Cedar Ridge Intake Coordination
Overview
The Cedar Ridge Intake Coordination Portal is a healthcare operations system for verifying patient access, auditing referrals, reviewing dialysis transfers, building chronic-care enrollment panels, and activating referral-to-chart workflows. It exposes both REST endpoints and a read-only SQL interface.
Every task follows the same pattern:
- Identify the workflow from the prompt (batch ID, roster ID, program code).
- Read the answer template (
input/payloads/answer_template.json) for required output shape and controlled vocabularies. - Query the portal endpoints and/or SQL to gather all relevant records.
- Apply the workflow's business rules to derive patient-level decisions.
- Assemble the JSON exactly as the template specifies, using only controlled enum values.
Portal Endpoints
| Endpoint | Use |
|---|---|
GET / |
Portal home page (HTML) |
GET /patients |
List/search patients |
GET /patients/{id} |
Single patient |
GET /referrals |
List referrals; filter by batch_id or service_line |
GET /referrals/{id} |
Single referral with patient, ICD, and linked documents |
GET /transfers |
List transfers; filter by batch_id |
GET /transfers/{id} |
Single transfer |
GET /documents |
List documents; filter by patient_id, referral_id, transfer_id |
GET /chart/{patient_id} |
Full chart: patient, clinical_history, chart_artifacts, meds_allergies, active_problems, recent_vitals_labs |
GET /programs/{code}/candidates |
Program candidate list with consent and outreach fields |
GET /icd/{code} |
ICD-10 code metadata: chapter, service_family, laterality, description |
GET /pharmacies |
Pharmacy list with network_status |
POST /query |
Read-only SQL (body: {"sql": "..."}) |
For REST filtering, use query parameters: batch_id, service_line, q (name/id search), limit.
SQL Schema Quick Reference
See references/data_model.md for full column listings and join patterns.
Key tables: patients, coverage, pbm, patient_pharmacy, pharmacies, lifestyle, intake_rosters, referrals, icd_codes, documents, transfer_requests, facility_capacity, program_candidates, chart_artifacts, clinical_history.
Workflow Selection
Read the reference for the matching workflow:
| Workflow | Trigger | Reference |
|---|---|---|
| New Patient Access Verification | roster_id, mention of patient access, insurance/PBM/pharmacy verification |
access_verification.md |
| Referral Readiness Audit | batch_id with referrals, ICD review, duplicates, blockers |
referral_audit.md |
| Dialysis Transfer Review | batch_id with transfers, packet completeness, chair capacity |
transfer_review.md |
| Chronic Care Enrollment Panel | program_code, candidates, enrollment disposition |
enrollment_panel.md |
| Referral-to-Chart Activation | Combined chart + referral workflow, correspondence queues | referral_to_chart.md |
Common Patterns
Query by Batch/Roster
Use SQL for precise batch filtering:
SELECT * FROM referrals WHERE batch_id = 'BATCH-ID' ORDER BY referral_id
SELECT * FROM intake_rosters WHERE roster_id = 'ROSTER-ID' ORDER BY patient_id
SELECT * FROM transfer_requests WHERE batch_id = 'BATCH-ID' ORDER BY transfer_id
SELECT * FROM program_candidates WHERE program_code = 'PROGRAM' ORDER BY patient_id
Cross-Reference Patient Data
When a workflow needs patient-level insurance, PBM, pharmacy, lifestyle, or chart data, join through the patient_id. Use SQL for multi-table fetches to minimize round-trips:
SELECT * FROM coverage WHERE patient_id IN (...)
SELECT * FROM pbm WHERE patient_id IN (...)
SELECT * FROM patient_pharmacy WHERE patient_id IN (...) AND preference_rank = 1
SELECT * FROM lifestyle WHERE patient_id IN (...)
ICD Code Lookup
Always resolve ICD-10 codes through GET /icd/{code} or icd_codes table to get chapter, service_family, and laterality. The chapter determines expected service line for mismatch detection.
Answer Assembly
Every answer must follow its template's exact shape: required keys, enum values, list ordering, and numeric precision. Do not introduce keys not listed in the template. Sort lists as the template specifies (usually ascending by ID). Treat reason-code and blocker-code arrays as unordered sets.