Healthcare fraud is big business and ferreting it out is, too. Depending on the year, it’s estimated that the DOJ’s efforts via its Health Care Fraud and Abuse Control Program returns from $4 to $10 for every dollar spent.
But now, per an August 13, 2026, memorandum released by the Assistant Attorney General overseeing the DOJ’s newly formed National Fraud Enforcement Division, referred to as “NFED”, the government is about to supercharge the existing Health Care Fraud Strike Force model with additional resources, data analytics support, and technology.
From the viewpoint of catching criminals, that’s fantastic. After all, it’s estimated that up to 10% of national healthcare spending, which translates to up to $700 billion annually, ends up in the pockets of fraudsters.
But from the viewpoint of the huge majority of physicians and medical groups, there’s growing concern that the algorithmic approach announced by the DOJ might catch high-performing, completely legitimate practitioners by mistake . . . like dolphins caught up in tuna nets.
The Algorithm Said You Did It
Overall, the government plans to find and prosecute fraudsters and recover billions by throwing significant resources at the effort. By the time you read this, the DOJ should have nearly 500 attorneys and staff tasked to the National Fraud Enforcement Division, which is to work in concert with the other DOJ divisions, federal agencies (think IRS, OIG, Postal Inspectors), as well as state and local agencies.
But most interesting, and potentially troubling, is the DOJ’s announcement that it will use data analytics, supported by a cross-disciplinary team of experts in data science and cutting-edge technology, to target healthcare fraud including, in their words, “telemedicine fraud, Medicare and Medicaid fraud, controlled substance diversion, home health and hospice schemes, and companies and individuals that deceptively market unsafe products and services.” This is, essentially, a reference to supercharging the DOJ’s existing healthcare fraud data analytics team.
The problem is that data analytics, undoubtedly aided by AI, spots anomalies, such as billing volume that deviates from peer group norms.
But what’s the appropriate peer group? And, deviation from a norm is a statistical description, not a clinical one.
In fact, some highly accomplished physicians clearly deviate from what might generally be described as their “peer group” and perform an incredible number of medically necessary procedures that others in their specialty might rarely, if ever, perform. Will, for example, a genuinely gifted surgeon be AI-branded as a potential fraudster to be investigated? Considering that the data analysts will have no way of knowing the clinical aspects of a physician’s practice, it’s safe to assume that the government isn’t looking to award prizes for excellence.
Returning to the issue of peer group construction, there’s the question of how any comparison group would be risk adjusted. For example, adjustment for subspecialty, case acuity, patient complexity, and so on. It’s unlikely that any of this will be, or even can be, taken into account on the data analytics level. Will this result in a physician who treats sicker, more complex patients, and who attracts those sorts of cases because of his or her reputation among peers, appearing statistically the same as a fraudster?
There are additional likely defects in the process, but you get the point. Analytics is great at finding statistical outliers, but “outlier” and “criminal” are not synonyms.
The Action Required
So, what’s the best practice moving forward?
Maintain active compliance programs including regular training. A “plan”, whether on a shelf or on the cloud, is useless.
Focus significant attention on charting, documentation, and coding. Is there support for everything?
Regularly review your group’s relationships with those to whom you refer and those from whom you receive referrals for strict compliance with anti-kickback and self-referral (i.e., Stark) laws, both federal and state.
Bring in outside counsel and counsel-engaged experts to self-investigate on a periodic basis. Audit your own numbers the way DOJ would, not just for accuracy, but for utilization. Know about the red flags within your practice before an FBI agent does.
Remediate misconduct and self-disclose where appropriate.
Understand that beefed-up prosecutorial resources will likely lead to an increase in cases originating within the DOJ as well as an increase in the number of False Claims Act whistleblower cases in which the feds intervene. Maintain excellent relationships with your staff and vendors and take seriously any reports from them of impropriety; certainly, don’t take action that creates whistleblowers.
Reach out to me at markweiss@weisspc.com if you have questions about the implications for your practice of the DOJ’s supercharged enforcement efforts, their likely impact on operations, compliance programs, vendor relations, and potential civil and criminal exposure.
Related posts
