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

To set up a care coordination platform for referral tracking and post-visit follow-up in a health system, treat it as both a workflow redesign and a technology implementation. The goal is to make referrals visible end-to-end, automate follow-up tasks, and close the loop between primary care, specialists, hospital teams, and patients.

1) Define the use cases first

Start with the specific coordination problems you want to solve:

• Incoming referrals: PCP to specialist, urgent referral, e-consult
• Outgoing referrals: specialist back to PCP, community services, imaging, therapy
• Post-discharge follow-up: scheduling, medication reconciliation, symptom check-ins
• No-show and missed referral management
• Closed-loop communication: “Was the referral seen?” “Was the note received?” “Was action taken?”
• Care gaps: missed labs, overdue visits, unmet social needs

For each use case, define:

• Trigger
• Responsible role
• Expected turnaround time
• Escalation path
• Completion criteria

2) Map current workflows and ownership

Create a current-state process map across settings:

• Ambulatory clinics
• Hospital discharge team
• Referral center/call center
• Specialist offices
• Nursing care managers
• Social work/community health workers

Identify:

• Where referrals are entered
• How status is updated today
• Where delays occur
• Which steps are manual
• Who follows up if a patient doesn’t schedule or attend

Then assign clear ownership for each step. A referral system fails when “everyone” is responsible.

3) Define the core data model

Your platform should track each referral or follow-up as a work item with standardized fields such as:

• Patient identifiers
• Referring provider / referring location
• Receiving provider / department
• Reason for referral
• Priority and target date
• Referral status
• Appointment date/time
• Outcome/decision
• Documentation received
• Follow-up tasks
• Escalation flags
• Closure date and closure reason

For post-visit follow-up:

• Encounter type
• Discharge date
• Med changes
• Labs/imaging ordered
• Follow-up due date
• Outreach attempts
• Patient response
• Completed actions

4) Choose the operating model

Decide whether the platform will be:

A. Embedded in the EHR

Best if you want:

• Strong clinical integration
• Single sign-on
• Easier chart context
• Better documentation continuity

B. Standalone care coordination layer

Best if you need:

• Cross-EHR coordination
• Custom workflow rules
• Broader analytics and task routing

C. Hybrid

Often the best option:

• EHR remains source of truth for clinical documentation
• Care coordination platform handles tasks, routing, reminders, and tracking
• Data syncs both ways

5) Build the referral tracking workflow

A solid closed-loop referral process usually includes:

1. Referral created

   • Standardized order/referral form
   • Required fields and specialty-specific criteria

2. Referral triaged

   • Determine urgency, appropriateness, and service line
   • Route to correct clinic/provider

3. Patient outreach

   • Scheduling team contacts patient
   • Automated text/call/email reminders if allowed

4. Appointment scheduled

   • Record date/time and appointment type
   • Flag delays beyond target SLA

5. Visit completed or not completed

   • Mark attended, cancelled, no-show, rescheduled, declined

6. Consult outcome documented

   • Specialist note, recommendations, procedures, test results

7. Referring provider notified

   • Closed-loop alert or message back to PCP/team

8. Referral closed

   • Completion reason captured
   • If unresolved, route to exception queue

6) Build post-visit follow-up workflows

For discharge or clinic follow-up, create standardized pathways based on risk:

• High-risk patients: outreach within 24–48 hours
• Moderate-risk: follow-up within 3–7 days
• Routine: follow-up within 1–2 weeks

Typical tasks:

• Confirm medication pickup/adherence
• Review symptoms and red flags
• Confirm appointments were scheduled
• Ensure labs/imaging are completed
• Arrange transportation or community support
• Escalate if worsening symptoms

Use task templates by diagnosis or visit type so staff are not starting from scratch.

7) Automate where possible

Automation should reduce manual chasing, not replace clinical judgment.

Useful automations:

• Status-based task routing
• Reminder messages to patients
• Escalation if no appointment after X days
• Alerts for incomplete discharge follow-up
• Duplicate referral detection
• SLA breach notifications
• Work queue prioritization by risk

Examples:

• “If referral not scheduled within 5 business days, send task to referral coordinator.”
• “If discharged heart failure patient not reached in 48 hours, escalate to RN care manager.”
• “If specialist note not signed in 7 days, alert clinic supervisor.”

8) Integrate with the EHR and other systems

Integration is critical. Prioritize:

• Patient demographics and scheduling
• Orders/referrals
• Encounter and discharge data
• Clinical notes and consult notes
• Results/labs/imaging
• Messaging/inbasket
• Identity matching

Common standards:

• HL7/FHIR for interoperability
• APIs for scheduling and task status
• SSO and role-based access control

If you can’t integrate everything on day one, start with the most important data flows:

1. Referral order
2. Appointment status
3. Visit completion
4. Outcome documentation
5. Follow-up task closure

9) Design dashboards and queues

Make work visible to each role:

• Referring provider dashboard
• Referral coordinator queue
• Specialist triage queue
• Care manager follow-up queue
• Supervisor exception queue
• Operational dashboard for leaders

Key metrics to show:

• Referral volume by service line
• Time to triage
• Time to appointment
• Open referral aging
• No-show rate
• Closure rate
• Follow-up completion rate
• Outreach attempts
• SLA compliance
• Unresolved exceptions

10) Put governance in place

Establish a governance structure with:

• Clinical leadership
• Operational leaders
• IT/integration team
• Compliance/privacy
• Patient access/scheduling
• Care management
• Quality improvement

Define:

• Referral standards
• Escalation rules
• Access permissions
• Documentation requirements
• Exception handling
• Change management process

11) Address privacy, consent, and regulatory needs

Make sure the platform complies with:

• HIPAA and local privacy rules
• Minimum necessary access
• Audit trails
• Messaging consent rules
• Data retention requirements
• Business associate agreements with vendors

If you use texting or email, confirm:

• Patient consent
• Approved communication content
• Opt-out handling
• No sensitive info in unsecured channels if prohibited

12) Pilot before scaling

Don’t launch systemwide immediately.

Pilot with:

• One specialty clinic
• One discharge unit
• One patient population
• A limited number of coordinators

During the pilot, test:

• Referral routing accuracy
• Data completeness
• Alert fatigue
• Queue design
• Staff workload
• Patient response rates

Refine workflows based on real use before expanding.

13) Train users by role

Training should be role-specific:

• Referring clinicians: how to submit complete referrals
• Coordinators: triage, outreach, closure
• Specialists: documenting outcomes and recommendations
• Nurses/care managers: follow-up protocols
• Supervisors: dashboards, exceptions, metrics

Include:

• What to do
• What not to do
• Escalation steps
• Sample cases
• Quick reference guides

14) Measure success

Track outcomes tied to operational and clinical impact:

• Percent of referrals closed loop
• Average days to scheduled visit
• Average days to specialist review
• Post-discharge contact within target window
• Readmission reduction
• Reduced referral leakage
• Patient satisfaction
• Staff time saved
• Fewer duplicate referrals

Recommended implementation sequence

1. Pick one high-value referral workflow
2. Map current state and define owners
3. Standardize data fields and statuses
4. Integrate with EHR scheduling/referrals
5. Build queues, alerts, and dashboards
6. Pilot with one team
7. Train and refine
8. Expand to more service lines
9. Add post-visit follow-up automation
10. Use analytics for continuous improvement

Common pitfalls to avoid

• Too many custom statuses
• No clear owner for unresolved items
• Poor EHR integration
• Building alerts without action queues
• Not defining closure criteria
• Over-automating before workflows are stable
• Ignoring patient outreach preferences
• Launching without a governance structure

If you want, I can also give you:

1. a reference architecture for this platform,
2. a workflow diagram,
3. a sample data model, or
4. a 90-day implementation plan.