Trending Topics

Billing/Coding for Hospital Medicine

Michelle L. Booth-Kowalczyk, APRN, C.N.P. discusses the importance of accurate documentation and the impacts of inaccurate documentation, including defining the components of documentation that make up a DRG and what the DRG determines, and more.
March 25, 2022
Michelle L. Booth-Kowalczyk, APRN, C.N.P.


I get asked a lot of things like, “Why do you know this stuff,” and “Why do you care?” Because, you know, we don't really get taught about it in school much at all. And it started back when I worked in hospital medicine. 

I was a quality chair. And for the direction of healthcare, we're really looking at the length of stay readmission rates, bundled payments now even. And I feel like, as one provider in a very large institution, I don't have great ability to have a direct effect on someone's length of stay. I don't necessarily get to make that directly impact all the time, but documentation is something that I personally have control over. And it can make a huge impact on the patient's case. 

I delved into that to try and improve the documentation on hospital medicine. And then cardiology is now benefiting from that. I say benefiting no disclosures. No one will pay me to talk about this stuff, but I'm open to it if anyone is looking. 


We're gonna go through just basically why documentation is helpful and necessary and the different components that make up the DRGs, and what that DRG really determines. And then we'll briefly talk about billable components as well. 

So, I want to just get an idea of the audience of what type of charting systems you use. Mayo, about a year and a half ago, transitioned into Epix, and it has dramatically changed our documentation. And so it made me realize how important that is. 


So, there are two main parts of the documentation. 1. capturing how sick your patients are, and 2. support level of billing of note. The functional purpose of note as well: express to others your thought and patient trajectory of care, and justify your medical decisions. And then there are the coders, who translate all of our information into codes. I recognize not everybody has this luxury. Kudos to you if you're doing your own coding. 

Some basics here: so, there are thousands and thousands and thousands of diagnoses that describe our patients. There are a select few of them that have been chosen to be given some extra weight in their classified into either MCCs or CCs. They can be present at any time during the hospitalization and technically only need to be captured once. It always helps things be more legitimate when they're captured more. 


Medicare determines the DRGs based on the MCCs and CCs, so typically it's a three-tiered system. There are a couple of DRGs that are just two-tiered. You either have CC or MCC, or you don't. But the majority of them are, you know, the DRG with the addition of an MCC or addition of a CC or none. You only need one to be able to capture that DRG That by no means means that you shouldn't be capturing all the other things that the patient has going on. And then there's the severity of illness and risk of mortality. And these are Vizient factors. 


I'm not sure if Vizient used to be UHC. It's how we benchmark against other institutions. It takes a DRG, and it overlays a bunch of predetermined factors for that DRG to determine what, for that patient, the risk of mortality or severity of illness is something that you typically want to optimize. And it really looks at how sick the patient is for expected mortality. It looks at how sick the patient is when they come through the door. This is why admission notes are so important. 

And you really want to capture all those co-morbid things on admission, and why sometimes surgery loses out when they don't write admission notes because they aren't capturing those things. If you're a rock star and you pick up a patient five days in and you're like, “They'd never documented the cardiogenic shock,” or, something huge, you can do that and say present on admission. 

And as long as there are indicators that support it, you know, it'll be captured that way. 


Chronic problems don't need to be. These are some common MCCs and CCs in practice. The bolded italics terms are what classify them into those different categories. Acute is a big one, and it often seems redundant. I do a lot of education with our residents. When I'm working duo, a super common one I see is, you know, a patient who's coming in with a GI bleed and they'll have documented GI hemorrhage, but not acute blood loss anemia. 

And so I'll tell them that. And they look at me like I have three heads. Of course, they've got acute blood loss anemia. They've hemorrhaged. But it has to be explicitly stated to be captured. 

Severe malnutrition. That's a big one. And a high audit one, we'll talk a little bit more about it. But you want to make sure you have the right documentation to support it. So for those who use a WDWNd physical exam, you know, I'd recommend taking out the well, well-nourished because sometimes that gets left in. And then it clashes with a severe malnutrition diagnosis. 

Just a couple more examples. For example, if someone if someone's coming in altered, really try to classify that into delirium or they might be encephalopathic in some way, altered mental status does not code to what you want it to code. It does not give you the severity that you're looking for. And then like flash pulmonary edema. 

That's something that translates to you and me, but not in the coding world. The coding world is very black and white. So, you have to say the word acute. These are some just general top misdiagnoses. This isn't based on any information other than the probably thousands of chart reviews I've done between hospital medicine and cardiology and then working DUO and seeing charting across our institution. 


Acute heart failure is universally missed through the entire institution because of the way it needs to be specified. Also commonly missed diagnoses, acute blood loss, acute anemia.

Comorbidities (electrolyte imbalance, fluid status, ventilation) are combined with primary diagnosis using a calculator for billing. Saying medical comorbidities is a good way to be able to still capture from your HPI. And then another thing to remember that, you know, it's just because someone's normotensive and they had hypertension and they're on three agents. They don't have a history of hypertension, they have controlled hypertension.  We need to make it clear in our documentation what is controlled vs uncontrolled comorbidities.


Ventilated on admission day, as you can imagine, is a big risk of mortality. Someone who is ventilated on admission is sicker than someone who's not. And some of them are broad categories. So, I'll show you late how some of them fall into different categories. 

This is how when we do our calculators, this is what would come up for heart failure. And these are the three DRGs. Again, there's the 291 that's with an MCC. And then it goes to with CC and then with neither. And you can see the different levels. And we have like these calculators where you put in the information and choose the different indicators and then it gives you a seizure with like doing a bunch of calculations. Then it spits out numbers, but it gives weight to all the different factors. 

This is an example of what types of diagnoses fall within those factors, so like fluid and electrolyte abnormalities, I'd never want you to put that as a diagnosis. But it's a factor in these are all the things that go into that factor. They are not all CCs. That's also confusing. That's how the Vizient works, though. 

Make sure to document the source of sepsis, and CMS and documentation use different codes for sepsis (have to see slides to see what they are). 


So, there's obviously a large gap between coding language and our clinical language and how we speak. I feel like I'm banging my head against the wall sometimes when I'm in these meetings. Like, what do you mean that AFib RVR doesn't translate like it? It doesn't. 

You have to have like AFib with tachycardia. RVR doesn't mean anything in the coding world. There are all kinds of ICD-10 codes out there. Make sure to document the source of sepsis; CMS and documentation use different codes for sepsis.

The SIRS with a suspected infectious source. I mean, the practice has gone typically to adopting the Sepsis-3 guidelines and definition, which is very different than the CMS documentation version. So, CMS still does sepsis, severe sepsis, septic shock, and then in Sepsis-3, that's not how it translates. There is chatter. Nothing is confirmed. But there is chatter coming that CMS will be adopting the Sepsis-3, which is going to have huge impacts across institutions. Nothing is official yet. But I would suspect that most institutions are trying to prepare for how they're going to address that because it's going to be, you know, a huge difference from what we've learned for documentation for a long time and loss of revenue. I'll just say that. But some special highlights – again, heart failure is a big one. 

I think the biggest thing with heart failure is you have to remember to document three things. You have to document the chronicity of it. Is it acute, chronic, or acute on chronic? Is it systolic diastolic or is it combined? And then you have to use the word failure if, in fact, that's what it is. 

It's more generous than I would have thought, especially working in the hospital, but endorsed by our circulatory failure division chair. And so really try to think about this when you're seeing patients. It gives a lot more latitude for when patients are considered acute. 

  • AFib. This was just released on October 1st, and I'm still kind of sorting through it and have been asked my opinion a little bit. But the gist of it is they're looking at reclassifying AFib, basically, based on how you're treating it. 

So, if you've got someone who has chronic AFib, who is just on rate-controlling meds, and there's no plan for cardioversion, and you're not actively pursuing converting them to normal sinus, then that would be permanent. They would be looking at a permanent code there. 

  • If you've got someone who's on antiarrhythmic therapy, then they're wanting to see more of the persistent, which is different than how I would have thought of it. So, I think I had always documented, you know, paroxysmal AFib when someone maybe, you know, had a prior history of AFib and now they're on an antiarrhythmic and they’re normal sinus. And so they're pushing for that to be a little bit different. 

For clarity, I would probably recommend stating something like persistent atrial fibrillation on a name your antiarrhythmic I can't endorse anything. Currently, a normal sinus rhythm. So, to really paint the picture of who that patient is. The good news is that everything other than unspecified AFib, I believe, is now a CC, so, that's huge because everybody has AFib. That's good. 

  • Respiratory failure. And this is also another, you know, high audit one. You want to make sure that your documentation supports the diagnosis. When you're documenting acute hypoxic respiratory failure on admission and your physical exam says that they're resting comfortably in bed and that their respirations are even and nonlabored that that that's conflict and it puts you at risk for, you know, compliance audit type things. So really trying to paint the picture of what you're seeing and what they're struggling with and what you're thinking they're feeling from. 
  • Chronic hypoxic respiratory failure that's just a kind of low-hanging fruit. So there are clearly there are ABG guidelines as well that I sadly don't have in my brain at the moment. But so many of our patients, they're not necessarily coming in with an acute flare, but they have severe COPD or whatever that case may be, and they use home oxygen. 

So, they know we didn't get an ABG because they're not struggling in that way. But just know if they use home oxygen or BiPAP, no CPAP. If they have CPAP with an O2 bleed in, then that counts. But if they use BiPAP or home O2, then that's by definition chronic hypoxic respiratory failure. And it's an MCC, so you want to capture it. 

  • Pneumonia.
  • You know, Dr. Kashiwagi mentioned the age cap has kind of lost its edge. And in the coding world, it codes exactly to community-acquired pneumonia, which is a completely different DRG. So it's not even like a different. The level of DRG from respiratory infection, it's a completely separate DRG, and so when you can be specific about pneumonia, you want to be. 

This is just an example of how the different codes line up. You can see the difference in length, expected length of stay, and reimbursement. This is an exaggeration, but I'd seen it of how you could potentially paint two different patients and a huge difference. Right. And the reimbursement and the length of stay by being more specific. So we're going to practice here. 

  • EXAMPLE: So, this is Connie. She's an 87-year-old, who comes from an SNF concern for sepsis. She was febrile and hypotensive and all those things. You can see her medical comorbidity is no past medical history. Very typical patient for most of us. You can see she got some leukocytosis and sodium is a little low, creatinine is up, her lactate is, by our standards, normal, but by sepsis world not. 

They do some appropriate things. Antibiotics are probably not appropriate. I should change. Those guidelines have changed. So, that's her. And so you go to write your diagnosis list showing a fair amount of oxygen coming to a non ICU. 


1. Make sure your diagnoses match your clinical care and vice versa and 2. it's more than just the coding, after a while due to coding templates every patient starts to look the same and have the same diagnoses listed, it's important to add some granularity to better differentiate patients clinically and take care of them


Billing is pretty straightforward, specifically inpatient billing. I say that because we have billing people who help us, but there are progress notes, your admission notes, consult notes, discharge codes. These are for in-hospital patients only. 


And then critical care codes. I think the most important part is to understand the different components of a note and how you support the billing level that you are placing. I'll say for HPI typically on admission notes they’re not a problem because we go into all kinds of detail, you know when the patients are rambling on but know on like progress notes that you want to prove that you talked to the patient. 

So, saying no acute events overnight, is not the best practice. We're typically asking how they're feeling and following up on whatever symptoms they had been having. A couple of lines about what you guys talked about and how they're feeling.

In cardiology, we often comment on vessels. You know, we comment on their JVP, we comment on extremities, and common on their edema. And all of that is cardiovascular on old school dictations of physical exams. That doesn't mean that if they have a traumatic head, you shouldn't say that. But just knowing being aware of what falls under what and then for review systems, it is completely kosher to use a blanket statement if you're afraid that you are not going to hit 10 separate systems. 

Typically, I think most providers and their HPI talk about pertinent positives and negatives but don't necessarily say 10 systems worth of negative. You can say this is an OK thing to do. For data, you just want to make sure you're giving yourself credit like we take care of really sick people. So, when you review an EKG tracing, instead of just pulling the results in, it adds a lot more benefit to everybody.


I would say that the best practice is - you can support your billing codes just on decent documentation for the majority of the codes. If you're using prolonged visit codes, SH-4’s IH5, IH6, then you need to specify the time that you're spending. The NPPA and the physician time shouldn't be combined. 

So, if I bill on time, they don't even look at the physician billing. Some other institutions where their coders can separate it. They'll allow you to combine it, but it can't be overlapped. So, you and your consultant can't go into a room, spend 30 minutes in there, and then combine that time into 60 if that makes sense. 


There are phrases/shorthands, things that you can use. Discharge codes. You want to make sure you're billing on your expiration summaries. Those sometimes are more work than regular summaries. Technically, your discharge code can be done on the day before discharge, if that's where all of the work happened. And then they're leaving early morning the next day. 

I would say, in the hospital, a lot of dismissals are DDMs, Ms. The amount of coordination of care to get that person discharged go over their changes and complete all the paperwork, as is typically a fair amount of workload. Critical care codes. 

So if you work in an ICU, you use these all the time, but on general care and PCU units, you can use them as well. They are meant for more critical care. If you use Epic, you can bill your addendum, if you've signed your note and already billed your note for the day. And so, it documents it nicely with time and such. 


You just want to make sure you're again justifying that code. So, I was called urgently to the bedside because of X, I ordered X, Y, Z stat. This was my exam. This is my differential. This is what I think's going on. You have to kind of show the workload there. And it's cumulative time. 

So I work a fair amount of nights, and I may see a patient and have about fifteen minutes of contact for something acute. Inevitably, I'm seeing them again later. And so then, when I do that second addendum, if it equals 30 minutes, then I'll bill it at that. 


So, post-test, your documentation of your patient's diagnoses directly affects expected mortality rates. True or false? 

The answer is TRUE.. 

So, the MS-DRG affects which of the following?

The answer is E. A and B. These are people's favorite. Where there's like to choose between a bunch of different but overall reimbursement, expected length of stay, the severity of illness. All. three of those just A and B are just A and C. 

So confusing the reality, it does reimbursement, expected length of stay, the severity of illness, and risk of mortality that comes from some of those factors that I was telling you about, which don't necessarily get played in two specifically into the DRG. All still very important to capture, so it doesn't actually matter, you just write them all and then the right people capture him in the way that they should.

Watch Video


GIBLIB is the premier streaming service with a vast catalog of unrivaled continuing medical education content covering surgical videos, medical lectures, courses, and transcripts. Learn from the world’s best doctors, from the top leading institutions, including Mayo Clinic, UCSF, Cedars-Sinai, Keck Medicine of USC, and many more.