The case was brought by the data analysis firm Integra Med Analytics.
Moore’s law, the concept that the cost of storing data falls by half about every two years, is proving to be true. With data storage and data processing costs plummeting year over year, massive data analyses are within reach for more and more entities, including a new breed of whistleblower — those who analyze data to smoke out suspected patterns of fraud.
A company that has been a pioneer in the field of data whistleblowing is Integra Med Analytics. We have covered them previously on RACmonitor and Stat. Integra is a company that purports to apply statistical analyses of data sets to detect fraud and then files cases under the False Claims Act (FCA), a law that allows private party whistleblowers to file a case on the government’s behalf, alleging that the government is being defrauded (and then share in 15 to 30 percent of any eventual recovery). While most prototypical FCA whistleblowers are corporate insiders who witness their company or a company they work with committing fraud against the government, Integra is different. Instead of being a first-hand witness to fraud, Integra’s core business model is to identify fraud through mathematical modeling – and later, where possible, to fill in the gaps with information gleaned from the work of private investigators.
Although Integra’s efforts at rooting out fraud through data analysis have been unsuccessful thus far, a recent decision by the U.S. Department of Justice (DOJ) to join one of their FCA whistleblower cases may prove to be a watershed moment for both Integra and other similar whistleblowers.
Before turning to the New York whistleblower case Integra successfully persuaded the government to join, let us first revisit the three earlier healthcare fraud cases Integra filed under the whistleblower provisions of the FCA that were unsuccessful. In these cases, Integra accused hospital chains in California, Texas, and North Carolina of defrauding the Medicare program by overcharging for certain hospital visits. Using its data analytics specialty, Integra analyzed publicly available hospital discharge data for the three hospital chains and identified abnormally high rates of comorbidity diagnoses, such as encephalopathy, respiratory failure, and severe malnutrition, being applied to its Diagnosis-Related Groups (DRGs). Because Medicare pays hospitals based on which DRG they either treated, in the case of an illness, or performed, in the case of a surgery, and increases the amount it will reimburse for a patient with comorbidities or complications, Integra alleged that the hospital chains at issue improperly overcharged Medicare by applying comorbidity diagnoses to its patients’ DRGs in rates that far exceeded the prevalence of such conditions.
The courts ruled against Integra on defendants’ motions to dismiss in the North Carolina and Texas cases, with the Texas ruling later being affirmed by the Fifth Circuit Court of Appeals. Although Integra successfully survived defendant’s motion to dismiss its California case, the result was later overturned by the Ninth Circuit Court of Appeals, ending that matter. These dismissals were primarily based on two arguments: 1) that Integra could not provide enough detail about the specifics of the fraud; and 2) that the fraud was uncovered from publicly available documents, thereby making relevant the FCA’s “public disclosure bar,” which states that a whistleblower lawsuit cannot be based on information that is broadly, publicly available and takes no specialized expertise to interpret.
It is worth noting that the government did not join any of Integra’s three unsuccessful whistleblower lawsuits alleging DRG fraud. In an FCA action, the government has the option to join a whistleblower’s lawsuit, known as government intervention, and take the case forward alongside the whistleblower. If the government joins a case, the whistleblower is eligible to receive an award of 15 to 25 percent of the recovery, and the chances of the case succeeding rise dramatically. If the government declines to intervene in an action, the whistleblower award range increases to 25 to 30 percent, but the likelihood of success declines dramatically, with over 90 percent of non-intervened whistleblower cases resulting in zero recovery.
Fortunately for Integra, its luck with government intervention decisions changed last week, when the DOJ intervened in the company’s whistleblower case against a group of New York nursing homes alleging improper inflation of Resource Utilization Groups, or RUGs, the system Medicare uses to reimburse nursing homes. RUGs measure factors like how many minutes of nursing and therapy services a patient is expected to need to determine the complexity of a patient’s case, with patients needing more or more complex care generally drawing higher Medicare reimbursement amounts than patients who required less (or less complex) care. By performing statistical analyses of the claims data for the New York nursing home chains, Integra found that “ultra high,” the most severe and expensive RUGs, were through the roof at several facilities. Integra also ran several statistical models to control for other, innocent explanations for these anomalous rates – and finding none, concluded that they were likely the result of fraud, then filing this whistleblower case in 2017.
Integra’s whistleblower complaint is a detailed breakdown of the statistical methods the company used and why innocent explanations cannot account for the grossly inflated rates. For several of the problematic trends described in the complaint, Integra determined they had the alleged probability of less than 1 in 100,000,000. By Integra’s estimate, the Medicare program was harmed to the tune of $129 million by these frauds.
Earlier this month, after a three-year investigation, the DOJ joined Integra’s whistleblower case and filed its own complaint. Unlike Integra’s complaint, the government’s complaint-in-intervention did not provide an in-depth tour of statistical methods. Instead, it laid out the fruits of the government’s lengthy investigation, providing detailed allegations of the defendants’ knowledge of the fraud.
This case provides a perfect example of the type of public-private partnership envisioned by the False Claims Act at work, and of the benefit a data-driven whistleblower can bring to the government. Healthcare fraud is endemic, and the government is under-resourced. Integra, an entrepreneurial whistleblower outsider harnessing public data to identify anomalous trends, alerted the government to a series of suspect nursing home providers, thereby saving the resources the government would have needed to expend to perform its own external audits and statistical modelling. Thanks to Integra’s analyses, which DOJ apparently found persuasive, the government can now focus its precious resources on prosecuting, rather than detecting, this alleged fraud.
Over the past few years, in the newsrooms of most major media outlets, a new breed of investigative journalist known as “data journalist” has emerged: reporters who acquire, analyze, and present the ever-increasing flow of available data to create and augment news stories. It therefore should be no surprise that a new breed of whistleblower – the data whistleblower – also seems to be taking shape. Like the editor at a major news organization cultivating data journalists, the government should leverage the resources data whistleblowers like Integra are beginning to provide in the form of fraud detection, using analyses of data sets to uncover heretofore undisclosed frauds. Data whistleblowers are force multipliers for overworked and resource-starved government prosecutors and investigators, who can help in their fight to redirect hundreds of millions of dollars away from the hands of fraudsters and into the public welfare, where it belongs.