This article was initially available on our podcast. Click here to listen.
Grouping together providers for analytics can sometimes be challenging. There’s a key that everybody typically uses to link records. That key is visit number or encounter ID or whatever your system uses for that name. That’s not an issue for connecting the records. In other words, connecting charges, remittances, payments, and all these things can use that key. So we’re not talking about whether or not we can link records together. This is a little bit of a different issue where when someone is trying to do reporting, 835s can sometimes screw you up a little bit.
Analyze the 835s
If it’s a single provider, meaning it’s a single practice, even if 50 physicians are in the course with one tax ID, all the 835s come into one place in the system. So even if the 835s have some mistakes, they’re all received, go into that one system, loaded, and could be analyzed. It doesn’t matter if all of the information is correct in those records because they all drop into one centralized place, and you know they’re all related to that one practice or entity.
However, this can be not easy if a billing company wants to analyze. The reason is that the payee name (or the provider) is not always the same in the 835s. Payers frequently screw it up. There are a lot of different iterations of this.
Review the variations
For example, we might see seven different names used by insurance companies when they send in the 835s. That means if you’re trying to look at some analysis on a practice level, you want to see how “Cardiothoracic Surgeons of Omaha” are performing, you’re going to see seven different practices, all with some variation of that. If it happens to include something like “associates,” like “Cardiothoracic Associates of New York,” every once in a while, they truncate “associates” to “assoc.” or even “ass.”. So you end up with a cardiothoracic ass. I’m not kidding. We see this a lot. You got to find humor in some of this stuff, too, to make it fun.
The problem with that, of course, is if you have a lot of different names for the practice that’s coming in and you’re trying to group and look at the performance of the individual practice, it gets hard to look at that because it breaks into too many different chunks. You can’t analyze it all together.
Practices and sub-practices
The alternative is to use something like the NPI or some ID that groups them but who’s memorized all the ten-digit numbers for all of these practices. That’s completely bonkers.
Suppose you’re a hospital system, for example. In that case, where you have multiple practices or groups or something like that that all roll up underneath that system, they’ll get split up into different practices and sub-practices, and you can’t see the performance of the entire practice. You won’t have data at a practice level, and you won’t be able to do the analysis you’re looking to do.
There is a solution to this. If in the 835, they’re screwing up the practice name or the entity name, which happens a lot, there’s going to be a lot of different iterations of that. You can use the ID (typically, the type II NPI, the practice level NPI) to group them and then essentially do a reverse lookup backward.
Check by practice name
What I mean by that is, if you have a lookup table where you have the tax ID or the type II NPI for practice, and that’s, again, cardiology (“Associates of Miami” or something), that one type II NPI with that single name, you can then create a column in your data that overwrites or adds a column that then you don’t have all those different variations of the practice name. You have “Cardiology Associates of Miami.” That way, when you run some analysis, and you’re trying to group by the practice name, and you want to see analysis at that practice level, you can use that column with the correct practice name in there to run your analysis.
Easy solution. It just requires overwriting. We generally prefer non-destructive solutions, so adding a column that is called “Normalized practice name” or “Practice name,” something that tells you, “Hey, this is our grouped practice name that we’re going to use.” And then, you run all the analysis off of that field.