Why fraud in niche healthcare services has historically been overlooked
Detecting fraud in niche healthcare services presents a variety of challenges. These claims tend to be low-volume, highly variable, and often lack sufficient historical data. Additionally, patterns across providers can be difficult to see without combining internal and external data. Because of these challenges, health plans have typically focused on high-cost, high-volume areas; however, AI-based solutions provide an opportunity to overcome these challenges.
Uncovering fraud in niche healthcare services with Shift Payment Integrity
Specialized services, such as home meal delivery, are often overlooked by insurers. Reasons may include low volume, limited data, complexity, and lack of cross-provider insights.
Shift Payment Integrity addresses these challenges by combining internal and external data sources with AI-powered analytics. Its models detect patterns across providers, patients, and billing practices, even for low-volume or niche services.
-
Detection of overbilling patterns that would be invisible without external data
- Efficient discovery of lower volume, higher complexity fraud schemes
The power of AI: spotting anomalies at scale
To illustrate, consider this example shared by Juandiego Marquez, Shift Lead Data Scientist - US Healthcare. In this case, Shift identified healthcare providers overbilling for home-delivered meals to highly overlapping patient groups.
- Using entity resolution algorithms, Shift linked providers who were owned or represented by the same individual, even when records were inconsistent or contained misspellings.
- Shift incorporated and analyzed provider review data with AI-powered sentiment analysis and natural language processing to uncover trends of low ratings and quality of care through online reviews of home-delivered meal service providers.
- AI-driven pattern recognition then analyzed billing behavior, specialties, patient histories, and demographics to establish accurate peer groups and detect anomalies.
This approach allowed insurers to uncover suspicious activity across multiple provider networks, revealing overbilling patterns that would have been nearly impossible to detect using only internal datasets.
If you don't see the video above click the button below to adjust your cookie settings
Manage CookiesLearn more from our team of 200+ data scientists
This video is part of a series of interesting cases presented by our insurance-focused team of data scientists. For more examples of Shift Technology's AI-driven results, browse the AI in Action library
-1.png)
