CORPUZ-2024 / mssp_analytics

CMS MSSP County-Level Analytics · v5 spec · Modules A · B · C
— github
Loading build manifest…

Risk score YoY delta distribution

AVG_RISK_SCORE delta by county · source: derived

Risk score growth vs. expenditure efficiency ratio

X: YoY delta (%) · Y: PER_CAPITA_EXP / AVG_RISK_SCORE ($/unit) · source: derived

OC-cut stability — risk score across data cuts

Operational cut → Final · a large cut-to-final delta is a late-coding pattern (single-source HCC proxy) · source: derived

Top flagged counties — RADV risk composite ranking

Composite_score = 0.4 × YoY_delta + 0.4 × Exp_growth_ratio + 0.2 × V28_coding_exposure_index
Illustrative weights — pending empirical validation against RADV audit outcome data. Sensitivity: test 0.5/0.3/0.2 and 0.33/0.33/0.33 per v5 spec.
Key finding

Counties with YoY risk score growth >10% AND expenditure growth exceeding the national trend factor show HCC concentration patterns consistent with RADV audit targeting criteria. For ESRD enrollment type, a positive V28 delta signals CKD coding specificity opportunity — CHF+CKD combinations rewarded under V28 when documented at Stage 4/5. Aged Non-Dual counties showing negative V28 delta face vascular disease and metabolic HCC compression unrelated to actual health change.