Fraud Detection of Claims


Case IV:

Advanced analytics helps payer identify $45M in overpaid claims from fraud, waste and abuse

The US spent $2.7 trillion on healthcare in 2013, more than 17 percent of GDP. According to a Commonwealth Fund study, despite the high spending on health care, Americans have poor population health compared to other high-income countries (indicated by low life expectancy at birth, chronic conditions, obesity rates, infant mortality, etc.). One of the reasons is fraud, waste and abuse, which diverts significant resources away from necessary care. The US healthcare system loses more than $200 billion every year in fraud, waste and abuse, nearly 10 percent, of annual healthcare spending.

The Government Accountability Office (GAO) has deemed Medicaid to be highly vulnerable to fraud, waste and abuse. This could be due to services or drugs not covered, not medically necessary, provided but incorrectly billed, or billed for but never provided.

A leading national healthcare payer
Identify claims overpayment and opportunities to reduce cost

Unsupervised learning methods identified a recovery opportunity of $45M from overpayments, and established cost containment processes

A leading multi-billion dollar healthcare payer, with a growing government business supporting Medicare and Medicaid, wanted to identify claims overpayments and opportunities to better contain costs.

Claims may be overpaid due to fraud, waste or abuse by providers, pharmacies, members or payers’ internal claims processing errors. With variations in member demographics, health conditions, condition severities and treatment patterns, it is critical to compare claims with other similar claims when looking for irregularities or aberrancies.

An unsupervised learning framework was developed to identify overpayments

  • Defined the causes of fraud, waste and abuse, and identified the focus areas with high impact
  • Developed over 50 hypotheses to define key internal and external data elements to analyze
  • Used multi-level unsupervised clustering framework and business rules, to create homogenous segments across member, provider and procedure details 4.Evaluated millions of claims to identify patterns or thresholds of aberrant dosage and pricing, while controlling for clinically homogenous segments
    The aberrancy identification framework flagged claims likely to be overpaid and helped in achieving unbiased leads for foragers and Cost Containment Units (CCU). An independent clinical validation showed identification of fraud, waste or abuse with 60% accuracy compared to 10%-20% accuracy of the existing processes. The entire framework was automated for a scalable and efficient process

The Payer was able to identify more than $45 million dollars in recoverable overpaid claims by:

Developing business rules leading to systemic changes to hold possible overpayments

Identifying 20 times more claims with dosage and pricing aberrancies

Optimizing the recovery process through the recommended prioritization framework

More than 75% of the identified claim overpayments were due to higher dosage or units billed than actually serviced

Nearly 25% of the identified overpaid claims were systemic inconsistencies

More than $45M of recoverable opportunity identified from overpayments

 Identified 20x more claims with dosage and pricing aberrancies

 Faster time to insight to adjust actions

Efficiency improvement going beyond flagging overpaid claims to also recommend reasons to foragers

Scalable approach to quickly expand lines of business and procedures