Artificial intelligence can help identify anomalies in the data and prevent them from reoccurring

Coordination of Benefits has a clear mission to maximize recoveries by identifying instances of other health insurance coverage. Subrogation maximizes recoveries by finding other parties responsible for the payment. Data Mining has a mission to identify overpayments, but the way the service finds the overpayments isn鈥檛 as clear cut as it is for Coordination of Benefits and Subrogation. Data Mining, a post-payment payment integrity service, detects errant medical claims in sets of data.鈥疎rrant claims typically include duplicate payments and claims with billing issues that lead to overpayments. 

Identifying and preventing overpayments

野花社区 identifies errant claims by scouring the data. We use artificial intelligence, including machine learning and algorithms, to identify incorrectly paid claims and anomalies in the data. We also have subject matter experts on our data mining team to review and confirm the findings of the technology. All findings are presented to our clients for approvals.

To help prevent the same type of overpayments from reoccurring, our Data Science team reviews the anomalies identified and determines which are the most important in terms of the overpayments they cause and whether action can be taken to correct and prevent them from reoccurring. If action can be taken, the team designs a programmable rule that can prevent the anomaly from occurring again. 

Finding overpayments others don鈥檛

Three of the country鈥檚 top five health plans use 野花社区鈥檚 Data Mining service. Collectively, 野花社区鈥檚 Data Mining service has identified over $145 million in commercial overpayments for these clients. What makes that number even more impressive is that typically our clients and other data mining vendors searched for overpayments before we had our turn. For one of these clients, we reviewed the data after six others had, and we still found overpayments. This means other vendors miss things we find. It also means we have to think outside the box when reviewing clients鈥 data.

Here鈥檚 what we find

The following are examples of overpayments 野花社区鈥檚 Data Mining team has found in our clients鈥 data:

  • According to one of our client鈥檚 systems, a provider鈥檚 contract contained atypical language stating the provider would be paid a per diem of $3,000 for a service that lasted longer than one day. 野花社区 pulled the paper contract and confirmed the per diem clause was intended only for extremely rare cases. We obtained the client鈥檚 permission to expand the recovery period from one to two years and were able to identify 306 claims that were paid incorrectly, amounting to approximately $4 million in recoveries for overpayments.

  • A national large insurer contracted with a high cost drug provider for 10% of charges and a 90% discount. However, the client鈥檚 system was programmed wrong, and they were paying 90% of charges with a 10% discount. The client was able to recover approximately $1.5 million from 17 incorrectly paid claims. 

  • Our client鈥檚 contract with a provider stipulated the provider was entitled to 100% of billed charges (less denied claim lines) if their exclusions didn鈥檛 hit the contracted threshold. To sidestep the stipulation, the provider filed claims individually that should have been filed together so that the threshold wasn鈥檛 reached and the provider received 100% of billed charges. By identifying this, we helped our client recover approximately $1.7 million on 213 claims. 

Let us take a look at your data. You might be surprised by what we find. Learn more.