The Healthcare Fraud Dilemma
Millions of claims are processed but, according to the FBI, 10% are fraudulent. (Financial Crimes Report to the Public, Fiscal Year 2010, retrieved from http://www.fbi.gov/stats-services/publications/fcs_report 2010)
How do you find the 10% when on the face of the claim they all look alike?
The Solution: Find the bad guys by finding the bad guys they learned how to be bad from
The COPS Process – Linking Fraudulent Behavior to Social Media
The COPs uses a highly advanced Cluster Optics Process (COPs) programming that quickly analyzes high volumes of claims data to immediately identify inappropriate relationships between physicians, pharmacies and patients, providing the watchful eyes that the healthcare industry so desperately needs. Using Social Network Connection Analysis, COPs intelligently gathers evidence form every social media channel including Facebook, Twitter, Vine, LinkedIn as well a publicly available information such as State Board of Pharmacy sanctions and criminal records. This investigative analysis would take hundreds of hours to compile manually. The COPs process reduces staff and costs and is an effective evidence gathering tool. The COPs software learns as it goes, so with every analysis, it produces more positive results.
How does it work?
Step 1: The software runs through thousands of scripts and quickly creates clusters to visually spot a problem.
Step 2: Inappropriate relationships are defined and analyzed
Step 3: Final Report is Produced for Investigator
BIG DATA is brought in to show:
- Facebook “likes” between prescriber and pharmacy owner
- Google Maps and street level picture
- Linked-in connections
- Arrest records or license sanctions
COMBINED WITH data in the files:
- Drug Mix (divertible drugs)
- Same Day reversals
- Family diversity
- Pharmacy type
Case Studies:
Case One used about 5,000 claims and produced a handful of clusters. It found a disbarred Illinois physician practicing in Connecticut with entire families allegedly under his care but most likely the victims of medical identity theft.
Case Two took 62,337 claims in a two-week period. Prescription drug claims were analyzed to determine if there was a co-conspiracy scheme occurring. Analysis took data from 5,380 pharmacies and 15,321 prescribers and within minutes determined 37 suspicious clusters.
Once suspicious activity is found through COPs, additional social media information is provided to asses probability that fraud is actually occurring in these cases. If the pharmacy continues to look suspicious, and fraud is likely to be found, PIC consultants can provide on-site audits of the pharmacy.