7 Referral Software Picks Medical Practices Use 2026

TL;DR

Referral software for a medical practice is the system that takes an outbound or inbound specialist request, keeps a current status, and shows a named owner when the loop is still open.

The seven products below are ReferralMD, AristaMD, Kyruus Compass, Salesforce Health Cloud, Relatient, Artera, and Treatspace. Pick the product whose record you can actually export and whose privacy owner will approve, not the product with the longest feature list.

None of these tools copies a completed referral into your scheduling, billing, and outreach stack with a human hold — that hand-off is a configured workflow, not a ranking prize.

Checked 2026-08-28 against each vendor's public product or pricing page.

Who this is for

This page is for a practice manager, referral coordinator, or operations lead at an office-based or multi-site practice that already has an EHR, a fax or Direct inbox, and a person who owns "was this referral scheduled?"

Red flags: do not buy a referral platform if you cannot identify the source order in the EHR, name who reviews PHI leaving the practice, or stop an automated patient message when the specialist slot changes. A closed-loop dashboard is not a clinical decision, a network-adequacy filing, or a HIPAA determination.

If a coordinator already tracks a handful of paper faxes, stay on the clipboard until volume or audit need is real. CDC also reports 1.0 billion physician office visits (2019 NAMCS), according to CDC NCHS FastStats, so size software to the referred slice, not every visit. Matching, eConsults, or leakage reporting is when this shortlist applies.

How we evaluated medical referral software

We scored each product on six operating tests: source-record identity, closed-loop status, privacy and transport, coordinator workload, public commercial terms, and export or API evidence — a buyer framework, not a clinical endorsement, HIPAA certification, or implementation quote.

The demo that matters is an awkward record: a missing insurance card, a specialist who never confirmed, a patient who should not get an SMS, and a cancelled slot. Ask for the status field, the owner, the stop button, and the export.

Where a vendor publishes a number, we quote the page and check date; where the page routes to a demo, we write "contact vendor." Week counts are planning denominators for a 20-referral pilot, not vendor SLAs.

Weights are this article's scoring rubric, not a vendor claim. Checked 2026-08-28.

Physicians citing burnout: 53% according to AMA (2024). Referral chase work is clerical, but it sits on the same calendar as documentation; a product that still requires three inboxes per order does not reduce that load.

Feature matrix: what a practice should actually test

A referral product wins when a coordinator can answer four questions from one record: who the patient is, what was ordered, whether the specialist accepted, and who owns the next chase. Matching directories and eConsult panels are useful, but they are not a substitute for status.

Capability labels follow each vendor's public positioning as of 2026-08-28. "Contact vendor" means no public list price on the product page we opened.

Adults with a doctor visit: 85.2% according to CDC NCHS (2024). Most of those visits never need a software-rated "best" referral suite; the buy is for the slice that leaves the office and must come back with a scheduled specialty slot.

Pricing and 12-month TCO (planning, not a quote)

ReferralMD is the only vendor on this list that publishes a free platform plus unit rates. The others sell through a demo. A 12-month total is therefore a worksheet, not a catalog number: platform + usage + coordinator hours + remaining fax lines.

Hours and weeks are local planning denominators for a 20-referral weekly pilot, not vendor performance claims. ReferralMD unit rates are from ReferralMD as of 2026-08-26. Remaining vendors: contact vendor.

A modeled ReferralMD usage line is arithmetic on published units (pages × $0.05, segments × $0.03, $2 per fax number), not a volume forecast. EHR-connected booking is an Enterprise row on the same page. For quote-only vendors, demand a written 12-month sheet that names users, transport, message units, and whether the EHR interface is in the base fee.

The 7 referral products, scored for a medical practice

1. ReferralMD

ReferralMD is a dedicated referral operating platform. Its own pricing page lists a $0 Free tier with unlimited users for closed-loop referral management, workflows, a provider CRM, DirectTrust addressing, and portals, then a Practice tier that keeps the $0 platform and adds usage: $0.05 per fax page, $0.13 per agentic document page, $0.25 per intake form, $0.03 per SMS segment, $2 per month per fax number, and $0.25 per minute for voice automation, according to ReferralMD. Enterprise is custom and covers EHR-integrated appointment booking, parent-child structure, and API/MCP access.

Best fit: a practice that wants a referral-specific system of record and can live on published usage rates until an EHR book-from-referral project is funded.

Limitations: EHR-integrated booking and custom APIs sit on Enterprise, and usage can exceed a cheap-looking $0 headline once fax and SMS volume is real. Confirm BAA, Direct, and retention in writing.

Pros

• Public $0 platform plus named unit rates, rare in this category.

• Closed-loop referral, tasks, and DirectTrust are on the free column of the vendor's own matrix.

Cons

• EHR booking and custom APIs sit on Enterprise.

• Usage lines (fax, SMS, intake) are easy to underestimate.

2. AristaMD

AristaMD is an eConsult network: a PCP (often with an AristaMD nurse assembling the packet) sends a case to a board-certified specialist panel and receives a recommendation, with the vendor stating a response inside 24 hours and more than half of cases back inside 4 hours on its eConsult pages. It is for avoiding an unnecessary face-to-face visit, not running every imaging and procedure referral in a multi-specialty group.

Best fit: primary-care groups that want specialist input without transferring the patient, especially for non-urgent, non-procedural questions.

Limitations: it is not a general inbound/outbound referral CRM, and its public ROI calculator uses planning inputs such as $150 per consult — not a published seat list, so the commercial term is still "contact vendor."

Pros

• Specialist panel and 24-hour response window are the product, not an add-on slide.

• Designed to keep care in primary care when the case allows.

Cons

• Wrong tool if you need closed-loop scheduling of every outbound order.

• License dollars are not on a public price list.

3. Kyruus Compass

Kyruus Compass is provider-matching and search infrastructure used by health systems to send patients to an in-network clinician who actually has capacity, which can reduce leakage for a practice sitting inside or beside a system. For an independent office, Kyruus is usually a system-led buy, not a 10-user SaaS checkout.

Best fit: employed or tightly affiliated practices whose parent already runs Kyruus for Find-a-Doc and referral routing.

Limitations: public pages do not publish a practice-level seat price, and matching is not the same as confirming the specialist completed the visit.

Pros

• Strong when the problem is "which in-network clinician, when."

• Built for system-scale directories rather than a single-office fax queue.

Cons

• Independent practices rarely buy this as a standalone checkout.

• Contact vendor for price, scopes, and whether your office is even a customer of record.

4. Salesforce Health Cloud

Salesforce Health Cloud is a CRM for care coordination: households, care plans, referrals as CRM objects, and the broader Salesforce automation surface. It wins when the practice already lives in Salesforce and needs referral as one more care-coordination object, not a fax inbox with status lights.

Best fit: groups that already have Salesforce administrators, a BAA path, and a named integration owner.

Limitations: Health Cloud, Experience Cloud, licenses, and implementation partners are quote-only. A CRM object is not an EHR order. Clinical content must stay in the chart; the CRM should hold the least PHI needed.

Pros

• Referral can sit next to the rest of the relationship graph.

• Salesforce APIs are documented, so an orchestration layer has something to call.

Cons

• Quote-only, and the stack around Health Cloud is easy to under-scope.

• Overkill if the only job is closed-loop specialty scheduling.

5. Relatient

Relatient sits in patient access and engagement: scheduling, reminders, and intake that often surround a referral rather than replace the referral ledger. Practices choosing Relatient usually feel the no-show and access problem more than the "which specialist" problem.

Best fit: groups whose referral leakage is actually a scheduling and reminder failure after the order is placed.

Limitations: public list pricing is not on the product homepage we opened (contact vendor). Confirm whether the referral object is first-class or whether you are bolting referral status onto an appointment reminder product.

Pros

• Strong when the missing piece is patient access after the order exists.

• Scheduling-adjacent, which is where many "referral" complaints actually live.

Cons

• Not the first pick for a dedicated closed-loop referral ledger.

• Contact vendor for price, BAA, and EHR write-back.

6. Artera

Artera (formerly WELL) is a patient-communication layer: two-way messaging, outreach, and conversational workflows across a health system's phones. Referral status can ride on those messages, but the system of record for the order should remain the EHR or a referral platform.

Best fit: practices whose coordinators already lose hours in SMS and voice tag, and whose privacy officer will approve a messaging vendor.

Limitations: contact vendor for price. Do not let an SMS thread become the only copy of a referral status. Keep a durable record, and keep a stop for patients who opted out.

Pros

• Built for the conversation the patient actually sees.

• Useful as the last mile on top of a referral ledger.

Cons

• Messaging is not a referral database.

• Quote-only commercial terms.

7. Treatspace

Treatspace is a referral-management network aimed at closing the loop between originating and receiving practices, with emphasis on specialist access and referral analytics — worth a shortlist spot only when both ends of the referral will live in the same network tool.

Best fit: originating practices that send a high share of referrals to a defined specialist panel that will also log outcomes.

Limitations: contact vendor for current packages. A network only works if the receiving offices actually update status; otherwise you have bought a second inbox.

Pros

• Category-native: referrals, not a general CRM or a marketing inbox.

• Useful when originating and receiving offices will share the loop.

Cons

• Network effects fail if the receiving office will not log.

• No public seat ladder on the pages we opened; contact vendor.

Worked example: one FHIR order becomes a coordinator queue

A three-physician primary-care office sends about 40 specialty referrals a month, of which 12 are GI and 8 sit more than 14 days without a scheduled slot. The EHR already stores the order as a FHIR ServiceRequest; HL7 defines ServiceRequest.status as a required 1..1 field with values such as draft, active, on-hold, revoked, completed, entered-in-error, or unknown, according to HL7 FHIR. In a proposed, configurable workflow, that status change would extract the patient identifier, specialty, and last update, skip any record missing a BAA-covered destination, and open a coordinator task linked to the source order; a person still decides whether to call the specialist, message the patient, or leave the order in the EHR. Prerequisites are a FHIR API or export the EHR vendor will enable, a status-value mapping, and a human review point before any patient-facing SMS. This is not a live customer result.

That is the step where US Tech Automations would sit if you configure it: trigger on the documented ServiceRequest.status change, write the exception, and stop when PHI or clinical judgment is required. See the agentic workflow builder for how that kind of trigger-to-queue design is specified, including the API prerequisite and the review gate.

Prior-auth electronic adoption: 35% according to CAQH (2024 Index). Referral packets and prior-auth packets travel the same manual last mile; a product that cannot attach the current order status to the chase will not fix that 35% problem by itself.

Related pages: appointment reminders, invoicing, payment reminders, and SMS marketing for medical practices.

Common mistakes

Buying "referral software" when the real failure is specialist capacity. No dashboard creates GI slots.

Treating eConsult volume as closed-loop referral volume. AristaMD can replace a visit; it does not schedule the colonoscopy.

Putting clinical notes in the CRM or the SMS vendor because the coordinator "needed the history in one place." Keep the chart in the chart.

Assuming a $0 ReferralMD platform is cheaper than a quote-only product without modeling fax and SMS units and the EHR interface you actually need.

Skipping the stop button. If a specialist cancels and the reminder still sends, you have created a patient-safety and complaint event.

Which job each product is actually built for

The seven products above solve five different jobs. Buying the wrong one for the job is the mistake underneath most of the list above.

When Zapier, Make, or n8n is the real alternative

The honest alternative is a coordinator-owned scenario in Zapier, Make, or n8n that watches an EHR export, a fax OCR folder, or a form, writes a row, and notifies Slack — keeping run histories, retries, error branches, and audit evidence when you configure them. You still own observability, idempotency (no duplicate task for the same order), escalation, access, retention, and maintenance.

A proposed US Tech Automations design differs mainly in where the controls sit: a named trigger (ServiceRequest.status or a vendor webhook), a required human review before patient messaging, and a durable exception queue the coordinator already uses. It would not replace the EHR, infer medical necessity, or go live without the API or export the EHR vendor documents. If your only need is "email me when a spreadsheet row is new," Zapier is enough.

When NOT to use US Tech Automations

Skip it when the EHR already closes the loop you need, when the specialist network will not share status, or when the only "automation" requested is unattended clinical triage. A simpler product on this list, or a single Zapier catch-hook into a shared inbox, wins those cases. US Tech Automations is a configurable layer above a system of record, not a new EHR and not a substitute for a privacy officer.

Key Takeaways

• Choose a referral ledger first (ReferralMD or Treatspace), an eConsult panel (AristaMD), a matcher (Kyruus), a CRM (Health Cloud), or a messaging/access layer (Relatient, Artera) — those are different jobs.

• ReferralMD is the only vendor here with a public $0 platform plus unit rates; everyone else is contact vendor as of 2026-08-28.

• Score source identity, status, privacy, and export. Ignore feature-count slides.

• A 20-referral pilot with two human review points beats a 90-day all-specialties rollout.

• Orchestration belongs after the ledger exists: trigger, exception, human hold.

FAQs

What is the best referral software for a small medical practice?

ReferralMD is the default first look because the platform is $0 with published usage rates; Treatspace is the look when both ends of the referral will share one network. Confirm BAA and EHR needs before you treat either as done.

Does ReferralMD really cost nothing?

The platform column is $0; fax, SMS, intake, and Enterprise EHR booking are not. Read the unit table on the vendor's pricing page and model your pages and segments.

Is AristaMD a replacement for outbound referral tracking?

No. It is an eConsult path for cases a specialist can advise without a face-to-face visit. Keep procedure and urgent orders on the EHR referral path.

Can we automate a patient text when a specialist slot opens?

Only from an approved template, after a coordinator or a documented rule confirms the slot, and only if the patient has a current communication permission. The workflow should stop when the order is cancelled.

Should Salesforce Health Cloud be on a three-physician shortlist?

Usually no. Put it on the list when Salesforce is already the relationship system and you have an admin and an integration owner. Otherwise you are buying a CRM program to solve a fax-status problem.

When is a no-code stitch enough?

When one export, one inbox, and one named person can own retries and access. The moment you need idempotent status, PHI minimization, and a durable exception queue across EHR and messaging, design the review gates on purpose — in Zapier, Make, n8n, or a configured US Tech Automations workflow — rather than hoping the first scenario is the control system.

For a limited operations build above whichever product you pick, US Tech Automations can map the source field, the exception owner, and the stop condition. Current packaging is on the pricing page.