Healthcare fraud costs tens of billions of dollars each year in the US alone, according to the FBI; this figure could soon be on the decline, though, as artificial intelligence (AI) tools are deployed to prevent and detect fraud.
As outlined by the Association of Certified Fraud Examiners (“ACFE”), there are three key factors—“The Fraud Triangle”—in explaining the key reasons that cause someone to commit occupational fraud, which, in combination, lead to fraudulent behavior: a perceived pressure or financial need, a perceived opportunity, and a rationalization of the behavior. In order to prevent fraud with technological innovation, AI must analyze behavioral data that might indicate pressure someone is facing and how they could rationalize fraud as one way to deal with those pressures. Yet the current organization of healthcare data may be a hindrance on efficient implementation of AI tools for fraud prevention:
Still, the success of AI tools for fraud detection will depend heavily on the ability to at least have all of the information available in one place, providing a deeper view into human behavior and processes from which different AI models can be trained. Healthcare’s transition to increasingly digital records could be a major boon in that effort.
Detecting fraud that has already happened tends to involve more analysis of transactional data. In healthcare, it often requires looking through information such as doctor’s notes, electronic health records, and insurance claims to look for mismatches or events that don’t fit larger patterns. Examples of the most common and challenging types of healthcare fraud that AI may tackle include: i) billing for services patients never received; ii) “upcoding” for more expensive treatments; iii) kickbacks and corruption.
Read more here.
11 Feb 2019