This article is initially available on our podcast. Click here to listen.
Have you ever heard of Simpson’s paradox? It would help if you had because it impacts you. Have you ever felt like you’re doing an excellent job in terms of your billing performance, whether it’s your billing department or your billing company or whatever it might be? Still, then somebody comes along and says how much better somebody else is performing? Some of these are vague statements like, “Oh, blah-blah-blah is doing better than you,” or “Why aren’t I getting something?”
Billing varying rates
We’ve seen this be specific. Some doc comes to their billing department or to their billing company or whatever it might be and says, “Hey, how come I’m only getting $1,500 per procedure when blah-blah-blah down the road is getting $2,000?” That’s a pretty specific example. They might be talking about something specific like an arthroscopy, chondroplasty, or PV ablation. I just picked random things. It could be anything.
How come X person or organization is getting something more significant than I’m getting? It doesn’t matter whether it’s a billing department or a billing company. You’re going to run in these kinds of situations.
Gross collection rate
Maybe, somebody comes back and says, “How come we’re getting a 40% gross collection rate, and blah-blah-blah is getting 45%?” Now, ignoring the fact that I hate gross collection rate (I think it’s got all kinds of problems with it), I’m just trying to give you an example where somebody comes back and says somebody is performing better or worse than somebody else. It has a quantitative measure for it, not some vague statement.
A better metric might be something like that. You get paid on 90% of the encounters with patients compared to practice Y spent on 92% of them. On a quantitative measure, you seem to be performing worse. At that 30,000-foot level, you are unless somebody drills in and figures out that there may be a problem with this. 92 is better than 90. Nobody can question that.
The sports comparisons
Let me give an example from the sports world. It’s not that I’m a basketball junkie. I’m not. It’s not one of my favorite sports. But let me give you an example of where this is coming into play where it can look like somebody is performing worse, who is performing better.
The 2011 NBA playoffs. I think it was the NBA Finals. Kevin Durant and Russell Westbrook are superstars. Kevin Durant is still a superstar. Suppose you look at how they performed in those finals. It has to be a particular game. The dataset is too small to be multiple games, so I think it’s one game. For those of you who don’t know basketball, you have two-point shots and three-point shots. If you make the two-point, you get two. If you make the three-point, you get three. Generally speaking, you make more two-point ones than the three-point ones because it’s an easier shot.
On two-point shooting, Kevin Durant made 57%, and Russell Westbrook made 48%. 57 to 48 – Kevin Durant wins. He shot much better than Russell Westbrook. Kevin Durant shot 17% on the three-point shots, which in general isn’t that great, but it’s a lot better than the 0% that Russell Westbrook shot. It’s not that he didn’t shoot any, but he missed all of the shots. So 17 to 0. Kevin Durant also shot better on three-point shots.
Now, this gets into sort of a mixed conversation. This is exhaustive. You only have two- and three-point shots. We’re going to ignore free throws and things like that. To illustrate, for those of you who say, “What about free throws, or technicals, or other kinds of things?” Two points and three points. That’s all of it.
More on percentages
Kevin Durant had a better percentage on two-point, had a better percentage on three-point. But combined, Kevin Durant had 45%, and Russell Westbrook had 46%. How is that even possible? Kevin Durant shot better on both of the categories, but overall he shot worse. How? This is what’s called Simpson’s paradox. We’re not getting in too much detail.
The basic concept is that you can’t look at the aggregate numbers because it will screw you up. You have to look at the individual segments. Otherwise, on its surface, Kevin Durant, we think, played worse, but he played better. He shot better in both of those categories.
Why does it matter to you? If you’ve got a more demanding workload from a revenue cycle management standpoint, it may make it look like you’re performing worse than someone you’re being compared to, even if you are performing better. It may be that you’re worse in some category.
Where does your organization thrive?
Like maybe, some organizations are good at worker’s comp, and others are not. Some are good at working Medicare and others are not. Some are good at working out-of-network and so on. So even if you are better in all categories, not just one of those but all sorts, you may still quantitatively perform worse on an aggregate percentage basis.
An example of this could be surgeries. You’re much better at getting surgeries paid than somebody else. You get a higher percentage of those produced. You may even get a higher percentage of the office visits paid. And on the aggregate, you perform worse. It’s what it seems like, even though in both of those categories, you perform better.
The net of this is, you can’t compare gross numbers. It would help if you segmented into the things that are important, that drive performance. Otherwise, it’s not an apples-to-apples comparison. It can make good billing organizations look like underperforming ones when, in fact, that’s not the case.
Understand how Simpson’s paradox works to make sure that your performance monitoring is not only accurate but representative of how you’re performing. Fun stuff! I love statistics stuff. So check out Simpson’s paradox!