While payers grapple with $240 million in annual healthcare fraud, waste and abuse (FWA), analysts find AI will greatly improve detection rates. What’s holding the healthcare industry back?

Artificial intelligence (AI) can identify early indicators of healthcare FWA before it occurs, making technology a very powerful tool for healthcare payers. Better detection also reduces false positives by as much as ten times. Brighterion collaborated with PYMNTS analysts to learn more about what is needed to increase adoption rates of AI-based fraud prevention in healthcare.

Based on interviews with 47 leading healthcare executives, PYMNTS reported in The Unlocking AI playbook: healthcare edition that while advanced AI has made some progress in healthcare, legacy technology systems seem to endure, despite obvious limitations.

Here are the key findings:

AI’s use in the healthcare sector is limited and confined to large organizations

Only 4.3 percent of healthcare firms report using AI, while older legacy systems are much more commonly used. Twenty-two percent of firms with more than $500 million in annual revenue use AI, while none below that threshold do so. Legacy systems are far more prevalent, with almost 64 percent using business rules management systems (BRMS) and 32 percent using data mining.

Healthcare firms are highly interested in adopting advanced AI

The research shows that 34 percent of healthcare organizations are “very” or “extremely” interested in implementing advanced AI. That increases to 55.6 percent among those with over $500 million in annual revenue.

Healthcare firms believe advanced AI could improve FWA detection

The most interest was in using advanced AI to better target fraud. The three benefits that firms believe are most important are stopping fraud before it happens, reducing payments fraud and reducing fraud management personnel. The most commonly cited benefit, however, is reducing false positives (65.6 percent), the long-term result of improved fraud detection.

 The most common technologies used by payers – and their results

The report reveals some interesting facts about technology currently used to verify healthcare payments. Along with BRMS (64 percent) and data mining (32 percent), healthcare payers also use case-based reasoning (17 percent), fuzzy logic (8.5 percent) and deep learning and neural networks (4.3 percent). In addition to these legacy systems, another 4.3 percent don’t use any technology to investigate and detect fraud.

The unlocking AI playbook reveals an obvious gap in the detection and prevention of healthcare FWA. While the research showed a great interest in AI investment, perhaps the hesitation is how to get started.

How to use The unlocking AI playbook: healthcare edition

Fraud, waste and abuse continues to rise each year, costing healthcare payers, governments and ultimately the patients who use the services. The unlocking AI playbook shows how using AI to combat FWA can reduce false positives, stop fraud before it happens, increase revenues and reduce charge offs. Download it, have a read and let us know if you have any questions.

 The unlocking AI playbook: healthcare edition was researched and written by the analysts at PYMNTS.