Run #73 — FrontPath — MRF Quality Dashboard

CMS Official Schema Validation FAILED (exit 1)

Validator output

error Could not find a schema version named "1.0.0". Available versions are:
v2.0.0
v0.1
v0.2
v0.3
v0.3.1
v0.3.2
v0.4.0
v0.4.1
v0.5.0
v0.5.3
v0.7.0
v0.8.0
v0.8.1
v0.9.0
v0.9.1
v0.10.0
v0.10.1
v0.11.0
v0.12.0
v0.13.0
v1.0.0
v0.10.3
v1.0.4
v1.0.6
v1.0.7
v1.1.0
v1.1.1
v1.1.2
v1.2.0
v1.3.0
v1.3.1
v1.3.2
v1.3.3
v1.4.0
v1.5.0
v1.6.0
v1.6.1
v1.6.2
v2.0.1
v2.1.0
v2.2.0
v2.2.1

Errors (3)

file_freshness

File is 182 days old (last_updated_on exceeds the 90-day threshold)

cms_schema_validation

CMS official schema validator FAILED (exit code 1). File does not conform to the TIC in-network-rates schema.

zero_rates

1870 zero-dollar rates (0.23%) — CMS schema requires negotiated_rate > 0 (exclusiveMinimum)

Warnings (5)

billing_code_format

75332 CPT codes do not match expected format

duplicate_billing_codes

17476 billing codes appear in more than one in_network item (96.2%)

rate_spread_by_class

billing_class='institutional' / negotiated_type='negotiated': P95/P50 spread is 12.5x (threshold: 10x, N=63,373 (1,000 sampled), high confidence)

rate_spread_by_class

billing_class='professional' / negotiated_type='fee schedule': P95/P50 spread is 13.8x (threshold: 5x, N=743,377 (1,000 sampled), high confidence)

per_code_rate_spread

183583 rate contexts have a max/min ratio exceeding the type-specific threshold (20x professional / 50x facility, min 3 occurrences required). Each context is a unique combination of all 10 rate-key dimensions. n= shows how many distinct provider rates exist for that exact context.

Code	Code Type	Negotiated Type	Billing Class	Arrangement	Setting	Min	Median	Mean	Max	Ratio	n
`J0461`	CPT	fee schedule	professional	ffs	—	$0.02	$8883.19	$9666.98	$20901.50	1045075.0×	4
`J0461`	CPT	fee schedule	professional	ffs	—	$0.02	$8883.19	$9666.98	$20901.50	1045075.0×	4
`J0461`	CPT	fee schedule	professional	ffs	—	$0.02	$8883.19	$9666.98	$20901.50	1045075.0×	4
`J0461`	CPT	fee schedule	professional	ffs	—	$0.02	$8883.19	$9666.98	$20901.50	1045075.0×	4
`J0461`	CPT	fee schedule	professional	ffs	—	$0.02	$8883.19	$9666.98	$20901.50	1045075.0×	4
`J0461`	CPT	fee schedule	professional	ffs	—	$0.02	$8883.19	$9666.98	$20901.50	1045075.0×	4
`J0461`	CPT	fee schedule	professional	ffs	—	$0.02	$8883.19	$9666.98	$20901.50	1045075.0×	4
`J0461`	CPT	fee schedule	professional	ffs	—	$0.02	$8883.19	$9666.98	$20901.50	1045075.0×	4
`J0461`	CPT	fee schedule	professional	ffs	—	$0.02	$8883.19	$9666.98	$20901.50	1045075.0×	4
`J0461`	CPT	fee schedule	professional	ffs	—	$0.02	$8883.19	$9666.98	$20901.50	1045075.0×	4

Info (1)

percentage_rates

173 percentage rates (0.0%) — values represent % of a reference rate, not dollar amounts; excluded from spread analysis

Recommended Actions

1.

schema file_freshness

File is 182 days old (last_updated_on exceeds the 90-day threshold)

P1
2.

schema cms_schema_validation

CMS official schema validator FAILED (exit code 1). File does not conform to the TIC in-network-rates schema.

P1
3.

pricing zero_rates

1870 zero-dollar rates (0.23%) — CMS schema requires negotiated_rate > 0 (exclusiveMinimum)

P1
4.

pricing rate_spread_by_class

billing_class='institutional' / negotiated_type='negotiated': P95/P50 spread is 12.5x (threshold: 10x, N=63,373 (1,000 sampled), high confidence)

P2
5.

pricing rate_spread_by_class

billing_class='professional' / negotiated_type='fee schedule': P95/P50 spread is 13.8x (threshold: 5x, N=743,377 (1,000 sampled), high confidence)

P2
6.

pricing per_code_rate_spread

183583 rate contexts have a max/min ratio exceeding the type-specific threshold (20x professional / 50x facility, min 3 occurrences required). Each context is a unique combination of all 10 rate-key dimensions. n= shows how many distinct provider rates exist for that exact context.

P2
7.

code_coverage billing_code_format

75332 CPT codes do not match expected format

P3
8.

code_coverage duplicate_billing_codes

17476 billing codes appear in more than one in_network item (96.2%)

P3

Metrics

header_missing_fields

header_conditional_issues

file_age_days

182

items_total

772880

items_missing_required_pct

0.0

items_empty_rates

0

prices_total

806931

prices_missing_required_pct

0.0

prices_missing_field_breakdown

prices_missing_service_code

0

prices_invalid_billing_class

0

rates_without_providers

0

negotiation_arrangements

ffs

772880

billing_code_types

CPT	766682
MS-DRG	6120
RC	78

expired_prices

0

invalid_expiration_format

0

provider_references_in_file	121
provider_group_ids_referenced	121
unresolved_references	0
resolution_rate_pct	100.0
npis_validated	0
invalid_npi_count	0
npi_validity_rate_pct	None
invalid_npi_examples
eins_validated	0
invalid_ein_count	0
ein_validity_rate_pct	None
invalid_ein_examples
empty_npi_groups	0
groups_without_tin	0
npis_in_multiple_groups	0
unique_npis

unique_codes_total

18174

duplicate_codes

17476

duplicate_pct

96.16

by_code_type

CPT	17386
MS-DRG	765
RC	23

unknown_code_types

format_invalid_by_type

CPT

75332

codes_not_in_reference

reference_not_loaded

most_frequent_codes

Type	Code	Occurrences
CPT	85576	205
CPT	95819	170
CPT	95816	170
CPT	95813	170
CPT	73610	169
CPT	73080	169
CPT	73630	169
CPT	72110	169
CPT	72100	169
CPT	72020	169
CPT	73110	169
CPT	72120	169
CPT	72040	169
CPT	73120	169
CPT	73562	169
CPT	72170	169
CPT	72070	169
CPT	73590	169
CPT	76942	168
CPT	77003	168

total_prices_checked

806931

total_rates

806758

per_diem_rates

0

percentage_rates

173

negative_rates

0

zero_rates

1870

extreme_high_rates

77

extreme_low_rates

0

rate_distribution

sample_n	806758
sample_k	5000
confidence	high
p5	8.4195
p25	43.3475
p50	232.985
p75	955.09
p95	3103.135500000001
p99	8325.133800000014

by_billing_class

Class / Type	Count	Median	p25	p75	p95	Confidence
institutional/negotiated	63,373	1558.8	1042.2	3613.8	19500.0	high
professional/fee schedule	743,377	179.6	36.0	712.0	2478.7	high
professional/negotiated	8	90.0	43.0	90.0	96.5	low

negotiated_types

negotiated	63381
fee schedule	743377

unique_rate_contexts

1786981

rate_key_dimension_validity

invalid_negotiated_type	0
invalid_negotiated_types_seen	{}
invalid_setting	0
invalid_settings_seen	{}
invalid_severity_of_illness	0
severity_on_non_apr_drg	0
institutional_with_service_codes	0
invalid_service_code_format	0
billing_code_modifier_too_long	0

Scoring Methodology embedded in this report at generation time

Weighted sum of four structural dimensions, normalized to a 0–100 scale.

Normalized Weights

Schema Integrity	30%
Provider Mapping	15%
Code Coverage	15%
Pricing Sanity	40%

Confidence Bands

High	≥90
Usable With Caution	≥75
Limited Reliability	≥60
Not Usable	<60

Score Caps

Condition	Cap	Reason
Raw Json Errors Only	74.0	Native JSON syntax errors in the unpatched source file. File must be re-exported by the payer; scoring reflects auto-patched data only.
Cms Validation Failure Only	65.0	CMS official schema validator reports the file does not conform to the TIC spec.
Both Raw Json Errors And Cms Failure	59.0	Both native JSON syntax errors and CMS schema validation failure present.

Rate Context Key — 14-Tuple Field Coverage

Every rate in a CMS TIC MRF file is described by a 14-field tuple. Fields 1–10 form the rate-context key used to group and compare rates across the system. Fields 11–12 (provider, expiration date) are validated separately and excluded from the grouping key for analytical reasons. Each of the four scoring dimensions validates a distinct slice of this tuple — together they cover all 14 fields.

Field	Validated by
1 billing_code_type	Schema (required field) + Code Coverage (enum + format validation)
2 billing_code_type_version	Schema (required field)
3 billing_code	Schema (required field) + Code Coverage (format, duplicates, reference lookup)
4 billing_code_modifier	Pricing (modifier length, key normalization)
5 service_code	Pricing (POS format, normalization, institutional-class check)
6 negotiated_type	Pricing (CMS TIC enum validation)
7 billing_class	Schema (CMS TIC enum validation) + Pricing (spread thresholds)
8 negotiation_arrangement	Schema (CMS TIC enum validation) + Pricing (FFS vs bundle/capitation gating)
9 severity_of_illness	Pricing (APR-DRG only, valid values 1–4)
10 setting	Pricing (CMS TIC enum validation)
11 provider (NPI/EIN)	Provider Mapping (Luhn checksum, IRS prefix, group resolution) — excluded from grouping key
12 expiration_date	Schema (date validity, far-future sanity) — excluded from grouping key
13 additional_generic_notes	not validated (free-text)
14 negotiated_rate	Pricing (negative/zero/extreme-value checks, spread analysis)

Fields 1–10 are the grouping key. Each unique combination is a distinct rate context — rates with different modifiers, POS codes, or arrangements land in separate buckets and are never compared against each other.
Provider (field 11) is excluded from the key: the spread check is cross-provider by design. Partitioning by provider produces singleton buckets and eliminates the spread signal.
Expiration date (field 12) is excluded because it is a contract lifecycle attribute, not a clinical context. Rates for the same service should be comparable regardless of when they expire.
service_code (field 5) arrays are flattened and normalized before keying: '1' → '01', and a rate with ['11','22'] contributes to both the '11' and '22' buckets so rates are compared apples-to-apples by place of service.

Validates required fields, enum values, conditional requirements, and date validity per the CMS TIC in-network-rates schema. Also checks file freshness and expiration date sanity.

Scoring Method

Penalty-based deductions from 100, capped per category.

Penalty / Component	Formula
per_missing_required_header_field	`5`
per_header_conditional_issue	`2`
freshness_warn	`5`
freshness_error	`10`
item_missing_fields_pct	`×5 (cap 30)`
empty_rates_pct	`×0.5 (cap 5)`
price_missing_fields_pct	`×10 (cap 30)`
rates_without_providers_rate	`×200 (cap 20)`
expired_prices_pct	`×0.5 (cap 5)`

Thresholds

file freshness warn days	45
file freshness error days	90
expiry far future years	3

Verifies that all provider_group_id references in in_network items resolve to an entry in the provider_references array. Validates NPI integrity via Luhn checksum and EIN integrity via IRS-issued 2-digit prefix.

Scoring Method

Weighted component sum (not purely penalty-based).

Component	Formula
provider_resolution (60%)	`resolution_rate% × 0.60`
npi_validity (30%)	`(100 − invalid_npi_pct × 5) × 0.30`
ein_validity (10%)	`10 − (invalid_ein_pct × 0.1) [0% invalid → 10 pts, 100% invalid → 0 pts, linear]`

Checks

`provider_group_resolution`	Ratio of provider_group_id references that resolve to a known provider group. Unresolved references mean rates cannot be attributed to any provider.
`npi_validity`	Luhn checksum applied to every NPI value in the file.
`ein_validity`	IRS-issued 2-digit prefix validation on EIN values.
`empty_npi_groups`	Provider groups with an empty npi list.
`missing_tin`	Provider groups with no TIN (type + value pair).
`duplicate_npis`	Informational: NPIs appearing in more than one provider group.

Tracks every (billing_code_type, billing_code) pair and flags unrecognized CMS TIC code types, format violations for CPT/HCPCS/NDC, and duplicates (same code appearing in multiple in_network items).

Scoring Method

Penalty-based deductions from 100.

Penalty / Component	Formula
per_unknown_code_type	`3 pts each (cap 20)`
format_invalid_pct	`× 0.5 (cap 10)`
duplicate_code_pct	`× 2 (cap 20)`
codes_not_in_reference_pct	`× 0.5 (cap 30) — only when reference set is loaded`

Recognized Code Types

AP-DRG, APC, APR-DRG, APS-DRG, CDT, CPT, CSTM-ALL, EAPG, HCPCS, HIPPS, ICD, LOCAL, MS-DRG, NDC, R-DRG, RC, S-DRG

Detects invalid rates (negative, zero, extreme-value) and distribution anomalies (per-class P95/P50 spread, per-code max/min ratio, flat-rate distributions). Exact counts are used for all validity checks (negative, zero, extreme, dimension validity). Percentile-based checks (spread, IQR) use reservoir sampling — k=5 000 global, k=1 000 per (billing_class, negotiated_type) bucket — so memory stays bounded on large files. Per-code max/min spread is exact (all rates seen, no sampling).

Scoring Method

Penalty-based deductions from 100.

Penalty / Component	Formula
negative_rate_pct	`× 5 (cap 20)`
zero_rate_pct	`× 3 (cap 15)`
extreme_rate_pct	`× 5 (cap 25) — ffs only`
class_spread_excess	`(spread − threshold) × 2, max across (billing_class, negotiated_type) buckets (cap 15)`
per_code_high_spread_count	`× 0.1 (cap 15)`
invalid_negotiated_type_pct	`× 3 (cap 10) — rates silently dropped`
invalid_setting_pct	`× 1 (cap 5) — silently defaults to wildcard`
invalid_severity_pct	`× 1 (cap 5) — silently normalised to ''`
institutional_with_service_codes_pct	`× 1 (cap 5) — extra key variation`
invalid_service_code_pct	`× 2 (cap 5) — encode raises ValueError`

Thresholds

extreme high by billing class	`professional`: 25000.0`institutional`: 2000000.0`both`: 2000000.0`default`: 500000.0
extreme low	0.01
spread warn p95 over p50 by class	`professional`: 5`institutional`: 10`both`: 10`default`: 5
per rate context max min ratio	`professional_codes`: 20`facility_drg_codes`: 50
flat rate iqr p75 threshold pct	5.0
flat rate min rates to check	100
spread min n to flag	50
per code min n to flag	3

Notes

per diem and percentage rates are not dollar amounts and are excluded from all spread and distribution analysis.
Zero rates are treated as errors per CMS schema (exclusiveMinimum: 0).
Extreme-value flags (negative, zero, extreme high/low) are only applied to ffs negotiation_arrangement rates — bundle and capitation rates have different magnitude ranges and lack calibrated thresholds.
P95/P50 spread and flat-rate IQR checks are keyed by (billing_class, negotiated_type) so fee schedule and negotiated distributions are evaluated independently.
flat_rate_distribution is suppressed for negotiated_type='fee schedule' buckets — a flat distribution is correct behaviour for a fee schedule, not a defect.

The dashboard assigns a persistent mrf_key to each MRF. All validation runs of the same MRF share the same key so the score-history chart reflects how a file has changed over time — even if the payer re-exports it at a new URL. This key is not stored in the report JSON; it is computed at ingest time.

Key derivation (priority order)

Tier	When used	Input string
1	Both `reporting_entity_name` and `plan_id` present	`plan\|<entity>\|<plan_id_type>\|<plan_id>`
2	`plan_id` absent — falls back to URL (entity name alone is not safe: a payer publishes multiple distinct plans under the same entity)	`<file_location URL>`

The key is a 16-character MD5 hex digest of the input string (case-insensitive, whitespace-stripped). Tier 1 is stable across monthly file re-exports because the plan identity in the MRF header does not change when the payer fixes a bug and re-publishes. Tier 3 is URL-dependent and will produce a new key if the file path changes.

This run:

mrf_key	`0672c48ec3767cc2`
Entity	FrontPath
Key tier	Tier 2 URL hash

Geographic analysis is a supplemental feature computed on demand after scoring completes. It does not affect any scoring dimension — it is an observational overlay to help assess the breadth and distribution of in-network providers.

Data Sources

`NPPES`	CMS National Plan and Provider Enumeration System — monthly full-replacement CSV. Maps each NPI to its primary registered ZIP code.
`ZCTA centroids`	GeoNames US postal code file. Maps each 5-digit ZIP to a (latitude, longitude) centroid for map placement.

Process

Extract all NPIs from the MRF file (npi and npi_numbers fields).
Resolve each NPI to its primary practice ZIP via NPPES lookup.
Aggregate provider count per ZIP code.
Map each ZIP to a (lat, lon) centroid via ZCTA table.
Render as a weighted heatmap — intensity proportional to provider count per ZIP.

Limitations

NPIs not found in NPPES (recently issued, test NPIs, EINs used as NPIs) are excluded from the map and reduce the geocoding match rate.
Location reflects the provider's NPPES-registered primary address, not necessarily where they accept this specific plan.
Map viewport auto-fits to the bounding box of ZIP codes covering 90% of total provider count, dropping sparse geographic outliers.

FrontPath

CMS Official Schema Validation FAILED (exit 1)

Errors (3)

Warnings (5)

Info (1)

Recommended Actions

Metrics

Schema Integrity

Provider Mapping

Code Coverage

Pricing Sanity

Scoring Methodology embedded in this report at generation time

Overall Score

Schema Integrity

Provider Mapping

Code Coverage

Pricing Sanity

Dashboard: MRF Identity Key ingest

Provider Geographic Coverage supplemental