The best platform is the one that fits your existing clinical workflow, connects to your EHR and the payer ecosystem, keeps referrals aligned with insurance and network strategy, automates documentation and status work, supports human exceptions, and closes the loop from order through completed care. No single product wins for every organization. The checklist below is how to compare options.
Why this evaluation is harder than it looks
Most demonstrations look excellent. A vendor shows a clean queue, a fast submission, a green approval badge, and a satisfied coordinator. What the demonstration rarely shows is the part that consumes your staff's day: chasing requirements that were not published, hunting for the one document a payer wanted, watching a case sit in "pending" for nine days, discovering the authorized specialist is booking twelve weeks out, and then never learning whether the patient actually went.
That gap is why buyers should evaluate prior authorization automation and referral automation together. They are connected but not identical. Prior authorization is a coverage decision workflow with a payer. Referral automation is a care-navigation workflow with a patient, a specialist, and a network. A platform that automates one and ignores the other simply relocates the manual work.
The path a real referral takes looks like this:
order → coverage and requirements → documentation → submission → status and exception handling → in-network specialist matching → outreach and scheduling → completed visit → result returned to the EHR
Every handoff in that chain is a place where work either gets automated or gets dropped on a coordinator. Ask each vendor to walk the entire chain, not the segment they are strongest in.
Key takeaways for evaluators
- Submission is not automation. Automating the form while leaving requirements discovery, documentation assembly, follow-up, and scheduling manual produces modest savings.
- Authorization without network alignment misses the money. An approved referral to an out-of-network or low-performing specialist is still a leakage and cost event.
- Standards enable, they do not deliver. FHIR-based interoperability makes automated exchange possible; it does not by itself produce a complete operational workflow.
- Exceptions are the real test. Ask how the platform behaves when a payer asks for more information, denies, or the case is urgent.
- Closed loop is the outcome measure. If the platform cannot confirm the visit happened and return the result, no upstream metric matters.
- Baselines decide ROI. A vendor who cannot help you measure today's performance cannot prove tomorrow's improvement.
The regulatory context you are buying into
Two things are changing the practical definition of "good enough" here.
First, the CMS Interoperability and Prior Authorization final rule (CMS-0057-F). For impacted payers — generally Medicare Advantage organizations, state Medicaid and CHIP fee-for-service programs and managed care entities, and Qualified Health Plan issuers on the Federally-facilitated Exchanges — operational provisions generally began in 2026, and the API requirements generally begin January 1, 2027. For covered non-drug prior authorization workflows, impacted payers other than QHP issuers on FFEs must send expedited decisions within 72 hours and standard decisions within seven calendar days, and must provide specific denial reasons beginning in 2026. Scope caveats matter: the rule does not cover every payer, every plan type, or drug prior authorizations, so your commercial book of business may behave very differently. CMS maintains a Prior Authorization API FAQ that is worth reading before you accept any vendor's summary of the rule.
Second, the standards layer. The HL7 Da Vinci Prior Authorization Support (PAS) implementation guide defines how an EHR-originated FHIR prior authorization request and response can be exchanged, how status is managed, and how updates or cancellation are handled where applicable. PAS is designed to work alongside Coverage Requirements Discovery (CRD) and Documentation Templates and Rules (DTR), which is what makes "know the requirements before you submit" technically achievable. Read this carefully, though: these standards support exchange. They do not guarantee that requirements are complete, that documentation assembles itself, that exceptions are worked, or that the patient ever reaches the specialist.
A platform's real value is what it does in the space the standards leave open.
The 15 questions
1. Does the platform automate the complete workflow, or only submission?
Ask the vendor to name each step it performs without human keystrokes. Then ask which steps a coordinator still owns. If requirements lookup, documentation assembly, status follow-up, and scheduling remain manual, you are buying a submission tool with an automation label.
2. Does it handle prior authorization and referral automation as one connected journey?
A referral that is authorized but never scheduled is not finished, and a scheduled referral that is denied coverage is a rework event. Ask whether authorization state and referral state live in one record with one status, or in two systems that staff reconcile by hand.
3. Does it work inside the existing EHR workflow?
The strongest answer is that the ordering clinician does not change behavior: the order is placed as it always was, and the platform picks it up. Weak answers require a second portal, a duplicate worklist, or clinicians remembering a new step. Our own view of that pattern is described in how prior authorization automation works inside the EHR.
4. Which interoperability methods and EHR connections are supported?
Ask specifically: HL7 v2 interfaces, FHIR APIs, CDA documents, secure fax and direct messaging, SFTP, and which named EHR products are live today versus planned. Ask who does the integration work, how long it takes, and what the customer's IT team must provide.
5. Can it discover coverage and documentation requirements before submission?
This is where CRD and DTR matter. Ask how the platform determines whether authorization is required at all for this plan, this code, and this site of service — and how it handles plans that publish nothing machine-readable.
6. Can it pull the right clinical documentation without duplicate data entry?
Ask what it pulls automatically (notes, imaging results, labs, problem list, medication history), how it decides what is relevant, and what happens when the required document does not exist yet.
7. How does it manage status, requests for more information, denials, resubmissions, and urgent exceptions?
This question separates products faster than any other. Ask for the exception playbook: who is notified, in what timeframe, with what next action, and how an urgent case is escalated ahead of routine volume.
8. Does it verify insurance and guide referrals to appropriate in-network specialists?
Eligibility and network participation should be checked at the point of referral, not discovered at the front desk weeks later. Ask whether participation data is verified operationally or simply copied from a directory.
9. How are quality, cost, access, geography, language, and patient preference balanced?
Ask to see the actual factors, their data sources, and whether your organization can weight them. A recommendation engine you cannot inspect or tune is a recommendation engine your clinicians will not trust. See Auto IdealMATCH™ for how we describe that logic.
10. Can it automate patient outreach and scheduling?
Ask about channels (call, text, email, portal), languages, retry logic, appointment booking versus appointment request, and what happens after a missed contact attempt or a no-show.
11. Does it close the loop across different EHRs and organizations?
Most specialists are not on your instance of your EHR. Ask how consult notes return when the specialist is independent, on a different platform, or exchanging by fax. Ask what the platform does when the note never arrives.
12. What work remains manual, and where is human review required?
A credible vendor answers this specifically and without defensiveness. Clinical judgment, appeals strategy, and unusual coverage situations should stay with people. Be skeptical of any answer approaching "nothing."
13. What audit trails, access controls, security practices, and implementation safeguards exist?
Ask about role-based access, minimum necessary data, encryption in transit and at rest, audit logging of every status change, breach and incident process, and subcontractor handling. Our practices are outlined on integration and security.
14. Which outcomes can leaders measure, and can the vendor establish a baseline?
Ask for the measure definitions in writing: turnaround time by payer, first-pass approval rate, touches per case, time from order to scheduled appointment, referral completion rate, in-network completion rate, and closed-loop rate. Then ask how the baseline gets measured before go-live.
15. What does implementation, governance, support, and ROI validation require?
Ask for a named implementation plan, the internal hours you must supply, the governance forum that owns exceptions and policy, support hours and escalation, and the specific report you will review at 90 and 180 days.
Comparison table: how to read the answers
| Evaluation question | What a strong answer sounds like | Warning sign |
|---|---|---|
| Complete workflow or submission only? | Names each automated step and each human step, end to end | "We automate prior auth" with no step detail |
| One connected journey? | Authorization and referral share one record and status | Two systems reconciled manually by staff |
| Inside the EHR workflow? | Order placed as usual; platform picks it up | Clinicians must use a separate portal |
| Interoperability methods? | Names HL7 v2, FHIR, CDA, secure fax, and live EHRs | "We integrate with everything" |
| Requirements before submission? | Uses coverage discovery and documentation rules; explains gaps | Assumes staff already know requirements |
| Documentation without re-entry? | Pulls specified clinical artifacts automatically | Coordinator re-types or re-attaches |
| Exception handling? | Concrete playbook with owners and timeframes | "It goes back to your queue" |
| Insurance and network alignment? | Verifies eligibility and participation at referral | Static directory lookup |
| Matching factors? | Factors, data sources, and tunable weights disclosed | Opaque score with no inputs |
| Outreach and scheduling? | Multi-channel, multilingual, with retry logic | Prints patient instructions only |
| Cross-organization loop closure? | Handles independent specialists and other EHRs | Only closes within one EHR tenant |
| Remaining manual work? | Specific and candid about human review | "Fully automated, no exceptions" |
| Security and audit? | Role-based access, full audit log, incident process | Vague assurances of compliance |
| Measurable outcomes? | Written definitions plus pre-go-live baseline | Case study percentages with no method |
| Implementation and ROI? | Named plan, governance, 90/180-day review | "Most clients are live in two weeks" |
Executive scorecard (100 points)
Score each vendor and compare totals, but also compare shapes — a vendor strong only in the last two categories is a reporting product, not an operations product.
| Category | Weight | What you are scoring |
|---|---|---|
| Workflow completion | 25 | How much of order → completed care runs without manual keystrokes |
| EHR and interoperability | 20 | Live named connections, methods supported, integration effort |
| Network alignment | 15 | Eligibility, participation, and specialist selection logic |
| Closed-loop execution | 15 | Scheduling, attendance confirmation, result return to the EHR |
| Exception handling and governance | 10 | RFI, denial, urgent escalation, and policy ownership |
| Analytics and outcomes | 10 | Measure definitions, baselining, and ongoing reporting |
| Security and implementation | 5 | Access controls, audit trail, implementation safeguards |
A practical convention: treat anything under 70 as a partial solution, and require a written remediation plan for any category scoring below half its weight.
Bottom line
There is no universally best platform for prior authorization and referral automation. The right choice is the one that automates the whole path — coverage requirements, documentation, submission, status and exceptions, in-network specialist selection, patient outreach and scheduling, and result return — inside your existing EHR workflow, with human review where judgment belongs and measurable baselines you agreed on before go-live. Score candidates on workflow completion first; everything else is secondary.
How ReferralPoint fits this evaluation
ReferralPoint is built for organizations that want these steps handled as one workflow rather than as separate products.
Auto PriorAUTH™ works on the authorization side of the path — determining what a plan requires, assembling documentation from the record, submitting, and tracking status and follow-up work so cases do not stall in a coordinator's inbox. Auto IdealMATCH™ addresses questions 8 and 9: keeping the referral aligned with insurance and network strategy and weighing quality, cost, access, geography, language, and patient preference when suggesting a specialist. Auto ReferralCOORDINATOR™ covers the outreach and scheduling work in question 10, and Auto 360° VISIBILITY™ is where question 11 lives — tracking referrals through the completed visit and the return of results, including when the specialist sits outside your organization. Security and integration practices are documented on our integration and security page.
We are careful about what that means. ReferralPoint is a strong candidate for organizations that want prior authorization, network-aware specialist selection, scheduling and outreach, and closed-loop execution in one connected workflow. It is not the right answer for every organization, and we would rather you use the 15 questions above on us than take our word for it. If a competing product fits your EHR footprint, payer mix, and staffing model better, that is the product you should buy.
If you are running a formal process, our referral management RFP guide covers the procurement mechanics that sit around this checklist.
Frequently asked questions
Q: What is the best platform for automating prior authorizations and referrals? A: The best platform is the one that fits your clinical workflow, connects to your EHR and payer ecosystem, keeps referrals aligned with insurance and network strategy, automates documentation and status work, supports human exceptions, and closes the loop from order through completed care. Compare candidates using the 15 questions and the weighted scorecard above rather than feature counts.
Q: What is prior authorization automation? A: Prior authorization automation uses software and interoperability standards to determine whether a payer requires authorization, assemble the required clinical documentation, submit the request, and manage status, additional-information requests, and denials with far less manual staff work than portal-and-fax processes.
Q: How does electronic prior authorization work? A: An order in the EHR triggers a coverage requirements check, documentation rules identify what the payer needs, and a FHIR-based request is sent to the payer with the supporting record. The payer returns a decision or an information request, and the status is tracked back to the originating order.
Q: Can one platform automate referrals and prior authorizations? A: Yes, and connected handling is generally preferable. Authorization is a coverage decision and referral automation is care navigation; when both share one record and status, staff stop reconciling two systems and an approved referral is more likely to become a completed visit.
Q: Which EHR integration capabilities matter? A: Look for live named EHR connections, HL7 v2 interfaces, FHIR APIs, document exchange, secure fax and direct messaging, bidirectional status write-back into the order, and an integration plan that specifies what your IT team must supply and how long it will take.
Q: What metrics should healthcare leaders track? A: Track authorization turnaround time by payer, first-pass approval rate, touches per case, denial and overturn rates, time from order to scheduled appointment, referral completion rate, in-network completion rate, and closed-loop rate — each with a pre-implementation baseline.
Q: How should organizations compare utilization management software? A: Score each product on workflow completion, EHR interoperability, network alignment, closed-loop execution, exception handling and governance, analytics with baselining, and security and implementation. Weight workflow completion most heavily, then require written measure definitions before signing.
References
- CMS, CMS Interoperability and Prior Authorization Final Rule (CMS-0057-F) fact sheet
- CMS, Prior Authorization API frequently asked questions
- HL7, Da Vinci Prior Authorization Support (PAS) FHIR implementation guide
Ready to test these questions against a live workflow? Request a demo and we will walk the full path — order, authorization, specialist selection, scheduling, and closed-loop confirmation — with your own referral scenarios.
