HEALTHCARE COST & UTILIZATION PROJECT

User Support

Do Your own analysis
Explore Expert Research & Limited Datasets

NIS Description of Data Elements

 
DRG - DRG in use on discharge date
 
Documentation Sections:
General Notes
Uniform Values
State Specific Notes
General Notes
 

The Diagnosis Related Group (DRG) appropriate for the date of discharge is assigned by the Medicare DRG Grouper algorithm during HCUP processing.

Changes in DRG assignment

With Version 33 of the DRGs (effective October 2015), CMS revised the DRG software to use International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) diagnoses and procedures. Prior to that the DRG software was based on International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes.

With Version 25 of the DRGs (effective October 2007), CMS revised the DRG scheme (up to 540 DRGs) into over 700 Medicare-severity DRGs (MS-DRGs). First, DRGs were consolidated into 335 base MS-DRGs. Of these, 106 were split into two subgroups, and 152 were split into three subgroups, to arrive at 745 total MS-DRGs. The subgroups were based on the presence of complications or comorbidities (CCs) or major CCs (MCCs).

Diagnosis and Procedures Used for DRG Assignment

Beginning in the 1996 data, the DRG grouper can handle a maximum of 50 diagnosis and 50 procedure codes. Only diagnoses and procedures that are valid on the date of discharge are used by the grouper for DRG assignment.

In the 1988-1995 data, the DRG grouper cannot handle more than 15 diagnoses and 15 procedures. Therefore, the following rules were used when more than 15 diagnoses or 15 procedures were available:

  • the principal diagnosis/procedure (regardless of validity) is retained in DX1/PR1. No secondaries are shifted into the principal position.
  • the first 14 valid (by HCUP standards) additional diagnosis or procedure codes are passed to the HCFA DRG grouper.

Different Definitions of Diagnosis and Procedure Validity

HCUP validation of diagnosis and procedure codes allows a window of time around the official ICD-10-CM/PCS coding changes (usually October 1), for anticipation of or lags in response to official ICD-10-CM/PCS coding changes. Prior to October 1, 2015, the ICD-9-CM coding system was used and the same timeframe was followed. During the 1988-1997 HCUP data processing, a six-month window (three months before and three months after) was allowed. Beginning in the 1998 data, a year window (six months before and six months after) was allowed.

The DRG Grouper rules differ in two ways:

  • diagnosis and procedure codes must be valid on the date of discharge to be used for assigning the DRG; and
  • some valid diagnoses (E-codes) are ruled by the DRG Grouper to be invalid if entered as a principal diagnosis.

This inconsistency between the definition of a valid diagnosis or procedure is obvious when a discharge has a valid principal diagnosis under HCUP standards, but the assigned DRG is 470 "Ungroupable." Consider a discharge with DX1="V300" on October 1, 1989. The diagnosis code "V300" is considered valid by HCUP standards because until September 30, 1989 "V300" is a valid ICD-9-CM code. The DRG Grouper does not recognize the "V300" code on October 1, 1989 and therefore groups the record to "Ungroupable," DRG=470 and MDC=0.

Changes in DRG Grouper Logic

Until the eighth DRG version (before October 1, 1990), the first step in the determination of the DRG had been the assignment of the appropriate MDC based on the principal diagnosis. Beginning in October 1990, there are two types of exceptions:

  • The principal diagnosis is not the initial data element in DRG assignment when the initial step in DRG assignment is based on a procedure. If a patient has a liver transplant (DRG 480), a bone marrow transplant (DRG 481) or tracheostomy (DRG 482 and 483), then the patient is assigned to these DRGs independent of the MDC assigned from the principal diagnosis.
  • Assignment to MDC 24 (multiple trauma) and MDC 25 (patients with HIV infection) is based on BOTH principal diagnosis and procedure.

The Need for a Valid Discharge Date

The DRG grouper needs a valid discharge date because DRG versions change at specific points in time. If the discharge date was invalid or not available from a data source, a temporary discharge date (for use only by the DRG grouper) was created based on the discharge quarter and year according to the following rules:

  • Discharge year (YEAR) is always nonmissing.
  • Discharge quarter (DQTR) ranges from zero to 4, where zero indicates that the quarter was missing or invalid.
Discharge Quarter (DQTR) Temporary Date (MM/DD/YY) passed to DRG Grouper
1 01/01/YY
2 04/01/YY
3 07/01/YY
4 10/01/YY
0 07/01/YY

Labels

Labels for the DRGs are provided as an ASCII file in HCUP Tools: Labels and Formats.

 
Top
 
Uniform Values
VariableDescriptionValueValue Description
DRGDRG in use on discharge datennnDRG value
 
Top
 
 
State Specific Notes

California

One discharge in 1991 with an invalid principal diagnosis code (DXV1=1) and at least one non-missing secondary diagnosis code (DX2, etc.) had the incorrect DRG and MDC assigned because of an error in HCUP processing. The DRG should have been 470; and the MDC should have been equal to 0.

No other years are affected.



Iowa

Beginning in data year 2001, the Iowa Hospital Association prohibits the release of two types of discharges:

  • HIV Infections (defined by MDC of 25) and
  • Behavioral Health including chemical dependency care or psychiatric care (defined by a service code of BHV).

These discharges were not included in the source file provided to HCUP and are therefore not included in the HCUP files.

For 2011 data only, HCUP data are modified to comply with confidentiality requirements of the Iowa data Partner:

  • Discharges from skilled nursing care or skilled swing beds have the MDC and DRG set to missing
  • Discharges that are assigned to MDC 25 during HCUP processing are deleted from the SID.


Massachusetts

Some 1989-1990 discharges with a missing principal diagnosis code (DX1=" ") and at least one non-missing secondary diagnosis code (DX2, etc.) have the incorrect DRG and MDC assigned because of an error in HCUP processing. The DRG should be 470; and the MDC should be equal to 0. The following number of records are affected: 1 record in 1989 and 1 record in 1990.

Some 1988-1991 discharges with an invalid principal diagnosis code (DXV1=1) and at least one non-missing secondary diagnosis code (DX2, etc.) have the incorrect DRG and MDC assigned because of an error in HCUP processing. The DRG should be 470; and the MDC should be equal to 0. The following number of records are affected:

  • for 1988, 34 records;
  • for 1989, 30 records;
  • for 1990, 44 records; and
  • for 1991, 33 records.

Beginning with 1992 discharges, DRG and MDC were processed correctly.



Washington

Some 1988-1992 discharges with an invalid principal diagnosis code (DXV1 = 1) and at least one non-missing secondary diagnosis code (DX2, etc.) have the incorrect DRG and MDC assigned because of an error in HCUP processing. The DRG should be 470; and the MDC should be equal to 0. The following number of records are affected:

  • for 1988, 184 records;
  • for 1989, 68 records;
  • for 1990, 13 records;
  • for 1991, 1 record; and
  • for 1992, 1 record.

Beginning with 1993 discharges, DRG and MDC were processed correctly.



Wisconsin

According to source documentation, the principal and secondary procedures for one hospital (DSHOSPID="056" and HOSPID=55155) are incorrect in the fourth quarter of 1997. System problems at the hospital caused the last procedure coded on the medical record to be stored as the principal procedure. No secondary procedures were recorded. This affects the DRG, DRG10, MDC, and MDC10 assignment.

Some 1989-1992 discharges with an invalid principal diagnosis code (DXV1=1) and at least one non-missing secondary diagnosis code (DX2, etc.) have the incorrect DRG and MDC assigned because of an error in HCUP processing. The DRG should be 470; and the MDC should be equal to 0. The following number of records are affected:

  • for 1989, 23 records;
  • for 1990, 4 records;
  • for 1991, 1 record; and
  • for 1992, 10 records.

Beginning with 1993 discharges, DRG and MDC were processed correctly.


 
Top
 
 

Internet Citation: HCUP NIS Description of Data Elements. Healthcare Cost and Utilization Project (HCUP). September 2008. Agency for Healthcare Research and Quality, Rockville, MD. www.hcup-us.ahrq.gov/db/vars/drg/nisnote.jsp.
Are you having problems viewing or printing pages on this website?
If you have comments, suggestions, and/or questions, please contact hcup@ahrq.gov.
Privacy Notice, Viewers & Players
Last modified 9/17/08