In most of the articles and press pieces that speak of health care fraud, the topic is usually covered in broad terms regarding the type or activity that has taken place: over-billing to Medicare/Medicaid, services billed and not delivered, over subscription of medication (by members or Pharmacies), and the list goes on and on.

What is not evident at first glance is that the details and patterns associated with the various types of fraudulent behavior differ from region to region and continue to change over time. For example, it is generally accepted that fraud schemes begin in more populated regions and migrate to more rural areas over time. What is usually not mentioned, however,  is how the schemes themselves alter to adjust to new regions that they are moving into (accounting for different policies or regulations, for example). And more importantly, the schemes get smarter, having learned from their previous geography, what they can and cannot get away with.

As a result, it becomes exceedingly important and respectively more difficult for different health care companies (states, countries, insurance plans, etc.) to adjust to the new changes that a scheme has gone through.

Catching and even recognizing fraud is not as simple as saying, "a provider defrauded Medicaid by submitting false prescriptions." How does that provide a pattern to look for? And even if it did, now that it is a known pattern, one can be fairly certain that it will change to avoid easy detection.

Existing technologies can look for specific patterns, but what happens when the pattern changes? This raises the question that we should be thinking about in the industry: How can we more quickly adapt to identify a known scheme that has changed its spots?