A robust data analytics platform can pull out trends and patterns of potential fraud, but special investigative units also need to rely on "street smarts" to target fraudulent providers, according to Doug Cedras, former director of corporate and financial investigations at Blue Cross Blue Shield of Michigan (BCBSM).
Cedras, who currently works as an independent consultant, shared several ways in which data analytics can help direct fraud investigations in an interview with All Analytics. He said BCBSM began with a rule-based approach, but quickly graduated to a more advanced system that looked at providers who were arrested, criminally convicted or administratively sanctioned over a five-year period.
This approach, combined with aggregating six different data systems into "one source of truth," helped the insurer identify patterns within certain provider groups, allowing investigators to catch fraudulent payments earlier in the cycle. He added that more payers are shifting toward advanced analytics by integrating new algorithms and demographics into their system to facilitate "machine learning."
Cedras added that while analytics was critical to narrowing investigations, BCBSM's team of former law enforcement officials used "street smarts" by knocking on doors and talking to providers to parse out those who were engaged in fraud schemes and those who were merely committing coding errors.
Public and private payers are using data analytics to uncover sophisticated fraud schemes, but humans remain as an important aspect in fraud enforcement. Predictive analytics played a major role in last year's historic healthcare fraud bust that ensnared 243 people.
To learn more:
- here's the All Analytics interview
Feds, states turning to predictive analytics to prevent fraud
In fight against fraud, humans just as important as machines
Predictive analytics helps fraud fighters detect sophisticated schemes [Special Report]