The Medicaid Eligibility Automation Playbook for Therapy Practices

Why Medicaid eligibility automation is the highest-ROI therapy AI workflow

Medicaid is 70%+ of revenue in many ABA practices, 40%+ in pediatric SLP, and a meaningful slice of mental health Medicaid managed care plans. Manual eligibility verification eats 2-4 hours per patient — front-desk staff calling state portals, navigating IVR trees, transcribing coverage and benefit information into the EHR, and re-running every reauthorization cycle (typically 90-180 days). For a 10-clinician practice with 800 active Medicaid clients, that’s 1,600-3,200 hours of admin time per year — easily 1-2 FTEs. AI automation drops the same work to under 2 minutes per patient and frees the admin team to focus on prior auth narratives and denial appeals.

Step 1: Establish the HIPAA-compliant foundation

Before any PHI touches the automation pipeline, every link must be inside a signed BAA. That includes the AI platform, the clearinghouse, the EHR, and any orchestration tools.

• Sign BAA with AI platform (e.g. Frontdesk)
• Sign BAA with clearinghouse (Availity, Office Ally, Trizetto, Change Healthcare, or Waystar)
• Confirm BAA with EHR (SimplePractice, TherapyNotes, Jane App, CentralReach, TheraNest, Fusion Web Clinic)
• PHI encrypted at rest (AES-256) and in transit (TLS 1.3) across every hop
• Role-based access controls (clinical vs admin vs vendor)
• Granular audit logs of every PHI access event

Step 2: Wire the 270/271 transaction pipeline

The 270/271 transaction is the standard X12 EDI exchange for healthcare eligibility. The AI submits a 270 with patient demographics + Medicaid ID; the payer returns a 271 with coverage, benefits, copays, deductibles, and authorization references.

• Patient demographics pulled from EHR (SimplePractice, TherapyNotes, etc.)
• Medicaid member ID parsed from intake form or insurance card OCR
• 270 request submitted via clearinghouse (Availity, Office Ally, Trizetto, Change Healthcare, Waystar)
• 271 response parsed and normalized
• Coverage data written back into EHR insurance record
• Exceptions flagged for human review

Step 3: Handle the exceptions intelligently

Routine eligibility (active coverage, valid auth, no lapses) auto-completes. Exceptions — coverage lapses, missing prior auth, secondary coverage conflicts, name mismatches — must be flagged for human review.

• Coverage lapsed: trigger patient outreach + re-enrollment guidance
• Prior auth expiring within 30 days: trigger renewal workflow
• Name mismatch / data quality issue: flag for admin review
• Secondary coverage detected: prompt admin to update COB
• Plan-level service exclusions: flag against service codes in EHR

Step 4: Run reauthorization tracking on autopilot

Most Medicaid plans require reauthorization every 90-180 days. AI tracks every active client’s auth window and triggers reauthorization workflows automatically.

• Auth window tracked per client per service code
• 30-day prior alert for reauthorization workflow
• Patient outreach for any required updated documentation
• EHR auth record updated when new auth received
• Denied auth: routed to clinical team for appeal

Step 5: Wire into ABA prior authorization for the deepest unlock

ABA practices have the most complex Medicaid auth — multiple service codes (97151, 97153, 97155, 97156, 97158), hours-per-week caps, and frequent reauthorization. Automating this layer is the highest-leverage move.

• Track per-client service-code authorizations and remaining units
• Alert at 80% utilization for proactive reauth submission
• Surface gaps where actual billed units don’t match auth
• Sync auth status into CentralReach, Rethink, Catalyst, AccuPoint, Lumary, Total ABA, Theralytics, or TheraBe

Step 6: Measure the ROI

Practices typically see clear ROI within 60-90 days. The metrics that move:

• Manual eligibility time: 2-4 hrs/patient → <2 min
• Denied claims from coverage issues: down 30-50%
• Reauthorization on-time rate: up to 95%+
• Admin FTE reclaimed: 0.5-2 per 10 clinicians