-pass plus human-verify workflow you just read about behind every one of them. A typical practice is live in 1 to 2 weeks, at up to 70% below the cost of hiring locally, and no one on our side goes out without a trained backup already inside your workflow, so a records request never sits because the one person who handles audits is out.
And the security piece your compliance officer will ask about: we are audited to SOC 2 Type II with zero exceptions and certified for ISO/IEC 27001:2022, HIPAA, and GDPR, with zero breaches in eight years. Every workstation runs inside a secure enclave on US-based servers, with screen captures and downloads blocked by policy, so PHI never sits on someone’s home laptop. Every client account carries a $5M E&O and cyber liability policy and a BAA signed before any work starts; the full detail lives in our HIPAA and security posture.
Put the routine and the people together, and a specific list of things simply stops happening.
Ready to Make Every 90837 Note Audit-Proof?
How We Permanently Fix the Process
A person alone is not the fix, and neither is a bot alone. The fix is a documented review workflow: every 90837 checked for real start and stop times, individualized content, and clear necessity before the claim goes out, your utilization tracked so the pattern is visible, and a set response for when a records request lands. Before we review a single claim for a new practice, we look at your current 90837 share and your note structure so we can see where the exposure actually is, and we build the review against that, not against a generic template.
From there the workflow becomes a living playbook rather than a habit in one clinician’s head. It records exactly how time has to be documented, what individualized content looks like for your specialty, how to read your utilization, and the step-by-step response to a records request. It is written down, kept current as payer rules change, and owned by the team. When your specialist is out, a trained backup works the same playbook the same way, so a weak note never slips through and a records request never waits for one person to come back.
That is the difference between hoping this year’s sample holds and fixing the process for good, and it is what a dedicated medical billing partner actually buys you. A staffer leaving used to mean the review lapsed and round-number times started creeping back into the chart. Under this model the workflow keeps running, the playbook stays, the backup steps in, and a 90837 audit stops being the thing that can reach back and reopen years of paid claims.
The Whole Thing in Four Sentences
Therapists get hit with extrapolated 90837 recoupment because auditors score a small sample and apply the failure rate across the whole claim history, and the notes fail on documentation, round-number times instead of a real clock, cloned content, a high utilization pattern that triggered the review, not on the care itself. Writing sixty-minute session, reusing a template, or treating a records request as routine paperwork all fail the same way. The fix is to put real start and stop times in every note, keep each note individualized, watch your own 90837 share, and treat a records request as a prepared defense. A multi-provider outpatient therapy group runs exactly this model with us today, names withheld, no patient data shown.
If you want to check us out before talking to anyone: our security posture is independently auditable, we are an MGMA 2026 Corporate Member, and 800+ providers run back office work with us.
Ready to make every 90837 audit-proof? Try us risk free: two weeks, your real 90837 documentation, dedicated specialists reviewing every note before it goes out, and if it does not earn the handoff, you walk away. From here down is the sales part, and it is short: here is exactly what it costs.
One Flat Weekly Rate. 45 Hours of Coverage.
No hourly meters, no setup fees, no long-term contracts. Your dedicated team member covers your desk 45 hours every week, and a trained backup steps in at no charge whenever they are out.
One dedicated remote specialist owning your 90837 documentation review and audit response, solo therapist or single-site behavioral health practice
5+ remote specialists covering psychotherapy billing and pre-audit review across a multi-provider group practice or several sites
10+ remote specialists, multi-location behavioral health network, MSO, or PE-backed platform running documentation review across many clinicians
45 hours of coverage for less than others charge for 40.
Standard US full-time year: 40 hrs x 52 weeks = 2,080 hours, the federal basis for computing hourly pay per the U.S. Office of Personnel Management. A Staffingly plan: 45 hrs x 52 weeks = 2,340 hours a year, that is 260 additional hours included in your flat rate. $399/week x 52 = $20,748 a year / 2,340 hours = $8.87 per hour. Typical US market rates for healthcare virtual assistants run $9.50 to $13.00 per hour for 40 hours of coverage.
Make Every 90837 Note Hold Up This Month
You have seen the whole method. The pilot proves it on your own 90837 notes, with a review your clinicians can watch every day.
Start My 2-Week Free TrialRequest Information
Single specialty or multi-site? One payer or many? Tell us your situation and we will map the right coverage within 24 hours.
Frequently Asked Questions
Where the Claims on This Page Come From
Sources & References
- American Medical Association CPT Psychotherapy Coding. Official CPT definitions and time thresholds for 90832, 90834, and 90837, including the documented 53-minute requirement for 90837. ama-assn.org
- CMS Comparative Billing and Audit Resources. Federal guidance on billing-pattern profiling, documentation requirements, and how sampled findings are extrapolated across a claim history. cms.gov
- MGMA Practice Operations and Coding Compliance Resources. Benchmarks and guidance on documentation, coding accuracy, and audit response for medical group and behavioral health practices. mgma.com
- AAPC Coding and Audit Guidance. Professional coding-body resources on psychotherapy code selection, time documentation, and defending against payer audits and downcoding. aapc.com
- HFMA Revenue Integrity and Audit Resources. Guidance on audit response, extrapolated recoupment, and the revenue impact of documentation gaps in the claim record. hfma.org