The Centers for Medicare & Medicaid Services (CMS) has outlays of approximately $1.2 billion per year. The Medicare, Medicaid, and Children’s Health Insurance Programs (CHIP) provide healthcare for one in four Americans.* There are a significant number of legislative mandates requiring compliance for these programs, such as requirements for audits and special reporting. With increasing administration complexity and growing healthcare costs, CMS has been proactive in adopting new practices and technologies to oversee funds effectively.
One program that has benefitted from new solutions is Disproportionate Share Hospital (DSH) payments. Federal law requires that state Medicaid programs make DSH payments to qualifying hospitals that serve a large number of Medicaid and uninsured individuals. For states to receive Federal Financial Participation (FFP) for DSH payments, the law requires states to submit an independent certified audit and an annual report to the Secretary describing DSH payments made to each DSH hospital. CMS has the responsibility to determine the appropriateness of DSH payments.
How DCCA Engaged
As the prime contractor on CMS’ Medicaid and CHIP Financial (MACFin) program, DCCA has modernized the existing legacy systems and introduced leading-edge artificial intelligence (AI), machine learning (ML) and robotic process automation (RPA) capabilities. For DSH payments, DCCA used AI and ML technologies to develop payment anomaly and forecasting models as a proof of concept (POC). The anomaly model uses historical data to determine payment outliers in audit data. These data are provided to CMS auditors for review. Rather than random audit sampling, this solution applies intelligence to focus resources on potentially fraudulent or improper payments. The forecasting model also provides an estimate of anticipated future payments.
In the original POC, DCCA identified the top 5% of outlier payments for review by CMS. Our delivery team will work in collaboration with CMS audit staff to research identified anomalies. While an identified anomaly does not necessarily confirm an incorrect or fraudulent payment, our AI-driven models identify a cohort of outliers that assist in providing targeted samples. These AI-driven models contain hundreds of viewpoints around these payments, which focus reviewers on reviewing more relevant samples. These solutions provide CMS additional confidence that they have all the tools required to oversee and administer funds with the highest level of fiscal integrity.
*CMS Annual Financial Report 2021