Medicare Part D Prescriber Risk & Compliance Analytics System

National scale analytical framework identifying high risk Medicare Part D prescribers through temporal machine learning

$31.70B
FY 2024 Improper Payments
72.46%
Predictive Accuracy (AUC)
33 fold
Improvement Over Random
26
Exclusions in Top 1,000
1.38M
Providers Analyzed
3.7 hrs
Processing Time
System Performance Metrics
72.5%
Temporal AUC
3.3%
Top 1% Precision
Key Achievements
26 Exclusions in Top 1,000
0.1892 Matthews Correlation
Top Risk Indicators (SHAP Importance)
Total Claims
9.8%
Opioid Share
9.6%
Beneficiary Risk Score
8.6%
Cost/Claim
8.0%
Brand Share
7.8%
Geographic Risk Distribution
Region Providers Exclusions Rate Risk Level
West 192,314 103 0.054% 1.82× baseline
Southwest 189,443 82 0.043% 1.46× baseline
Southeast 412,837 156 0.038% 1.28× baseline
Northeast 287,419 89 0.031% 1.05× baseline
Midwest 298,652 71 0.024% 0.81× baseline
High Risk Specialties
Pain Management
2.34× risk
Psychiatry
1.76× risk
Internal Medicine
1.37× risk
Exclusion Rates
Pain Mgmt: 0.089%
Psychiatry: 0.067%
Internal: 0.052%
Risk Stratification Performance
Risk Tier Score Range Providers Exclusions Precision Improvement
Critical 0.7 to 1.0 12,653 133 1.05% 10.5× baseline
High 0.5 to 0.7 15,341 81 0.53% 5.3× baseline
Medium 0.3 to 0.5 13,914 37 0.27% 2.7× baseline
Low 0.1 to 0.3 17,089 34 0.20% 2.0× baseline
Very Low 0.0 to 0.1 10,315 7 0.07% 0.7× baseline
Total Providers
69,312
Validation Set
Total Exclusions
292
2023 Testing Data
Top 1% Precision
3.3%
33 fold improvement
Baseline Rate
0.10%
Random Selection
Data Sources: CMS Medicare Part D Prescriber Data, Open Payments Database, HHS OIG LEIE Exclusions | Analysis Period: 2022 to 2023 | Total Data Volume: 26 GB across 5 federal databases | FY 2024 Medicare FFS Improper Payments: $31.70B (7.66%)
Note: This system identifies statistical patterns indicative of compliance risk. Determination of actual improper payments requires investigation by authorized agencies.