DME Program Integrity Analytics System

Dual-Method Detection Framework for Medicare DME Billing Analysis | 2021 to 2023 Data

Billing Anomalies Identified
$4.66B
Above state baseline norms
High-Risk Suppliers
2,246
5% of 44,918 analyzed
Detection Accuracy
94.2%
With 87.3% precision
Temporal Prediction Lift
3.4x
Vs random selection

Top 5 States by Billing Anomalies

System Performance Metrics

Model Validation Results

Detection Accuracy 94.2%
Precision Rate 87.3%
Recall Rate 71.6%
F1 Score 78.6%

Extreme Outliers Detected

Annual Trend Analysis (2021 to 2023)

Risk Score Distribution Across Suppliers

35,934
Low Risk (0 to 50)
6,738
Medium (50 to 75)
1,797
High (75 to 90)
404
Critical (90 to 99)
45
Extreme (99 to 100)
1,166x
Max Billing Anomaly Ratio
3.4x
Temporal Prediction Lift
3
OIG Fuzzy Matches (85%)
<4 hrs
Processing Time

Data Sources

Medicare DME by Supplier
198,621 records • 93 variables
Medicare DME by Service
1,454,474 combinations
OIG LEIE Database
81,914 excluded entities
CMS Geographic Variation
33,639 units • 247 metrics
Note: Identified anomalies represent statistical patterns warranting investigation; determination of fraud, waste, or abuse requires case-by-case review by appropriate authorities. The $4.66 billion figure represents billing amounts exceeding state baseline norms among flagged suppliers, not confirmed improper payments.