A New Tool In The Anti-Fraud Arsenal
Healthcare fraud is a cat-n-mouse game!
Fraudulent service providers will develop schemes to over bill for services provided, or even services not provided. Any scheme which passes through the payers’ rules-based systems will be exploited by the fraudsters until the payers catch on and add a new rule.
Claim Analytics provides a NEW approach.
A new approach is needed to provide a second layer of protection for insurers, an approach that can detect novel and undetected schemes that a rules-based approach will miss.
Claim Analytics technology is particularly strong in fraud detection in its ability to expose emerging fraudulent practices:
- Examines millions of submitted claims
- Identifies atypical claims
- Open-minded unbiased analysis
- Pre-payment identification of suspect claims
- Works with lightning speed
Our technology goes beyond traditional rules-based approaches By comparing each claim to every other claim, and each practitioner to every other practitioner, our technology goes beyond traditional rules-based approaches to fraud detection.
- This open-ended approach identifies virtually any type of atypical activity, rather than only those defined by a pre-determined set of rules.
- Claims departments can become more efficient by focusing their investigative and recovery resources on the outlier providers.
- Prepayment screening becomes more effective by measuring how typical or abnormal any given claim is.
- Different levels of tolerance can be allowed, depending on the profile of the service provider, before referring the claim for manual adjudication.
The claim history of each provider is compared to every other provider of the same specialty and outlier are quickly and easily identified. Isolating the atypical providers offers significant opportunities for both payment recovery and prepayment screening.
We look at the bigger picture.
Previous generation fraud detection tended to focus on the individual claim level. However, this doesn’t take into account who the claimant or practitioner is, and the nature of their claim history.
We use individual claims as building blocks to grow a bigger, clearer picture of activity at the level of individual practitioners and individual claimants. We don’t consider each claim in isolation we take into account the entire history of work for each practitioner, and for each patient, allowing our model to find and quantify more types of irregularities.
Atypical providers are easily identified and understood.
The models find similiarities and differences between all claims and categorize each into one of a small number of clusters. The distribution of claims by cluster for each service providers can be compared to the aggregate distribution. Atypical providers are then easily spotted and why they are atypical can be understood.
Predictive Modeling and Dental Fraud Detection