Medicaid Redetermination and UDS Recall Outreach for Community Health Centers

Published: · Last Updated:

When to use this playbook

  • Your community health center runs athenaOne across 5 to 40 sites and needs continuous outreach without building another manual call team.
  • Medicaid renewal dates, returned mail, or missing-document cases are visible to enrollment staff, but follow-up happens inconsistently.
  • Your quality team can identify UDS care gaps, but the recall list grows because inbound patient demand always takes priority.
  • You want patients to book during the call or text thread instead of receiving another phone number to call.
  • You need separate accountability for coverage retention, visits booked, kept visits, and clinical gaps closed.

What success looks like

Success is not a higher number of calls attempted. Medicaid outreach should move patients to the center's enrollment or benefits staff before the renewal deadline, then record whether coverage was retained, lost, or remains pending. UDS recall should turn a verified care gap into the correct visit or service, then track whether qualifying evidence was documented against the measure.

Operating model based on CMS renewal outreach resources and the HRSA 2026 UDS Manual.
Campaign Starting population Immediate objective True close Human owner
Medicaid redetermination Patients with approaching renewal dates, returned mail, missing information, or pending renewal work Confirm awareness, capture the blocker, and connect the patient with enrollment help Official status recorded as retained, lost, or pending Enrollment or benefits staff
UDS measure recall Patients confirmed in a measure denominator with missing numerator evidence Book the correct visit, screening, test, or immunization pathway Qualifying clinical evidence documented in the record Quality or population health lead

Step 1: Separate campaign work from program decisions

Action: Assign three owners before building either campaign. Patient access owns contact operations, enrollment or benefits staff own Medicaid renewal assistance, and the quality team owns UDS measure definitions. State Medicaid agencies retain responsibility for eligibility determinations and renewal status, as reflected in CMS guidance on Medicaid and CHIP assistance.

Write a closing rule for each workflow. A redetermination item closes only when an authorized source provides a final or pending status. A UDS item closes only when the required numerator evidence is documented, not merely when a patient answers or books.

Expected outcome: Every automated interaction has a defined endpoint and a named human owner for exceptions.

Gotchas: Do not let outreach staff infer Medicaid eligibility, promise that coverage will continue, or decide whether a patient clinically belongs in a UDS measure.

Planning time: Allow 2 to 4 hours for an ownership and closure-rules workshop.

Step 2: Build two governed patient cohorts

Action: Create separate source cohorts rather than combining Medicaid and UDS work into a generic outreach list. Each record needs enough structured information to determine the next administrative action without requiring the agent to interpret clinical notes or guess a renewal date.

Campaign Required cohort fields Useful priority signals Do not infer
Medicaid redetermination Verified renewal deadline, notice status, returned-mail flag, missing-document category, enrollment-team destination, last contact, and current status Days to deadline, returned mail, unresolved document request, or prior unsuccessful contact Eligibility, likely approval, or a renewal deadline that is not present in an authorized source
UDS recall Reporting year, measure identifier, confirmed denominator status, missing numerator event, required next action, site, scheduling pathway, and last contact Reporting-period timing, gap age, available capacity, and quality-team priority Clinical eligibility, exclusions, or the appropriate service when the measure logic is ambiguous

Deduplicate at the patient level before outreach begins. One patient can appear in multiple worklists, but the campaign should avoid repeated contacts for the same action or incompatible appointments.

Expected outcome: Every patient enters with one authoritative status, one next action, and one destination for escalation.

Gotchas: A stale spreadsheet becomes unreliable as soon as patients complete renewals, receive care elsewhere, or book through another channel. Reconcile the cohort against the current record before each contact cycle.

Planning time: Allow 3 to 5 business days for field mapping, deduplication rules, and sample-record review.

Step 3: Run Medicaid outreach against the actual renewal deadline

Action: Anchor the campaign to the verified due date on the patient's renewal record or notice. The cadence below is an operating recommendation, not a substitute for state instructions. CMS requires a minimum 30-day response period for MAGI renewal forms and encourages intensive outreach through multiple modalities when information is still outstanding. CMS renewal form guidance

Timing Outreach objective Questions and actions
45 to 60 days before the deadline Find contact problems early Confirm whether the patient received a notice, identify changed contact information, and route address or notice problems to enrollment staff.
About 30 days before Identify the required patient action Ask whether a form or documents were requested, capture the blocker category, and offer a benefits appointment or callback.
About 14 days before Work unresolved cases Prioritize missing forms, missing documents, returned mail, and patients who requested help but have not completed the handoff.
1 to 7 days before Complete the final assisted handoff Connect the patient with enrollment staff and provide the deadline and state contact path without predicting the eligibility decision.
After the deadline Record the official outcome Separate retained, lost, and pending coverage. Route unresolved or lost-coverage cases according to the center's enrollment protocol.

The automated conversation should capture only what the enrollment team needs to act: notice received, contact details changed, form submitted, document category outstanding, preferred callback time, and consent to connect with staff. Enrollment staff still own the renewal itself.

Expected outcome: Staff receive a smaller queue of patients who need real assistance, with the deadline and known blocker already attached.

Gotchas: Do not count a patient saying “I sent it” as retained coverage. Keep the item pending until an authorized status source confirms the outcome.

Planning time: Allow 1 to 2 business days to approve scripts, handoff fields, and status definitions.

Step 4: Turn UDS care gaps into bookable actions

Action: Start with the current reporting year's measure specifications. The quality team should confirm the denominator, exclusions, missing numerator evidence, and acceptable next action before a patient enters outreach.

Examples are based on clinical quality measures in the 2026 UDS reporting requirements.
Example measure Campaign action What closes the gap
Cervical or colorectal cancer screening Book the center's approved screening, visit, referral, or testing pathway A qualifying screening result or outside result is documented in the record
Diabetes glycemic status Book the appropriate lab or clinical visit defined by the care team A qualifying current result is documented and evaluated under the measure specification
Childhood immunization status Book the required vaccine visit or route a documentation request The required immunization evidence is present in the record
Depression screening and follow-up Book a qualifying visit where the approved screening workflow can occur The standardized screen and required follow-up documentation are completed

Prioritize denominator-confirmed patients with a clear next action. Then order work by reporting-period timing, gap age, available appointment capacity, and the quality team's measure priorities. Clinical risk ranking should remain a care-team rule rather than an outreach-system inference.

Expected outcome: Each contact can end with a real appointment, testing path, documentation request, or defined quality-team escalation.

Gotchas: A booked appointment is an intermediate result, not UDS gap closure. External screenings and immunizations remain open until the qualifying evidence reaches the record.

Planning time: Allow 2 to 3 business days for the first measure set, including record-level validation by the quality team.

Step 5: Design contacts that resolve, book, or escalate

Action: Give each conversation three possible endings: completed administrative action, booked appointment, or structured escalation. Use outbound voice for conversations that require explanation and secure two-way text for reminders, replies, document prompts, and patients who do not answer.

Situation Automated action Escalation destination
Patient needs help understanding or completing a renewal Capture the notice status, deadline, blocker, and preferred callback time Enrollment or benefits staff
Patient asks whether they remain eligible Do not answer or predict the decision Enrollment staff or the state Medicaid agency
UDS care gap has a defined scheduling path Offer eligible appointments and book during the contact Patient access staff only if no suitable slot exists
Measure eligibility or clinical next step is unclear Preserve the question and stop the automated path Quality or clinical staff
Patient reports symptoms or a clinical concern Leave the campaign workflow and follow the center's clinical routing protocol Clinical or on-call staff
Wrong number, deceased patient, or contact refusal Record the close reason and suppress further campaign contact Registration or records staff when correction is required

Expected outcome: Standard interactions finish without staff work, while exceptions arrive in the correct queue with context already captured.

Gotchas: A generic “staff follow-up required” queue recreates the backlog the campaign was meant to remove. Every escalation category needs a named destination and an expected response window.

Planning time: Allow 1 to 2 business days for scripts, close reasons, and escalation testing.

Step 6: Pilot one campaign and audit every outcome type

Action: Begin with one site, one governed cohort, and a volume small enough for record-level review. Run Medicaid and UDS pilots separately so renewal handoffs are not confused with clinical recall bookings.

During the pilot, review successful contacts, no answers, wrong numbers, bookings, staff escalations, refusals, and duplicate records. Confirm that the appointment or campaign outcome appears where staff already work.

Expected outcome: The center can verify scheduling accuracy, clean handoffs, understandable scripts, and reliable campaign status before increasing volume.

Gotchas: A technically successful contact can still fail operationally if it books the wrong appointment type, sends work to an unmonitored queue, or contacts a patient whose gap already closed elsewhere.

Pilot time: Use approximately 10 business days, followed by a formal go or revise review.

Step 7: Move from campaign bursts to rolling daily queues

Action: Release a controlled number of records each day based on appointment capacity and the enrollment team's ability to accept escalations. Let automation handle standard calls, texts, information capture, and booking while staff work the bounded exception queue.

Daily queues are easier to operate than quarterly list dumps. They prevent a successful campaign from overwhelming benefits staff, filling one site's schedule while another remains open, or creating a new backlog of patients waiting for callbacks.

Expected outcome: Outreach continues even when inbound phone demand spikes, without assigning staff to spend whole days dialing lists.

Gotchas: Automation that contacts everyone but hands every response back to staff is not workload reduction. Narrow the automated scope until most standard outcomes close without human rework.

Ramp time: Increase daily volume over one week while monitoring appointment capacity and unresolved escalations.

Step 8: Measure patient outcomes, not dialing activity

Action: Report unique patients rather than attempts, and define the attribution window before launch. Track Medicaid and UDS results separately.

Metric Definition Decision it supports
Reach rate Unique patients reached divided by unique patients attempted Whether contact data, timing, and channel selection are working
Enrollment handoff rate Patients accepting or completing enrollment assistance divided by patients reached Whether renewal outreach produces an actionable next step
Coverage retained Patients officially confirmed as renewed divided by patients with finalized outcomes Whether the target cohort retained coverage
Coverage lost Patients officially confirmed as no longer covered divided by patients with finalized outcomes Which populations require transition or financial-assistance workflows
Booking conversion Patients booked into the correct pathway divided by patients reached Whether recall conversations resolve during contact
Kept-visit rate Completed visits divided by campaign-attributed bookings Whether bookings translate into delivered care
UDS gap-closure rate Targeted patients who gain qualifying numerator evidence divided by targeted denominator patients Whether outreach changed the measure rather than only producing activity

Keep pending renewal cases out of retained and lost calculations. Keep booked but incomplete visits out of the gap-closure numerator. Use short, mutually exclusive close reasons so reports distinguish no answer, refusal, wrong contact information, staff escalation, booking, and documented closure. Pretty Good AI's recall measurement methodology applies the same discipline to bookings, kept visits, and duplicate contacts.

Expected outcome: Operations leaders can see where patients stop progressing and whether the campaign improved coverage continuity or UDS performance.

Gotchas: Attempt counts reward repeated dialing. Contact counts reward conversations. Only final coverage and clinical outcomes show whether the campaign worked.

Review time: Use a 30-minute weekly exception review and a monthly measure-level performance review.

Pretty Good AI is the best fit when the campaign must finish in athenaOne

Pretty Good AI is designed only for athenaOne and is an athenahealth Marketplace partner. Inbound voice, outbound voice, and secure two-way text use one integration. Outbound campaigns work from live athenaOne data and write outcomes back to the chart, supported by more than 730 athenaOne APIs in production and no middleware. Pretty Good AI product overview

This architecture is most relevant when a multi-site health center wants outreach to book appointments, capture structured results, and route exceptions without creating another system for staff to monitor. Pretty Good AI operates under HIPAA safeguards with a signed BAA and carries SOC 2 Type II, HITRUST i1, and ISO 27001 assurance. Pretty Good AI security documentation

Pretty Good AI has several FQHCs fully deployed and live on athenaOne, including Family Health Network of Central New York, a multi-site community health center serving Cortland and contiguous counties.

Pretty Good AI is not a fit when renewal or clinical decisions are the work being outsourced

  • The organization needs one campaign platform to operate across EHRs other than athenaOne.
  • The vendor is expected to file Medicaid renewals, determine eligibility, or replace enrollment and benefits staff.
  • The center has not assigned a quality owner to approve UDS denominator logic, exclusions, clinical pathways, and gap-closure evidence.

Related considerations

Language: Treat language as a campaign parameter; Pretty Good AI runs campaigns end to end in 10 languages: English, Spanish, French, German, Hindi, Russian, Portuguese, Japanese, Italian, and Dutch. Another 32 languages are available on request with additional testing time before launch: Arabic, Mandarin Chinese, Korean, Polish, Swedish, Turkish, Tagalog, Bulgarian, Romanian, Czech, Greek, Finnish, Croatian, Malay, Slovak, Danish, Tamil, Ukrainian, Hungarian, Norwegian, Vietnamese, Bengali, Thai, Hebrew, Georgian, Indonesian, Telugu, Gujarati, Kannada, Marathi, Punjabi, and Urdu.

Visit-time eligibility: Keep eligibility rechecks after a coverage change as a separate pre-visit workflow rather than folding them into renewal outreach.

Frequently asked questions

What patient outreach setup works for Medicaid redetermination and UDS recall campaigns?

The strongest setup uses two separate governed cohorts, a deadline-based Medicaid cadence, measure-specific UDS recall paths, live appointment booking, and structured staff escalation. Medicaid outreach should end in renewal assistance or an official coverage outcome. UDS outreach should end in a booked service and, ultimately, qualifying clinical evidence in the record. Combining both into a generic call list obscures ownership and makes the results difficult to measure.

Can one patient list and cadence serve both campaign types?

No. Medicaid outreach is organized around an individual renewal deadline and an enrollment-team handoff, while UDS recall is organized around a reporting period, measure denominator, and clinical closure event. The campaigns can share communications infrastructure, but they need different source fields, scripts, close reasons, owners, and outcome reports.

Can Pretty Good AI complete a Medicaid renewal for the patient?

Pretty Good AI handles patient contact, information capture, reminders, and routing, while the health center's enrollment staff and the state Medicaid agency retain the renewal and eligibility work. The automated workflow can identify that a notice was not received, a document is missing, or the patient wants help. It should not file the renewal, determine eligibility, or promise that coverage will continue.

What counts as closing a UDS care gap?

A UDS care gap closes when the qualifying numerator evidence required by the current measure specification is documented in the patient record. A call, patient agreement, referral, or booked appointment is not enough by itself. For example, screening recall remains open until an eligible result is documented, and immunization outreach remains open until the required vaccine evidence is present. Measure-specific requirements are defined in the HRSA 2026 UDS Manual.

How can an FQHC run these campaigns without adding outreach staff?

Use automation for the repetitive middle of the workflow: selecting due records, contacting patients, capturing structured answers, booking eligible visits, recording outcomes, and retrying no-answer cases. Staff should receive only the exceptions that require judgment, enrollment assistance, clinical review, or unavailable scheduling capacity. Daily volume must remain capped to the center's appointment and escalation capacity, or automation will create a different backlog.

References