Hospital Discharge Coding

My apologies for the long absence.  I've been ridiculously busy getting ready for next week and have been on the road too much.

The real reason I haven't written is that I've had a bizarre attack of self-conciousness...normally, this blog is a purely solipsistic exercise and I do not think much about people actually reading it.  However, three times in the past two weeks I have had people on the phone or in person suddenly say, "Wait a second!  You're that guy who posts on the coding lists/MGMA lists/PedTalk/your blog!"  Fortunately, they've warmed up considerably once they realize they "know" me, which is quite flattering...or, they're awfully polite.  Probably the latter.

My lack of input isn't due to a lack of content.  For example, let's explore everyone's favorite topic, hospital discharge coding.  OK, 1/2 of my readers just left.  For those of you who stayed, Q asked me the following back in June:

The June 2009 AAP Coding Newsletter brings up a good question about
whether practices are billing for hospital discharge services properly.
They're losing money if they report Subsequent Hospital Care of a Normal
NB (99462) when they should bill Discharge Day Management (99238 or 99239).

How're our clients doing?  Total Hospital Visits vs total number of
99238's and 9's combined should be a fair gauge,  yeah?

Yeah, I think so.  If I follow the logic properly, there should be at least one Discharge Day Management for each kid who goes into the hospital.  You might rack up a couple Subsequent Days, of course, for a multi-day stay, but it's nice to know what the ratio is.

Why is this important?  A 99462 is worth ~.83 RVUs in 2009 while the 99238/9 are worth 1.83 and 2.65 respectively.  In other words, somewhere between 2-3x as much.

Igor went to work (I love it when everyone else does the shoveling) and came back with some interesting preliminary info. We took a big sample of PCC clients and counted up their hospital visits (9943X and 9946X) and counted up their discharges (99238/9) and sorted them.  I subjectively grouped the clients into thirds based on hospital volume and discovered this:

Number of Hospital

Visits (Annual)

 Percentage of

Discharge Codes

 0-75 (Avg: 40)  76%
 76-161 (Avg: 114)  73%
 161-1100 (Avg: 298)  64%

So...those who do fewer hospital visits are better at coding discharges?  No so fast, I suspect.  I'll bet that many of those smaller groups, by virtue of doing far fewer hospital visits, are less likely to have those kids with long stays?  I don't know, I have to work with Igor some more.

What was REALLY interesting, though, was the variance in discharge use.  We clearly have some clients whose coding we need to go fix (a job for Q!) practice, for example, had over 650 hospital visits and only 19 discharge codes.  Another practice had 21 hospital visits (solo doc)...and one discharge.  Meanwhile, more than 10% of our clients had more discharges than total visits!  How does that happen?  Anyone?

I can look at revenue effects if anyone is interested.  Q: let's make some client phone calls and get these practices some well earned money!

Off to CCHIT on Wednesday.