This article original available on podcast, click here to listen podcast.
There are a ton of different types of compliance analytics to perform. In another podcast, maybe we will go through many other kinds: what payers are doing, what the government is doing to enforce compliance. They are using data and analytics to do that and some artificial intelligence.
Today, we will share one example. We brought in to do some consulting for a huge provider group doing more than 100 million in revenue. Also we will not even share the specialty because we want to be careful not to violate any confidentiality, not only for legal reasons, but we do not want to cause any problems for anybody. We were doing a bunch of different types of analytics. What popped out immediately was one of their physicians was doing all level 5.
Different levels in coding
Let me take a step back and explain levels in coding. In many different areas of healthcare, there are sometimes other levels of a particular procedure. For evaluation and management (E&Ms), you may have a one-to-five scale. For new patient encounters, follow-up visits, types of consults, critical care, and other things, you might have two or three levels of that same type of holiday. Example, for something in the laboratory, you might have more or less complex, so you do different levels of a particular kind of test. That is often going to be the case.
- What you will find is that there should be a distribution amongst those. It does not mean that it is going to be a normal distribution, but it should be that there are some low-level, some mid-level, and some high-level. There should be some variation. You can find some patterns.
- In this particular case, with this physician, when we looked at a histogram. A histogram was just the perfect way to demonstrate this. We were sitting down with the chief operating officer (the COO). I turned my laptop towards him and showed him the histogram. Level 1 was 0, level 2 – 0, level 3 – 0, level 4 – 0, level 5 – 100%. There was a giant spike. It was so glaringly obvious. I did not have to say anything. I just showed him the picture and nothing else.
- The only thing he said was, “I need the name of that doctor.” Of course, we gave them the name, and they had to launch an internal audit. Again, we are not entirely sure what the outcome of that was because that was their internal. We did not email that information to them.
Our goal is not to be a big brother and report them.
Significance of compliance analytics
Having compliance analytics in place to catch these types of problems is essential because that can expose you to a lot of liability. It is pretty much impossible to do all level 5. That cannot happen as it is statistically impossible. It is effectively considered fraud to bill something like that. That means you can have financial liability; you can have legal liability. There can be a whole range of issues that got tied up in that type of situation.
It is just a straightforward example. Many different types of compliance analytics can ensure that you are not having a problem. We will go through some more of those in future podcasts. However, It is just a straightforward one you can put in place, which is, for any procedure where you know there should be some distribution, a simple check to see if somebody is billing all one particular level. Some providers bill all level 3, for example. There is no 1, 2, 4, or 5. It may not be as bad a problem as billing all level 5, but it is still the same problem. From a legal standpoint, you run into the same issue.
Put in place some compliance program where you are using data to check for these types of problems.