In the near future, two industry trends are likely to converge and then create a significant push by the federal government to beef up its investigations of fraud, waste, and abuse on the part of Medicare Advantage (MA) plans.
First is the surging popularity of MA plans, driven in part by patient demand for their low premiums and innovative, new benefits. Next is a flood of Covid-19-related claims as patients, payers and providers wade through medical bills and clinical records to reconcile billions of dollars in medical expenses.
Taken together, these factors will create a new reality in which payers will be forced to sift through increasingly huge volumes of clinical records to identify potential fraud, waste, and abuse (FWA), as well as confirm bill accuracy to properly compensate providers.
MA plans: Higher visibility, greater audit risk
Over the last decade, enrollment in MA, the private plan alternative to traditional Medicare, has more than doubled. In 2021, more than 26 million Americans were enrolled in MA plans, accounting for 42% of the total Medicare population, and $343 billion (46%) of total federal Medicare spending, net of premiums. The average Medicare beneficiary had access to 33 MA plans in 2021, the most in the last decade.
MA plans enter into contracts with the U.S. Centers for Medicare and Medicaid Services (CMS) to offer various benefits to enrollees, and are paid under a capitated system, in which plans receive a predetermined dollar amount each month for each enrollee from CMS. Importantly, monthly capitated payments are risk-adjusted for all individual enrollees to reflect their health status and project an appropriate level of monthly spending for Medicare-covered services.
To help ensure the accuracy and integrity of monthly payments to MA plans, CMS conducts risk adjustment data validation (RADV) audits with the goal of recovering any improper payments in which risk-adjustment diagnoses were not supported by data in medical records.
In 2021, OIG announced its intention to release of a series of several RADV audits of MA plans over the coming years that could find “tens of millions of dollars” in recovered funds as a result of improper payments. RADV audits help to ensure that MA advantage companies do not accidentally or intentionally submit claims for payments that are not supported by patients’ risk-adjusted diagnosis codes.
In April of this year, 21 people were charged in a number of schemes involving $150 million in alleged false billings and theft from federal Covid-19 assistance programs. In one case, defendants allegedly offered Covid-19 testing to get people to provide their personal identifying information and a saliva or blood sample, then took the information and samples to submit fraudulent claims to Medicare for unnecessary, far more expensive tests or services. Other defendants allegedly exploited telehealth policies put into place during the pandemic, misappropriated money intended for frontline medical providers, and manufactured and distributed fake vaccination record cards.
Most assuredly, these will not be the last cases of Covid-19-related fraud that result in federal charges.
Increasing audit efficiencies
Traditionally, payers have relied on expensive and time-consuming chart reviews to find and extract from patient records important unstructured data, such as information that indicates whether various Covid-related tests are medically necessary. Now, however, as an alternative to chart reviews, payers are increasingly looking to natural language processing (NLP), an artificial-intelligence-based technology that enables computers to “read” and understand text by simulating humans’ ability to interpret language, but without the limitations of human bias and fatigue.
NLP gives organizations the power to retrospectively analyze longitudinal health data to find a particular piece of clinical information about a single patient or identify subsets within populations that require further exploration. As CMS continues to target MA plans for FWA audits, NLP will play an increasingly valuable role in helping payers pinpoint instances of FWA before they get to the audit stage.
How NLP improves audit performance
To perform well in RADV audits, or avoid them in the first place, MA plans need solutions that enable users to efficiently and accurately identify details that support accurate risk adjustments. To assess risk and identify potential FWA, auditors must access and collect key pieces of information from thousands of pages of medical records, Following are three ways MA plans can leverage NLP to improve FWA audit performance:
- Detect patterns: Among the hallmarks of fraud is a pattern of repeatability in the data, such as a large number of patients meeting the same prior authorization requirements, for example. NLP helps payers detect these patterns that lack the natural variability found in legitimate patient records.
- Identify outliers: Similarly, payers can use NLP to spot unusual data that may be indicative of fraud, such as expensive tests for which there is no medical justification in the record. By analyzing unstructured data to identify anomalies within a patient record, NLP can quickly verify the presence, or lack of, critical data.
- Improve scale: Humans are limited in their ability to perform a high volume of chart reviews in a small window of time, but NLP automates the process, enabling massive gains in scale. With some medical records running into the thousands of pages, advanced NLP tools can drive significant savings in time and money.
Seniors’ demand for Medicare Advantage plans shows no signs of slowing down any time soon, and that means CMS is likely to continue accelerating its FWA audits of MA plans, particularly given the prevalence of Covid-19-related fraud. MA plans can prepare now for the coming flood of claims and audits by leveraging AI-based tools like NLP to improve efficiencies and accelerate the accurate identification of potential FWA.