Home  |   Subscribe  |   Resources  |   Reprints  |   Writers' Guidelines

February 4 , 2008

Build Your Own DRG
By Judy Sturgeon, CCS
For The Record
Vol. 20 No. 3 P. 6

To understand the diagnosis-related group (DRG) payment system, it is necessary to accept that it doesn’t make sense from a clinical standpoint. Some DRGs are affected by age, while others aren’t. Some are affected by surgical procedures performed during the encounter; others only change DRG for transplants and tracheostomies.

A newborn with numerous health problems may end up in a DRG that reads “Normal newborn,” while a two-day stay newborn may legitimately be assigned to the DRG claiming to be for term neonates with major problems. A pregnant patient can require one month in the hospital prior to a vaginal delivery and end up in a DRG that reads “Vaginal delivery without complicating diagnosis.” Age, gender, and discharge status may affect the final DRG—but, then again, they may have no effect on it at all.

A review of the grouper logic—and examples of how a DRG is determined—can help clarify some of the seemingly unreasonable assignments, such as those previously mentioned. In its fundamental makeup, this system is not based on severity of illness. While the new complication and comorbidity (CC) system offers allowances for greater and lesser severity in secondary diagnoses, the base DRGs are still essentially a reflection of reported charges and/or costs of patient types. The dollar amount to be paid for each is based on annual statistics for each group—hence the name.

Each DRG is assigned an average length of stay and a relative weight that helps calculate hospital payment. The key word is average—the entire system is designed to pay the provider an average amount of money for patients with similar types of diagnoses and procedures rather than a payment based on the charges or severity of illness for that specific patient. In theory, some cases will make money and some will lose money, but it should balance out over the year.

Determining the principal diagnosis—the reason, after study, that caused the patient to be admitted—is the first step in identifying the initial major diagnostic category (MDC) to which the admission will most likely be assigned. These categories are the basic building blocks of the DRG system and include 25 that are based on body systems or medical specialties, one for odd stuff (surgical procedures totally unrelated to the reason for admission and invalid or ungroupable cases) and a separate one for transplants and other overriding surgical scenarios. The MDC is generally divided into a surgical side and a medical side, but even that is not a hard-and-fast rule. Some MDCs don’t have a medical side (the weird one); others skip the surgical side (newborns).

Once the MDC has been tentatively established, procedures performed on the patient may move the encounter to the DRG’s surgical side. These procedures don’t necessarily have to take place in a major operative suite to qualify for a surgical DRG (eg, coronary stent, excisional debridement, endoscopic lung biopsy). Neither does the use of the operating room (OR) guarantee that a surgical DRG will be assigned (eg, some incisions and drainages, excisions, suturings, biopsies). Historical costs, charges, and comparative lengths of stay have statistically determined the surgical DRGs, not the procedure’s location.

Secondary conditions that have statistical or financial significance may change the DRG if past statistics indicate the need. The 2007 and earlier versions of the grouper were based on whether 75% of patients who had a condition stayed at least one day longer in the hospital than patients without that diagnosis. Version 25 for fiscal year 2008 has turned this concept on its ear by also considering if the problem rates as a 3 or 4 in the severity-adjusted all patient refined-DRG system and by further classifying them as simple CCs or major CCs (MCCs).

Sometimes, age changes the DRG classification for an admission. There were many DRGs reserved for neonates and patients aged 0 to 17 in fiscal year 2007 and earlier versions. There were even separate DRGs for patients with diabetes over or under the age of 35. However, the 2008 Medicare severity DRGs have eliminated age-specific DRGs with the exception of those specifically for neonates.

There are gender-specific DRGs for male and female reproductive systems. Even discharge disposition may move the patient to a different DRG. Neonatal DRGs depend on regular home discharge vs. death or transfer to another acute care facility. Cardiac DRGs may be affected by the death of a patient who had an acute myocardial infarction (MI) during the encounter.

Last, but certainly not least, there is a hierarchy that must be followed when more than one circumstance competes for the DRG title. Within each MDC, there are diagnosis and procedure codes that override others. MDCs themselves are subject to hierarchy: There are so-called pre-MDCs that pull rank on all others and include cases involving transplant, tracheostomy, extracorporeal membrane oxygenation, or heart-assist system implants. There are even DRGs that can span multiple MDCs.

With a basic understanding of the grouper logic, it is possible to figure out the correct DRG for an admission without using grouper software. If a patient is admitted for workup of chest pain, no cause is found, and the patient is sent home without further ado, the DRG will be “Chest pain” from the medical side of the MDC for circulatory system disorders.

If the chest pain is found to be due to an acute MI and the patient survives without significant secondary problems, the DRG will change to “Acute MI, discharged alive, without CC/MCC” but remain on the medical side of the same MDC. Add a diagnosis of acidosis and move to the DRG version with CC. Add pneumonia or cardiogenic shock and move up even further to the version with MCC. Much like rock breaks scissors in the child’s game of hierarchy, one MCC overrides all simple CCs for DRG assignment. Should this patient die during the encounter but prior to any significant surgical intervention, the DRG will assign instead to “Acute MI, expired, with MCC.”

If the patient survives, a diagnostic cardiac catheterization will keep the admission on the medical side of the DRG fence but add a coronary angioplasty and the DRG changes to the surgical side. Surgical hierarchies will struggle to claim the patient who has additional procedures such as drug-eluting stents, pacemaker or defibrillator implants, and coronary bypass or heart-assist device implants. While these procedures and surgeries will change the patient’s DRG, they still remain in the circulatory system MDC.

Consider another MI patient who is admitted and stabilized, found to have a suspicious spot on the lung during a preprocedure chest x-ray, and undergoes a wedge resection of a pulmonary malignancy. This case then escapes the circulatory system MDC and is assigned to “Extensive OR procedure unrelated to principal diagnosis, with CC” in the “DRGs associated with all MDCs.” If the patient subsequently has cardiac failure and receives a heart transplant prior to discharge, the case jumps MDCs one more time to the DRG trio “Heart transplant: with MCC, CC, or neither” in the pre-MDCs.

The critical fact to remember when assigning or validating a DRG is that the codes determine the DRG, not the other way around. It makes no difference if the DRG description sounds completely wrong for the patient’s condition or if the length of stay is significantly out of range of the norm. The coder or auditor may not arbitrarily choose the DRG they like; the complex grouper logic must be followed, and the obvious complexity makes its software a most welcome tool for the task. If documentation to support the diagnoses and procedures is present and the codes are assigned correctly by coding guidelines, the DRG assigned by the grouper logic is also correct.

— Judy Sturgeon, CCS, is the hospital coding senior manager at The University of Texas Medical Branch in Galveston and a contributing editor at For The Record. While her initial education was in medical technology, she has been in hospital coding and appeal management for nearly 20 years.