Decoding insurance claims and medical fraud

Artificial intelligence can help insurance companies and third-party entities to process insurance claims. It can verify data, check for errors and fraud, or find correlations and trends.

We count on insurance to be there for us when we need it, and like anything else, it’s a system that can be overused and even abused. Artificial intelligence is taking on an important role in the prevention of inaccurate healthcare claims and in innovative claims management.

fraud-detected-2

At H2OWorld 2017, in Mountain View, CA, speaker Adam Sullivan of Change Healthcare explained the process of using machine learning to verify healthcare claims data, denied claims, and erroneous payments for hospitals/providers, as well as to predict procedures. He said, machine learning is a current application of AI based on the idea that we should really just be able to give machines access to data and let them learn for themselves.

Global use of machine learning to tackle medical fraud

Globally, medical fraud and abuse continue to be a major problem contributing to loss of revenue. According to the National Heath Care Anti-Fraud Association it costs US patients about $68 billion annually.

Machine learning has been implemented to help detect medical data entry errors or unearth illicit activities. In Chile, a study was conducted in 2006 to collect digital medical claims data by running it through a fraud detection system with records ranked by their fraud probabilities, acting as a pre-screen filter. The new data mining workflow estimated that the new fraud detection system rejected medical claims that contributed between 9.5% and 10.0% of the overall raw costs.

“According to The National Heath Care Anti-Fraud Association it costs US patients about $68 billion annually.”

Data mining can help third-party payers such as health insurance organizations to extract useful knowledge from thousands of claims and identify a smaller subset of the claims or claimants for further assessment and scrutiny for fraud and abuse. In this way, the data mining approach is part of a more efficient and effective IT-based auditing system. It should be noted that fraud detection is only one part of a bigger program of combating health care fraud, abuse, and waste (A.Rashidian et al., 2012). Fraud detection should note the pitfalls that health care delivery policies that can create an increase in the possibility of fraud and abuse (Capelleveen, 2012).

The mining of electronic health records (EHR) data provides opportunities for better detection of fraud and abuse. The innovations in machine learning and artificial intelligence brings attention to automated methods of claims data fraud detection.

Healthcare claims go global

Patients are participating in medical tourism, from eye surgery to plastic surgery, and understanding claims data, electronic healthcare records and the complexity of data portability is a global challenge.

As global and local healthcare becomes more community outpatient-based, the consumer and clinician continue to struggle in deciphering healthcare claims data. The new epatient as a consumer is generating global healthcare claims data, from the LASIK procedure in South Africa to outpatient chemotherapy at a community clinic in Estonia. AI is is being used to mine and process healthcare claims data.

medical-tourism-1

Epatients are looking for ways to cut and cover healthcare costs while managing claims data. The customer experience for healthcare coverage has gone global and also become more personalized. AI will enable a seamless automated buying experience, using chatbots that can pull on customers’ geographic and social data for personalized interactions.

Insurance carriers will also allow users to customize coverage for specific items and events (known as on-demand insurance). On-demand insurance, e.g. the insurtech startup Neosurance, uses an AI agent and cloud2cloud technology that allows targeted engagement with customers with appropriate policies based on their profile, location, context, and behavior. It also delivers “contextual insurances” to consumers managing medical coverage via smartphone, thus enabling on-demand engagement.

“AI will enable a seamless automated buying experience, using chatbots that can pull on customers’ geographic and social data for personalized interactions.”

The chase is on globally for value based care, focusing on value by rewarding better outcomes and lower spending. Patients are digital consumers needing digital tools that enable access to their personal information and help them to make better choices based on quality, cost, and convenience. The clinicians understand the shift from patient care to consumer care that brings value. They also understand the implications of AI in healthcare in improving precision care as a result of accurate claims data. Only when these concepts are demystified can we can bring a collaborative trust and transparency to the new digital consumer.

Sources:

  • Fraud statistics. Blue Cross.
  • Health Systems in Transition (HiT) profile of Norway. The Health Systems and Policy Monitor.
  • Joudaki, Rashidian, Minaei-Bidgoli, Mahmoodi, Geraili, Nasiri and Arab. Using Data Mining to Detect Health Care Fraud and Abuse: A Review of Literature. PMC.
  • Marr, Bernard. What Is The Difference Between Artificial Intelligence And Machine Learning?. Forbes.
  • Ortega, Figueroa and Ruz. A Medical Claim Fraud/Abuse Detection System based on Data Mining: A Case Study in Chile. ResearchGate.
  • Rashidian, Joudaki and Vian. No Evidence of the Effect of the Interventions to Combat Health Care Fraud and Abuse: A Systematic Review of Literature. PLOS.
  • Zagorin, Edmund. Artificial Intelligence in Insurance – Three Trends That Matter. TechEmergence.