Thursday, January 6, 2011

What the Heck is a DRG? And Why Should I Care About Case Mix?

So you want to be a coder. And not just that, you want to be a hospital coder because, on average, they make more money than physician coders. And you don't just want to be a hospital coder, you want to be an inpatient hospital coder because then you get to look at the whole chart and piece together the patient's clinical picture. If this is your goal, then everything you need to know you will not learn in school. And that's mainly because there is so much to learn and practical experience is key.

Most of all, if you want to be an inpatient coder, you need to know diagnosis-related groups (DRGs) because in hospitals, it's all about DRGs and case mix - and compliance. If you have no idea what I'm talking about, fear not - here's a primer on DRGs! I wish I could say I cover it all here, but this is just a beginning!

What is a DRG?
The ICD-9-CM coding system contains about 16,000 diagnosis codes and ICD-10-CM contains over 68,000 codes. Imagine trying to determine a payment amount for each individual condition. And that doesn't include accounting for procedures. The most logical solution is to create a system that allows for broader classification of conditions and services for easier comparison and assignment into payment categories. DRGs were created for this purpose. I look at DRGs as a way to "organize the junk drawer" where patients are grouped into different categories based on similar conditions and cost to treat the patient.

History
DRGs were first developed at Yale University in 1975 for the purpose of grouping together patients with similar treatments and conditions for comparative studies. On October 1, 1983, DRGs were adopted by Medicare as a basis of payment for inpatient hospital services in order to attempt to control hospital costs. Since then, the original DRG system has been changed and advanced by various companies and agencies and represents a rather generic term. These days, we have various DRG systems in use - some proprietary and some a matter of public record - all of which group patients in different ways. Two of the main DRG systems currently in use are the Medicare Severity DRG (MS-DRGs) and 3M's All Patient Refined DRGs (APR-DRGs). Different DRG systems are used by different payers.

How to Get a DRG
All DRG systems are a little different, but the basic premise is the same. DRGs are based on codes. In effect, DRGs are codes made up of codes. The following elements are taken into consideration when grouping a DRG:
  • ICD-9-CM diagnosis codes
  • ICD-9-CM procedure codes
  • Discharge disposition
  • Patient gender
  • Patient age
  • Coding definitions as defined by the Uniform Hospital Discharge Data Set (UHDDS) - in other words, the sequence of codes on the claim
Back in the 80s, DRGs were grouped manually using decision trees. These days, DRGs are grouped with the touch of a button and DRG groupers are a big part of encoding software. But I would be doing you a disservice if I didn't at least give you an idea of the grouper logic. As I mentioned, there are different DRG systems and probably the most popular is the MS-DRG system, so I will explain how MS-DRG grouper logic works.

MS-DRG Grouper Logic
The first step in assigning an MS-DRG is to classify the case into one of the 25 major diagnostic categories (MDC). These MDCs are based on the principal (first) diagnosis and, with a few exceptions, are based on body systems, such as the female reproductive system. Five MDCs are not based on body systems (injuries, poison and toxic effect of drugs; burns; factors influencing health status (V codes); multiple significant trauma; and human immunodeficiency virus infection). Organ transplant cases are not assigned to MDCs, but are immediately classified based on procedure, rather than diagnosis. These are called pre-MDC DRGs.

Once a case has been assigned into an MDC (with the exception of the transplant pre-MDCs), it is determined to be either medical or surgical. Surgical cases require more resource consumption (that's industry speak for "costs more!"), so they must be separated from the medical cases. If there are no procedure codes on the case (e.g., a patient with pneumonia may have no procedure codes), then it's simple - it's a medical case. But if the patient had a procedure, that procedure may or may not be considered surgical. For example, an appendectomy is quite clearly a surgical procedure. But something like suturing a laceration is not. It's all based on resource consumption - the cost of performing the procedure. In general, anything requiring an operating room is surgical.

Quick sidebar here - this is why skin debridement is such a hot topic in the world of coding compliance. Nonexcisional debridement (code 86.28) groups as a medical case. However, excisional debridement (code 86.22) groups as a surgical case and the change in reimbursement is rather drastic.

Okay, so now that we have our MDC and a designation as medical or surgical, we need to look at the other diagnoses on the claim. Right now, Medicare is able to process the first 9 diagnoses on the claim (even though 18 are reportable). These other diagnoses, depending on their severity, may be designated as complications and comorbidities (CCs) or major complications and comorbidities (MCCs). Medicare maintains lists of CCs and MCCs and updates them annually. CCs and MCCs are conditions that have been identified as significantly impacting hospital costs for treating patient with those conditions. For example, it's been determined that congestive heart failure without further specification does not significantly impact costs and it is not a CC/MCC. However, patients with chronic systolic or diastolic heart failure do have slightly higher costs, so those conditions are CCs. More so, patients with acute systolic or diastolic heart failure have even higher costs, so they are designated as MCCs. Are you beginning to see how slight changes in a physician's diagnostic statement impact coding and thus payment?

DRG Weights
Okay, so we know the MDC, whether the case is medical or surgical, and whether or not there are any CCs or MCCs. How does that translate into reimbursement? Well, if you're using an encoder (and if you code for a hospital, you will), you hit a button and presto! You have a DRG with a relative weight. Now if only you knew what that relative weight meant. The DRG relative weight is the average amount of resources it takes to treat a patient in that DRG. Huh?

Let me demonstrate. The baseline relative weight is 1 and represents average resource consumption for all patients. Anything less than 1 uses less than average resources. Anything above 1 uses more than average resources. So let's compare some respiratory MS-DRGs:
  • MS-DRG for lung transplant has a relative weight of 9.3350
  • MS-DRG for simple pneumonia (no CC/MCC) has a relative weight of 0.7096
  • MS-DRG for chronic obstructive pulmonary disease with an MCC has a weight of 1.1924
You can see how different combinations of codes lead to different MS-DRGs with different relative weights. In order to convert that into monetary terms, we multiply the relative weight by the hospital base rate. Now I'm sure you want to know how to get that hospital base rate. Me too. Well, up to a point. The base rate is exclusive to each hospital and takes a lot of historical, facility-specific data into account, like what they've been paid in the past, whether or not they are an urban or rural hospital, and how much the hospital pays out in wages. That's just more math than my poor little head can comprehend! So for the purposes of this exercise, let's pretend like this hospital - we'll call it Happyville Hospital - has a base rate of $5000. So if we multiply the relative weights above by $5000, our reimbursement for those cases, respectively, is $46,675, $3,548, and $5,962.

Case Mix
You just might be asked in an interview if you understand case mix. It's a good indication of whether or not someone really understands DRGs. And I have to admit, in my sometimes sadistic manner, I like seeing that look of glazed-over confusion on someone's face when I bring up case mix. But case mix is simple. It's the average relative weight for a hospital. So get out a big piece of paper for your hospital and start writing down the relative weights for every single case and then divide to get your average. Okay, so it's computerized now. But that's all case mix is - an average.

In the industry, we officially refer to case mix as the type of patients a hospital treats. Let's say at Happyville, we have a high volume of transplant cases plus a trauma center and a well-renowned cardiac program. These are all highly weighted types of cases and our overall case mix will be higher than say, Anytown Hospital down the street that has no trauma center, no transplant program, and basic cardiac services (they transfer all their serious cardiac cases to Happyville!). Happyville's case mix will be higher than Anytown's.

As a coder, you don't need to know what your specific hospital's case mix is at any given time. But knowing what impacts case mix is an indication that you know your stuff. First and foremost, case mix fluctuates. Most hospitals monitor case mix on a monthly basis because changes in case mix are a precursor to changes in reimbursement. Of course your CFO wants case mix to continue to rise, but that could be a red flag. And he certainly doesn't want case mix to fall. If case mix begins to decrease, the first place hospital administration usually looks is coding - after all, case mix is based on DRGs, which are based on codes. But there are lots of things that can impact case mix and many of them have nothing to do with coding, such as:
  • The addition or removal of a heavy admitting physician - especially specialty surgeons
  • Opening or closing a specialty unit
  • Changes in a facility's trauma level designation
  • Movement of cases from the inpatient setting to outpatient, and
  • Anything else that impacts the type of services the hospital provides
Your Life as an Inpatient Coder
As an inpatient coder your job is to make sure you get all the codes on the claim in the correct order so that the accurate DRG is assigned and the hospital gets paid appropriately. When I put it that way, it sounds so easy! The reality is, with more and more patients being treated as outpatients, those who are admitted as inpatients are sicker than they've ever been. And sicker means harder to code. For instance, the patient comes in with shortness of breath and the final diagnosis is acute exacerbation of COPD, stapholococcal pneumonia, and respiratory failure. How you code and sequence the case will determine the appropriate DRG and reimbursement. The good news is, you'll have an encoder to help you model the DRGs and see what pays what. The bad news is, you have to paw through the medical record to determine the true underlying cause of that shortness of breath.

So are you ready for the challenge? Are you ready to apply DRGs?

7 comments:

  1. Good one, explained a great deal for a beginner in DRG Coding person like me..

    ReplyDelete
  2. Awesome! I am a physician and was never taught this stuff in training. Now, as a hospitalist, I realize how much it really matters. Thanks for the great explanations and examples.

    ReplyDelete
  3. Where/How do you get these relative weights manually (without a computer program)?

    ReplyDelete
    Replies
    1. Chelsey - the relative weights are published each year in the final rule for IPPS. The final rule is a large document, but each MS-DRG is printed with relative weights in Table 5. Grouping is not discussed in these files. You can manually group with a DRG Definitions Manual, but pretty much everything is computerized now. I haven't seen a definitions manual in about 15 years! You can access the federal files at CMS' website. Here is the link to the Acute Inpatient PPS page - http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/index.html. From there, select the year you are looking for (right now we are in fiscal year (FY) 2012 and then select the file containing tables from the list. Relative weight is listed in the table as RW. Good luck!

      Delete
  4. Hi Kristi, I am working towards a certificate in coding and was researching some info. on DRG's, and came across your blog. I just wanted to say Thanks! and well done!....you explained in such a clear way(this is all new to me).
    I look forward to reading the rest of your blog and learning from you.

    ReplyDelete
  5. Great article. I found it very helpful. I have a question about the way Medicare and Medicaid determines payment. In reimbursment, is there a way to know what Medicare/Medicaid will pay at a given hospital for a patient who falls into a certain DRG? In other words, if a 70 yr old has an MI and stent in New Jersey, will that stay and procedure cost the same as for a patient (same age and gender) in California? I know it must be different, but what is it based on exactly? Thanks, Gina

    ReplyDelete
    Replies
    1. Hi Regina: I can't speak to Medicaid because those are state-run programs and every state sets their Medicaid programs differently. For Medicare, the DRG weight is multiplied by a hospital weight and the resulting figure is the reimbursement. The hospital weight is based on the hospital's case mix and wage index (how much they pay their employees) and is calculated using a very complex equation. If you grouped the same codes for a hospital in New York and California, you would get the same DRG and same DRG weight, but the hospital weight would be different and would be based on local factors. We call that the adjusted reimbursement rate and it's the reason different hospitals aren't paid the same for similar services. It wouldn't make sense for a hospital in rural Arkansas without a large case mix to be paid the same as an urban NYC hospital with complex cases like trauma or open heart procedures.

      Delete