Humata Health [What It Changes]

TL;DR

• Humata Health is an AI vendor that checks payer rules, bundles clinical evidence, and files prior-authorization requests before a claim is submitted, so a clinic is not waiting on a phone or fax queue for the “yes.”

• As of August 18, 2026, R1 agreed to acquire Humata Health and fold that authorization layer into Phare OS, the company’s pre-bill revenue operating system; the deal is expected to close by the end of the third quarter.

• The constraint that broke is not a new medical code. It is a hard clock: CMS now requires faster payer decisions and FHIR prior-authorization APIs, and EHR vendors have started shipping those APIs live.

• A two-truck HVAC shop, a 10-person agency, and a solo clinic all live the same operational pattern: work sits idle until an outside desk stamps approval. Humata Health is that stamp, automated, for medical services.

Key Takeaways

• R1 is buying the authorization step, not a new billing code set. Humata’s agents sit upstream of claims, inside Phare Intelligence and Payer Atlas.

• Vendor-stated Humata outcomes in the acquisition release are a 96% first-pass approval rate, 30% fewer write-offs, 83% fewer reschedules, and 45% fewer staff touches. Those are company figures, not an audit.

• Independent surveys still show a manual mess: physicians report about 40 prior authorizations per week, and payers still take almost half of medical prior-authorization requests by phone, fax, or mail.

• CMS-0057-F already started operational clocks in 2026 (72-hour urgent decisions, seven-day standard decisions, public metrics) and sets FHIR API go-live for January 1, 2027.

• Small clinics that never sign an R1 contract still feel this. Payer portals, EHR order entry, and document packets are the same three queues whether you have 12 exam rooms or two.

What Humata Health is, in one sentence

Humata Health is a physician-led software company whose agents read a payer’s authorization rules, assemble the clinical packet, and submit the prior-authorization request before a claim ever leaves the clinic.

If you run a two-truck HVAC shop, you already know this job. A compressor sits on the truck until a warranty desk says the unit is covered. Until that desk answers, the crew cannot finish the call, the customer waits, and the dispatcher reshuffles the afternoon. A 10-person marketing agency lives the same stall when a campaign cannot ship until brand-legal returns a redline. A solo-run clinic lives it when a scheduler cannot book an MRI until someone proves to an insurer that the scan is covered.

That stall is prior authorization. It is not a billing flourish. It is a gate in front of the visit, the infusion, or the procedure. Humata Health’s product is built to take that gate off the phone tree and put it on a software path: match the policy, bundle the notes, answer the attestation, file the request, and watch the status. The August 18, 2026 R1 deal matters because it attaches that path to a revenue-cycle operating system already sitting in large health systems, which is how a point tool becomes a default queue.

Small operators should care for a concrete reason. When a national billing vendor automates the proof-of-coverage step, payer portals and EHR order screens change around everyone else. Independent practices already saw Humata ship a standalone portal on June 30, 2026 aimed at clinics that still file by phone and fax. You do not need to become an R1 customer to inherit the new default: fewer faxes, more structured packets, and less patience for a staffer who only knows how to hold.

This sits in the same operational stack as healthcare automation work already in motion: intake, document routing, and exception queues. The rest of this hub explains the deal, the mechanism, the CMS clock, the honest limits, and what a small clinic should actually change in the next scheduling cycle.

What R1 announced on August 18, 2026

R1, a Chicago-based revenue-cycle company, announced on August 18, 2026 that it had entered into an agreement to acquire Humata Health. The GlobeNewswire wire and the Yahoo Finance reprint carry the same company text. Financial terms were not disclosed.

According to GlobeNewswire, Humata’s solution allows providers to achieve up to 96% first-pass approval. That same release states write-off reductions of 30%, rescheduled-appointment reductions of 83%, and staff-touch reductions of 45%. Treat those as vendor-stated operating results, not a regulator’s scorecard.

According to TechTarget, Phare OS launched in October 2025 and R1 expects to close the Humata purchase by the end of the third quarter of 2026. After close, the Humata team is slated to join R1’s R37 agentic-AI lab rather than remain a standalone brand with its own sales motion. Steve Albert, R1’s chief strategy and growth officer, told TechTarget the agentic layer is meant to connect “across hundreds of payers.” Joe Flanagan, R1’s CEO, called real-time authorizations a near-term goal in the release. Jeremy Friese, MD, Humata’s founder and CEO, framed the same deal as a way to stop authorization delays from draining clinics.

R1’s own “about” block in that wire says the company partners with 1,000 providers, including 95 of the top 100 U.S. health systems, and handles over 600 million payer transactions annually. The Phare OS product page repeats the 600 million-plus transaction figure, adds $76 billion-plus of net patient revenue flowing through the operating system, and lists 1,500-plus payer connections in Payer Atlas. The Payer Atlas page states 1,500-plus connections, 115,000-plus payer policies monitored in real time, and 36 million-plus accounts priced daily.

None of those scale numbers are a clinic’s daily reality. They are why a prior-authorization specialist is being bought rather than partnered: R1 already sits on the pre-bill path (authorization, utilization review, documentation, coding) and wants Humata’s policy-matching and packet-building agents inside Phare Intelligence instead of as a side car.

Sources: CMS press release; Phare OS; Humata June 30, 2026 release; GlobeNewswire.

How the mechanism actually works

Strip the product names and the job is four steps a human already does badly.

First, someone has to know whether this CPT, this drug, this imaging study, or this stay needs authorization at all, for this payer, on this plan, today. That list changes. Humata’s site describes PolicyLink, PolicyMatch, and Policy Intelligence as the layer that digitizes and matches those rules. R1 says that monitoring will feed Phare Intelligence and Payer Atlas, which already claims 1,500-plus payer connections.

Second, someone has to pull the clinical evidence the payer asked for: notes, labs, step-therapy history, imaging reports. Humata calls this AI-driven clinical bundling. The June 30, 2026 standalone portal for independent practices scores an uploaded packet against Medicare coverage determinations and commercial policies and returns a completeness grade before anything is sent.

Third, someone has to answer the payer’s attestation questions. Humata’s “Smart Attestation Answering” reads the bundle and drafts those answers. That is the part a scheduler currently copies from a chart into a portal while a patient waits in a hallway.

Fourth, someone has to submit, status, and chase. Humata lists automated statusing, post-auth monitoring, CPT-mismatch checks, and an audit trail. The company’s public site also states a vision of 90% touchless authorizations and two-minute human reviews. That vision is a product goal on Humata’s homepage, not a measured result in the acquisition release.

The industry term for software that takes a goal (“get this MRI approved”) and runs those steps with limited hand-holding is an agentic workflow. It is not a chatbot glued onto a fax server. It is a loop that reads a policy, gathers evidence, files a transaction, and only wakes a person when the payer asks for something the packet does not contain.

That loop only works if the clinic can actually reach the payer electronically. HL7 Da Vinci Coverage Requirements Discovery (CRD) is the implementation guide for asking, at order time, whether authorization is required. Documentation Templates and Rules (DTR) is the guide for pulling the right questions and documents. Prior Authorization Support (PAS) is the guide for sending the request and getting the decision back, built on FHIR R4. CMS recommended those three Da Vinci guides in the CMS-0057-F fact sheet. ONC is the federal office that keeps the interoperability stack (USCDI, FHIR, certification) from being a private dialect.

On August 17, 2026 — one day before the R1 announcement — Epic said Ochsner Health, Froedtert ThedaCare, Denver Health, and Summit Health were live with a CRD API against UnitedHealthcare, Aetna, Network Health, and other payers, with 16 more payers in testing. TechTarget’s write-up notes CMS is working with 29 early adopters, including Epic, Oracle Health, and athenahealth. Specialty groups already comparing those EHRs in Epic versus athenahealth workflows will see this first at order entry, not in a separate authorization shop.

Humata does not replace CRD. CRD answers “do I need auth?” Humata tries to finish “here is the packet, here is the attestation, here is the status.” R1 is buying the second job so Phare OS can claim a pre-bill path that starts at the order and ends at a cleaner claim.

Why this is happening now

Prior authorization is not new. What changed is the legal clock plus the channel.

According to the CMS press release on CMS-0057-F, the rule is estimated to produce $15 billion in estimated ten-year savings. Impacted payers (Medicare Advantage, Medicaid and CHIP FFS, Medicaid managed care, CHIP managed care, and QHP issuers on the federally facilitated exchanges, with some carve-outs) must send decisions within 72 hours for expedited requests and seven calendar days for standard requests, generally beginning in 2026. The same release delays FHIR API compliance to January 1, 2027 after comment.

According to the CMS fact sheet, impacted payers have 72 hours for urgent decisions and 7 calendar days for standard decisions. Beginning in 2026 they must give a specific denial reason even if the request arrived by fax or phone. Public prior-authorization metrics must be posted annually, with the first set due by March 31, 2026. The Prior Authorization API must list covered items and services, documentation needs, and an approve / deny / need-more-info response, and it must be live beginning January 1, 2027. HHS also announced HIPAA X12 278 enforcement discretion so a FHIR-only or FHIR-plus-X12 API can satisfy the rule. Drugs are carved out of several of these API and process provisions.

Traditional Medicare still uses a narrower prior-authorization and pre-claim-review set (certain hospital outpatient services, repetitive non-emergent ambulance, selected DMEPOS, and review-choice demonstrations), which CMS documents separately from the MA and Medicaid API mandate.

Sources: CMS-0057-F fact sheet; CMS press release.

Payers cannot meet a 72-hour urgent clock if half the inbound requests still arrive as a fax. Providers cannot meet a MIPS attestation if their EHR cannot hit a Prior Authorization API. That is the constraint that broke. R1 is not waiting for 2027. It is buying the agent that already files the packet so Phare OS customers can use the new pipes on day one.

The burden the product is aiming at

Vendor claims only matter against the mess they say they fix.

According to a KFF Health Tracking Poll fielded January 13–20, 2026 among 1,426 U.S. adults, 34% of insured adults name prior authorization as the single biggest non-cost burden in getting care. One in three insured adults (33%) call it a “major burden,” and 69% call it a burden of some kind. Among insured adults with a chronic condition, 39% pick prior authorization as the single biggest burden. Nearly 47% of insured adults say a service, treatment, or medication was denied or delayed in the past two years.

A separate KFF analysis of CMS contract-level data found Medicare Advantage insurers made nearly 53 million prior-authorization determinations in 2024 (52.8 million), or about 1.7 per enrollee. Insurers fully or partially denied 4.1 million requests (7.7%). Only 11.5% of those denials were appealed, and 80.7% of appeals were partially or fully overturned. Traditional Medicare logged just over 625,000 requests in fiscal year 2024, with a 22.9% denial rate, on a much smaller service set. About 99% of MA enrollees face prior authorization for some services.

According to the 2025 AMA prior authorization physician survey of 1,000 practicing U.S. physicians, practices complete 40 prior authorizations per physician per week and spend 13 hours each week completing them. That survey also reports 95% of physicians saying prior authorization delays care, 79% saying it can lead to treatment abandonment, 26% reporting a serious adverse event, and 88% saying it leads to higher overall utilization. Forty percent of respondents have staff who work exclusively on prior authorization. Sixty percent are concerned that AI will increase denial rates.

The American College of Physicians toolkit still cites the older 2022 AMA wave (45 requests per physician per week, 14 hours weekly) and a $2,161 to $3,430 annual cost per full-time primary-care physician from an older study. Humata’s June 30, 2026 ambulatory release quoted that 45-per-week figure. Use the 2025 AMA 40-per-week and 13-hour figures when you need the current physician-survey baseline; keep the 45 only as what Humata cited that day.

According to AHIP’s June 2025 prior-authorization report, 45% of medical-service prior-authorization requests are still submitted manually by phone, fax, or mail (47% for prescription drugs). That same survey of member plans says 93% of commercial medical claims are not subject to prior authorization at all, that 97% of commercial medical prior-authorization requests are ultimately approved, and that about 20% of medical prior authorizations are approved in real time. All responding plans said they do not use AI or algorithms without clinician review to deny medical-necessity requests.

Those two stories can both be true. Payers say most claims never need authorization and most requests are eventually approved. Physicians say the 7% that need a file still consume 13 hours a week and delay care. Humata’s 96% first-pass claim is an attempt to collapse “eventually approved” into “approved on the first packet.”

According to the HHS Office of Inspector General (April 2022, OEI-09-18-00260), 13 percent of denied Medicare Advantage prior-authorization requests in the sampled week met Medicare coverage rules. Eighteen percent of denied payment requests met coverage and billing rules. The report’s point is not that every denial is wrong. It is that a non-trivial share of “no” answers would have been “yes” under original Medicare, often because of extra MAO clinical criteria, missing-documentation flags that the chart already answered, or manual error.

The 2025 DataSpring Index (the renamed CAQH Index) puts a $21 billion industry savings opportunity on leftover manual administrative work. That is a cross-workflow automation gap, not a Humata KPI, but it is the same pile of phone-and-fax labor Humata says it is removing.

Source: R1 / Humata acquisition release on GlobeNewswire. Remainder column is arithmetic on those four figures only.

USTA analysis: what 45% fewer staff touches would mean against the AMA week

Do not treat the next table as a Humata guarantee. It is a desk calculation that uses two sourced inputs already cited above.

Input A: the 2025 AMA survey’s 13 hours per week that physicians and staff spend completing prior authorizations. Input B: Humata’s vendor-stated 45% reduction in staff touches from the R1 release. If those two numbers described the same clinics — they do not, and that is the caveat — then 13 × 0.45 = 5.85 hours of authorization labor displaced per physician per week, leaving 13 × 0.55 = 7.15 hours.

USTA analysis. Inputs: AMA 2025 survey (13 hours) and GlobeNewswire (45%). The product is not a measured result. The AMA sample is 1,000 U.S. physicians; Humata’s 45% is a vendor operating claim from unnamed sites.

The useful read is directional. Even if you believe the vendor cut, more than half the week’s authorization labor is still there. Automation moves the phone queue into an exception queue. It does not delete medical-necessity review, gold-card exceptions, drug authorizations carved out of CMS-0057-F, or the appeal when a first-pass miss lands.

A second sourced comparison, not a new statistic: AHIP’s 97% eventual approval on commercial medical prior authorizations versus Humata’s 96% first-pass claim. If both were measured on the same book of business, the product’s pitch is speed and packet quality, not a higher final-yes rate. The pain KFF and AMA document is the wait and the reschedule, which is why the release also claims 83% fewer rescheduled appointments.

What a small clinic should actually change

You do not need Phare OS to feel this. You need three operational moves.

Map which of your orders already hit a payer API at schedule time. If you are on Epic, the August 17, 2026 CRD go-live is the pattern. If you are on another certified EHR, ask whether Coverage Requirements Discovery is on the 2026 roadmap, because the 2027 API mandate does not care about your vendor’s marketing name.

Stop sending thin packets. Humata’s ambulatory portal is a completeness grade against policy. A clinic can do a cheaper version of that today: a checklist per high-volume CPT that names the labs, notes, and step-therapy evidence the top three payers demand. The patient-engagement layer matters here because a missing clinical note is often a missing patient questionnaire.

Keep a human exception lane. AHIP says plans do not auto-deny medical necessity with unreviewed AI. AMA respondents still fear the opposite. OIG found 13% of sampled MA denials already met Medicare rules. A clinic that lets an agent file every request with no review will eventually file a wrong packet faster. Teams already routing intake documents through US Tech Automations workflows can treat a Humata-style authorization step as a model swap on the same document path, not a rebuild.

A clinic that already extracts payer letters in US Tech Automations can feed those packets into an authorization agent without standing up a second intake queue. Operators who run exception queues in US Tech Automations will still need a person when a payer asks for evidence the chart does not hold.

If you want to see that pattern on a builder instead of a slide, map the authorization step onto an existing workflow.

Honest limits

The purchase agreement is not a close. TechTarget reports a third-quarter 2026 target. Forward-looking language in the R1 newsroom copy is caveated as such.

The 96%, 83%, 45%, and 30% figures are Humata’s, repeated by R1. The release does not name the sites, specialties, payer mix, or measurement window. Texas Health Resources and UHealth quotes on Humata’s site are qualitative overtime and fax-referral comments, not those percentages.

CMS-0057-F does not cover drugs in the Prior Authorization API, the Patient Access prior-authorization add, or the 72-hour / seven-day clocks as applied to medications. Infusion and specialty-drug authorization — a Humata product line on its homepage — still sits in a different regulatory bucket.

Independent practices that never join R1 may get a Humata-branded portal for a while and an R1-branded module later. Integration work, BAAs, and EHR connectors do not vanish because a press release used the word “touchless.”

Gold-carding (skipping authorization for clinicians with high approval rates) is a state-level experiment the ACP toolkit notes in Michigan, Louisiana, Texas, and West Virginia, with mixed operational reports. Humata lists “Intelligent Gold-Carding” as a feature. A feature name is not a gold-card statute.

Signal vs Speculation

Demonstrated fact (sourced): As of August 18, 2026, R1 agreed to acquire Humata Health, with a stated aim of folding Humata’s authorization agents into Phare OS and R37, targeting a third-quarter close. Humata states up to a 96% first-pass approval rate and cuts of 30% (write-offs), 83% (reschedules), and 45% (staff touches). CMS-0057-F is in force, with 2026 process clocks and a January 1, 2027 API deadline. Epic has four health systems live on CRD. KFF, AMA, AHIP, and OIG describe a system where most requests are eventually approved, a large minority of remaining work is still manual, patients name authorization as the top non-cost burden, and a slice of MA denials already met Medicare rules.

Our read (12–36 months, small and mid-size operators): If the close happens on the stated timetable and Phare OS customers turn the module on, large-system authorization shops will file more first-pass packets through APIs and keep humans on exceptions. Independent clinics will not get Phare OS. They will get payer portals that increasingly expect structured FHIR payloads, EHR prompts that know authorization is required at order time, and less tolerance for fax-only staff. The clinics that win are the ones that already have a document packet and an exception queue. The clinics that lose are the ones whose “authorization department” is one person who knows three phone trees. Drug authorization stays messy because the federal API rule mostly left it out. Do not budget as if 96% first-pass is your number unless you measure it on your own payer mix.

Glossary

• Humata Health: Physician-led vendor whose agents match payer authorization policies, bundle clinical evidence, answer attestations, and submit prior-authorization requests.

• Prior authorization: A payer’s requirement that a service, item, or (separately) a drug be approved before it is covered.

• Phare OS: R1’s revenue operating system for pre-bill work (authorization, utilization review, documentation, coding) and later modules such as audit and denials.

• Payer Atlas: R1’s connectivity and policy-intelligence layer; the product page claims 1,500-plus payer connections and 115,000-plus policies monitored.

• FHIR Prior Authorization API: The CMS-0057-F requirement that impacted payers expose a FHIR interface for coverage rules, documentation needs, and approve / deny / more-info responses, generally by January 1, 2027.

• Coverage Requirements Discovery (CRD): Da Vinci FHIR guide for asking, inside the EHR at order or scheduling time, whether prior authorization is required.

• First-pass approval: A request approved on the first submission, without a return for missing evidence or a phone chase.

• Gold-carding: A policy that lets clinicians with high historical approval rates skip prior authorization on specified services.

FAQs

What is Humata Health?

Humata Health is an AI-powered prior-authorization company that matches payer rules, bundles clinical evidence, and submits requests without a staffer sitting on a phone or fax queue. It sells to health systems, payers, and, as of June 30, 2026, independent practices through a standalone portal. The company describes itself as physician-led and “Built for Yes” on its site.

Who is buying Humata Health, and when?

R1 announced an agreement to acquire Humata Health on August 18, 2026, and TechTarget reports a close targeted for the end of the third quarter of 2026. Terms were not disclosed. After close, Humata’s team is described as joining R1’s R37 lab and connecting into Phare OS.

Does this matter if my clinic does not use R1?

Yes. Payer APIs and EHR order-entry checks are changing on a federal clock, not on R1’s customer list. Epic’s August 17, 2026 CRD launch and CMS’s January 1, 2027 API deadline move the “do I need auth?” question into the chart. Humata already opened a portal for independent practices that still file by phone and fax. Your staff will face structured packets even if the logo on the module is not R1.

What numbers in the press release are independently audited?

None of the 96%, 83%, 45%, or 30% Humata operating claims in the acquisition wire are presented as an external audit. Independent numbers you can cite without Humata are KFF’s 34% “single biggest burden,” AMA’s 40 prior authorizations and 13 hours per physician per week, AHIP’s 45% manual medical submissions, OIG’s 13% of sampled MA denials that met Medicare rules, and CMS’s $15 billion ten-year estimate for CMS-0057-F.

How does CMS-0057-F change daily work?

Impacted payers generally must decide urgent requests in 72 hours and standard requests in seven calendar days starting in 2026, give a specific denial reason, and publish metrics. By January 1, 2027 they must offer a FHIR Prior Authorization API. Eligible clinicians and hospitals will attest to sending at least one electronic prior-authorization request via that API in the 2027 reporting period. Drugs are excluded from several of these provisions.

Will AI start denying care on its own?

AHIP’s 2024 member survey says responding plans do not use AI or algorithms without clinician review to deny medical-necessity prior authorizations. AMA’s 2025 survey says 60% of physicians still worry AI will raise denial rates. Those are parallel facts. A clinic’s control is an exception queue with a named reviewer, not a press-release assurance.

What should we do this quarter?

Pick the ten CPT codes that cause the most reschedules. Write the evidence list each major payer wants. Route those packets through one queue instead of three inboxes. Ask your EHR vendor whether CRD is live. Measure first-pass approval and reschedule rate on those ten codes so you have a baseline before anyone sells you 96%.

If you already run document intake and exception routing, see how agentic workflows handle those queues on the homepage product path rather than standing up a second authorization island.

The R1–Humata Health deal is a large vendor buying the pre-claim proof step. The operational lesson for a small clinic is smaller: stop treating prior authorization as a phone skill, start treating it as a packet and an exception, and watch the 2027 API date the way you already watch a claim timely-filing calendar. Open the agentic workflow builder if you want that packet-and-exception pattern on a canvas instead of a policy PDF.