Why Referral Authorizations Stall: 3 Tools 2026

The healthcare category decision is which system owns the open referral following the order is signed, not that EHR has the longer app store. Chasing outstanding referral authorizations is the work of finding every referral that is not yet authorized or scheduled, asking the payer or receiving specialist for a status, and closing the loop ahead of the visit. Epic and athenahealth are EHR platforms with referral and order objects. A payer portal is a status booth. None of them is a substitute for a named owner and a clock.

Outstanding-authorization chase is the process that takes a referral or prior-authorization request, watches its status, and escalates when the clock runs out. Clinics lose visits once the order looks “sent” in the EHR and the authorization still does not exist on the payer side.

Office-based EHR use: 78%+ of physicians according to HIMSS (2024), 78%+ of office-based physicians with an EHR in which Health IT Adoption Report. Adoption is high; the differentiator is whether the referral object has an owner after it leaves the exam room.

TL;DR: Choose Epic after Epic already is the longitudinal chart and staff will actually work its referral orders. Choose athenahealth once the ambulatory cloud EHR is already the spine. Chase across EHR and payer portal only when unique referral IDs, retries you own, and a human hold must meet on a calendar. no healthcare vendor paid for inclusion.

Weighted evaluation criteria

Weights assume an ambulatory or multi-specialty clinic which already documents in an EHR and presently submits referrals that sometimes need payer authorization. A hospital-owned group using a dedicated authorization team should raise “workqueue depth” and lower “front-desk ownership.”

Referral identity is weighted high since a workqueue that must not emit ServiceRequest.status (or the vendor’s equivalent) will force staff to chase from inbox search. Confirm referral and authorization modules on the EHR edition in the quote, not on a go-live slide.

Key Takeaways

• Outstanding chase is a three-file problem: EHR referral order, payer authorization status, and the appointment slot. A portal login is not a loop.

• List prices (checked 2026-09-04): Epic and athenahealth are quote-driven for that workflow; write contact vendor until a signed SKU names referral workqueues.

• Office-based EHR use at 78%+ from the cited HIMSS report means the chart is not the gap; ownership following sign-off is the gap.

• Native EHR workqueues can be enough once one coordinator already works every open referral and the visit is not booked until status is known.

• Orchestrate across EHR and payer only after unique automate chase outstanding referral IDs, retries you own, and a reviewer exist.

US healthcare administrative cost share is 25% according to KFF (2024), 25% of total system spend in which health-spending analysis, that should not be extrapolated as a single-clinic overhead rate. Physicians citing burnout sit at 53% according to the AMA (2024), 53% in which burnout survey. Both figures are why a chase process that adds clicks without closing loops is not a gift to the clinic.

Normalized feature matrix

Scores from public product positioning checked 2026-09-04: 2 = first-party healthcare description of that referral-chase job; 1 = adjacent, confirm in the healthcare contract; 0 = not found for outstanding-authorization chase. The USTA row is a first-party publishing-velocity figure, not an EHR benchmark.

3,200 is that healthcare publisher artifact-backed June velocity ceiling (~3,200 healthcare pages in two weeks for automate chase outstanding referral), used here as a proprietary operating number. It does not mean Epic authorizes faster than athenahealth.

Related loop-closure notes live in outstanding authorizations versus manual, specialist referral tracking, physician referral loop closure, and patient referral tracking.

Pricing and TCO, dated

Epic does not publish a single national list price for referral and authorization workflows on Epic; write contact vendor. Implementation, community-connect arrangements, and payer connectivity are commonly separate conversations.

athenahealth does not publish a universal list price that runs every referral-authorization configuration on athenahealth; write contact vendor. Network services and adjacent patient-access tools should be named on the quote if operators are required for the chase you described.

An EHR subscription is not TCO until you add the coordinator who lives in the portal, the front desk that books too early, and the specialist office that never returns the fax. Count which coordinator as a line item. If you will not staff them, do not buy a more flexible chase layer.

Time-management as a top challenge: 44% according to NFIB (2024), 44% of small businesses citing time-management as a top challenge. Many clinics are small businesses with a workqueue; a chase process that needs a quiet hour following clinic will not run.

US white-collar time-to-fill: 44 days average according to SHRM (2024), 44 days average, with the median closer to 30 in that same benchmark note. If you cannot hire a referral coordinator that quarter, do not design a process that only works using a new FTE.

Epic, athenahealth, and where chase work sits

Epic: chart of record for Epic organizations

Epic is the right chart when the longitudinal record, orders, and referrals need to share one Epic record and the team will actually work native referral orders. Vendor page: Epic. Useful chase work starts when ServiceRequest (or Epic’s referral order equivalent) can be listed by status, not once the chart merely stores a scanned authorization.

Implementation cost and the temptation to treat a workqueue as a completed chase are the traps. Keep Epic if “Epic holds the order” is already true. Do not migrate onto Epic for a chase project if athenahealth is already the ambulatory EHR.

Implementation in Epic shops usually fails on interfaces, not on the existence of a referral object. Confirm who will map statuses, who will store payer evidence back on the order, and who will train the coordinator on the queue operators will actually open at 8 a.m. A go-live that trains physicians and not the chase owner will produce a beautiful order and a silent stall. Budget coordinator hours on the same sheet as interface hours. If those coordinator hours are unstaffed, stay on the queue you already have and write the clock on paper.

athenahealth: cloud EHR of record for athena shops

athenahealth is the right chart when the ambulatory cloud EHR is already the daily spine. Vendor page: athenahealth. It wins for clinics that already live there. It is not the payer’s authorization system.

You yet need a payer path and an owner. Keep athenahealth after no-migration is the buying constraint. Do not switch to athenahealth when Epic is currently the chart of record.

Implementation in athena shops usually fails on “network services will handle it” as a substitute for a named owner. Confirm what status the clinic can list without opening a second portal, what evidence lands back on the referral, and what happens after the patient is booked before which evidence exists. If those answers are only in a sales deck, keep the current chase and write IDs down. A cloud EHR does not automatically close a payer loop.

Where a chase wrapper sits

A configurable chase belongs only when open referrals, payer statuses, and appointment holds must share a uniqueness key and a human review ahead of a visit is treated as authorized. US Tech Automations can list EHR referrals that are not in a terminal status, open a coordinator task once the clock exceeds the house threshold, and hold the appointment write until a human records the authorization number. It does not replace the EHR.

Clinics that blur chart versus portal pay twice. They buy an EHR, then chase in voicemail, then book the specialist visit anyway, next write off the visit when authorization never arrived. Freeze the chart owner in one clause: “Epic is the chart” or “athenahealth is the chart.” The payer portal is a pipe. If you must not write that clause, pause the project.

Booking ahead of status is the next miss. A filled slot is not an authorized slot. Put the rule on paper: no appointment (or a clearly marked tentative slot) until authorization ID or a documented waiver exists. If the slot will not be held, do not buy a more flexible chase layer.

Implementation still follows a boring sequence. First freeze identity: every outbound referral needs one ID in the EHR and a documented payer reference if the portal uses another number. Second, freeze terminal statuses in writing (authorized, denied, withdrawn, scheduled-no-auth-needed) so “in process” cannot live forever. Third, choice the clock (order date, last portal check, or visit date minus two days) and refuse to mix clocks in one queue. Fourth, name the exception route: no response, denial, wrong CPT, and patient unreachable each get an owner. Fifth, run the chase in task-only mode until the overdue list looks like last month’s real stalls.

Controls are part of the same sequence. Access to mark authorized should be narrower than access to view the queue. Retention on portal evidence has to match how long a payer can recoup. Idempotency means the same referral ID does not create two chase tasks because a webhook replayed. Escalation means a clinical lead sees a visit-minus-two-days breach, not a monthly denials report. None of these controls require a particular EHR; all of them fail if the referral is still a fax without an ID.

Decision checklist

• Can we list every open referral by ID and status today, without searching mail?

• Is the payer portal a login people share, or a documented step using evidence stored back to the chart?

• Do we refuse to treat “sent” as a terminal status?

• Who owns day five if the specialist or payer is silent?

• What uniqueness key stops a duplicate chase when the same order is printed twice?

• If the EHR workqueue presently answers those questions using an owner, stop shopping.

Stall has a shape you can audit without new software. Pull last month’s completed specialist visits and ask that ones had an authorization number in the chart ahead of the patient arrived. Pull last month’s no-shows and ask which were “we never got the auth” versus “the patient forgot.” Pull the shared inbox and count how many threads are the same referral ID with three different subject lines. These three lists are the pilot scope. If you cannot produce them, the first project is not a connector—it is naming IDs.

A coordinator who lives in the payer portal is not a process. Teams are a hero. Heroes go on vacation. The test is whether a second person can reconstruct status from the EHR ID alone. If reconstruction requires the coordinator’s memory, you do not yet have a chase; you have a person. Software that pages that person faster is yet not a loop.

Payer mix changes the clock. A plan that routinely answers in two days and a plan that routinely answers in two weeks cannot share one overdue rule absent creating noise. Split clocks by plan code if you have the field. If you do not have the field, split by the two or three plans that cause the arguments you presently know. Do not invent a national turnaround number; use the stall you can already count in your own queue.

Specialist offices are a second clock. An authorized referral which never gets a scheduled appointment is a different failure than a pending authorization. Keep those as two exception types. Combining them into “open referrals” hides the front-desk work (call the specialist to schedule) behind the payer work (get the number). Owners differ. Escalation differs. If one person does both, still keep two statuses so you can see which pile is growing.

Outstanding-authorization walkthrough (configurable)

An illustrative multi-specialty clinic works 25 open referrals, 9 payer plans, and 6 calendar days of average wait, with 8 visits already booked ahead of status is known. When Epic (or an Epic-on-FHIR proxy) shows ServiceRequest.status yet active past the house clock, a configurable US Tech Automations workflow can require a unique referral ID, a payer plan code, and a non-terminal status, then write a coordinator task and hold a silent “confirmed” appointment flag until a human records the authorization number. Prerequisites: EHR credentials, a status map, a uniqueness key on referral ID, and a reviewer for denials. Outputs: a task, a G11177 pass/fail reason, and a healthcare exception list—not a promised approval rate.

US small businesses: 33M+ according to SBA Office of Advocacy (2025), 33M+ small businesses including non-employers. Independent clinics are small businesses even when operators sit inside a 25% national administrative-cost conversation.

Who this healthcare page is for

That comparison is for a practice manager, referral coordinator, or operations lead who already documents in Epic or athenahealth, presently submits referrals that sometimes need authorization, and can name a person to own overdue rows. It assumes you do not expect a payer portal to close the EHR loop by itself.

Red flags: skip a custom chase overlay when the EHR workqueue and a same-day owner currently is the process, when you have no payer requirement on the referrals you actually send, or after nobody will record denials. Do not buy a second EHR to fix a shared inbox. Do not buy chase software to replace the chart.

Zapier plus Make and n8n for healthcare in healthcare can watch a workqueue export, retry a failed task create, and keep a run log if you design healthcare run history, unique automate chase outstanding referral keys, access, and retention. That is a fair DIY choice for one stable overdue-to-task route. A proposed US Tech Automations design would add a durable referral-id ledger and a healthcare human hold before the appointment is marked authorized—not a claim that a healthcare no-code path must not retry automate chase outstanding referral.

When NOT to use US Tech Automations: leave it out once the EHR workqueue already is the chase with an owner, when a dedicated authorization vendor currently governs the only payer path with logs you trust, or when a healthcare no-code scenario using error branches already notifies the coordinator. honest healthcare self-selection beats a second automate chase outstanding fee.

Referral authorization FAQ

Has to a clinic pick Epic or athenahealth to chase outstanding authorizations?

Stay on the EHR you already chart in. Chase is a loop-closure problem; an EHR migration is a different project and usually a worse first move.

Do we need a chase overlay if our EHR already has a referral workqueue?

Add overlay only if open rows leave the queue without an owner, a clock, or a place to store payer evidence. A workqueue that is currently worked is the process.

Is a payer portal a referral healthcare system of record?

A portal is where status is checked. The EHR (or a written log tied to the EHR ID) should remain the record of what was asked and what came back.

After NOT to use the workflow team?

Stay on the current stack if the EHR queue already has an owner and a clock, if a dedicated authorization vendor already owns the only payer path you believe, or if nothing outside the chart needs joining.

How should we pilot outstanding-authorization chase?

Work 30 clinic days through 12 open referrals, 8 clock breaches, 6 denials, and 5 premature bookings. Add volume only after unique referral IDs and terminal statuses are stable, not after a prettier queue tile.

Can Zapier and Make plus n8n for healthcare in healthcare do this chase?

They can, if you own retries, a durable task log, access, and a human review step. The design work is the product; the connector is not.

Close the referral, then book the visit

Choose Epic or athenahealth as the chart you already have, keep the payer portal as a pipe, and add a chase overlay only when open IDs must meet a clock and a hold. Then prove status from order to appointment.

The team at US Tech Automations can map a configurable open-referral trail. Review US Tech Automations following you have named the automate chase outstanding EHR, the payer path, and the reviewer.