How do I set up a care coordination platform for referral tracking and post-visit follow-up in a health system?

Setting up a care coordination platform for referral tracking and post-visit follow-up in a health system is mostly an operating model + workflow design project, with technology supporting it. A good setup usually has 6 parts:

1. Define the use cases
2. Map the workflow and ownership
3. Choose the platform architecture
4. Integrate with core clinical systems
5. Build tracking, alerts, and reporting
6. Pilot, train, and scale

Below is a practical blueprint.

---

1) Define the scope and goals

Start by being specific about what the platform must do.

Common initial use cases

• Outbound referrals: PCP or specialist sends a referral to another service
• Referral status tracking: see whether referral was accepted, scheduled, completed, or closed
• Post-visit follow-up: ensure patient received instructions, meds, labs, imaging, care gaps, etc.
• Closed-loop referral management: referral does not end until outcome is documented
• High-risk transitions: post-discharge outreach, no-show follow-up, medication reconciliation
• Care gaps: missed screening, overdue visit, pending test results, unresolved consults

Define success metrics

Examples:

• Referral completion rate
• Time from referral order to scheduled appointment
• Time from visit discharge to follow-up contact
• No-show rate after outreach
• Percent of referrals with documented outcome
• Reduction in duplicated outreach and missed handoffs

---

2) Map the end-to-end workflow

A platform succeeds only if you know who owns each step.

Typical referral workflow

1. Referral created
   • Provider places referral order in EHR or sends request
2. Referral routed
   • Rules route by specialty, location, insurance, urgency, language, etc.
3. Referral reviewed
   • Intake staff checks completeness, clinical appropriateness, required records
4. Patient contacted
   • Scheduling team or care coordinator reaches out
5. Appointment scheduled
   • Slot is booked and confirmed
6. Visit completed
   • Specialist documents consult and recommendations
7. Outcome returned
   • Referring provider receives summary and next steps
8. Closed-loop verification
   • Referral marked complete only after outcome and follow-up are documented

Typical post-visit follow-up workflow

1. Visit/discharge event occurs
2. Task created automatically
   • e.g., follow-up call within 48 hours, labs in 1 week, specialist appt in 2 weeks
3. Coordinator/outreach team contacts patient
4. Needs assessed
   • symptoms, meds, understanding, barriers, transportation, equipment, social needs
5. Escalations triggered if needed
   • clinical symptom alerts, missed medication, urgent appointment scheduling
6. Task closed with documentation
7. Care team notified

Important ownership questions

• Who owns referral intake?
• Who owns patient scheduling?
• Who closes the referral?
• Who handles missed follow-up?
• Who escalates clinical concerns?
• Who documents in the EHR vs the coordination platform?

---

3) Decide on platform model

You usually have three options:

A. Build within the EHR

Best when: your EHR has strong referral management, tasking, and care management modules.

Pros

• Less duplicate documentation
• Better clinical context
• Easier provider adoption

Cons

• May be limited for multi-department workflows
• Harder to customize dashboards and automation

B. Use a dedicated care coordination platform

Best when: you need cross-system tracking, complex outreach, and specialized workflows.

Pros

• Better referral queue management
• More flexible workflow automation
• Stronger care management analytics

Cons

• Integration complexity
• Risk of being “another system to check”

C. Hybrid model

Best for many health systems.

• EHR remains source of truth for orders, notes, and clinical documentation
• Coordination platform manages tasking, routing, outreach, and status visibility

This is often the most practical approach.

---

4) Design the core data model

Your platform should track a few core entities.

Referral record

• Referral ID
• Patient ID
• Referring provider/service
• Receiving specialty/service
• Reason for referral
• Priority/urgency
• Diagnosis/problem list link
• Required records
• Status
• Timestamps
• Assigned coordinator
• Outcome

Follow-up task

• Task type
• Due date
• Assigned user/team
• Related encounter/referral
• Status
• Attempts made
• Patient response
• Escalation flag

Outcome record

• Appointment scheduled/completed
• Consult note received
• Recommendation acted upon
• Patient unreachable
• Referral declined
• Patient declined
• Completed externally
• Closed with reason

Barriers and needs

• Transportation
• Language
• Financial
• Housing
• Medication access
• Caregiver support
• Technology access

---

5) Build integrations with source systems

This is usually the hardest part.

Key systems to integrate

• EHR: orders, encounters, notes, demographics, problem list
• Scheduling: appointment availability and booking
• ADT feeds: admit/discharge/transfer events
• HIE/interoperability layer: external visit status and documents
• Labs/imaging systems: pending results and completion
• Patient portal/SMS/phone tools: outreach
• CRM/case management tools: if used for social needs or community referrals

Integration patterns

• HL7/FHIR for event and clinical data exchange
• API-based updates for status changes
• Batch imports for less urgent reporting
• Bidirectional sync where appropriate, but minimize data duplication

Integration principle

Keep one clear system of record for each data type:

• EHR = clinical documentation and orders
• Coordination platform = workflow state, outreach, task management
• Scheduling system = appointment truth

---

6) Design referral routing and rules

Automate as much as possible.

Routing rules examples

• Specialty based on diagnosis
• Location based on patient ZIP code
• Urgency based on triage criteria
• Insurance/network restrictions
• Age/sex-specific routing
• Language or accessibility needs
• Facility capacity or next available appointment

Exception handling

Create queues for:

• Incomplete referral
• Missing records
• Insurance authorization pending
• No appointment availability
• Patient unreachable
• Referral clinically inappropriate
• Duplicate referral

This prevents referrals from getting stuck silently.

---

7) Build closed-loop tracking

Closed-loop tracking is essential for referral completion.

Minimum closed-loop statuses

• New
• Under review
• Accepted
• Needs info
• Ready to schedule
• Scheduled
• Completed
• Returned to referring provider
• Declined
• Deferred
• Cancelled
• Unable to reach patient
• Closed

What “closed” should mean

A referral should not be closed just because it was sent. Close it only when one of these occurs:

• Visit completed and consult note returned
• Patient declined after documented outreach
• Referral abandoned after defined outreach attempts
• Referral rerouted or resolved elsewhere
• Clinical reason for closure documented

---

8) Set up post-visit follow-up workflows

Post-visit follow-up works best when triggered by event.

Trigger examples

• ED discharge
• Inpatient discharge
• Specialist visit completion
• Procedure completion
• High-risk medication start
• Missed appointment
• Abnormal lab result
• Care gap identification

Follow-up templates

Create standardized scripts and task templates, such as:

• 48-hour discharge call
• 7-day PCP follow-up
• 14-day specialty follow-up
• Medication pickup confirmation
• Symptom check and escalation
• Transportation assistance
• Social needs assessment

Escalation rules

If patient reports:

• Chest pain
• Shortness of breath
• Worsening symptoms
• Suicidal ideation
• Severe medication reaction

then the system should trigger immediate escalation to clinical staff or emergency instructions.

---

9) Build dashboards and reporting

Your leaders and teams need operational visibility.

Useful dashboards

• Referral volume by specialty
• Referral aging by status
• Completion rate by service line
• No-show rate
• Average time to first contact
• Average time to appointment
• Follow-up completion rate
• High-risk patient outreach completion
• Team workload by coordinator
• Reasons for referral closure

Operational views for users

• My queue
• Overdue tasks
• Needs info
• Today’s outreach
• Escalations
• Unresolved referrals older than X days

Clinical/administrative views

• Provider-level performance
• Specialty capacity issues
• Care gap closure
• Unmatched consults
• External referral leakage

---

10) Establish governance

You need a governance structure so the platform stays usable.

Governance team should include

• Clinical leadership
• Care coordination leadership
• Referral center/scheduling
• IT/integration
• Compliance/privacy
• Operations
• Quality improvement
• Patient access

Governance responsibilities

• Define workflow standards
• Approve status codes and data definitions
• Prioritize changes
• Review backlog and bottlenecks
• Resolve ownership disputes
• Monitor compliance and privacy

---

11) Address privacy, compliance, and security

Especially important for health systems.

Key considerations

• HIPAA minimum necessary access
• Role-based permissions
• Audit logs
• Secure messaging
• Patient consent for outreach methods
• Device and identity management
• Data retention and documentation standards

If the platform includes social needs or community referrals, make sure sharing rules are clearly defined.

---

12) Pilot before scaling

Do not launch enterprise-wide first.

Pilot approach

Choose:

• One specialty line
• One hospital discharge unit
• One ambulatory clinic
• One high-risk population

Pilot goals

• Test workflow clarity
• Validate integrations
• Check task ownership
• Measure time savings
• Identify alert fatigue
• Ensure staff adoption

What to refine after pilot

• Queue logic
• Status definitions
• Automation thresholds
• Dashboard design
• Training needs
• Exception handling

---

13) Train users and design for adoption

Even a great platform fails without operational adoption.

Training should cover

• What counts as a referral or follow-up task
• How to update status
• When to escalate
• How to document in the correct system
• How to use dashboards and queues
• Expected response times

Adoption tips

• Keep clicks low
• Pre-fill fields where possible
• Use standardized templates
• Avoid forcing duplicate documentation
• Make ownership obvious
• Provide daily queue reports

---

14) Recommended implementation sequence

A practical rollout plan:

Phase 1: Foundation

• Define workflows and status taxonomy
• Identify systems of record
• Build referral and task data model
• Set permissions and governance

Phase 2: Integration

• Connect EHR, scheduling, ADT, messaging
• Ingest referral events and discharge events
• Create task generation rules

Phase 3: Operational launch

• Launch pilot queue
• Train coordinators and clinicians
• Use dashboards for monitoring

Phase 4: Optimization

• Add automation and rules
• Expand to more specialties/sites
• Add social needs and external referral tracking
• Improve analytics and quality metrics

---

15) Common pitfalls to avoid

• Treating the platform as only software, not workflow redesign
• No single owner for referral closure
• Duplicate documentation in multiple systems
• Too many status codes
• Poor exception handling
• Launching without scheduling integration
• Not tracking external referrals
• Ignoring patient outreach preferences
• No escalation path for abnormal findings
• Not measuring backlog and aging

---

16) A simple target operating model

A workable model looks like this:

• Referral orders originate in EHR
• Coordination platform receives event and creates work item
• Referral center reviews and routes
• Specialty office schedules/completes visit
• Platform tracks status and due dates
• Care coordinators perform follow-up outreach
• EHR stores clinical notes and final outcome
• Dashboards monitor completion and delays

---

17) If you’re selecting software, ask these questions

• Does it support closed-loop referrals?
• Can it route referrals automatically by rules?
• Can it manage post-visit outreach tasks?
• Does it integrate with our EHR and scheduling system?
• Can it handle external referral status?
• Can it document outreach attempts and outcomes?
• Does it support role-based queues and escalation?
• Can it generate operational dashboards?
• Can it track referral aging and leakage?
• Can users work without duplicate charting?

---

If you want, I can also provide one of these:

1. a sample workflow diagram
2. a requirements checklist for vendor selection
3. a data model / table schema
4. a 90-day implementation plan for a health system