It seems simple. Health insurers receive a bill for medical care and make the payment. Yet, payment integrity has long posed a challenge in health care. Although health plans have made substantial progress, in 2014, they were still over- or under-paying claims 9.5% of the time. Consequences include health care provider abrasion, wasted spending and investment in resources to recoup excess payments.
The arsenal for identifying and recouping overpayments is varied. Network design, contracting, provider behavior modification, pre-authorization and clinical review audits all play a role. However, one area that is often overlooked is data mining.
Potential for automation and low provider abrasion make data mining appealing, but these programs have historically been plagued by a tendency to return false positives or responses that aren’t actionable. The reason? A failure to account for internal processes and systems during onboarding, and an inability to customize programs.
Overpayment identification is not the place for a one-size-fits-all solution.
Plans whose data-mining programs integrate custom concept development can achieve far more accurate results — higher than 95% — and a more favorable ROI. The key is starting with standard queries and then building in customization to ensure any overpayments identified are actionable.
Plans working to modernize their payment integrity data-mining programs should start with industry-standard concepts. Coordination of benefits, claims-processing platforms, and industry and provider trends are important to data-mining programs at any health plan. However, even this stage of preparation, data-mining teams should examine internal processes and systems that might override any patterns detected in payment data. Chasing down nonissues isn’t a good use of anyone’s time or resources.
Those early steps provide the foundation for customization. The idea behind custom concept development is to enable health plans to zero in only on overpayments they are willing and able to collect on. Analysts begin by looking at a plan’s policies for claims processing and contracts with members and providers. Stakeholders, including contracting, legal, compliance, medical and clinical teams, should have a hand in determining which types of overpayments are actionable. Provider exclusions, time-frame limitations and other parameters that might influence actionability should all be on the table.
As custom concepts are developed and tested, auditors should compare results against key information from claims-processing systems, pricing systems and other sources. The idea is to validate the findings and ensure overpayments meet the plan’s standards for action. This process lays the groundwork for automation.
Even after automation is established, parameters can change, so quality control is needed. A percentage of results from each production run should be reviewed and validated before deployment to ensure nothing has changed, results will continue to be accurate and ROI will be maintained.
As custom concepts are optimized and the data-mining program is refined, plans will reap the benefits of accuracy, reduced provider abrasion and alignment with a data-mining partner who understands the drivers behind payment accuracy issues and the messaging around payment adjustments. Every plan is different, and data-mining partners who understand that will deliver fewer false positives, better ROI and far fewer headaches.
Learn more about SCIO Health Analytics’ customized approach to overpayment identification and other solutions for health plans at sciohealthanalytics.com.