HCUP has four data elements related to race and ethnicity:

• Original data provided by the data source (HCUP Partner organizations):
  • RACE_X contains the original data values for race. For some data sources, there may be a value for RACE_X that is specific to Hispanic ethnicity.
  • HISPANIC_X contains the original data values for ethnicity, if provided. This data element became available in data year 1998.
• HCUP uniform variables:
  • RACE contains a uniform coding for race and ethnicity.
    • If the data source supplied information on race and ethnicity as separate data elements (RACE_X and HISPANIC_X, respectively), ethnicity takes precedence over race in setting the HCUP uniform values. For example, a patient who is Hispanic and Black is assigned to the category of Hispanic (RACE=3 based on information included in HISPANIC_X).
    • If information on ethnicity is not provided as a separate data element and, instead, provided as a separate value for RACE_X, then the category of Hispanic (RACE=3) is based on the information in RACE_X.
  • HISPANIC contains a uniform coding for race and ethnicity in which Hispanic patients are assigned to a category that includes information on the race. Using the same example of a patient who is Hispanic and Black, HISPANIC is assigned to the category of Black Hispanic (HISPANIC=2). It should be noted that the HISPANIC data element does not retain information on the race of non-Hispanic patients. This data element became available in data year 2015.

Not all HCUP Partner organizations provide information on race and ethnicity. The availability of this information may differ by year. Additionally, the reporting of information on race and ethnicity can vary by hospital. For additional information, refer to the HCUP data element, RACE_PCT_MISS_HOSP, which identifies the percentage of hospital discharges where race and ethnicity information is missing or invalid. This data element is available beginning data year 2020.

Beginning in 2008, Arizona allows up to four separate responses to be recorded in the race field.

Central Distributor only - because of confidentiality restrictions, the race of the patient (RACE) is suppressed on some records.

Prior to 1998, Georgia does not separately classify Hispanic (RACE = 3). No documentation was available about how these were coded. Beginning in 1998, Georgia revised their coding scheme for race of the patient. See table below.

In the source data files provided by Hawaii, the coding of race of the patient was different for each Hawaiian hospital. During HCUP processing, the hospital-specific values were recoded into the values of RACE_X described below.

Due to an error in processing the 2015 data file, all non-Hispanic patients were coded to invalid (.A) instead of Not Hispanic (value 0).

Beginning in 2008, Kentucky supplies Race and Ethnicity. Kentucky codes Ethnicity as a separate data element.

Beginning in 2008, Maine supplies Race and Ethnicity. Maine codes Ethnicity as a separate data element.

Beginning in 1993, Maryland reported Hispanic ethnicity as a separate variable. If patient ethnicity was coded as Spanish/Hispanic origin, patient race was set to Hispanic (RACE = 3) during HCUP processing.

Prior to 1993, Maryland did not report Hispanic ethnicity as a separate variable or category of race. Hispanic ethnicity (RACE = 3) is not coded in the 1988-1992 HCUP Maryland data. The source documentation available for Maryland did not indicate which race code(s) were used for Hispanic ethnicity.

Beginning October 2014, Massachusetts uses the full list of standard codes from the CDC ethnicity code set and those listed below: https://www.cdc.gov/nchs/data/dvs/Race_Ethnicity_CodeSet.pdf

Small numbers of Hawaiian and Pacific Islanders (less than 1.7%) were included in "Invalid" in the 2006-2008 Massachusetts SEDD.

Small number of Hawaiian and Pacific Islanders (less than 1%) were included in "Invalid" in the 2006-2008 Massachusetts SID.

Beginning October 2006, Massachusetts significantly changed how race and ethnicity are reported. The information is reported in seven separate fields (Race1, Race2, RaceOther, Ethnicity1, Ethnicity2, EthnicityOther, and Hispanic Indicator). The data elements RaceOther and EthnicityOther are sparsely populated and were not retained on the HCUP files. The remaining fields were retained as follows:

RACE_X contains the concatenated values of Race1 and Race2. RACE_X = Race1 | Race2

Beginning 2006, Hispanic_X contains the Hispanic Indicator values. Beginning 2013,

Hispanic_X contains the concatenated values of Hispanic Indicator, Ethnicity1, and Ethnicity2.

Hispanic_X = Hispanic Indicator | Ethnicity1 | Ethnicity2

Information on patient race and ethnicity is no longer retained on the file beginning data year 2013.

In the middle of 2007, Nebraska hospitals changed the coding of race for Whites and Native Americans.

• Prior to June 2007, a value of 1 in RACE_X indicated the patient was White and a value of 5 indicated the patient was Native American.
• Beginning in June 2007, a value of 1 in RACE_X indicated Native American and a value of 5 indicated the patient was White.

Due to an error in HCUP data processing for the partial year 2007 and 2008-2010, RACE_X for values 1 and 5 were incorrectly mapped into the uniform variable RACE and misidentified Whites and American Indians. Beginning with 2011 data, the RACE data element correctly identifies Whites and American Indians.

Race is coded on only 8% of Nebraska discharges in 2005. Nebraska informed HCUP that RACE values are not consistent with coding guidelines.

Beginning with the 2019 data year, New Hampshire provides race in three separate fields (Race Code 1, Race Code 2, Race Code 3).

The fields were retained as follows:
RACE_X contains the concatenated values of Race Codes 1-3. RACE_X = Race1 | Race2 | Race3.

Beginning in 2017, NY uses the standard codes from the CDC Race and Ethnicity Code Set: www.cdc.gov/nchs/data/dvs/Race_Ethnicity_CodeSet.pdf

RACE_X contains the hirerarchical CDC Race code up to the 9 value level and concatenated values of up to seven Race values. RACE_X = Race1 | Race2 | Race 3 | Race4 | Race5 | Race 6| Race 7

Hispanic_X contains the hirerarchical CDC Ethnicity code concatenated values of up to seven Ethncity values. Hispanic_X = Ethnicity1 | Ethnicity2 |Ethnicity 3 | Ethnicity4 | Ethnicity5 |Ethnicity 6 |Ethnicity 7

North Carolina resumed providing the race and ethnicity variable beginning in 2010.

Beginning in Q4 2010, the race and ethnicity data values changed.

North Carolina did not provide RACE in 2009 as the coding was considered unreliable.

From 1998-2002 patient's race (RACE and RACE_X) was set to missing for records from South Carolina freestanding surgery centers because the data was not reliable. Beginning with 2003, race information is retained in RACE_X and RACE on all ambulatory surgery records. Analysts should examine facility-specific frequencies before using the data.

For 2015-2017, RACE has been dropped from the Utah file. The source supplied field, RACE_X, remains available.

For 2012, a large hospital system stopped the collection of information on Hispanic ethnicity of the patient and just collected information on the patient's race. We recommend checking hospital-level statistics before using RACE or HISPANIC_X. For 2013-2014, the hospital system continued this practice. All records for hospitals included in this hospital system have had RACE values set to missing for 2013-2014. Original hospital RACE values are contained in RACE_X.

For 2013, a few facilities had mapping issues with race and ethnicity reporting. The issues were mainly reported in the first quarter of 2013.

Race and ethnicity information is available on the Wisconsin SEDD beginning with 2005 data.

Race and ethnicity were available on the 2004 SASD, so any records in the 2004 SEDD that originated in the Wisconsin Ambulatory Surgery file will include information on race and ethnicity.