Healthcare Cost and Utilization Project (HCUP) Fast Stats

The national estimates are drawn from the HCUP National (Nationwide) Inpatient Sample (NIS). The NIS is based on data from community hospitals, which are defined as short-term, non-Federal, general, and other hospitals, excluding hospital units of other institutions (e.g., prisons). The NIS includes obstetrics and gynecology, otolaryngology, orthopedic, cancer, pediatric, public, and academic medical hospitals. Excluded are community hospitals that are also long-term care facilities such as rehabilitation, psychiatric, and alcoholism and chemical dependency hospitals. Beginning in 2012, long-term acute care hospitals (LTACs) are also excluded from the sampling frame. However, if a patient received long-term care, rehabilitation, or treatment for psychiatric or chemical dependency conditions in a community hospital, the discharge record for that stay will be included in the NIS.

The NIS is sampled from the HCUP State Inpatient Databases (SID). Beginning with the 2012 data year, the NIS is a 20 percent sample of discharges from all community hospitals that participate in the corresponding data year. For data years 1988 through 2011, the NIS was a 20 percent sample of community hospitals and included all discharges within sampled hospitals. The national estimates were developed using the NIS Trend Weight Files for consistent estimates across all data years (e.g., LTACs were removed from analysis using trend weights).

The unit of analysis is the hospital discharge (i.e., the hospital inpatient stay) or an emergency department (ED) visit, not a person or patient. This means that a person who is admitted to the hospital or visits the ED multiple times in one year is counted each time as a separate discharge from the hospital or a separate visit in the ED. Discharge and ED visit counts exclude transfers to another acute care hospital or ED. If the hospital inpatient stay or ED visit includes more than one diagnosis code for opioid-related hospital use, the encounter is only counted once.

All stays and visits are counted one time only, regardless of the number of relevant diagnosis or procedure codes that appear on the record. For instance, when identifying opioid-related inpatient stays and ED visits, a record may include more than one of the opioid-specific codes; in such a case, the record is only included once in the counts.

On October 1, 2015, the United States transitioned from ICD-9-CM¹ to ICD-10-CM/PCS². The graphics demarcate this transition with statistics reported using ICD-9-CM coding identified as “ICD-9-CM” on the graphs and statistics reported using ICD-10-CM/PCS coding identified as “ICD-10-CM/PCS” on the graphs. The 2015 rates of stays per 100,000 population and average statistics for hospitalization type are based on the first three quarters of data with ICD-9-CM codes only (January 1, 2015 to September 30, 2015). The number of inpatient stays by hospitalization type in 2015 is not reported because the statistics are not based on full year data. Statistics for all other characteristics include data for the full 2015 calendar year since these statistics are non-clinical, and therefore not impacted by the transition to ICD-10-CM/PCS.

Notable increases or decreases may be observed in the statistics across the ICD-9-CM to ICD-10-CM/PCS transition that are more reflective of definitional changes rather than changes in hospital utilization. Compared with the ICD-9-CM time period, some definitions of hospitalization type for ICD-10-CM/PCS may be more narrowly or more broadly defined. More information on the impact of ICD-10-CM/PCS is available on the HCUP User Support (HCUP-US) web page for ICD-10-CM/PCS Resources.

¹ International Classification of Diseases, Ninth Revision, Clinical Modification
² International Classification of Diseases, Tenth Revision, Clinical Modification/Procedure Coding System

Population-based rates are presented for inpatient stay trends overall and by age, sex, community-level income, and hospitalization type. Rates are not reported by expected payer because currently there is no data source for national population insurance estimates that align with HCUP’s definition of expected primary payer. The rate of stays includes the HCUP number of stays in the numerator and the U.S. resident population in the denominator (with a multiplier of 100,000). For age, sex, and community-level income, the denominator is consistently defined with the numerator (i.e., rates for females use HCUP counts and population counts specific to females). For hospitalization type, the denominator represents the total U.S. resident population. Population data are obtained from Claritas, a vendor that produces population estimates and projections based on data from the U.S. Census Bureau. Claritas estimates intercensal annual household and demographic statistics for geographic areas. Rates are only reported from 2002 forward because the population denominators for age, sex, and community-level income were unavailable prior to 2002.

The NIS includes information on total hospital charges for an inpatient stay. Charges represent the amount a hospital billed for the entire hospital stay, excluding professional (physician) fees. Total hospital charges are converted to costs using HCUP Cost-to-Charge Ratios (CCRs) based on hospital accounting reports from the Centers for Medicare & Medicaid Services (CMS). Costs reflect the actual expenses incurred in the production of hospital services, such as wages, supplies, and utility costs. For each hospital in the NIS, a hospital-wide cost-to-charge ratio is used. The average cost per stay is calculated using discharges with nonmissing total costs. Costs are not imputed if total charges are not reported on the discharge record. Costs are only reported from 2000 forward because HCUP Cost-to-Charge Ratios are unavailable prior to 2000.

The data element for total hospital charge (TOTCHG) was replaced in the 2023 NIS by the data element (TOTCHG_2023) that included an adjusted total hospital charge that accounted for the hospital charges of discharges in states that were not available to participate in the 2023 NIS. Costs reported in these tables for 2023 were based on the TOTCHG_2023 data element. Caution should be used when comparing 2023 cost to prior years. Additional relevant information can be found in the 2023 NIS Introduction (https://hcup-us.ahrq.gov/db/nation/nis/NISIntroduction2023.pdf).

The actual average cost per stay is inflation adjusted using price indexes for the Gross Domestic Product (GDP) from the U.S. Department of Commerce Bureau of Economic Analysis (BEA). We used the BEA Interactive Data query tool to request National Data, GDP & Personal Income, Section 1 Domestic Product and Income, Table 1.1.4. Price Indexes for Gross Domestic Product. Price indexes for data years 1994-2014 were obtained on June 23, 2015. Price indexes for subsequent data years were obtained at later dates to coincide with Fast Stats updates. The adjustment used 2010 as the index base so that updates to the trends could retain a consistent base.

The length of stay (LOS) is the number of days that the patient stayed in the hospital. It is calculated by subtracting the admission date from the discharge date. Same-day stays are therefore coded with a length of stay of 0. The average LOS is calculated using discharges with nonmissing LOS.

In-hospital mortality is determined by the discharge disposition of the patient from the hospital. The numerator of the mortality rate is the number of patients within a reporting category (e.g., within a specific diagnosis category) who died in the hospital. The denominator is based on the total number of discharges in the reporting category. Discharges missing discharge disposition are excluded from the numerator and denominator of the in-hospital mortality rate.

National Inpatient Stays:
Age refers to the age of the patient at admission. Discharges missing age are excluded from results reported by age.

National Emergency Department Visits:
Age refers to the age of the patient at admission to the ED. ED visits missing age are excluded from results reported by age.

Opioids Hospital Use:
Age refers to the age of the patient at admission. Discharges or visits missing age are excluded from results reported by age.

It should be noted that beginning with the transition to the ICD-10-CM/PCS coding system on October 1, 2015, a noticeable increase in opioid-related inpatient stays was observed for adults aged 65 years and older in most States. Additional information regarding this pattern is available on the HCUP-US web page for ICD-10-CM/PCS Resources in the report “Preliminary Case Study: Exploring How Opioid-Related Diagnosis Codes Translate from ICD-9-CM to ICD-10-CM“.

Hurricane Impact:
Age refers to the age (in years) of the patient at admission. Discharges or visits missing age are excluded from results reported by age. Age is grouped into three categories: 0-17 years, 18-64 years, and 65+ years. Less than 0.05 percent of records are missing information on age.

All nonmale, nonfemale responses are set to missing. Discharges and visits with missing values for sex are excluded from results reported by sex.

The “expected payer” data element in HCUP databases provides information on the type of payer that the hospital expects to be the source of payment for the hospital bill. Information is reported by the following expected primary payers: Medicare, Medicaid, private insurance, and self-pay/no charge. Patients identified as self-pay/no charge have an expected primary payer of self-pay, no charge, charity, or no expected payment. More information on expected payer coding in HCUP data is available in HCUP Methods Series Reports by Topic “User Guide – An Examination of Expected Payer Coding in HCUP Databases” (multiple documents; updated annually). Discharges missing expected payer are excluded from results reported by expected payer.

Discharges with the following expected primary payers are not reported by expected payers: other Federal, State, and local programs; missing; or invalid. These excluded discharges represent approximately 3 percent of all annual discharges.

The total reflecting the number of discharges across all expected payers (including those groups not presented in the graphs) is provided in the underlying data tables (“Show Underlying Data Tables”) by expected payer. These totals are the same as the counts obtained for the all inpatient stays characteristic selection.

For comparison against the total described above for all expected payers, the Excel download file also provides the sum of the displayed expected payers (i.e., the sum of the expected payer counts of discharges across the expected payers that are displayed in the graphs).

Community-level income is based on the median household income of the patient’s ZIP Code of residence. Quartiles are defined so that the total U.S. population is evenly distributed across four groups. Over time, the data element in the NIS for community-level income has changed definitions. The cut-offs for the quartile designation are determined annually using ZIP Code demographic data obtained from Claritas, a vendor that produces population estimates and projections based on data from the U.S. Census Bureau. Claritas estimates intercensal annual household and demographic statistics for geographic areas. The value ranges for the national income quartiles vary by year. Income quartile is missing if the patient is homeless or foreign. Records missing the income quartile are excluded from results reported by community-level income.

For Inpatient data, information by community-level income is only reported from 2002 forward because of inconsistent definitions over time in the income-related data elements in the NIS.

Each adult discharge is assigned to a single hospitalization type hierarchically, based on the following order: maternal, mental health/substance use, injury, surgical, and medical. All discharges are categorized in one of the five mutually exclusive hospitalization types based on the principal diagnosis for the hospital stay.

The definitions for the maternal, mental health/substance use, and injury hospitalization types have changed over time; they are identified differently under ICD-9-CM and ICD-10-CM:

• Beginning with quarter 4 of 2015 (2015 Q4): Major Diagnostic Category (MDC) or Clinical Classifications Software Refined (CCSR) for ICD-10-CM default categorization scheme for the principal diagnosis
• Through quarter 3 of 2015 (2015 Q3): A principal diagnosis based on either ranges of ICD-9-CM codes or the Clinical Classifications Software (CCS) for ICD-9-CM

The definitions for the surgical and medical hospitalization types remain the same under both the ICD-9-CM and ICD-10-CM coding systems. They are defined based on the Diagnosis-Related Group (DRG).

It should be noted that beginning in December 2020, statistics by hospitalization type using the ICD-10-CM coding system (since quarter 4 of 2015) have been updated to reflect new definitions provided below for each of the five hospitalization types. Thus, there is a one-time change in previously released statistics beginning quarter 4 of 2015 through as late as quarter 1 of 2019, depending on the availability of a State’s data when the statistics were last updated prior to 2020. Many of the current and previously released statistics are similar, but statistics for some hospitalization types have changed substantially, specifically for mental health/substance use. The new definition of mental health/substance use is based on major diagnostic category (MDC). As a result, there are specific ICD-10-CM diagnosis codes that are no longer included. For example, discharges with a principal diagnosis code indicating alcoholic cirrhosis of the liver, Alzheimer’s disease, and poisoning by narcotics and psychodysleptics are no longer assigned to the hospitalization type for mental health/substance use. Additional information on the amount of change between current and previously released statistics for each hospitalization type by age group and expected payer is provided in the “ICD-10-CM Definition Changes” worksheet of the Excel download file.

Maternal discharges are defined using the following Clinical Classifications Software (CCS) for ICD-9-CM categories for data years 2015 and earlier or Major Diagnostic Category (MDC) beginning with data year 2016. MDC was assigned without using “present on admission” information on the record because not all HCUP data sources provide present on admission indicators.

• Beginning with Data Year 2016: MDC 14, Pregnancy, Childbirth and the Puerperium
• For Data Years 1994-2015: Principal diagnosis CCS 176-196

For data years 2015 and earlier, the CCS-based definition of maternal may result in slightly different counts of discharges when compared with other ways of classifying diagnosis codes. For example, compared with using MDC from 2015 and earlier, the CCS approach assigns 0.9 percent fewer cases to “maternal” because a maternal discharge is classified into a mental health CCS or a substance use CCS when the diagnosis code includes a mental health or substance abuse condition along with a maternal condition (e.g., drug dependence in pregnancy).

Neonatal discharges are defined using the following Clinical Classifications Software (CCS) for ICD-9-CM categories for data years 2015 and earlier or Major Diagnostic Category (MDC) beginning with data year 2016. MDC was assigned without using “present on admission” information on the record because not all HCUP data sources provide present on admission indicators.

• Beginning with Data Year 2016: MDC 15, Newborns and Other Neonates with Conditions Originating in the Perinatal Period
• For Data Years 1994-2015: CCS 218-224

For data years 2015 and earlier, the CCS-based definition of neonatal may result in slightly different counts of discharges when compared with other ways of classifying diagnosis codes. For example, compared with using MDC from 2015 and earlier, the CCS approach assigns 0.1 percent fewer cases to “neonatal” because a neonatal discharge is classified into a substance use CCS when the diagnosis code refers to a drug effect on the fetus or neonatal drug withdrawal.

The definition for adult discharges related to mental health/substance use has changed over time:

• Beginning 2015 Q4: MDC 19, Mental Diseases and Disorders, or 20, Alcohol/Drug Use or Induced Mental Disorders
• For 2007-2015 Q3: Principal diagnosis CCS 650-663, 670
• For 2003-2006: Principal diagnosis CCS 65-75
• For National Inpatient NIS Data Years 1994-2006: CCS 65-75

MDCs were assigned without using “present on admission” information on the record because not all HCUP data sources provide present on admission indicators.

Beginning with the 2017 data year, the Iowa SID includes records for behavioral health patients treated in chemical dependency or psychiatric care units. Prior to 2017 data, these records were prohibited from release, and therefore not reported in Fast Stats.

Injury discharges are identified by either: a) a principal diagnosis based on ranges of ICD-9-CM codes for data years 2015 and earlier, or b) a combination of the Clinical Classifications Software Refined (CCSR) for ICD-10-CM default categorization scheme for the principal diagnosis and individual ICD-10-CM diagnosis codes, for data years 2016 and later.

The definition for adult discharges related to injury has changed over time:

•  Beginning with Data Year 2016: CCSR INJ001-INJ027, INJ032 and ICD-10-CM diagnosis codes in the T84 series (used only for discharges from 1/1/2016-9/30/2016)
•  Beginning 2015 Q4: Default for principal diagnosis CCSR INJ001-INJ027, INJ032
•  Through 2015 Q3: Principal ICD-9-CM diagnosis codes 800-909.2, 909.4, 909.9, 910-994.9, 995.50-995.59, 995.80-995.85

The above definition of injury through 2015 Q3 includes five ICD-9-CM diagnosis codes (965.00, 965.01, 965.02, 965.09, and 980.0) that are also included under two CCS diagnosis categories (660 and 661) used for the definition of the mental health/substance use hospitalization type for ICD-9-CM. Because of the hierarchical ordering used to assign discharges to hospitalization type, discharges with one of these five principal ICD-9-CM diagnosis codes are assigned to the mental health/substance use hospitalization type and not the injury hospitalization type.

Excluded Codes

It should be noted that ICD-9-CM and ICD-10-CM diagnosis codes related to complications of surgical or medical care, or adverse events or anaphylactic shock resulting from medication, anesthesia, or food are not used in the definition of the injury hospitalization type.

Surgical discharges are identified by a surgical diagnosis-related group (DRG). The DRG grouper first assigns the discharge to a MDC based on the principal diagnosis. For each MDC, there is a list of procedure codes that qualify as operating room procedures. If the discharge involves an operating room procedure, it is assigned to one of the surgical DRGs within the MDC category; otherwise it is assigned to a medical DRG.

All other discharges that are not identified as maternal, neonatal, mental health/substance use, injury, or surgical are identified as medical discharge.  If the DRG indicates the information on the record is ungroupable (i.e., not identifiable as medical or surgical), then the discharge is assumed to be medical. This rarely occurs (less than 0.1 percent of total discharges).

The national estimates are drawn from the HCUP Nationwide Emergency Department Sample (NEDS). The NEDS is the largest all-payer emergency department (ED) database in the United States, yielding national estimates of hospital-owned ED visits. Hospital-owned EDs are limited to community hospitals, which are defined as short-term, non-Federal, general, and other hospitals, excluding hospital units of other institutions (e.g., prisons). The NEDS includes specialty, pediatric, public, and academic medical hospitals. Excluded are long-term care facilities such as rehabilitation, psychiatric, and alcoholism and chemical dependency hospitals.

The NEDS is sampled from the HCUP State Emergency Department Databases (SEDD), which capture information on ED encounters that do not result in an admission (i.e., treat-and-release visits and transfers to other hospitals), and the HCUP State Inpatient Databases (SID), which contain information on patients initially seen in the ED and then admitted to the same hospital. The NEDS approximates a 20 percent stratified sample of U.S. hospital-owned EDs. The NEDS is stratified on the following hospital characteristics: U.S. Census region, trauma center designation, urban-rural location of the hospital, ownership, and teaching status. The NEDS is currently available for data years 2006-2023. Since the initial release of the NEDS, State participation has grown from 24 States in 2006 to 40 States and the District of Columbia in 2023.

The unit of analysis is the hospital discharge (i.e., the hospital inpatient stay) or an emergency department (ED) visit, not a person or patient. This means that a person who is admitted to the hospital or visits the ED multiple times in one year is counted each time as a separate discharge from the hospital or a separate visit in the ED. Discharge and ED visit counts exclude transfers to another acute care hospital or ED. If the hospital inpatient stay or ED visit includes more than one diagnosis code for opioid-related hospital use, the encounter is only counted once.

All stays and visits are counted one time only, regardless of the number of relevant diagnosis or procedure codes that appear on the record. For instance, when identifying opioid-related inpatient stays and ED visits, a record may include more than one of the opioid-specific codes; in such a case, the record is only included once in the counts.

On October 1, 2015, the United States transitioned from ICD-9-CM¹ to ICD-10-CM/PCS². The 2015 data in HCUP Fast Stats include three quarters of information based on ICD-9-CM coding, whereas the fourth quarter is based on ICD-10-CM/PCS coding. Users may observe discontinuity in trends analyses that span the October 1, 2015 transition date.

Notable increases or decreases may be observed in the statistics across the ICD-9-CM to ICD-10-CM/PCS transition that are more reflective of definitional changes rather than changes in ED utilization. Compared with the ICD-9-CM time period, some definitions of ED visit type for ICD-10-CM/PCS may be more narrowly or more broadly defined. More information on the impact of ICD-10-CM/PCS is available on the HCUP User Support (HCUP-US) web page for ICD-10-CM/PCS Resources.
¹ International Classification of Diseases, Ninth Revision, Clinical Modification
² International Classification of Diseases, Tenth Revision, Clinical Modification/Procedure Coding System

Population-based rates are presented for ED visit trends overall and by age, sex, community-level income, and patient location. Population-based rates are not reported by expected payer because currently there is no data source of national population insurance estimates that aligns with HCUP’s definition of expected primary payer. The rate of ED visits includes the HCUP number of ED visits in the numerator and the U.S. resident population in the denominator (with a multiplier of 100,000). For all characteristics, the denominator is consistently defined with the numerator (i.e., rates for females use HCUP counts and population counts specific to females). Population data are obtained from Claritas, a vendor that produces population estimates and projections based on data from the U.S. Census Bureau. Claritas estimates intercensal annual household and demographic statistics for geographic areas.

National Inpatient Stays:
Age refers to the age of the patient at admission. Discharges missing age are excluded from results reported by age.

National Emergency Department Visits:
Age refers to the age of the patient at admission to the ED. ED visits missing age are excluded from results reported by age.

Opioids Hospital Use:
Age refers to the age of the patient at admission. Discharges or visits missing age are excluded from results reported by age.

It should be noted that beginning with the transition to the ICD-10-CM/PCS coding system on October 1, 2015, a noticeable increase in opioid-related inpatient stays was observed for adults aged 65 years and older in most States. Additional information regarding this pattern is available on the HCUP-US web page for ICD-10-CM/PCS Resources in the report “Preliminary Case Study: Exploring How Opioid-Related Diagnosis Codes Translate from ICD-9-CM to ICD-10-CM“.

Hurricane Impact:
Age refers to the age (in years) of the patient at admission. Discharges or visits missing age are excluded from results reported by age. Age is grouped into three categories: 0-17 years, 18-64 years, and 65+ years. Less than 0.05 percent of records are missing information on age.

All nonmale, nonfemale responses are set to missing. Discharges and visits with missing values for sex are excluded from results reported by sex.

The “expected payer” data element in HCUP databases provides information on the type of payer that is expected to be the source of payment for the ED bill. Information is reported by the following expected primary payers: Medicare, Medicaid, private insurance, and self-pay/no charge. Patients identified as self-pay/no charge have an expected primary payer of self-pay, no charge, charity, or no expected payment. More information on expected payer coding in HCUP data is available in HCUP Methods Series Reports by Topic “User Guide – An Examination of Expected Payer Coding in HCUP Databases” (multiple documents; updated annually). ED visits missing expected payer are excluded from results reported by expected payer.

ED visits with the following expected primary payers are not reported by expected payer: other Federal, State, and local programs; missing; or invalid. These excluded ED visits represent approximately 4 percent of all annual ED visits.

The total reflecting the number of ED visits across all expected payers (including those groups not presented in the graphs) is provided in the underlying data tables (“Show Underlying Data Tables”) by expected payer by ED visit type. These totals are the same as the counts obtained for the overall characteristic selection for the respective ED visit type.

For comparison against the total described above for all expected payers, the Excel download file also provides the sum of the displayed expected payers (i.e., the sum of the expected payer counts of ED visits across the expected payers that are displayed in the graphs).

Community-level income is based on the median household income of the patient’s ZIP Code of residence. Quartiles are defined so that the total U.S. population is evenly distributed across four groups. Over time, the data element in the NIS for community-level income has changed definitions. The cut-offs for the quartile designation are determined annually using ZIP Code demographic data obtained from Claritas, a vendor that produces population estimates and projections based on data from the U.S. Census Bureau. Claritas estimates intercensal annual household and demographic statistics for geographic areas. The value ranges for the national income quartiles vary by year. Income quartile is missing if the patient is homeless or foreign. Records missing the income quartile are excluded from results reported by community-level income.

For Inpatient data, information by community-level income is only reported from 2002 forward because of inconsistent definitions over time in the income-related data elements in the NIS.

Patient location is based on the six-category, county-level scheme developed by the National Center for Health Statistics (NCHS) to study the relationship between urbanization and health:

• Large central metropolitan: Counties in metropolitan statistical areas (MSAs) of 1 million or more population that contain the entire population of the largest principal city of the MSA, have their entire population contained in the largest principal city of the MSA, or contain at least 250,000 inhabitants of any principal city of the MSA
• Large fringe metropolitan (suburbs): Counties in MSAs of 1 million or more population that did not qualify as large central metropolitan counties
• Medium metropolitan: Counties in MSAs of populations of 250,000 to 999,999
• Small metropolitan: Counties in MSAs of population less than 250,000
• Micropolitan: Counties in micropolitan statistical areas
• Noncore: Nonmetropolitan counties that did not qualify as micropolitan

The micropolitan and noncore categories are combined into a single category (“Rural”) in order to preserve results when cell sizes are too small. For rates prior to 2014, the NCHS classification is based on population density from the 2000 Census. Starting in 2014, the NCHS classification is based on population density from the 2010 Census. Newborn hospitalizations missing patient location are excluded from the results reported by patient location. The 2023 NEDS replaced the six-category data element identifying the detailed metro status designation of the county of the patient’s residence with a new consolidated data element that distinguishes only two categories: metropolitan and non-metropolitan. The trends in the ED Visits dashboard hence reports statistics by patient location for only these two categories for the 2023 data year. The non-metropolitan category is named “Rural” in this table for consistency with prior data years.

Starting in 2024 HCUP databases, the classification of counties is specific to the 2023 revision released by NCHS. Users should exercise caution when analyzing time trends by patient location.

Patient location categories may include out-of-state patients because the classification is based on patient residence county but reported by the State for the hospital. Additional information on the patient location classification system is available at NCHS Urban-Rural Classification Scheme for Counties.

The national estimates are drawn from the HCUP National (Nationwide) Inpatient Sample (NIS). The NIS is based on data from community hospitals, which are defined as short-term, non-Federal, general, and other hospitals, excluding hospital units of other institutions (e.g., prisons). The NIS includes obstetrics and gynecology, otolaryngology, orthopedic, cancer, pediatric, public, and academic medical hospitals. Excluded are community hospitals that are also long-term care facilities such as rehabilitation, psychiatric, and alcoholism and chemical dependency hospitals. Beginning in 2012, long-term acute care hospitals (LTACs) are also excluded from the sampling frame. However, if a patient received long-term care, rehabilitation, or treatment for psychiatric or chemical dependency conditions in a community hospital, the discharge record for that stay will be included in the NIS.

The NIS is sampled from the HCUP State Inpatient Databases (SID). Beginning with the 2012 data year, the NIS is a 20 percent sample of discharges from all community hospitals that participate in the corresponding data year. For data years 1988 through 2011, the NIS was a 20 percent sample of community hospitals and included all discharges within sampled hospitals. The national estimates were developed using the NIS Trend Weight Files for consistent estimates across all data years (e.g., LTACs were removed from analysis using trend weights).

The unit of analysis is the hospital discharge (i.e., the hospital inpatient stay) or an emergency department (ED) visit, not a person or patient. This means that a person who is admitted to the hospital or visits the ED multiple times in one year is counted each time as a separate discharge from the hospital or a separate visit in the ED. Discharge and ED visit counts exclude transfers to another acute care hospital or ED. If the hospital inpatient stay or ED visit includes more than one diagnosis code for opioid-related hospital use, the encounter is only counted once.

All stays and visits are counted one time only, regardless of the number of relevant diagnosis or procedure codes that appear on the record. For instance, when identifying opioid-related inpatient stays and ED visits, a record may include more than one of the opioid-specific codes; in such a case, the record is only included once in the counts.

Beginning with data year 2020, COVID-19-related hospitalizations may be identified by any-listed ICD-10-CM diagnosis code of “U071” (2019 novel coronavirus disease) on the discharge record. Per coding guidelines, the use of diagnosis code “U071” is based on documentation by the provider or documentation of a positive COVID-19 test result.  The ICD-10-CM diagnosis code for COVID-19 was implemented beginning April 1, 2020.  There may be other ICD-10-CM codes that reflect conditions related to COVID-19 stays.

On October 1, 2015, the United States transitioned from ICD-9-CM¹ to ICD-10-CM/PCS². The 2015 rates of stays per 100,000 population are based on the first three quarters of data with ICD-9-CM codes only (January 1, 2015 to September 30, 2015). The number of inpatient stays by diagnosis in 2015 is not reported because the statistics are not based on full year data. More information on the impact of ICD-10-CM/PCS is available on the HCUP User Support (HCUP-US) web page for ICD-10-CM/PCS Resources.

¹ International Classification of Diseases, Ninth Revision, Clinical Modification
² International Classification of Diseases, Tenth Revision, Clinical Modification/Procedure Coding System

A change in the ICD-10-CM coding guidelines on which diagnosis to code as the principal diagnosis has caused a discontinuity in the ranking of inpatient stays related to chronic obstructive pulmonary disease (COPD) and pneumonia. Effective October 1, 2016 precedence was given to the COPD diagnosis. A subsequent revision effective October 1, 2017 bases precedence on the discretion of a clinical coder.

The principal diagnosis is that condition established after study to be chiefly responsible for the patient’s admission to the hospital. Diagnoses in the NIS are reported using ICD-9-CM codes through September 30, 2015 and ICD-10-CM codes beginning October 1, 2015. There are approximately 14,000 ICD-9-CM diagnosis codes and over 70,000 ICD-10-CM diagnosis codes.

The principal diagnosis is defined using the Clinical Classifications Software (CCS) for ICD-9-CM for data years 2015 and earlier, and the Clinical Classifications Software Refined (CCSR) for ICD-10-CM default categorization scheme for the principal diagnosis beginning with data year 2016. Results are reported by the CCS or CCSR category and list the top 10 most common principal diagnoses for each data year. The top 10 ranking is based on the weighted number of stays. Because of the transition from ICD-9-CM to ICD-10-CM/PCS on October 1, 2015, the total number of inpatient stays in 2015 is not reported. The 2015 rate of stays per 100,000 population is based on the first three quarters of 2015 data (Q1-3) only.

Results can be displayed with maternal and neonatal stays included or excluded from the ranking. This option is provided because maternal and neonatal discharges account for nearly a fourth of all hospital discharges in a year and the majority are low complexity, low cost stays. Maternal and neonatal stays are defined differently across data years. For data years 2015 and earlier, the principal diagnosis CCS 176 through 196 are used for maternal and CCS 218 through 224 are used for neonatal. Beginning with data year 2016, Major Diagnostic Category (MDC) 14 (Pregnancy, Childbirth, and the Puerperium) and MDC 15 (Newborns and Other Neonates with Conditions Originating in the Perinatal Period) are used.

Population-based rates are presented for inpatient stay trends overall and by age, sex, community-level income, and hospitalization type. Rates are not reported by expected payer because currently there is no data source for national population insurance estimates that align with HCUP’s definition of expected primary payer. The rate of stays includes the HCUP number of stays in the numerator and the U.S. resident population in the denominator (with a multiplier of 100,000). For age, sex, and community-level income, the denominator is consistently defined with the numerator (i.e., rates for females use HCUP counts and population counts specific to females). For hospitalization type, the denominator represents the total U.S. resident population. Population data are obtained from Claritas, a vendor that produces population estimates and projections based on data from the U.S. Census Bureau. Claritas estimates intercensal annual household and demographic statistics for geographic areas. Rates are only reported from 2002 forward because the population denominators for age, sex, and community-level income were unavailable prior to 2002.

National Inpatient Stays:
Age refers to the age of the patient at admission. Discharges missing age are excluded from results reported by age.

National Emergency Department Visits:
Age refers to the age of the patient at admission to the ED. ED visits missing age are excluded from results reported by age.

Opioids Hospital Use:
Age refers to the age of the patient at admission. Discharges or visits missing age are excluded from results reported by age.

It should be noted that beginning with the transition to the ICD-10-CM/PCS coding system on October 1, 2015, a noticeable increase in opioid-related inpatient stays was observed for adults aged 65 years and older in most States. Additional information regarding this pattern is available on the HCUP-US web page for ICD-10-CM/PCS Resources in the report “Preliminary Case Study: Exploring How Opioid-Related Diagnosis Codes Translate from ICD-9-CM to ICD-10-CM“.

Hurricane Impact:
Age refers to the age (in years) of the patient at admission. Discharges or visits missing age are excluded from results reported by age. Age is grouped into three categories: 0-17 years, 18-64 years, and 65+ years. Less than 0.05 percent of records are missing information on age.

All nonmale, nonfemale responses are set to missing. Discharges and visits with missing values for sex are excluded from results reported by sex.

The “expected payer” data element in HCUP databases provides information on the type of payer that the hospital expects to be the source of payment for the hospital bill. Information is reported by the following expected primary payers: Medicare, Medicaid, private insurance, and self-pay/no charge. Patients identified as self-pay/no charge have an expected primary payer of self-pay, no charge, charity, or no expected payment. More information on expected payer coding in HCUP data is available in HCUP Methods Series Reports by Topic “User Guide – An Examination of Expected Payer Coding in HCUP Databases” (multiple documents; updated annually). Discharges missing expected payer are excluded from results reported by expected payer.

Discharges with the following expected primary payers are not reported by expected payer: other Federal, State, and local programs; missing; or invalid.

Community-level income is based on the median household income of the patient’s ZIP Code of residence. Quartiles are defined so that the total U.S. population is evenly distributed across four groups. Over time, the data element in the NIS for community-level income has changed definitions. The cut-offs for the quartile designation are determined annually using ZIP Code demographic data obtained from Claritas, a vendor that produces population estimates and projections based on data from the U.S. Census Bureau. Claritas estimates intercensal annual household and demographic statistics for geographic areas. The value ranges for the national income quartiles vary by year. Income quartile is missing if the patient is homeless or foreign. Records missing the income quartile are excluded from results reported by community-level income.

For Inpatient data, information by community-level income is only reported from 2002 forward because of inconsistent definitions over time in the income-related data elements in the NIS.

The national estimates are drawn from the HCUP National (Nationwide) Inpatient Sample (NIS). The NIS is based on data from community hospitals, which are defined as short-term, non-Federal, general, and other hospitals, excluding hospital units of other institutions (e.g., prisons). The NIS includes obstetrics and gynecology, otolaryngology, orthopedic, cancer, pediatric, public, and academic medical hospitals. Excluded are community hospitals that are also long-term care facilities such as rehabilitation, psychiatric, and alcoholism and chemical dependency hospitals. Beginning in 2012, long-term acute care hospitals (LTACs) are also excluded from the sampling frame. However, if a patient received long-term care, rehabilitation, or treatment for psychiatric or chemical dependency conditions in a community hospital, the discharge record for that stay will be included in the NIS.

The NIS is sampled from the HCUP State Inpatient Databases (SID). Beginning with the 2012 data year, the NIS is a 20 percent sample of discharges from all community hospitals that participate in the corresponding data year. For data years 1988 through 2011, the NIS was a 20 percent sample of community hospitals and included all discharges within sampled hospitals. The national estimates were developed using the NIS Trend Weight Files for consistent estimates across all data years (e.g., LTACs were removed from analysis using trend weights).

The unit of analysis is the hospital discharge (i.e., the hospital inpatient stay) or an emergency department (ED) visit, not a person or patient. This means that a person who is admitted to the hospital or visits the ED multiple times in one year is counted each time as a separate discharge from the hospital or a separate visit in the ED. Discharge and ED visit counts exclude transfers to another acute care hospital or ED. If the hospital inpatient stay or ED visit includes more than one diagnosis code for opioid-related hospital use, the encounter is only counted once.

All stays and visits are counted one time only, regardless of the number of relevant diagnosis or procedure codes that appear on the record. For instance, when identifying opioid-related inpatient stays and ED visits, a record may include more than one of the opioid-specific codes; in such a case, the record is only included once in the counts.

On October 1, 2015, the United States transitioned from ICD-9-CM¹ to ICD-10-CM/PCS². The 2015 rates of stays per 100,000 population are based on the first three quarters of data with ICD-9-CM codes only (January 1, 2015 to September 30, 2015). The number of inpatient stays by diagnosis in 2015 is not reported because the statistics are not based on full year data. More information on the impact of ICD-10-CM/PCS is available on the HCUP User Support (HCUP-US) web page for ICD-10-CM/PCS Resources.

¹ International Classification of Diseases, Ninth Revision, Clinical Modification
² International Classification of Diseases, Tenth Revision, Clinical Modification/Procedure Coding System

Operating room (OR) procedures are identified using the Procedure Classes for ICD-9-CM for data years 2015 and earlier and the Procedure Classes Refined for ICD-10-PCS beginning data year 2016. The Procedure Classes tools identify procedures as diagnostic or therapeutic and whether they would be expected to be performed in an operating room. OR procedures are identified using all-listed procedures (principal and secondary) that are available on the discharge record.

The definition of an OR procedure varies over time. With the Procedure Classes for ICD-9-CM, procedure codes are considered to be valid OR procedures based on the definition of a major surgery in the Medicare Severity Diagnosis Related Group (MS-DRG) grouper. The MS-DRG classification scheme relied on physician panels that classified ICD-9-CM procedure codes according to whether the procedure would be performed in a hospital OR in most hospitals. Beginning with the Procedure Classes Refined for ICD-10-PCS v2021.2, ICD-10-PCS procedure codes are determined to be valid OR procedures based on the AHRQ Quality Indicator (QI) software. Changes to either the definition of an OR procedure or the coding system may affect how some procedures are ranked. For example, circumcision is no longer reported as a most common OR procedure beginning data year 2016.

Results are reported using the Clinical Classifications Software (CCS) for ICD-9-CM for data years 2015 and earlier, and the Clinical Classifications Software Refined (CCSR) for ICD-10-PCS procedure codes beginning with data year 2016. Results are reported by the CCS or CCSR category and list the top 10 most common operations for each data year. Counts for procedures are de-duplicated within a discharge record: if a particular CCS or CCSR procedure occurs multiple times during the same hospital stay, it is counted only once. Because there can be multiple procedures reported on a single hospital stay, one discharge record may contribute to the count for more than one operating room procedure. The top 10 ranking is based on the weighted number of stays. Because of the transition from ICD-9-CM to ICD-10-CM/PCS on October 1, 2015, the total number of inpatient stays in 2015 is not reported. The 2015 rate of stays per 100,000 population is based on the first three quarters of 2015 data (Q1-3) only.

Results can be displayed with maternal and neonatal stays included or excluded from the ranking. This option is provided because maternal and neonatal discharges account for nearly a fourth of all hospital discharges in a year and the majority are low complexity, low cost stays. Maternal and neonatal stays are defined differently across data years. For data years 2015 and earlier, the principal diagnosis CCS 176 through 196 are used for maternal and CCS 218 through 224 are used for neonatal. Beginning with data year 2016, Major Diagnostic Category (MDC) 14 (Pregnancy, Childbirth, and the Puerperium) and MDC 15 (Newborns and Other Neonates with Conditions Originating in the Perinatal Period) are used.

Population-based rates are presented for inpatient stay trends overall and by age, sex, community-level income, and hospitalization type. Rates are not reported by expected payer because currently there is no data source for national population insurance estimates that align with HCUP’s definition of expected primary payer. The rate of stays includes the HCUP number of stays in the numerator and the U.S. resident population in the denominator (with a multiplier of 100,000). For age, sex, and community-level income, the denominator is consistently defined with the numerator (i.e., rates for females use HCUP counts and population counts specific to females). For hospitalization type, the denominator represents the total U.S. resident population. Population data are obtained from Claritas, a vendor that produces population estimates and projections based on data from the U.S. Census Bureau. Claritas estimates intercensal annual household and demographic statistics for geographic areas. Rates are only reported from 2002 forward because the population denominators for age, sex, and community-level income were unavailable prior to 2002.

National Inpatient Stays:
Age refers to the age of the patient at admission. Discharges missing age are excluded from results reported by age.

National Emergency Department Visits:
Age refers to the age of the patient at admission to the ED. ED visits missing age are excluded from results reported by age.

Opioids Hospital Use:
Age refers to the age of the patient at admission. Discharges or visits missing age are excluded from results reported by age.

It should be noted that beginning with the transition to the ICD-10-CM/PCS coding system on October 1, 2015, a noticeable increase in opioid-related inpatient stays was observed for adults aged 65 years and older in most States. Additional information regarding this pattern is available on the HCUP-US web page for ICD-10-CM/PCS Resources in the report “Preliminary Case Study: Exploring How Opioid-Related Diagnosis Codes Translate from ICD-9-CM to ICD-10-CM“.

Hurricane Impact:
Age refers to the age (in years) of the patient at admission. Discharges or visits missing age are excluded from results reported by age. Age is grouped into three categories: 0-17 years, 18-64 years, and 65+ years. Less than 0.05 percent of records are missing information on age.

All nonmale, nonfemale responses are set to missing. Discharges and visits with missing values for sex are excluded from results reported by sex.

The “expected payer” data element in HCUP databases provides information on the type of payer that the hospital expects to be the source of payment for the hospital bill. Information is reported by the following expected primary payers: Medicare, Medicaid, private insurance, and self-pay/no charge. Patients identified as self-pay/no charge have an expected primary payer of self-pay, no charge, charity, or no expected payment. More information on expected payer coding in HCUP data is available in HCUP Methods Series Reports by Topic “User Guide – An Examination of Expected Payer Coding in HCUP Databases” (multiple documents; updated annually). Discharges missing expected payer are excluded from results reported by expected payer.

Discharges with the following expected primary payers are not reported by expected payer: other Federal, State, and local programs; missing; or invalid.

Community-level income is based on the median household income of the patient’s ZIP Code of residence. Quartiles are defined so that the total U.S. population is evenly distributed across four groups. Over time, the data element in the NIS for community-level income has changed definitions. The cut-offs for the quartile designation are determined annually using ZIP Code demographic data obtained from Claritas, a vendor that produces population estimates and projections based on data from the U.S. Census Bureau. Claritas estimates intercensal annual household and demographic statistics for geographic areas. The value ranges for the national income quartiles vary by year. Income quartile is missing if the patient is homeless or foreign. Records missing the income quartile are excluded from results reported by community-level income.

For Inpatient data, information by community-level income is only reported from 2002 forward because of inconsistent definitions over time in the income-related data elements in the NIS.