Avoid Fraudulent Medical Claims With AI-Based Fuzzy Logic

Uncovering and reducing health care fraud in the insurance industry has never been more important – or more difficult. The economic uncertainty driven by the COVID-19 pandemic has resulted in an increase of fraudulent activity and health care schemes. For example, according to a 2021 study conducted by CliftonLarsonAllen,  telehealth fraud (one of the top five fraud areas in the U.S. in 2021) was a nationwide scheme costing $6 billion to the federal and private health insurance sectors in the first quarter of this year.

In addition to increasing fraud, many health insurance companies are overburdened with the effects of the COVID-19 pandemic while simultaneously being challenged to meet the demands of their customers.

As such, many insurers simply pay low-value, high-volume claims because it's not cost effective to investigate them. For example, a suspicious $500 claim for unnecessary orthotics is more cost effective to pay than to spend time looking at. But why not let artificial intelligence investigate these suspicious low-value claims, alleviating the burden placed on insurers and elevating them to take on higher-value tasks?

AI automation, for almost no cost and no human intervention, can compare billions of claim documents and provide a score for the degree to which any person or claim is fraudulent. This makes AI far more effective than investigators alone at uncovering fraud. Leveraged in all aspects of claims management, AI-based fuzzy logic provides a unique opportunity to uncover previously unidentified fraud, ensuring bogus medical claims are flagged and that fraud is effectively stopped before it occurs.

How Does Fuzzy Logic Uncover Health Care Fraud?

In health insurance, fraud and risk are behaviors that differ from what is “normal” within a group of similar people or similar claims. For example, fraud may look like a doctor’s writing opioid prescriptions too frequently in comparison to their peers.

What makes fraud so difficult to identify is the fact that it arises from outlying human behavior – i.e., being different than one’s peers. This is because the notion of being different is an ambiguous concept, as one can be slightly different, moderately different, etc. To avoid fraud, all the degrees of difference must be distinguished from the all the different groups of “normal” – i.e., all the groups of individuals or claims that possess similar characteristics - that occur in an insurance business.

Ultimately, the degrees of difference that exist in fraudulent claims and activity can be expressed mathematically using AI-based fuzzy logic. Fuzzy logic is a mathematical method of solving problems using qualitative data to produce quantitative outputs. Ambiguous terms, such as “slightly” or “moderately” are referred to as “fuzzy.” Fuzzy logic factors in terms such as these and expresses them mathematically to produce quantitative outputs or decisions. An AI-based fuzzy logic system can autonomously choose the mathematical representations of the “fuzzy” terms and then perform billions of comparisons to assign a score for the degree to which any claim or person is different from their peers - a score that depicts a significant degree of difference is a flag for fraud.

It is no secret that fighting fraud is a tough job. Insurers are challenged to keep pace with the growing level of fraud in health care schemes while simultaneously being pressured to meet the demands of their customers. However, fraudulent claims don’t have to go undetected. When we let AI do what it is uniquely suited to do, investigators are better positioned to eliminate fraud and its effects across insurance businesses.

Ultimately, when leveraged in all aspects of claims management, AI-based fuzzy logic will uncover previously unidentified fraud, alleviating the burden placed on health insurers today and enabling them to effectively service their customers, remain competitive and remain profitable.

Gary Saarenvirta is founder and CEO of Daisy Intelligence. He may be contacted at gary.saarenvirta@innfeedback.com.

© Entire contents copyright 2021 by InsuranceNewsNet.com Inc. All rights reserved. No part of this article may be reprinted without the expressed written consent from InsuranceNewsNet.com.