06-reference/research

behavioral health billing firm ehr count

2026-10-05·research-brief·source: deep-research·by Ray Data Co (deep-research synthesis)·! medium
rcmbehavioral-healthababeachheadventure-exploration

The multi-EHR beachhead is half dead: provider MSOs run exactly one EHR, third-party billing firms name four to thirteen, and not one firm anywhere can evidence its count with a named customer

Sensitivity note: this brief names a family member's affiliated holding company and its captive biller. Local vault only, per the living-person rule in [[hcls/network|HCLS network]].

The question

Verbatim: "How many distinct EHRs does a typical behavioral-health billing company or MSO actually operate against? Establish the number for three to five named firms — the multi-EHR beachhead argument collapses if it is one."

Context: this is the #1 follow-up promoted out of [[2026-10-03-behavioral-health-billing-ai-bundled-vs-standalone]], which held the multi-EHR beachhead hypothesis at LOW confidence on purpose and stated plainly that nothing in its pass established the number is greater than one. Acronyms: EHR = electronic health record; EMR = electronic medical record (used interchangeably by these vendors); PM/PMS = practice management system; MSO = managed service organization; ABA = Applied Behavior Analysis; SUD = Substance Use Disorder; RCM = revenue cycle management; ATS = applicant tracking system; BPO = business process outsourcing; FTE = full-time equivalent.

A note on what the question conflated. "Billing company or MSO" is one phrase covering two structurally different buyers, and they return opposite answers. The brief reports them separately because that split is the finding.

What we already know (from the vault)

What the web says

Evidence standard. Every number in the tables below that I label self-verified, I extracted myself by curl with tag-stripping or by Playwright on 2026-10-05, not from a sub-agent paraphrase. Three sub-agents swept roughly 68 firms and ~600 URLs; I re-pulled the load-bearing pages. Evidence classes are the ones the queue row demanded: MARKETING = the firm's own page asserts it, no customer named; NAMED-CUSTOMER = a case study, portal link or operational artifact ties a specific organization to a specific system; GENERIC = "any EHR" with zero systems named, which fails as evidence.

Deliverable table 1 — third-party billing / RCM firms (the buyer the hypothesis needs)

Firm Named EHRs supported Count Source URL Marketing claim vs evidenced by named customers
PRRS (Professional Revenue Recovery Solutions) Kipu, Alleva, Sunwave Lightning Step, BestNotes, Sigmund AURA, ICANotes, Cerner/Oracle, Welligent, TheraNest, SimplePractice, PIMSY, Procentive 13 (self-verified distinct names on page; 11 behavioral-native) https://revenuerecoverysolutions.com/integrations MARKETING. Zero named customers. Its own page states behavioral-health engagements are under confidentiality, names withheld
Staffingly, Inc. AccuPoint, AlohaABA, CentralReach, CR Essentials, Ensora, ICANotes, Kipu, MeasurePM, Motivity, Noteable, NPAWorks, Qualifacts, Rethink, Ritten, SimplePractice, Theralytics, TheraNest, TherapyNotes, Valant, WebABA, Welligent 21 self-verified on one page (page's own claim: "27 platforms… 16 behavioral health EHRs and 11 ABA platforms") https://staffingly.com/behavioral-health/services/platforms/ MARKETING. Programmatic-SEO shape: 11 ABA platform pages share 51-56% twelve-word shingle overlap. Also a per-FTE staffing/BPO model, not a percentage-of-collections biller
ClaraRCM SimplePractice, TherapyNotes, Valant, Kareo/Tebra, Kipu, Alleva, CentralReach, Netsmart, Qualifacts 10 self-verified https://clararcm.com/mental-health-behavioral-health-billing/ MARKETING, with an honest hedge on the same page: "Exact setup depends on the client's software, permissions, clearinghouse, payer workflow, and scope of services"
Missing Piece Billing & Consulting CentralReach, AccuPoint, Rethink, Catalyst, AlohaABA 5 self-verified https://yourmissingpiece.com/faqs/ MARKETING. Verbatim: "We work with all major ABA platforms, including CentralReach, AccuPoint, Rethink, Catalyst, AlohaABA, and more." Zero named customers
Plutus Health CentralReach, Rethink, AccuPoint, Theralytics (+ Artemis ABA, which is Plutus's own product) 4 third-party, self-verified (was 7 in 2024-04 and 2025-07 Wayback captures) https://www.plutushealthinc.com/aba-therapy-billing-solutions MARKETING, and shrinking. Total ABA, NPAWorks, PracticeSuite, MeasurePM appear nowhere on its live site
Exact Billing Solutions (ICBD captive biller) CollaborateMD (plus generic "EMR") 1 https://beta.hirebase.org/company/icbd/jobs/collections-manager-1 (self-verified via Playwright; curl 403) NAMED-CUSTOMER-ADJACENT, weak provenance. Third-party aggregator's structured extraction of EBS's own Collections Manager posting (posted 2025-10-21): Technologies 2 — CollaborateMD · EMR. EBS's own site at exactbilling.com names zero systems (self-verified: only "Impeccable billing software systems")

Deliverable table 2 — provider MSOs and multi-site groups (the other half of the question)

Organization Named EHRs supported Count Source URL Marketing claim vs evidenced by named customers
Action Behavior Centers (400+ locations, 9 states) CentralReach 1 https://www.actionbehavior.com/resource/post/why-we-use-centralreach-to-track-progress-in-aba-therapy NAMED-CUSTOMER (first-party). Self-verified: a dedicated page titled "Why We Use CentralReach"; 39 CentralReach mentions, zero other vendor named. Verbatim: "At Action Behavior Centers, we use CentralReach, a trusted ABA data platform"
Acorn Health (70 locations, 7 states) CentralReach 1 https://centralreach.com/blog/how-acorn-health-achieved-92-same-day-note-conversion-by-fixing-documentation-at-the-source/ ; https://acornhealth.com/ NAMED-CUSTOMER (strongest in the set). Self-verified: vendor case study names the org 17 times, "a large ABA practice with 70 locations," and mentions zero other vendors; Acorn's own site-wide nav carries href="https://members.centralreach.com/#login">Client Login
BlueSprig Pediatrics (154 center pages in own sitemap) CentralReach 1 https://www.bluesprigautism.com/ NAMED-CUSTOMER (first-party). Self-verified in site-wide header markup: href="https://members.centralreach.com/#login"
Centria Autism (~120 locations) Proprietary "CareConnect" / "Centria EMR system" 1 (in-house) https://boards-api.greenhouse.io/v1/boards/centriaautism/jobs?content=true NAMED-CUSTOMER (operational artifact). CareConnect in 85 of 1,266 own job postings; commercial vendor hits: zero
ABA Centers of America (ICBD) Proprietary ABATECH / internal "ABA Connect" 1 (in-house) https://bhbusiness.com/2023/08/02/aba-centers-of-america-to-double-its-footprint-without-pe-backing/ NAMED-CUSTOMER (trade press + negative control). "Barnett also owns a technology company called ABATECH that builds software — including an electronic health record"; 202 live ICBD job postings contain zero commercial vendor names
Autism Learning Partners (16 states) CentralReach 1 own job-posting text, two independent mirrors (403 → Playwright) NAMED-CUSTOMER (operational artifact)
Hopebridge (100+ centers, 12 states) Rethink 1 https://www.hopebridge.com/blog/building-bridges-episode-aba-ot-teamwork/ NAMED-CUSTOMER (first-party podcast transcript; auto-generated, lowercase "rethink")
LEARN Behavioral (17 states) Catalyst (DataFinch) at the Autism Spectrum Therapies subsidiary; one unnamed "Trellis specific" system 1 named + 1 unnamed https://autismspectrum-learnbehavioral.icims.com/jobs/46583/ INFERENCE, NOT EVIDENCE. Catalyst was the only vendor hit across 364 live postings on 7 brand-specific ATS subdomains. Do not state "LEARN runs several EHRs" as fact
Behavioral Health Group (110+ locations, 20+ states) SAMMS (current) + Methasoft (legacy) 2, same vendor own UKG careers board full-text search; https://tsenta.com/jobs/behavioral-health-group-desktop-support-technician-... NAMED-CUSTOMER. But Netsmart owns both Methasoft (via Netalytics, 2023) and SAMMS — this is a within-vendor migration, not two independent platforms
Pinnacle Treatment Centers (130+ locations); BayMark Health Services Methasoft (Kipu claimed only by a Kipu salesperson's LinkedIn) 1 confirmed each https://discovery.hgdata.com/product/methasoft WEAK. Third-party technographic data only. Both orgs' own boards (400+ BayMark postings, Pinnacle's board) name zero vendors

The other findings that matter

Convergences and contradictions

Synthesis for RDCO

The answer is one number for one buyer and many for the other, and the half of the hypothesis the parent brief actually named is dead. Provider MSOs and acquisition-grown behavioral-health groups run exactly one EHR. Action Behavior Centers publishes a page explaining why it uses CentralReach. Acorn Health appears in a CentralReach case study at 70 locations with no other vendor mentioned and hands families a CentralReach-hosted portal. BlueSprig does the same in its site-wide header. Centria and ABA Centers of America skipped the commercial market entirely and built their own. The single two-system case, Behavioral Health Group, is migrating between two products Netsmart owns. This is the only place in the entire research where NAMED-CUSTOMER-class evidence exists, and it points one direction. Report it as written: the MSO beachhead is a kill, not a qualified maybe. The parent brief's phrasing — "a BUYER that is structurally multi-EHR" — does not describe this buyer, and the roll-up-sprawl intuition underneath it has no public support at all, in either direction.

The third-party billing company is a different animal and the number there is genuinely greater than one: PRRS names 13, Staffingly 21 on a single page, ClaraRCM 10, Missing Piece 5, Plutus 4. That is a real distribution, self-verified, and it is enough to say the surviving half of the hypothesis is not refuted. But three things make it a much thinner opening than "the number is greater than one" sounds. First, the modal firm names zero, which means most of this market will not even commit to a list, and about 48 of the 68 firms swept name nothing. Second, not one firm anywhere can evidence its count with a named customer — zero NAMED-CUSTOMER artifacts across two independent sweeps, with PRRS stating the confidentiality reason outright. Published counts also routinely exceed named counts (18+ vs 12, 25+ vs 0, 30+ vs 0, 100+ vs 26), two firms' lists shrank between 2025 and 2026, and one eight-item list is byte-identical boilerplate shared across two firms and includes one firm's own product. Third, and most corrosive, the mechanism is logins, not integrations. BestNotes documents third-party biller access as a free per-tenant login; PRRS says plainly it needs no API, ETL or middleware. A firm's EHR count therefore tracks its client roster, which is why "any EHR" is the honest marketing answer. The pain a multi-EHR product would be sold against is swivel-chair labor spread across many small shops, not integration engineering concentrated in one budget line.

Put the two tables side by side and the beachhead has an ownership problem the parent brief could not see. The buyer with budget, scale and a procurement process runs one EHR and is therefore perfectly servable by its incumbent's bundle — which is exactly the trap the parent brief established on the supply side. The buyer that is structurally multi-EHR is a Tier-2 billing shop whose multi-EHR condition costs it labor rather than capital, that will not publish what it runs, and that in the ABA segment is small. I hold the MSO kill at high confidence and "billing firms run 4-13 systems" at medium — medium rather than high because every single data point in table 1 is MARKETING-class, several sit on heavy programmatic-SEO domains, and a published list is a claim, not an audit.

For the family venture on its 3-6 year horizon, this narrows the thread rather than reopening it. Combined with the parent brief, the public record now says: the function is bundled on the supply side, the buyer with money is single-EHR, and the multi-EHR buyer is small and does not pay for integration. The one honest remaining shape is selling labor-displacement to billing firms as a per-seat or per-claim tool that rides existing logins — which is a BPO-productivity product, not a system-of-record play, and it competes with offshore labor rather than with CentralReach. Exact Billing Solutions is a weak design partner for the original hypothesis specifically because of what this pass found: one named system, and its largest in-group client runs a proprietary EHR that no third-party product can target. That does not make the conversation the founder owns worthless — it makes it a conversation about labor economics inside a captive biller rather than a validation of a multi-EHR beachhead.

Why this is in the vault

This closes the #1 open follow-up from [[2026-10-03-behavioral-health-billing-ai-bundled-vs-standalone]] by putting a number on the one variable that brief deliberately left at LOW confidence, and it kills the MSO half of the beachhead hypothesis with first-party and vendor-case-study evidence, so the family RCM/EOB exploration in [[hcls/problem-log|HCLS problem log]] stops carrying an unproven buyer-side premise forward. It also replaces the "roll-up EHR sprawl" intuition with its actual status (no public evidence either way) and supplies the mechanism — per-tenant logins, not integrations — that any future product shape in this lane has to be priced against.

Open follow-ups

Related

Sources

Vault:

Web, all retrieved 2026-10-05. (self-verified) = re-extracted directly by the author of this brief via curl with tag-stripping, or via Playwright where curl was 403'd.

Billing / RCM firms with named multi-system lists:

Mechanism evidence:

GENERIC-NO-NAMES (fail as evidence, listed because the negatives are the finding):

Provider MSOs and multi-site groups:

Exact Billing Solutions (ICBD):

Corroborating job postings:

Framing, not evidence:

Blocked, dead or not reached, so no verdict drawn: beta.hirebase.org and billingparadise.com and operantbilling.com 403 to curl (resolved via Playwright or Googlebot UA); indeed.com, linkedin.com, ziprecruiter.com gated; CareerBuilder and jobs.aapc.com Exact Billing postings 410/404 with no Wayback capture; billingsolutions.net parked (acquired by Medusind, June 2024); rev-ignition site down and never behavioral-health; coronishealth.com/?p=1077 "fully integrated with Kipu" claim 404s and reproduces on no live or archived page — not relied on; a Kipu salesperson's LinkedIn claim that BayMark runs Kipu — not relied on; OnlineJobs.ph posting #1672899 now HTTP 410 — not relied on. No paywalls were encountered. No firm was contacted, emailed, or sent a form.

Research-budget limitation worth recording: the session's WebSearch budget (200/200) was exhausted by the sub-agent fan-out before I could run a final independent falsification search for an organization publicly claiming multi-EHR operation. The negatives above rest on ~600 URLs swept across three sweeps, not on an exhaustive search.