Why Prior Authorization Automation Deserves a Serious RFP Process
Prior authorization remains one of the highest-friction workflows in healthcare operations. The Council for Affordable Quality Healthcare (CAQH) 2023 Industry Insights Report found that practices and systems spend an average of 14 hours per week per FTE on PA-related administrative work. For a 50-provider practice, that’s roughly equivalent to hiring 3.5 full-time staff members just to manage PAs. The American Medical Association (AMA) 2023 Health System Trends Study estimated that the aggregate time and cost burden of prior authorization in the U.S. healthcare system exceeds $31 billion annually.
But the efficiency gain from automation isn’t universal. A poorly chosen PA automation tool can actually make operations worse by creating new bottlenecks, poor EHR integration, or a system that requires more human oversight than your current manual process.
That’s why a structured RFP process matters. It forces you to articulate your current pain points, define what success looks like, and evaluate vendors against consistent criteria rather than being swayed by a slick demo or a low price quote.
Part 1: Define Your Current State and Baseline Metrics
Before you talk to any vendor, you need a clear baseline. This is where most practices stumble. They jump straight to vendor conversations without understanding their own workflow, which means they can’t evaluate whether a tool will actually solve their problem.
Start here: Spend a week documenting your PA workflow from initiation to approval.
Who submits PA requests (providers, clinical staff, coders)? How are requests captured (EHR order, phone call, paper, email)? Who reviews requests before submission (nurse, PA, provider)? Which payers require PAs for your top 20 procedure codes? How many PAs do you manage monthly? What’s your approval rate (what percent require multiple submission attempts)? What’s your average time-to-approval? How many requests expire or lapse because of follow-up delays? How many denials do you get, and what are the top denial reasons?
These metrics will become your success measures. If you’re currently managing 400 PAs per month with 35% of them requiring multiple submissions and a 5-day average approval time, your RFP should target a tool that reduces multi-submission rates to <15% and approval time to 2–3 days.
Document the cost of your current state. Most practices underestimate this. If you have one full-time PA coordinator at $55,000 base salary, that’s roughly $75,000 fully loaded. If they’re coordinating 400 PAs per month, that’s about $225 per PA in direct labor cost. If a vendor can reduce that to $75 per PA through automation, the ROI math becomes clear.
Part 2: Define Your Must-Have Capabilities
Not all PA automation tools are created equal. Some are designed for large health systems with dedicated revenue cycle departments. Others are built for 10-provider practices. Some focus narrowly on PA submission and tracking, while others integrate with broader revenue cycle workflows. The first step in your RFP is defining which capabilities are absolute must-haves for your operation.
Payer Coverage and Intelligence
The single biggest blocker for PA automation is payer coverage data. If your tool doesn’t know which payers require PAs for which procedures, it can’t automate the decision of whether to submit one.
Ask vendors: How comprehensive is your payer coverage database? How many payers do you cover (major commercial, Medicare, Medicaid, regional)? How frequently do you update payer policies (monthly, quarterly)? Do you track procedure-level PA requirements or just plan-level? Can you integrate payers that aren’t in your standard database using a custom rules engine?
For most practices, anything less than 80% coverage of your top 20 payers is risky. If your tool only covers Blue Cross and UnitedHealth but you have significant patient populations in Medicaid or regional payers, you’ll be manually managing those PAs anyway.
EHR Integration Depth
A PA automation tool that can’t read orders and clinical context from your EHR creates manual data entry work. You want a tool that integrates tightly with your EHR and can extract:
- Clinical diagnosis codes (ICD-10)
- Procedure codes (CPT) and modifiers
- Clinical indicators that inform PA decisions (prior surgeries, medication history, imaging results)
- Patient insurance information
- Provider credentials and taxonomy codes
Ask vendors: Do you offer direct integration with my EHR (Epic, Cerner, NextGen, athenahealth, etc.)? Is it real-time bidirectional integration or batch file transfers? Can you read structured clinical data from the EHR or just basic demographics? Do you offer a web portal option if direct integration isn’t available? How long does integration setup typically take? Are there additional implementation costs?
Poor EHR integration is the most common source of PA automation failure. If the tool requires staff to manually copy-and-paste clinical information from the EHR into the PA submission system, you’ve added work, not reduced it.
Automated PA Submission Capability
This is the core value proposition. Some tools simply organize and track PAs but still require staff to manually submit them to payers. Others submit PAs electronically directly to payers’ systems.
Ask vendors: Do you submit PAs electronically to payers, or do you require manual submission (fax, phone, web portal)? If electronic submission, what percentage of your target payers accept EDI PA submissions? For payers that don’t accept electronic submission, what’s your workflow (can you submit via fax or web portal automatically)? How do you handle phone-based PAs? Can you track submitted PAs and flag ones that aren’t acknowledged within expected timeframes?
Electronic submission is a significant time-saver, but only if coverage is comprehensive. A tool that can electronically submit to 60% of payers is better than nothing, but you’re still manually managing 40% of your volume.
Real-Time Status Tracking and Escalation
Once you’ve submitted a PA, the work isn’t over. You need to track status, identify approvals that are missing documentation, and escalate denials for review.
Ask vendors: Do you pull PA status from payers in real-time, or do you rely on manual status checks? Can you automatically flag PAs that have been pending for longer than expected? How do you handle partial approvals or approvals with restrictions? Do you send automated alerts to staff when a PA is approved, denied, or requires additional documentation? Can you integrate denial reasons into a trend analysis dashboard?
The best PA tools give you visibility into every PA’s status without requiring staff to log into payer portals and manually check each one.
Exception Routing and Clinical Judgment Integration
Not every PA decision is algorithmic. Some situations require clinical judgment, provider review, or coordination with other departments. A good PA tool recognizes this and routes exceptions intelligently.
Ask vendors: How do you handle PAs that fall outside standard approval criteria? When a PA is flagged for exception routing, how does it reach the right person (provider, nurse, prior auth coordinator)? Can you define custom escalation rules (e.g., “orthopedic surgeries over $50K require provider sign-off before submission”)? How does your system handle appeals after denial?
A tool that routes 80% of PAs directly to payers and escalates only 20% for review is significantly more efficient than one that escalates 60% due to inflexible rules.
Denial Management and Analytics
Denials are where a lot of operational value gets left on the table. Many practices simply accept denials without understanding the pattern.
Ask vendors: Do you categorize denial reasons and track trends? Can you identify “soft denials” that are overturnable vs. clinical denials that require care plan changes? Do you flag payers that have unusually high denial rates for certain procedure types? Can you automate appeals for soft denials (missing documentation, incorrect modifier, etc.)? Do you integrate with your revenue cycle system to flag denials that are impacting claim payment?
A mature PA automation tool should give you actionable denial insights that help you reduce future denials, not just track historical ones.
Part 3: Technical Integration Questions
Beyond PA-specific capabilities, you need to understand how the tool integrates with your broader tech stack.
Data Interoperability
Ask vendors: How do you exchange data with our EHR? (HL7, FHIR, SFTP, API?) What happens if our EHR goes down—do PAs get queued until the connection is restored? How do you handle patient data security (encryption in transit, HIPAA compliance, data retention)? Can you integrate data from outside imaging centers or specialist offices that aren’t in our EHR?
Reporting and Analytics
What reports do you offer out of the box (PA volume by payer, approval rates, time-to-approval, denial rates)? Can we create custom reports based on our metrics? Do you offer a dashboard for executive visibility? How frequently do you refresh data (real-time, daily, weekly)? Can we export data for analysis in our own tools?
Implementation Timeline and Support
Ask vendors: What’s your typical implementation timeline from contract signature to go-live? How much staff time are we expected to contribute? Do you provide training? What’s your first-year support model (dedicated implementation team, standard support, self-service portal)? What’s your average time-to-value—when do clients typically see measurable improvement?
Realistic vendors will tell you implementation typically takes 3–6 months depending on EHR integration complexity and organizational readiness. If a vendor promises go-live in 4 weeks, be skeptical.
Part 4: Evaluate ROI Claims and Require Proof
Vendors will give you impressive ROI projections. Some are realistic. Many are inflated. Here’s how to pressure-test their claims.
Ask for Reference Data: Request case studies from practices similar to yours in size, specialty, and payer mix. Contact those references directly and ask: How long was your actual implementation? What was the real time-to-value? Did the tool deliver on the promised ROI? What were the main surprises or challenges?
Require Proof of Payer Coverage: Don’t accept the vendor’s claim that they cover “500+ payers.” Ask for a list of your specific payers and documentation of their PA requirements. If your tool doesn’t cover the payers that represent 70% of your patient volume, the ROI is severely limited.
Model Staffing Reductions Conservatively: If a vendor promises that you can reduce PA staff by 50%, ask them to model this based on your specific volume, payer mix, and case complexity. A tool might be able to reduce a high-volume, commodity-procedure practice’s manual work by 60%, but a complex specialty practice with lots of exceptions might only see 25% reduction.
Quantify Time Savings at the Task Level: Rather than accepting “automation saves 30 hours per week,” ask vendors to break this down: How much time is saved per PA submission? Per status check? Per denial? Per appeal? This helps you estimate realistic savings based on your specific workflow.
Factor in Implementation Costs: Many RFPs focus on annual subscription cost but ignore implementation. A tool at $15,000/year with $30,000 in implementation costs is more expensive than a $20,000/year tool with $5,000 in implementation. Calculate total cost of ownership over 3 years and compare.
Part 5: Red Flags in Vendor Pitches
Over the years, I’ve seen certain vendor claims that consistently indicate problems ahead.
Red Flag #1: “Our tool is fully automated—it requires almost no staff.” Reality: Every PA automation tool requires some staff oversight. You need someone monitoring exceptions, reviewing escalations, chasing missing documentation, and appealing denials. If a vendor suggests you can eliminate your PA function entirely, they don’t understand your business. A realistic claim is “Our tool reduces your PA staff needs by 30–50% and shifts remaining staff to higher-value activities like denial management.”
Red Flag #2: “We support all major EHRs out of the box.” This usually means they offer a general HL7 integration that works but requires significant IT effort to customize. Ask specifically: Do you have a native connector for my EHR, or would we be using a generic integration? How long does integration setup typically take?
Red Flag #3: “Our payer network covers 90% of all U.S. healthcare.” Coverage rates vary wildly by region and patient population. A tool that covers 90% of national payers might cover only 65% of your payers if you have significant Medicaid or regional populations. Always ask for your specific payers.
Red Flag #4: “Implementation is turnkey—just activate and go.” PA automation requires workflow redesign, staff training, and ongoing tuning. If a vendor downplays the change management component, they likely haven’t done many successful implementations.
Red Flag #5: “Our pricing is per-transaction, so you only pay for what you use.” Per-transaction pricing makes cost unpredictable and can become expensive quickly if your PA volume grows. Flat annual fees are more predictable and align vendor incentives with your success (they want high volumes because they’re not charging per submission).
Part 6: Competitive Landscape Context
Several reputable vendors operate in the PA automation space. Here’s how to think about them as you’re evaluating options:
Cohere Health focuses heavily on payer connectivity and has strong relationships with large payers for real-time PA status integration. They excel for practices with complex, high-volume PA needs and strong IT infrastructure. Cost is higher; implementation is more involved.
AKASA offers a hybrid approach combining automation with human-in-the-loop review for complex cases. Good for practices that value a lighter-touch, more consultative implementation. Strong at denial management analytics.
Infinx (backed by Optum) emphasizes end-to-end revenue cycle automation, which can be good if you need to integrate PA with coding, billing, and AR workflows. Risk is over-engineering if you only want PA automation.
Rhyme operates at the lower end of the market and targets smaller practices with simpler PA needs. Good for commodity procedures; may not scale well to complex cases.
Honey Health offers tightly integrated PA automation within a broader platform that includes refill management, data fetching, and fax triage. Strength is operational simplicity for practices wanting to automate multiple administrative workflows simultaneously. Worth evaluating if you have pain points beyond just PAs (for example, if you also need prior authorization management).
The best vendor for your practice depends on your size, payer mix, EHR platform, and whether you want to solve just PA or a broader set of administrative workflows.
Part 7: Structure Your RFP and Timeline
Here’s a pragmatic timeline for vendor evaluation:
Week 1–2: Discovery & Baseline Definition Document your current PA workflow, metrics, pain points, and success criteria. Define must-have capabilities and nice-to-haves. Identify 4–6 vendor candidates to approach.
Week 3–4: RFP Distribution and Response Send your RFP to vendors. Ask for responses within 2 weeks. Request reference lists and detailed documentation of payer coverage, EHR integrations, and implementation timeline.
Week 5–6: Reference Calls Contact 2–3 references from each vendor. Ask detailed questions about implementation experience, actual ROI, challenges, and whether they’d choose the same vendor again.
Week 7–8: Live Demos Request live demos from your top 2–3 vendors. Don’t just watch their sales demo—ask them to walk through your specific workflow using your actual payer list and procedure codes. See how they handle exceptions.
Week 9: Scoring and Deliberation Score vendors against your criteria (payer coverage, EHR integration, ease of use, cost, support). Weight criteria by importance (payer coverage is probably more important than dashboard aesthetics). Identify your top choice and runner-up.
Week 10: Negotiation Negotiate contract terms, implementation timeline, support level, and pricing. Get specific commitments on payer coverage for your patient population and integration timeline for your EHR. Build in performance guarantees (e.g., “If payer coverage doesn’t reach 75% of your patient volume within 60 days, we’ll refund implementation costs”).
Realistic Expectations: The First Six Months
Once you’ve signed a contract, set realistic expectations for what’s ahead.
Months 1–2: Technical Setup Integration work, data mapping, workflow configuration, testing. This is the highest-touch phase and requires your IT and PA staff to collaborate closely with the vendor.
Month 3: Soft Launch You run the tool in parallel with your existing process for a full month. Staff submits PAs through both systems, comparing outputs and catching errors. This is where you identify gaps in payer coverage, EHR integration issues, and workflow tweaks.
Months 4–6: Full Transition You turn off the old system and run fully on the new tool. You optimize based on real-world usage, refine escalation rules, and tune denial appeal workflows.
Expect the first 60–90 days post-launch to feel slower than your baseline as staff adjust to new workflows and the system learns your patterns. By month 4–5, you should see measurable improvement in approval times and reduction in manual work.
The Forward Question
Once you’ve implemented PA automation, the question isn’t whether to automate other workflows—it’s which one to tackle next. Most practices that successfully automate PAs find that operational friction moves to adjacent functions: How are we handling refill requests? Who’s triaging inbound faxes? How are we managing eligibility verification before patient visits?
The best practices I’ve worked with treat PA automation as the first step in a broader operational modernization. They use the confidence gained from that implementation to tackle other high-friction, high-volume tasks with similar automation frameworks.
Meta Description: RFP guide for prior authorization automation software. Evaluate payer coverage, EHR integration, submission capability, and denial management before buying.
