Data Requirements
CMS TEAM Analyzer runs downstream of Tuva-conformed data. Optional enrichment improves precision, but the core package can run with standard claims, encounter, eligibility, patient, practitioner, and terminology models.
Required Tuva Models
| Logical source | Default model | Purpose |
|---|---|---|
| Medical claims | core__medical_claim | Anchor detection, claim assignment, spend, exclusions, post-acute utilization, post-episode monitoring |
| Encounters | core__encounter | Inpatient episode support, facility attribution, discharge context |
| Eligibility | core__eligibility | Enrollment coverage, Medicare attributes, geography, dual status |
| Patient | core__patient | Age, demographics, mortality, patient geography |
| Practitioner | core__practitioner | Surgeon/provider names and specialties |
| MS-DRG terminology | terminology__ms_drg | MS-DRG to MDC mapping for readmission logic and clinical grouping |
Each source can be overridden with a dbt var, such as team_medical_claim_model or team_eligibility_model.
Required Claim Content
The package expects the medical claim model to include the standard data needed for episode analytics:
- Claim identity and line identity
- Person and member identifiers
- Claim type, encounter type, and encounter group
- Claim, line, admission, discharge, paid, and file dates
- DRG, revenue center, HCPCS, and modifier fields
- Admit source, admit type, discharge disposition, place of service, and bill type
- Rendering, billing, and facility identifiers
- Paid, charge, cost share, and total cost amounts
Required Package Seeds
CMS TEAM Analyzer ships reference seeds for:
- TEAM episode categories
- Inpatient MS-DRG triggers
- Outpatient HCPCS triggers
- Outpatient HCPCS to MS-DRG normalization
- Excluded readmission MS-DRGs and MDCs
- Excluded Part B drugs and hemophilia clotting factor HCPCS
- TEAM discount factors
- Public CMS PY1 regional target prices
- Ineligible geography/payment exclusion states
- Optional OPPS status indicator mapping
Optional Inputs
| Input | Why it matters |
|---|---|
| Participant-specific CMS target prices | Highest-fidelity target price comparison for the client's actual participant/site context |
| OPPS status indicator mapping | Enables OPPS pass-through device exclusion without a client-specific claim field |
| Claim value code and value amount fields | Enables NTAP value-code exclusion support |
| Beneficiary census tract enrichment | Supports CDI/SVI/ADI-style risk and equity analysis inside the client's PHI environment |
| LIS, MDS/LTI, and hospital reference data | Improves future risk-adjustment fidelity |
| Quality score or final reconciliation inputs | Enables future scenario analysis; not required for current package analytics |
Spend Basis
Set team_spend_basis to the payment field that best matches the client's use case:
| Value | Use when |
|---|---|
paid_amount | Starting with standard Tuva claim payment data |
cclf_net_paid_amount | The CCLF connector surfaces x_cclf_net_paid_amount |
standardized_allowed_amount | The client supplies x_standardized_allowed_amount |
For CCLF feeds, allowed_amount is not a reliable full-episode spend basis because it is not populated consistently across claim settings and is generally not complete for institutional claims. Use paid_amount or a connector-derived CCLF net-paid proxy unless your project has created a complete standardized payment field.
CMS standardized allowed amounts are not available in standard CCLF output. They must be supplied by the client or produced by a separate methodology.