Skip Standard Navigation Links
CDC Home
Safer, Healthier People
CDC HomeCDC SearchHealth Topics A-Z
   
  WONDER Home FAQ Help Contact Us Search  
 

Compressed Mortality File:

Years 1968-1978 with ICD-8 Codes, 1979-1998 with ICD-9 Codes and 1999-2010 with ICD-10 Codes

Summary

The Compressed Mortality File (CMF) is a county-level national mortality and population database spanning the years 1968-2010. Compressed Mortality data are updated annually. The number of deaths, crude death rates or age-adjusted death rates can be obtained by place of residence (total U.S., Census region, Census division, state, and county), age group, race (years 1968-1998: White, Black, and Other; years 1999-present: American Indian or Alaska Native, Asian or Pacific Islander, Black or African American, and White), Hispanic origin (years 1968-1998: not available; years 1999-present: Hispanic or Latino, not Hispanic or Latino, Not Stated), gender, year of death, and underlying cause of death (years 1968-1978: 4 digit ICD-8 codes and 69 cause-of-death recode; years 1979-1998: 4-digit ICD-9 codes and 72 cause-of-death recode; years 1999-present: 4-digit ICD-10 codes and 113 cause-of-death recode), and urbanization level of residence for years 1999-present (per the 2006 or the 2013 NCHS Urban-Rural Classification Scheme for Counties). For more information, see Compressed Mortality File Topics below and http://www.cdc.gov/nchs/data_access/cmf.htm.

  • Beginning in 1989, additional confidentiality restrictions apply to NCHS vital statistics data. As of May 23, 2011, all sub-national data representing zero to nine (0-9) deaths or births are suppressed. Corresponding sub-national denominator population figures are also suppressed when the population represents fewer than 10 persons. With the release for the Compressed Mortality data for years 1999-2008 on February 15,2012, additional privacy constraints apply to infant mortality statistics for infant age groups and live births denominator population figures. See Assurance of Confidentiality for more information.
  • Due to changes in the classification for underlying cause-of-death adopted in 1979 and also in 1999, WONDER has separate query screens for CMF data for years 1968-1978, years 1979-1998, and for years 1999 and later.

Source:

Compressed Mortality File is produced by the National Center for Health Statistics (NCHS), at the Centers for Disease Control and Prevention (CDC). See Data Source Information.

In WONDER: You can produce tables, maps, charts, and data extracts. Obtain death counts, crude death rates and age-adjusted death rates, standard error and 95% confidence intervals for rates for user-specified geographic, cause of death, and demographic criteria. Mortality measures can be obtained for several levels of geographic detail: national, Census region, Census division, state, county and urbanization level. The population estimates used as the denominator for rate calculations are also shown.

Variables:  You can limit and index your data by any and all of these variables: 

  1. Location - U.S., state, county
  2. Year - 1968-1978, 1979-1998 and 1999-2010
  3. Age Group - <1 year of age, 1-4 years, 5-9 years, 10-14 years, 15-19 years, 20-24 years, 25-34 years, 35-44 years, 45-54 years, 55-64 years, 65-74 years 75-84 years, and 85years and over;
    or infant age groups:  < 1 day, 1-6 days, 7-27 days, 28-364 days
  4. Race - 1999-present:  American Indian or Alaska Native, Asian or Pacific Islander, Black or African American, White;
    1968-1998:  Black, White, Other
  5. Hispanic Origin - 1999-present:  Hispanic or Latino, Not Hispanic or Latino, Not Stated
    1968-1998:  not available.
  6. Gender (Sex) - Female, Male
  7. Cause of Death - underlying cause of death, 1999-present:  ICD-10; 1979-1998:  ICD-9; 1968-1978:  ICD-8.
  8. Injury Intent and Mechanism - for analysis of injury mortality
  9. Urbanization - 1999-present: pick between the 2006 or the 2013 NCHS Urban-Rural Classification Scheme for Counties; 1968-1998: not available.

Statistical measures:  The following statistical measures are available as query results:

  1. Death Counts
  2. Crude Rates
  3. Age-Adjusted Rates
  4. 95% Confidence Intervals for Rates
  5. Standard Errors for Rates
  6. Percent of Total

Suppression and Unreliability: 

  • Sub-national data representing fewer than ten (0-9) persons are suppressed for year 1989 and later years. The suppression constraint affects death counts, birth counts, rates and associated confidence intervals and standard errors, as well as corresponding population figures. See Assurance of Confidentiality for more information.
  • Crude death rates and age-adjusted death rates are marked as "unreliable" when the death count is less than 20.

Compressed Mortality Topics: Compressed Mortality Data Request
Data Source Information
Additional Information
Age Adjustment of Death Rates
Assurance of Confidentiality
Contact for Data Questions
Frequently Asked Questions about Death Rates
International Classification of Diseases (ICD)
     ICD 10th revision
     ICD 9th revision
     ICD 8th revision
Locations: About County Geography Changes
Mortality Data
     Infant Mortality
     CMF Archives
Population Estimates
     Population Migration due to Hurricanes in 2005
Race Reporting
Suggested Citation

Additional Information

Suggested Data Source Citations

United States Department of Health and Human Services (US DHHS),
Centers for Disease Control and Prevention (CDC),
National Center for Health Statistics (NCHS),
Compressed Mortality File (CMF) on CDC WONDER Online Database.

  • The current release for years 1999 - 2010 is compiled from:
    CMF 1999-2010, Series 20, No. 2P, 2013
  • The current release for years 1979 - 1998 is compiled from:
    CMF 1968-1988, Series 20, No. 2A, 2000 and
    CMF 1989-1998, Series 20, No. 2E, 2003
  • The current release for years 1968 - 1978 is compiled from:
    CMF 1968-1988, Series 20, No. 2A, 2000
  • The archive release for years 1999 - 2009 is compiled from:
    CMF 1999-2009, Series 20, No. 2O, 2012
  • The archive release for years 1999 - 2008 is compiled from:
    CMF 1999-2008, Series 20, No. 2N, 2011
  • The archive release for years 1999 - 2007 is compiled from:
    CMF 1999-2007, Series 20, No. 2M, 2010
  • The archive release for years 1999 - 2006 is compiled from:
    CMF 1999-2006, Series 20, No. 2L, 2009
  • The archive release for years 1999 - 2005 is compiled from:
    CMF 1999-2005, Series 20, No. 2K, 2008
  • The archive release for years 1999 - 2004 is compiled from: 
    CMF 1999-2004, Series 20, No. 2J, 2007
  • The archive release for years 1999 - 2003 is compiled from:
    CMF 1999-2003, Series 20, No. 2I, 2006
  • The archive release for years 1999- 2002 is compiled from:
    CMF 1999-2002, Series 20, No. 2H, 2004
  • The archive release with data for years 1999 - 2001 is compiled from:
    CMF 1999-2001, Series 20, No. 2G, 2004
  • The archive release with data for years 1979 - 1999 is compiled from:
    CMF 1968-1988, Series 20, No. 2A, 2000 and
    CMF 1989-2000, Series 20, No. 2C 2001
The suggested citation including the original series for the data is shown below each table, chart or map.

For more information on archive data from previous releases, see CMF Archives.

Note:  The 1989 data records for age groups 10-14 years and 25-34 years were revised on May 9, 2007 due to a data discrepancy in CMF 1989-1998, Series 20, No. 2E, 2003. Previously, 27 deaths in the 10-14 year age group were incorrectly recorded as deaths in the 25-34 age group.

Contact For data questions that are not addressed in this document, e-mail nchsquery@cdc.gov.


Mortality Data: 

The mortality data on the Compressed Mortality File are based on information from all death certificates filed in the fifty states and the District of Columbia. Deaths of nonresidents (e.g. nonresident aliens, nationals living abroad, residents of Puerto Rico, Guam, the Virgin Islands, and other territories of the U.S.) and fetal deaths are excluded. Mortality data from the death certificates are coded by the states and provided to NCHS through the Vital Statistics Cooperative Program or coded by NCHS from copies of the original death certificates provided to NCHS by the State registration offices. For more information, see Technical Appendix from Vital Statistics of United States: 1999 Mortality.

  • About cause of death classification: 
    • Cause of death on the CMF is the underlying cause-of-death, which is defined by the World Health Organization (WHO) as "the disease or injury which initiated the train of events leading directly to death, or the circumstances of the accident or violence which produced the fatal injury." Underlying cause-of-death is selected from the conditions entered by the physician on the cause of death section of the death certificate. When more than one cause or condition is entered by the physician, the underlying cause is determined by the sequence of conditions on the certificate, provisions of the ICD, and associated selection rules and modifications.
    • Underlying cause of death is classified in accordance with the International Classification of Disease. Deaths for 1968-1978 are classified using the Eighth Revision (ICD-8). Deaths for 1979-1998 are classified using the Ninth Revision (ICD-9). Deaths for 1999 and beyond are classified using the Tenth Revision (ICD-10).
    • The ICD-9 and ICD-10 classification systems are quite different. The ICD-9 system has a 4-digit numeric structure and about 5,000 categories for classifying causes of death. The ICD-10 system has a 4-digit alphanumeric coding structure and about 8,000 categories for classifying causes of death. Comparison of ICD-9 and ICD-10 shows that new chapters have been added to the ICD, old chapters have been rearranged, causes of death have been regrouped, and titles have changed. The differences between ICD-9 and ICD-10 make direct comparisons of cause of death difficult and result in discontinuities in cause-of-death trends. NCHS conducted a comparability study to measure the discontinuities between the Ninth and Tenth Revisions and computed comparability ratios for major causes of death. For further information on comparability of mortality causes between ICD-9 and ICD-10, please reference Comparability of Causes of Death Between ICD Revisions. For information on comparability of mortality causes between ICDA-8 and ICD-9, please see Estimates of Selected Comparability Ratios Based on Dual Coding of 1976 Death Certificates by the Eighth and Ninth Revisions of the International Classification of Diseases.
    • For deaths due to injuries and poisonings that occurred during 1968-1998, the External cause is coded (E800-E999) rather than the Nature of Injury (800-999). The letter "E" is included in the code in CDC WONDER.
    • Beginning with data for 2001, NCHS introduced categories *U01-*U03 for classifying and coding deaths due to acts of terrorism. The asterisks before the category codes indicate that they are not part of the International Classification of Diseases, Tenth Revision (ICD-10). Description of the specific 4-digit codes can be found at NCHS Classifications of Diseases, and Functioning & Disability: Appendix I. Deaths classified to the terrorism categories are included in the categories for Assault (homicide) and Intentional self-harm (suicide) in the 113 cause-of-death list. Additional information on these new categories can be found at Classification of Death and Injury Resulting from Terrorism.
    • Reporting anomalies:
      • About Georgia reporting anomalies in 2008-2009:  Circumstances in Georgia for the year 2008 have resulted in unusually high death counts for the Place of Death category “Place of death unknown” for years 2008-2009, and for the ICD-10 cause of death code R99, “Symptoms, signs and abnormal clinical and laboratory findings, not elsewhere classified” in year 2008. Caution should be used in interpreting these data. For more information, see Deaths: Final Data for 2008.
      • About New Jersey reporting anomalies in 2009:  Circumstances in New Jersey for the year 2009 have resulted in unusually high death counts for the ICD-10 cause of death code R99, "Symptoms, signs and abnormal clinical and laboratory findings, not elsewhere classified" and therefore unusually low death counts in other ICD-10 codes, most notably R95 "Sudden Infant Death Syndrome" and X40-X49 "Unintentional poisoning." Caution should be used in interpreting these data. Deaths: Final Data for 2009.
      • About Allen Parish, Louisiana reporting anomalies in 2006-2008:  Deaths for Allen Parish, Louisiana (FIPS code 22003] in years 2006 through 2008 are under reported due to problems with registering the deaths with the Louisiana Vital Statistics Office.
      • About cause of death classification changes:  Changes to cause of death classification affect reporting trends. For more information, see Changes in ICD-10 codes.
  • About race and ethnicity reporting: 
    • Race and Hispanic origin are reported separately on the death certificate in accordance with standards set forth by the Office of Management and Budget. The American Indian or Alaska Native race category includes: North, Central, and South American Indians, Eskimos, and Aleuts. The Asian or Pacific Islander race category includes Chinese, Filipino, Hawaiian, Japanese, and Other Asian or Pacific Islanders.
    • Hispanic origin was not reported on the death certificate for some deaths. On the mortality file, missing Hispanic origin information is coded as “not stated”. There is no corresponding population figure for this group. Therefore, deaths with Hispanic origin not stated are excluded when death rates are calculated by Hispanic origin.
    • Information included on the death certificate about the race and Hispanic ethnicity of the decedent is reported by the funeral director as provided by an informant, often the surviving next of kin, or, in the absence of an informant, on the basis of observation. Race and ethnicity information from the census is by self-report. To the extent that race and Hispanic origin are inconsistent between these two data sources, death rates will be biased. Studies have shown that persons self-reported as American Indian, Asian, or Hispanic on census and survey records may sometimes be reported as white or non-Hispanic on the death certificate, resulting in an underestimation of deaths and death rates for the American Indian, Asian, and Hispanic groups. Bias also results from undercounts of some population groups in the census, particularly young black males, young white males, and elderly persons, resulting in an overestimation of death rates. In " Quality of death rates by race and Hispanic origin:  A summary of current research, 1999," the authors estimate that the misclassification and under-coverage result in overstated death rates for the white and black populations (1% and 5%, respectively) and understated death rates for other population groups (American Indians, 21%; Asian or Pacific Islanders, 11%; and Hispanics, 2%). See also The validity of race and Hispanic Origin reporting on death certificates in the United States.
    • For 1979-2002, all 50 States and the District of Columbia collected race data on the death certificates using four single-race categories (American Indian or Alaska Native, Asian or Pacific Islander, Black, and White) in accordance with the 1977 OMB standards, allowing only a single race to be reported. Beginning with the 2003 data year, some States began collecting race data in accordance with the 1997 OMB standards, allowing one or more of five race categories to be reported. In order to provide uniformity and comparability of mortality data during the transition from the single-race format to the multiple-race format, NCHS is ‘‘bridging’’ the race responses of those for whom more than one race is reported (multiple race) to one of the single-race categories. The bridging procedure is similar to the procedure used to bridge multiple-race population estimates. Multiple-race decedents are imputed to a single race (White, Black, American Indian or Alaska Native, or Asian or Pacific Islander) according to their combination of races, Hispanic origin, sex, and age indicated on the death certificate. The imputation procedure is described in detail at NCHS Procedures for Multiple-Race and Hispanic Origin Data.
    • For more discussion of race and ethnicity data in CMF, see Race and Ethnicity Questions.
  • About "Not Stated" age or ethnicity:
    • Data for the "Not Stated" age category or the "Not Stated" Hispanic Origin category cannot be combined with any other specified age group or Hispanic Origin categories.
    • The "Not Stated" group is included in the "All Ages" or the "All" Hispanic Origin categories.
    • Death rates are not calculated for the "Not Stated" groups because there are no corresponding population denominator data for these groups.
  • About the "unknown county" in Georgia for years 1981-1991: 
    • For the years 1988 through 1991, if there were three or fewer deaths for a given Georgia county of residence (of deaths occurring in Georgia) with HIV infection (ICD codes *042-*044, 795.8) cited as a cause-of-death (underlying or non-underlying cause), these records were assigned a "missing" place of residence code (location code (FIPS code 13999).
    • These deaths do not appear in county death rates, but these deaths are included in the state and national death rates.

Population Data: 

The population estimates on the Compressed Mortality File are U.S. Census Bureau estimates of U.S. national, state, and county resident populations. The estimates for 1968, 1969, 1971-1979, 1981 - 1989, 1991- 1999 and 2001-2009 are intercensal estimates of July 1 resident populations. The 1970, 1980, 1990, 2000 and 2010 population estimates are April 1 modified census counts. The 2001-2009 population estimates are revised intercensal estimates of the July 1 resident population (released by NCHS on 10/26/2012). The population estimates for 1991-2010 have bridged-race categories. For more information, see notes below and Population Information.

Sources of the population data, by year and type of estimate for the current 1968-1978, 1979-1998 and 1999-2010 CMF
Date of estimate National Region, Division, State County
July 1, 1968 - July 1, 1969 National intercensal estimates Obtained by linear extrapolation from the corresponding July 1, 1970 and July 1, 1971 state population estimates Obtained by linear extrapolation from the corresponding July 1, 1970 and July 1, 1971 county population estimates
April 1, 1970 Sum of county modified census counts Sum of county modified census counts County modified census counts
July 1, 1971 - July 1, 1979 National intercensal estimates Sum of county intercensal estimates County intercensal estimates
April 1, 1980 Sum of county modified age-race-sex census counts Sum of county modified age-race-sex census counts County modified age-race-sex census counts
July 1, 1981 - July 1, 1989 National intercensal estimates State intercensal estimates County intercensal estimates
April 1, 1990 Sum of county modified age-race-sex census counts Sum of county modified age-race-sex census counts County modified age-race-sex census counts
July 1, 1991 - July 1, 1999 Sum of bridged-race county intercensal estimates Sum of bridged-race county intercensal estimates Bridged-race county intercensal estimates
April 1, 2000 Sum of bridged-race county modified age-race-sex census counts Sum of bridged-race county modified age-race-sex census counts Bridged-race county modified age-race-sex census counts
July 1, 2001 - July 1, 2009 Sum of bridged-race county revised intercensal estimates Sum of bridged-race county revised intercensal estimates Bridged-race county revised intercensal estimates
April 1, 2010 Sum of bridged-race county modified census counts Sum of bridged-race county modified census counts Bridged-race county modified age-race-sex census counts
Note:  all population estimates are estimates of the resident population of the United States.

  • About national, state, and county population estimates:
    There are national, state, and county population estimates on the CMF; for some years they are consistent with each other (sum to the same totals), for some years they are not. For 2001 - 2008, the state and county estimates on the CMF are consistent with each other, but not with the national estimates. For these postcensal years, the state and county estimates for all postcensal years are replaced with estimates from the most recent postcensal series when the file is updated each year. By contrast, the national estimates for the postcensal years are not revised when the file is updated annually, so that the national population estimates on the CMF will be the same as those used by NCHS to calculate published death rates. In order to replicate state or county death rates for year 2001 and later calculated using earlier versions of the CMF, please use the appropriate archived data.
  • About revisions of postcensal estimates:
    Changes were implemented in the postcensal estimates methodology used for Vintages 2007, 2008, and 2009. Additionally, Hurricanes Katrina and Rita in 2005 necessitated special modification of population estimates for the affected areas beginning with Vintage 2006. The methodology changes implemented for Vintage 2007-Vintage 2009 affect comparison of population estimates across the three vintages, as well as comparison of the estimates from these vintages with those from earlier vintages. At the national level, the methodology changes implemented for Vintage 2009 resulted in an upward shift of the Vintage 2009 postcensal population estimates when compared to those from the Vintage 2008 series. At the State and county level, some race and age groups experienced substantial changes (comparison of July 1, 2008 estimates from the Vintage 2008 series with the July 1, 2008 and July 1, 2009 estimates from the Vintage 2009 series). The net impact of the methodology changes implemented for Vintage 2008 was a downward shift of the Vintage 2008 postcensal population estimates when compared to those from the Vintage 2007 series. The net impact of the methodology changes implemented for Vintage 2007 series was a downward shift when compared to the Vintage 2006 estimates.

    For the archive CMF 1999-2009, archive CMF 1999-2007 and the archive CMF 1999-2006:  The 2005 population estimates for Alabama, Louisiana, Mississippi and Texas reflect population changes that occurred after Hurricane Katrina and Rita in 2005. To accommodate geographic shifts of the Alabama, Louisiana, Mississippi, and Texas populations resulting from Hurricanes Katrina and Rita in 2005, the U.S. Census Bureau developed adjustments in the methodology for state and county population estimates. See methodology for more information.

    For the archive CMF 1999-2005:  The 2005 population estimates for Alabama, Louisiana, and Mississippi do not reflect population changes that occurred after Hurricane Katrina. The July 1st 2005 population estimates for Alabama, Louisiana, and Mississippi do not reflect population changes that occurred after Hurricane Katrina on August 29, 2005. The populations of these three states and of Orleans parish in Louisiana, Baldwin and Mobile counties in Alabama, and George, Harrison, and Rankin counties in Mississippi were reduced following Hurricane Katrina; East Baton Rouge parish in Louisiana had an increased population. As a result, 2005 death rates for these areas (except East Baton Rouge) calculated using the population estimates on this file will be too low; the death rates calculated for East Baton Rouge will be somewhat too high. Death rates for Orleans parish are especially affected. At the time of the CMF 1999-2005 release, alternative population estimates for these areas were being developed by the National Cancer Institute.

  • About bridged-race population estimates:
    The population estimates on the CMF for 1991 and later are bridged-race population estimates. Race bridging refers to making data collected using one set of race categories consistent with data collected using a different set of race categories, to permit estimation and comparison of race-specific statistics at a point in time or over time. Census 2000 permitted individuals to report more than one race which resulted in “multiple-race” population counts. The death certificate currently used by most States only permits one race to be reported for decedents (single-race categories). Thus, the multiple-race data collected on the 2000 census are not comparable with the single-race mortality data. Therefore, multiple-race population estimates have been bridged to single-race categories. For more information see U.S. Census Populations with Bridged-race Categories.
  • About live births and "under 1 year of age" population estimates:
    • For years 1999 and later, the number of live births and estimates of the population under 1 year of age are both available as denominators in rate calculation on the CMF. The number of live births and the population estimate for the "Under one year of age" group differ slightly. NCHS live-birth data are used as the denominators in death rate calculations for "Infant Age Groups" so that neonatal mortality, post neonatal mortality, and infant mortality rates can be calculated. The population under one year of age is used as the denominator in death rate calculations involving the “Under 1 year” age group. Prior to August 2006, the number of live births was used for all rate calculations involving the “Infant Age Groups” and the "Under 1 year" age group. For years before 1999, the number of live births is used for all rate calculations involving both the “Infant Age Groups” and the "Under 1 year" age group. For more information, see Mortality for Infants.
    • Population estimates for the age groups “Under 1 year” and “1-4 years” did not appear on some Census population files For 1968-89 state and county estimates and for 1970 and 1980 national estimates, the estimates for the “0-4 years” age group were multiplied by 0.8 to obtain estimates for the “1-4 years” age group; estimates for the “under 1 year” age group were then obtained by subtraction.
  • About rounding:
    For years 1968-69 and 1971-79, and 1981-1989, NCHS used national population estimates rounded to the nearest 1,000 to calculate published death rates. Estimates were rounded after aggregation across race, sex, or age groups. The national population estimates on the CMF for 1968-69 and 1971-79 are rounded to the nearest 1,000 in accordance with this practice. However, while NCHS rounded after aggregating across age, race, and/or sex groups, the population estimates in the CMF for aggregate groups are obtained by summing already rounded population estimates. As a result, national death rates for aggregate groups calculated using the rounded estimates on the CMF may differ slightly from those published by NCHS. Population estimates for years 1981-89 on the CMF are unrounded; hence, death rates calculated using the CMF also may differ slightly from those published by NCHS.
  • About "Not Stated" age or ethnicity: Beginning with the August 2009 release of CMF data on WONDER online databases, data for the "Not Stated" age group or the "Not Stated" Hispanic Origin category cannot be combined with any other age group or Hispanic Origin categories, except for "All" age groups or "All" Hispanic Origin categories. Deaths for the "Not Stated" groups are not distributed among the other groups. Rates are not calculated for the "Not Stated" groups because there are no corresponding population denominator data for these groups.
    • You can select data for "All" age groups, or only for the "Not Stated" age group, or for any combination of the other age groups. Data for the "Not Stated" age group cannot be combined with the other specific age groups. Deaths of persons with "Unknown" or "Not Stated" age are included in "All" counts and rates, but are not distributed among age groups, so are not included in age-specific counts, age-specific rates or in any age-adjusted rates.
    • You can select data for "All" Hispanic Origin categories, or only for the "Not Stated" category, or for any combination of the other categories. Data for the "Not Stated" Hispanic Origin category cannot be combined with the other specific categories. Deaths of persons with unknown or "Not Stated" Hispanic Origin are included in "All" counts and rates, but are not distributed among ethnicity categories, so are not included in the specific Hispanic Origin group counts or rates.


Compressed Mortality Data Request

Output You can produce tables, maps, charts, and data extracts. Obtain death counts, crude rates and age-adjusted rates, select specific disease and demographic criteria to produce cross-tabulated mortality measures. Data are organized into three levels of geographic detail:  national, state (including multi-state regions) and county. The population estimates used as the denominator for rate calculations are also shown.
Variables You can limit and index your data by any and all of the variables.
How? The Request screen has sections to guide you through the making a data request as step-by-step process. However, to get your first taste of how the system works, you might want to simply press any Send button, and execute the default data request. The data results for your query appear on the Table screen. After you get your data results, try the Chart and Map screens. Or export your data to a file (tab-delimited line listing) for download to your computer.
For more information, see the following:
Quick Start Guide
Step 1, Organize and label results
Step 2, Select location and time
Step 3, Select demographics
Step 4, Select cause of death
Step 5, Select rate options
Step 6, Other options
'By-Variables' Select variables that serve as keys (indexes) for organizing your data. See How do I organize my data? for more information.
Note:   To map your data, you must select at least one geographical location as a "By-Variable" for grouping your data, such as State or County.
Help Click on any button labeled "Help", located to the right hand side of the screen at the top of each section. Each control's label, such as the "Location" label next to the Location entry box, is linked to the on-line help for that item.
Send Sends your data request to be processed on the CDC WONDER databases. The Send buttons are located on the bottom of the Request page, and also in the upper right corner of each section, for easy access.


Step 1. Organize table layout:

There are three basic steps you may take in this first section. However, you may also simply click the "Send" button to request the default query, which reports national death counts and crude death rates by age group for all causes and all years.

  1. Pick indexes or cross-tabulation for your results.
    Select up to five categorical variables that serve as indexes or keys for grouping your data. See Group Results By below for hints.
  2. Select statistical measures reported in your results.
    The following summary statistical measures are always available as query results:
    1. Death Counts
    2. Crude Rates
    These optional measures are also available:
    1. Age-Adjusted Rates (optional, see "Step 5" for age-adjusted rates)
    2. 95% Confidence Intervals for Rates (optional, see "Step 1" for crude rates and "Step 5" for age-adjusted rates)
    3. Standard Error for Age-Adjusted Rates (optional, see "Step 5" for age-adjusted rates)
    4. Percent of Total (optional, see "Step 1")
  3. Provide a title for your results.
    Enter any desired description in the Title box, to display above your results. Titles are optional.


Group Results By...

Select up to five variables that serve as keys for grouping your data. For example, you could select to group (summarize, stratify, index) your data by State and by County.

How?    See How do I organize my data? for more information.

Hints:   

  1. When age-adjusted rates are calculated, you cannot group the data by Age Group.
  2. About charts:
    You cannot make charts when your data has more than two By-Variables.
  3. About maps:
    To make a map, you must request data with a geographic location variable, such as State or County, as a "By-Variable." Then click the Map tab.

Death Counts

The death counts in the data represent deaths that occurred in the 50 United States and the district of Columbia, for the legal place of residence of the decedent. See Mortality data for more information.

Sub-national data representing fewer than ten persons (0-9) are suppressed for year 1989 and later years. See Assurance of Confidentiality for more information.


Crude Rates

Crude Rates are expressed as the number of deaths reported each calendar year per the factor you select. The default factor is per 100,000 population, reporting the death rate per 100,000 persons.

Crude Rate = Count / Population * 100,000

See Frequently Asked Questions about Death Rates .

Hints:
  • Rates calculated with population estimates are per 100,000 persons by default. However, infant mortality rates are calculated per 1,000 live births by default. See Step 5 to select the factor for rate calculations.
  • Select the precision for rate calculations in the Other Options section. When the rate calculated for a small numerator (incidence count) is zero, you may increase the precision to reveal the rate by showing more numbers to the right of the decimal point.
Notes:
  • Rates for small populations should be interpreted with caution.
  • Sub-national data representing fewer than ten persons (0-9) are suppressed for year 1989 and later years. See Assurance of Confidentiality for more information.
  • Rates are marked as "unreliable" when the death count is less than 20.
  • Crude rates are helpful in determining the need for services for a given population, relative to another population, regardless of size. Crude rates are influenced by the underlying age distribution of the state's population. Even if two states have the same age-adjusted rates, the state with the relatively older population (as demonstrated by having a higher median age) will have higher crude rates because incidence or death rates for most cancers increase with increasing age.
  • The population estimates for the denominators of incidence rates are race-specific (all races, white, black, and other races combined) and sex-specific population estimates. The population estimates are aggregated from the most detailed level selected. For example, if you have requested data for the nation grouped by state and by county, then the populations are the county-level population estimates aggregated to the state and national summaries. See Population Denominator Data Sources below for more information.
  • The population for "Infant age groups" is the number of live births in the given time period. See Mortality for Infants for more information.
  • Prior to August 2006, the number of live births had been substituted for the "under 1 year" age group population estimates. This substitution affected the "under 1 year" and "all ages" death rates slightly, when compared to published tables that rely solely on population estimates.

Age-Adjusted Rates

Age-adjusted rates are requested in the "Select rate options" section of the Request form.

Age-adjusted death rates are weighted averages of the age-specific death rates, where the weights represent a fixed population by age. They are used to compare relative mortality risk among groups and over time. An age-adjusted rate represents the rate that would have existed had the age-specific rates of the particular year prevailed in a population whose age distribution was the same as that of the fixed population. Age-adjusted rates should be viewed as relative indexes rather than as direct or actual measures of mortality risk. However, you can select other standard populations, or select specific population criteria to determine the age distribution ratios. See Frequently Asked Questions about Death Rates for more information.

The rates of almost all causes of death vary by age. Age adjustment is a technique for "removing" the effects of age from crude rates, so as to allow meaningful comparisons across populations with different underlying age structures. For example, comparing the crude rate of heart disease in Florida to that of California is misleading, because the relatively older population in Florida will lead to a higher crude death rate, even if the age-specific rates of heart disease in Florida and California are the same. For such a comparison, age-adjusted rates are preferable. Age-adjusted rates should be viewed as relative indexes rather than as direct or actual measures of mortality risk.

The Compressed Mortality online database and NCHS age-adjusts death rates using the direct method. That is, by applying age-specific death rates (Ri) to the U.S. standard population age distribution.

R' = S i ( Psi / Ps ) R i

where Psi is the standard population for age group i and Ps is the total U.S. standard population (all ages combined).

In the direct method, a standard age distribution is chosen and the age-specific death rates are weighted according to the standard. A reasonable choice for the standard is the U.S. total population (all races, both genders) for the year under study. To permit comparison of death rates from year to year, a standard population is used. Beginning with the 1999 data year, NCHS adopted the year 2000 projected population of the United States as the standard population. This new standard replaces the 1940 standard population that was used by NCHS for over 50 years. The new population standard affects the level of mortality and to some extent trends and group comparisons. Of particular note are the effects on race comparison of mortality. For detailed discussion, see: 

Anderson RN, Rosenberg HM. Age standardization of death rates: Implementation of the year 2000 standard. National Vital Statistics Reports; volume 47 number 3. Hyattsville, Maryland. National Center for Health Statistics. 1998.

Beginning with 2003 data, the traditional standard million population along with corresponding standard weights to six decimal places were replaced by the projected year 2000 population age distribution (see 2000 Standard Population below). A forthcoming report will describe the change in more detail. The effect of the change is negligible and does not significantly affect comparability with age-adjusted rates calculated using the previous method.

Age-Adjusted Rates Hints:

  • Age-Adjusted Rates are optional, see Rate Options to select age-adjusted rates.
  • Rates are calculated per 100,000 population by default. See Rate Options to select the factor for rate calculations.
  • Select the precision for rate calculations in the Other Options section. When the rate calculated for a small numerator (incidence count) is zero, you may increase the precision to reveal the rate by showing more numbers to the right of the decimal point.
  • Age-adjusted rates cannot be calculated when the data are grouped by Age Group.
  • Age-adjusted rates are not calculated when only one age group is selected (the effect is a ratio of one).
  • Age-adjusted rates are not available for "Infant age groups" because the populations for these age groups are the number of live births in the given time period, the same population denominator for each infant age group.
  • The selection must include both age groups "5 - 9 years" and "10 - 14 years" in order to calculate age-adjusted rates for this population. The combined age groups are required because the reference standard population has the age group "5 - 14 years."
  • The selection must include both age groups "15 - 20 years" and "20 - 24 years" in order to calculate age-adjusted rates for this population. The combined age groups are required because the reference standard population has the age group "15 - 24 years."

Notes:

  • Sub-national data representing fewer than ten persons (0-9) are suppressed for year 1989 and later years. See Assurance of Confidentiality for more information.
  • Rates are marked as "unreliable" when the death count is less than 20.
  • If a "non-standard" population is selected for age-adjusted rates, then the actual population estimates for the specified year that are on the Compressed Mortality File are used to determine the specific age-distribution ratios (or weights) used in the calculation.
  • The following standard populations (see tables below) are used for computing age-adjusted rates:

Year 2000 Standard Population for the United States
 
   Age Number
 All ages 274,633,642
 Under 1 year 3,794,901
 1-4 years 15,191,619
 5-14 years 39,976,619
 15-24 years 38,076,743
 25-34 years 37,233,437
 35-44 years 44,659,185
 45-54 years 37,030,152
 55-64 years 23,961,506
 65-74 years 18,135,514
 75-84 years 12,314,793
 85 years and over4,259,173

      *  Based on year 2000 projected population.

Year 2000 Standard Million Population for the United States
Numbers and All Ages Proportions (Weights) *
 
   Age Number Weight
 All ages 1,000,000 1.000000
 Under 1 year 13,818 0.013818
 1-4 years 55,317 0.055317
 5-14 years 145,565 0.145565
 15-24 years 138,646 0.138646
 25-34 years 135,573 0.135573
 35-44 years 162,613 0.162613
 45-54 years 134,834 0.134834
 55-64 years 87,247 0.087247
 65-74 years 66,037 0.066037
 75-84 years 44,842 0.044842
 85 years and over 15,508 0.015508

      *  Based on year 2000 projected population.
Note that these weights only apply to the all ages population,
the weights are calculated dynamically when age groups are selected.

Year 1970 Standard Million Population for the United States
Numbers and All Ages Proportions (Weights) *
 
   Age Number Weight
 All ages 1,000,000 1.000000
 Under 1 year 18,102 0.018102
 1-4 years 66,314 0.066314
 5-14 years 200,508 0.200508
 15-24 years 174,406 0.174406
 25-34 years 122,569 0.122569
 35-44 years 113,614 0.113614
 45-54 years 114,265 0.114265
 55-64 years 91,480 0.091480
 65-74 years 61,195 0.061195
 75-84 years 30,112 0.030112
 85 years and over 7,435 0.007435

      *  Based on the year 1970 population.
Note that these weights only apply to the all ages population,
the weights are calculated dynamically when age groups are selected.

Year 1940 Standard Million Population for the United States
Numbers and All Ages Proportions (Weights) *
 
   Age Number Weight
 All ages 1,000,000 1.000000
 Under 1 year 15,343 0.015343
 1-4 years 64,718 0.064718
 5-14 years 170,355 0.170355
 15-24 years 181,677 0.181677
 25-34 years 162,066 0.162166
 35-44 years 139,237 0.139237
 45-54 years 117,811 0.117811
 55-64 years 80,294 0.080294
 65-74 years 48,426 0.048426
 75-84 years 17,303 0.017303
 85 years and over 2,770 0.002770

      *  Based on the year 1940 population.
Note that these weights only apply to the all ages population,
the weights are calculated dynamically when age groups are selected.

95% Confidence Intervals for Rates

You can request 95% confidence intervals calculated for mortality rates. The method for confidence intervals calculated for 100 or more deaths differs slightly from the method for confidence intervals calculated for 99 or fewer deaths.

How? Click the check box to indicate the desired measure. Confidence intervals for crude rates are requested in "section 1" of the Request form. Confidence intervals for age-adjusted rates are requested in the "Select rate options" section of the Request form, after you click one of the buttons to calculate age-adjusted rates in this same section.

Notes:

  • The method for confidence intervals calculated for 100 or more deaths:
    The lower 95% confidence interval is the crude death rate minus (1.96 times the standard error of the rate). The upper 95% confidence interval is the crude death rate plus (1.96 times the standard error of the rate).
    LCI = R - 1.96 * S (R)
    UCI = R + 1.96 * S (R)
  • The method for confidence intervals calculated for 99 or fewer deaths:
    The lower 95% confidence interval is the crude death multiplied by the lower 95% confidence limit factor for a death rate based on a Poisson variable of the number of deaths. The upper 95% confidence interval is the crude death rate multiplied by the upper 95% confidence limit factor for a death rate based on a Poisson variable of the number of deaths. See Vital Statistics of the United States: Mortality, 1999: Technical Appendix Table S.
    LCI = R * L (0.95, D)
    UCI = R * U (0.95, D)
  • Where:
  • For more information, refer to


Standard Error for Age-Adjusted Rates

You can request standard errors calculated for age-adjusted mortality rates.

How? Click the check box to indicate the desired measure in the "Select rate options" section of the Request form, after you click one of the buttons to calculate age-adjusted rates in this same section.

Notes:


Percent of Total

You can request to show percent of total as an optional statistical measure in "Step 1." This option adds a "Percent of Total" column to the results table, and each row shows the percentage that the number of deaths represented in the row contributes to the total number of deaths for that table. This measure is not available when suppression constraints apply to the data in the results table, such as when the deaths are suppressed for any single row in any categorical section in the table.


Step 2. Select location:

Data are available for the United States by Region, Census Division, State, County. Select the location(s) for the query. Any number of locations can be specified here.

How? Hints:
  • The default is all values (the United States).
  • The Advanced mode lets you easily pick several items from different parts of the list. Items are not selected until you click the "Move" button in Advanced mode. You may also enter values by hand, one code per line, in the Advanced mode. Use the Finder to see the correct code format. For example, 02 is the Alaska state code.
  • The "plus" symbol, "+" indicates that you can open the item, to see more items below it.
  • The results to a search are shown in blue, and indicated by ">".


Region

Regions are multi-state groups. For regional data, you can group by Region, or you can select any combination of individual regions. There are two types of regions available, Census Regions and Health and Human Services (HHS) Regions.

How?   Notes:  
  • Region is based on the person's legal state of residence at the time of death.
  • The Regions are identified by both name and codes in data extracts.


Census Regions

The United States is split into 4 Census Regions: Northeast, Midwest, South and West. The states that comprise each region are shown below.
How?   Notes:  
  • Census Region is based on the person's legal state of residence at the time of death.
  • The Regions are identified by both name and codes in data extracts.

State abbreviation and name   FIPS code
________________________________________ 
Northeast Census Region: 
CT    Connecticut               09
ME    Maine                     23
MA    Massachusetts             25
NH    New Hampshire             33
NJ    New Jersey                34
NY    New York                  36
PA    Pennsylvania              42
RI    Rhode Island              44
VT    Vermont                   50
________________________________________
Midwest Census Region: 
IL    Illinois                  17
IN    Indiana                   18
IA    Iowa                      19
KS    Kansas                    20
MI    Michigan                  26
MN    Minnesota                 27
MO    Missouri                  29
NE    Nebraska                  31
ND    North Dakota              38
OH    Ohio                      39
SD    South Dakota              46
WI    Wisconsin                 55
________________________________________ 
South Census Region: 
AL    Alabama                   01
AR    Arkansas                  05
DE    Delaware                  10
DC    District of Columbia      11
FL    Florida                   12
GA    Georgia                   13
KY    Kentucky                  21
LA    Louisiana                 22
MD    Maryland                  24
MS    Mississippi               28
NC    North Carolina            37
OK    Oklahoma                  40
SC    South Carolina            45
TN    Tennessee                 47
TX    Texas                     48
VA    Virginia                  51
WV    West Virginia             54
________________________________________ 
West Census Region: 
AK    Alaska                   02
AZ    Arizona                  04
CA    California               06
CO    Colorado                 08
HI    Hawaii                   15
ID    Idaho                    16
MT    Montana                  30
NV    Nevada                   32
NM    New Mexico               35
OR    Oregon                   41
UT    Utah                     49
WA    Washington               53
WY    Wyoming                  56
________________________________________ 


Census Division

Census Divisions are multi-state groups, sub-sets of Census Regions. You can group by Census Division, or select any combination of individual Census divisions.
How?   Notes:  


HHS Regions

The Department of Health and Human Services (HHS) groups the 50 states, the District of Columbia, and the U.S. territories into ten reporting regions, referred to as the HHS regions. Any number of locations can be specified here.

How?   Notes:  
  • The CMF does not have data for Puerto Rico or any of the U.S. territories. Therefore, data for Puerto Rico and the Virgin Islands are not included in HHS Region 2; data for Guam and American Samoa are not included in HHS Region 9.
  • HHS Region is based on the person's legal state of residence at the time of death.
  • When the data are exported, separate columns show both the label and the code for each value. To see the full list of labels and code values, request data grouped by this region for the "All" and export the results.


   Health and Human Services (HHS) Regions List of States
  1 Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, Vermont
  2 New Jersey, New York (data for Puerto Rico, Virgin Islands are not included in CMF)
  3 Delaware, District of Columbia, Maryland, Pennsylvania, Virginia, West Virginia
  4 Alabama, Florida, Georgia, Kentucky, Mississippi, North Carolina, South Carolina, Tennessee
  5 Illinois, Indiana, Michigan, Minnesota, Ohio, Wisconsin
  6 Arkansas, Louisiana, New Mexico, Oklahoma, Texas
  7 Iowa, Kansas, Missouri, Nebraska
  8 Colorado, Montana, North Dakota, South Dakota, Utah, Wyoming
  9 Arizona, California, Hawaii, Nevada (data for American Samoa and Guam are not included in CMF)
10 Alaska, Idaho, Oregon, Washington


State

For state level data, you can select any combination of individual states. Or group by State and leave the Location Finder selection at the default (all locations or the 50 United States and the District of Columbia).
How?   Notes:  
  • The state coded represents the person's place of legal residence at the time of death.
  • The states and the District of Columbia are identified by both state name and Standard Federal Information Processing (FIPS) codes in data extracts. See About FIPS Codes below.


County

County-level data are available for the fifty states and the District of Columbia. In this step, you can select any combination of individual counties or leave the Location Finder at the default setting of “All Counties."
How?   Notes:  
  • The county coded represents the person's place of legal residence at the time of death.
  • The counties and the District of Columbia are identified by both county name and Standard Federal Information Processing (FIPS) codes in data extracts.
  • About FIPS Codes:   The FIPS State and county codes were established by the National Bureau of Standards, U.S. Department of Commerce in 1968. This standard set of codes provides names and codes for counties and county equivalents of the 50 States of the United States and the District of Columbia. Counties are considered to be the "first order subdivisions" of each State, regardless of their local designation (county, parish, borough, census area). Washington, D.C.; the consolidated government of Columbus City, Georgia; the independent cities of the States of Maryland, Missouri, Nevada, and Virginia; and the census areas and boroughs of Alaska are identified as county equivalents. The system is standard throughout the Federal Government. The State codes are ascending, two-digit numbers; the county codes are ascending three-digit numbers. For both the State and county codes, space has been left for new States or counties. Some changes in the FIPS codes have occurred since 1968. See Locations: About County Geography Changes for information on how these changes affect the Compressed Mortality data.
  • About County Changes:   Changes in county geography occur from time to time. Creation and deletion of counties results in missing data for those counties for some years. County boundary changes can result in substantial increases or decreases in the population of the affected counties and hence impact birth and death counts, population estimates, and birth and death rates for those counties. See Locations: About County Geography Changes for information on how county geography changes affect CMF data.


Step 3. Select demographics, year, and urbanization level:
Limit your data for any of the following data elements:
  1. Age Group - infant age groups are also available.
  2. Gender - Female, Male
  3. Race - American Indian or Alaska Native, Asian or Pacific Islander, Black or African American, White.
    For data before Vintage 2006, these 3 categories are available:  Black / African American, Other Races Combined, White.
  4. Hispanic Origin - Hispanic or Latino, Not Hispanic or Latino, Not Stated;
    available in CMF Vintage 2006 and later releases only.
  5. Year - 1999-200, or 1979-1998, or 1968-1978
  6. Urbanization - classifies population density and other factors;
    choose between the 2013 or the 2006 NCHS Urban-Rural Classification Scheme for Counties,
    available in CMF Vintage 2003 and later vintages only.

Age Groups

Use the radio buttons to select either the Age Group list or the Infant Age Groups list. From the selected list you can choose “All Ages” or any combination of the individual age groups in the list. “All Ages” from the Age Group list is the default.

  1. Age Group list - The following age groups are available from this list:  all age groups, under 1 year, 1 - 4 years, 5 - 9 years, 10 - 14 years, 15 - 19 years, 20 - 24 years, 25 - 34 years, 35 - 44 years, 45 - 54 years, 55 - 64 years, 65 - 74 years, 75 - 84 years, 85 years and over, age not stated.
  2. Infant Age Groups list - The following age groups are available from this list:  all infant age groups, under 1 day, 1 - 6 days, 7 - 27 days, 28 - 364 days.

How?  

Notes:  

  • When an age group(s) is selected from the Infant Age Groups list, the number of live births (not the number of infants in the age category(ies)) is used as the denominator in rate computations. Thus, the denominator used for infant age group rate calculations is the same regardless of how many infant age groups are selected. When a rate is computed for an aggregate of several years, the number of live births for each year in the time period are summed together. For more information about infant age groups, see Mortality for Infants .
  • For years 1999-2009:  when an age group(s) is selected from the Age Group list, Census Bureau population estimates for each age group are used as the denominator in rate computations. When a rate is computed for an aggregate of several years, the population estimates for each year in the time period are summed together.
  • For years 1965-1998:  When an age group(s) is selected from the Age Group list, Census Bureau population estimates for each age group are used as the denominator in rate computations, except for the “under 1 year of age group”. For the “under 1 year of age” group, the number of live births is used instead of the population estimate for this age group. When a rate is computed for an aggregate of several years, the population estimates for each year in the time period are summed together.
  • Deaths with "Unknown" or "Not Stated" age are included in "All Ages" counts and rates, yet are not included in age-specific counts, age-specific rates or in any age-adjusted rates. See About "Not Stated" age or ethnicity for more information.
  • Prior to August 2006, the number of live births had been substituted for the "under 1 year" age group population estimates for years 1999 and later.
  • See Age-Adjusted Rates for a discussion on the use of age-groups in calculating age-adjusted rates. Note that some of the age groups available in the data are not available in the standard population tables used to calculate age-adjusted rates. Also, age-adjusted rates are not available for infant age groups.
  • The age groups are identified by two columns, labels and codes, in data extracts.

Year

Pick any combination of years desired. Data are available for 1999 - 2010 in the Compressed Mortality online database with ICD 10 codes. Data are available for 1979 - 1998 in the Compressed Mortality online database with ICD 9 codes. Data are available for 1968 - 1978 in the Compressed Mortality online database with ICD 8 codes. The archive databases only contain through the final year available at the time of the release.

How?   See "How do I select items from the list box?

Gender

Select All Genders, Male, or Female

How?   See "How do I select items from the list box? Note:   The genders are identified by two columns, labels and codes, in data extracts.

Race

Select All Races or any combination of values. CMF 1999-2010, archive CMF 1999-2009, archive CMF 1999-2008, archive CMF 1999-2007 and archive CMF 1999-2006 have 4 bridged-race categories: American Indian or Alaska Native, Asian or Pacific Islander, Black or African American, White. CMF 1979-1998, archive CMF 1979-1998 and CMF 1965-1978 have 3 race categories: Black / African American, Other Races Combined, White.

How?   See "How do I select items from the list box? Notes:  

Hispanic Origin

Select All or any combination of 3 values:  Hispanic or Latino, Not Hispanic or Latino, Not Stated. Hispanic Origin data are not available for CMF 1968-1978 or CMF 1979-1998 .

How?   See "How do I select items from the list box? Notes:  

Urbanization

Use the radio buttons to select either the 2006 or the 2013 NCHS Urban-Rural Scheme for Counties. Select All Categories or any combination of values:  Large Central Metro, Large Fringe Metro, Medium Metro, Small Metro, Micropolitan (non-metro), NonCore (non-metro). Each county is assigned to one of these six categories. Each death is associated with a category based on the county of the person's legal residence.

How?  
  1. Click the Radio Button above the box, to pick your preferred list.
  2. See "How do I select items from the list box?," to limit your data to selected categories in the list.

Notes:  



Step 4. Select cause of death:
The cause of death is indicated by the International Classification of Disease (ICD) code in the underlying cause of death section on each death certificate. Use the radio buttons to select one of these 3 lists of causes of death:
  1. The default selection is the ICD chapters, subchapters, and codes list (ICD-8 Code List for years 1968-1978, ICD-9 Code List for years 1979-1998, ICD-10 Codes list for 1999 and later).
  2. The Selected Causes of Death or the ICD cause-of-death recode lists are also available (ICD-8 69 groups for years 1968-1978, ICD-9 72 groups for years 1979-1989, and ICD-10 113 groups for 1999 and later).
  3. The Injury Intent and Mechanism lists are available for year 1999 and later.
How?  
  1. Click the Radio Button above the box, to pick your preferred list.
  2. Select one or more items from the list to limit your data. The default value for any list is all causes of death.
    • See How do use a Finder? to learn more about search options, expanding or collapsing selected items, and displaying details for selected items in the Finder.
    • The "plus" symbol, "+" indicates that you can open the item, to see more items below it.
    • The results to a search are shown in blue, and indicated by ">".
    • The Advanced mode let you easily pick several items from different parts of the list. Items are not selected until you click the "Move" button in Advanced mode.
    • You may also enter values by hand, one code per line, in the Advanced mode. Use the Finder to see the correct code format. For example, "I20-I25" is the ICD-10 code for ischaemic heart diseases, and "GR113-059" is the Selected Causes of Death ICD-10 Group code for acute myocardial infarction.
    • See Finder Tool help for more hints.
Hints:  
  • When you select a code, note that some deaths are coded to the 3 digit code, and not the more detailed designation. Thus some deaths may be excluded from your data selection if your criteria are limited to only 4 digit codes. For example, the ICD-9 code 412 is the correct specification; 412.0 will not pick up deaths coded 412.
  • The causes of death are identified by two columns, labels and codes, in data extracts.
Notes:  
  • About the International Classification of Diseases:
    The mortality data on the CMF are compiled in accordance with the World Health Organization (WHO) regulations, which specify that member nations classify and code causes of death in accordance with the current revision of the International Statistical Classification of Diseases and Related Health Problems (ICD). The International Classification of Diseases is developed collaboratively between WHO and 10 international centers, for purposes of ensuring that medical terms reported on death certificates are internationally comparable and lend themselves to statistical analysis. The ICD has been revised approximately every 10 years since 1900 in order to reflect changes in understanding of disease mechanisms and in disease terminology.
  • About the 9th and 10th ICD revisions:
    The ICD-9 and ICD-10 classification systems are quite different. The ICD-9 system has a 4-digit numeric structure and about 5,000 categories for classifying cause-of-death. The ICD-10 system has a 4-digit alphanumeric coding structure and about 8,000 categories for classifying cause-of-death. Comparison of ICD-9 and ICD-10 shows that new chapters have been added to the ICD, old chapters have been rearranged, causes of death have been regrouped, and titles have changed. The differences between ICD-9 and ICD-10 make direct comparisons of cause-of-death difficult and result in discontinuities in cause-of-death trends. NCHS has conducted a comparability study to measure the discontinuities between the Ninth and Tenth Revisions and comparability ratios for major causes of death have been computed. Based on this work, comparability ratios have been calculated for selected causes. For further information please reference NCHS publications on Comparability of Cause-of-Death Between ICD Revisions.
  • For information on comparability of mortality causes between ICDA-8 and ICD-9, please see Estimates of Selected Comparability Ratios Based on Dual Coding of 1976 Death Certificates by the Eighth and Ninth Revisions of the International Classification of Diseases.
  • See also ICD 10th revision notes, ICD 9th revision notes, and ICD 8th revision notes.
  • Only those ICD codes that are used to classify causes of death in the National Vital Statistics System data are available in the Finder.
  • Deaths in the years 1999 and later are coded to the Tenth Revision of the International Classification of Diseases (ICD-10). There are 113 selected causes of death groups for the ICD-10 codes.
  • Deaths in the years 1979 - 1998 are coded to the Ninth Revision of the International Classification of Disease (ICD-9). There are 72 selected causes of death groups for the ICD-9 codes.
  • Deaths in the years 1968 - 1978 are coded to the Eighth revision of the International Classification of Disease Adapted for Use in the United States (ICDA-8). There are 69 selected causes of death groups for the ICD-8 codes.


ICD-10 Codes

Limit the data to any number of causes of death, for selected chapters, sub-chapters or codes.

How?   See Step 4. Select cause of death above. Check the Hints section.

Notes:  

  • The International Classification of Disease (ICD) 10th revision is used to represent the underlying cause of death for the years 1999 and later. The ICD system is organized by chapters, sub-chapters and codes.
  • ICD-10 uses a 4 digit alphanumeric coding scheme. Each of the 21 chapters in ICD-10 is classified to a letter or letters of the alphabet. Infectious disease codes in Chapter 1, for example, begin with an "A" or "B". Thus, Acute poliomyelitis is associated with the codes A80.0-A80.9 and Viral hepatitis is classified as B15.0-B19.9. Most cause of death codes in the ICD-10 system have 4-digits and consist of a letter, followed by two numbers, followed by a decimal point and another number. Some codes have only 3-digits and consist of a letter followed by two numbers.
  • About changes in the ICD-10 codes:  
    The valid ICD-10 codes used to classify underlying cause of death change over time. Effective with the 2003 data year, 1 code was introduced. Effective with the 2006 data year, 18 codes were introduced as valid causes of death, and 4 codes were discontinued. Effective with the 2007 data year, 4 codes were introduced as valid causes of death, and 2 codes were discontinued. Effective with the 2009 data year, 5 codes were introduced as valid causes of death, and 11 codes were discontinued.
    ICD-10 Changes in Years 2003-2009
    New causes of death in 2009
    ICD-10 Code Title
    A09.0 Other and unspecified gastroenteritis and colitis of infectious origin
    A09.9 Gastroenteritis and colitis of unspecified origin
    K52.3 Indeterminate colitis
    R26.3 Immobility
    R63.6 Insufficient intake of food and water due to self neglect
    New causes of death in 2007
    ICD-10 Code Title
    J09 Influenza due to identified avian influenza virus
    U04.9 Severe acute respiratory syndrome [SARS], unspecified
    X59.0 Exposure to unspecified factor causing fracture
    X59.9 Exposure to unspecified factor causing other and unspecified injury
    New causes of death in 2006
    ICD-10 Code Title
    B33.4 Hantavirus (cardio)-pulmonary syndrome [HPS][HCPS]
    G90.4 Autonomic dysreflexia
    I15.0 Renovascular hypertension
    I15.9 Secondary hypertension, unspecified
    K22.7 Barrett’s esophagus
    K85.0 Idiopathic acute pancreatitis
    K85.1 Biliary acute pancreatitis
    K85.2 Alcohol-induced acute pancreatitis
    K85.3 Drug-induced acute pancreatitis
    K85.8 Other acute pancreatitis
    K85.9 Acute pancreatitis, unspecified
    M31.7 Microscopic polyangiitis
    M79.7 Fibromyalgia
    P91.6 Hypoxic ischemic encephalopathy of newborn
    R29.6 Tendency to fall, not elsewhere classified
    R50.2 Drug-induced fever
    R50.8 Other specified fever
    W46 Contact with hypodermic needle
    New causes of death in 2003
    ICD-10 Code Title
    I27.2 Other secondary pulmonary hypertension

    Discontinued causes of death in 2009
    ICD-10 Code Title
    A09 Diarrhea and gastroenteritis of infectious origin
    F11.0 Mental and behavioral disorders due to use of opioids, acute intoxication
    F12.0 Mental and behavioral disorders due to use of cannabinoids, acute intoxication
    F13.0 Mental and behavioral disorders due to use of sedatives or hypnotics, acute intoxication
    F14.0 Mental and behavioral disorders due to use of cocaine, acute intoxication
    F15.0 Mental and behavioral disorders due to use of other stimulants, including caffeine, acute intoxication
    F16.0 Mental and behavioral disorders due to use of hallucinogens, acute intoxication
    F17.0 Mental and behavioral disorders due to use of tobacco, acute intoxication
    F18.0 Mental and behavioral disorders due to use of volatile solvents, acute intoxication
    F19.0 Mental and behavioral disorders due to multiple drug use and use of other psychoactive substances, acute intoxication
    K51.1 Ulcerative (chronic) ileocolitis

    Discontinued causes of death in 2007
    ICD-10 Code Title
    F10.0 Mental and behavioral disorders due to use of alcohol, acute intoxication
    X59 Exposure to unspecified factor

    Discontinued causes of death in 2006
    ICD-10 Code Title
    I25.2 Old myocardial infarction
    K85 Acute pancreatitis
    R50.0 Fever with chills
    R50.1 Persistent fever

  • About deaths due to acts of terrorism:  
    Beginning with data for 2001, NCHS introduced categories *U01-*U03 for classifying and coding deaths due to acts of terrorism. The asterisks before the category codes indicate that they are not part of the International Classification of Diseases, Tenth Revision (ICD-10). Description of the specific 4-digit codes can be found at NCHS Classifications of Diseases, Functioning and Disability: Appendix I. Deaths classified to the terrorism categories are included in the categories for Assault (homicide) and Intentional self-harm (suicide) in the 113 cause-of-death list. Additional information on these new categories can be found at NCHS Classifications of Diseases, Functioning and Disability:  Classification of Death and Injury Resulting from Terrorism. Terrorism related deaths in this data do not represent a final count of deaths resulting from the terrorist attacks on September 11, 2001, as this figure had not been determined. As of October 24, 2002, death certificates were issued for 2,957 of the estimated 3,028 individuals believed to have died as a result of the September 11, 2001 attacks. Of these, four were issued for terrorists and are classified as suicides. The criteria for issuing a death certificate for those believed to have died in the attacks differed by state, reflecting differences in state laws regarding death certification. Pennsylvania issued a death certificate for every individual, including the terrorists. Death certificates were not issued for any of the terrorists in Virginia or New York City. Virginia issued a death certificate only for those victims whose remains were identified. New York City issued a death certificate for those whose remains were identified or, if remains were not recovered, for those whose families applied for a death certificate. For more detailed information regarding New York City's processing of these deaths, see Deaths in World Trade Center Terrorist Attacks---New York City, 2001.
  • About deaths due to pulmonary hypertension:
    The World Health Organization (WHO) added code I27.2 in 2003. Prior to 2003, if the certifier listed “secondary pulmonary hypertension” the condition would be coded I27.0 because there was no code for secondary pulmonary hypertension. After code I27.2 was added in 2003, the number of deaths coded to I27.0 dropped significantly.


ICD-9 Codes

Limit the data to any number of causes of death, for selected chapters, sub-chapters or codes.

How?   See Step 4. Select cause of death above.

Notes:  

  • The International Classification of Disease (ICD) 9th revision is used to represent the underlying cause of death for the years 1979 - 1998. The ICD system is organized by chapters, sub-chapters and codes. For a further description of the ICD-9 codes see Volume II of the annual mortality volumes produced by the NCHS, such as Vital Statistics of the United States, 1988, Volume II - Mortality.
  • About Deaths due to Injuries and Poisoning in ICD-9: 
    For deaths due to injuries and poisonings that occurred during 1979-1998, the external cause is coded (E800-E999) rather than the Nature of Injury (800-999). The letter "E" is included in the ICD-9 code in CDC WONDER.
  • Most cause of death codes in the ICD-9 system have 4-digits and consist of three numbers, followed by a decimal point and another number. Some codes have only 3-digits and consist of three numbers.


ICDA-8 Codes

Limit the data to any number of causes of death, for selected chapters, sub-chapters or codes.

How?   See Step 4. Select cause of death above.

Notes:  

  • The International Classification of Disease Adapted for Use in the United States 8th Revision (ICDA-8) is used to represent the underlying cause of death for the years 1968 - 1978. The ICDA-8 system is organized by chapters, sub-chapters and codes. For a further description of the ICD-8 codes see Volume II of the annual mortality volumes produced by the NCHS, such as Vital Statistics of the United States, 1978, Volume II-Mortality.
  • For information on comparability of mortality causes between ICDA-8 and ICD-9, please see Estimates of Selected Comparability Ratios Based on Dual Coding of 1976 Death Certificates by the Eighth and Ninth Revisions of the International Classification of Diseases.
  • About Deaths due to Injuries and Poisoning in ICDA-8: 
    For deaths due to injuries and poisonings that occurred during 1968-1978, the external cause is coded (E800-E999) rather than the Nature of Injury (800-999). The letter "E" is included in the ICD-8 code in CDC WONDER.
  • Most cause of death codes in the ICDA-8 system have 4-digits and consist of three numbers, followed by a decimal point and another number. Some codes have only 3-digits and consist of three numbers.


ICD-10 113 Groups

Limit the data to any number of groups of selected causes of death.

How?   See Step 4. Select cause of death above.

Hint:   Group the data by "ICD-10 113 Groups" and also by "Cause of Death" to see the individual ICD codes included in each category.

Notes:  

  • NCHS developed the mortality tabulation lists for use in mortality analyses. The 113 selected causes of death list was designed to: 1) maintain continuity with the ICD-9 72 selected causes of death list, 2) to facilitate trend analysis, and 3) to separately identify causes of death that are of public health and medical importance.
  • Note that the ICD-10 113 groups have been adapted for the Injury Mechanisms categories. Some groups have been combined to make for inclusive categories, such as Motor Vehicle Traffic, Heart Disease and Tuberculosis. You may find it easier to select these broader categories of causes of death from the list under Injury Mechanism & All Other Leading Causes.


ICD-9 72 Groups

Limit the data to any number of groups of selected causes of death.

How?   See Step 4. Select cause of death above.

Hint:   Group the data by "ICD-9 72 Groups" and also by "Cause of Death" to see the individual ICD codes included in each category.

Note:   NCHS developed the mortality tabulation lists for use in mortality analyses. The 72 selected causes of death list was designed to: 1) maintain continuity with the ICD-9 69 selected causes of death list, 2) to facilitate trend analysis, and 3) to separately identify causes of death that are of public health and medical importance.



ICDA-8 69 Groups

Limit the data to any number of groups of selected causes of death.

How?   See Step 4. Select cause of death above.

Hint:   Group the data by "ICD-8 69 Groups" and also by "Cause of Death" to see the individual ICD codes included in each category.

Note:   NCHS developed the mortality tabulation lists for use in mortality analyses. The 69 selected causes of death list was designed to: 1) to facilitate trend analysis, and 2) to separately identify causes of death that are of public health and medical importance.



Injury Intent and Mechanism

Limit your data for any of the following data elements: 

  1. Injury Intent
  2. Injury Mechanism & All Other Leading Causes
About the External Cause of Injury Mortality Matrix:

For the analysis of injury mortality data, all causes of death have been classified by intent and by mechanism. The causes of death that are not related to injuries have been categorized as non-injuries, and are categorized in keeping with the 113 selected causes of death groups for ICD-10. The groups of injury mechanisms are different from those based on the "113 Selected Causes of Death" for ICD-10 codes. The groupings are based on the External Cause of Injury Mortality Matrix. In addition, some non-injury groups have been combined to make for broader categories, such as Heart Disease and Tuberculosis.

For more information, see:   External Cause of Injury Mortality Matrix.

Note:  Injury intent and Mechanism categories are not available for the ICD-8 codes in years 1968-1978.



Injury Intent

Limit the data to any number of categories.

How?  
  1. Click the Radio Button to pick your preferred list.
  2. Select one or more items from the list to limit your data. Use Ctrl + Click for multiple selections, or Shift + Click for a range.

Notes:  

  • Group the data by "Injury Intent" and also by "Cause of Death" to see the individual ICD codes included in each category.
  • Refer to External Cause of Injury Mortality Matrix for more information.
  • Injury intent and Mechanism categories are not available for the ICD-8 codes in years 1968-1978.


Injury Mechanism & All Other Leading Causes

Limit the data to any number of groups of selected causes of death.

How?  
  1. Click the Radio Button to pick your preferred list.
  2. Select one or more items from the list to limit your data. Use Ctrl + Click for multiple selections, or Shift + Click for a range.

Notes:  

  • Group the data by "Injury Mechanism" and also by "Cause of Death" to see the individual ICD codes included in each category.
  • NCHS has defined selected causes of death groups for analysis of injury mortality data. The groups of injury mechanisms are different from those based on the "113 Selected Causes of Death" for ICD-10 codes. The groupings are based on the External Cause of Injury Mortality Matrix. In addition, some non-injury groups have been combined to make for broader categories, such as Heart Disease and Tuberculosis.
  • Refer to External Cause of Injury Mortality Matrix for more information.
  • Injury intent and Mechanism categories are not available for the ICD-8 codes in years 1968-1978.
  • In order to allow analysis of Injury Mortality across a larger span of years, the ICD-9 codes that classify the underlying cause of death for years 1979 - 1998 have been categorized to be compatible with the ICD-10 External Cause of Mortality Matrix. The categories in WONDER differ slightly from the original ICD-9 External Cause of Mortality Matrix, as follows:


ICD-9 Codes and Updated Injury Mechanism Categories

ICD-9 Code Values Categories in the original
ICD-9 External Cause Mortality Matrix
Categories compatible with the
ICD-10 External Cause of Mortality Matrix
E990 Other specified and classifiable, legal intervention Fire or hot object or substance, legal intervention
E800 -  E807(.0,.1,.8,.9),    
E820 - E825(.0 - .5,.8,.9), E826(.2 - .8), E827 - E829(.2 - .9)
Transport, other, unintentional Other land transport, unintentional
E846 Other specified and classifiable, unintentional Other land transport, unintentional
E958.5 Motor Vehicle Traffic, suicide Other land transport, suicide
E988.5 Motor Vehicle Traffic, undetermined Other land transport, undetermined
E830 - E832 Drowning, Unintentional Other transport, unintentional
E847 - E848 Other specified and classifiable, unintentional Other transport, unintentional
E994 Other specified and classifiable, legal intervention Other transport, legal intervention
E958(.3) Natural or environmental, suicide Other specified classifiable, suicide
E958(.6) Transport, other, suicide Other specified, not elsewhere classified, suicide
E988(.3) Natural or environmental, undetermined Other specified classifiable, undetermined
E988(.6) Transport, other, undetermined Other specified, not elsewhere classified, undetermined



Step 5. Select rate options:
This section lets you select the multiplier for the rates, opt for age-adjusted rates and select the population for their calculation, and opt for 95% confidence intervals and standard errors. for age-adjusted rates.

Calculate Rates Per   Select the factor (multiplier) for your rates. By default, all ages rates are calculated per 100,000 persons, and rates limited to infant age groups are calculated per 1,000 persons.

How? See "How do I select items from the list box?

Include age-adjusted rates?   Crude rates are reported by default. To get age-adjusted rates, first indicate that you wish to calculate age-adjusted rates by clicking the preferred radio button. The options for the rates are displayed, and you can then choose the population to use for the weights in the calculation. You can also choose to show 95% confidence intervals and standard errors for the age-adjusted rates. For more information, see Age-Adjusted Rates and Frequently Asked Questions about Death Rates.

How? See How do I use a radio button?

Hints:

  • In Step 1, you may group your data by any variable, except for Age Groups. You are prevented from grouping the data by age groups when requesting age-adjusted rates, because any single age group alone yields a corresponding weight of 1.
  • In Step 3, you must select more than 1 age group when requesting age-adjusted rates.
  • Deaths with “unknown” age are not included in age-adjusted death rate calculations.
  • Age-adjusted rates are not available for Infant Age Groups.
Standard Population    If you picked standard populations for calculating age-adjusted rates, then the list of possible standard populations used to calculate these age-adjusted rates is shown. See Age-Adjusted Rates for more information.

How? See "How do I select items from the list box? Note: When using standard populations, your request criteria must include data for the following ranges of ages:  "5 - 14 years" and "15 - 24 years." Some of the age groups in the standard populations are aggregates of two age groups in the CMF data. When age-adjusted rates are calculated using the standard populations, the CMF deaths and populations counts are summed to match the standard population age groups. For example, the standard populations has a weight for the age group 15-24 years, so the CMF age groups 14-19 and 20-24 must be included in the request for age-adjusted rates (you cannot request age-adjusted rates only for persons 20 years and over).

Non-standard Population    If you picked to use a non-standard population for calculating age-adjusted rates, then the options for selecting the non-standard population are displayed. The non-standard population is derived from the population estimates in the CMF (those used to compute crude rates). You specify the year(s), gender(s), races(s), Hispanic origin(s), and location(s) to be included in the non-standard population. The selected population is the basis for the age-specific proportional weights to calculate these age-adjusted rates. See Frequently Asked Questions about Death Rates for more information.

How?

Note: If the same population is picked for your query criteria and your non-standard age-adjusted rate calculations, then the crude rates and age-adjusted rates are identical for those data rows that represent the non-standard population denominator.


Step 6. Other options:
Export Results   If checked query results are exported to a local file. More information on how to import this file into other applications can be found here.
How? See How do I use a checkbox?
Show Totals   If checked totals and sub-totals will appear in the results table.
How? See How do I use a checkbox?
Show Zero Values   If checked rows containing zero counts will appear in the results table. If unchecked, zero count rows are not displayed. However, this general feature is superseded by suppression constraints:
Sub-national vital statistics data representing fewer than ten persons (0-9) are suppressed for year 1989 and later years. See Assurance of Confidentiality for more information.
How? See How do I use a checkbox?
Precision   Select the precision for rate calculations. When the rate calculated for a small numerator (incidence count) is zero, you may increase the precision to reveal the rate by showing more numbers to the right of the decimal point.
How? See "How do I select items from the list box?," to limit your data to selected categories in the list.
Data Access Timeout   This value specifies the maximum time to wait for the data access for a query to complete. If the data access takes too long to complete, a message will be displayed and you can increase the timeout or simplify your request. If you can't complete a request using the maximum timeout, contact user support and we will try to run a custom data request for you.
How? See "How do I select items from the list box?," to limit your data to selected categories in the list.

Data Source Information

Mortality Data Sources The Compressed Mortality File is produced by the National Center for Health Statistics (NCHS) Office of Analysis and Epidemiology (OAE) at the Centers for Disease Control and Prevention (CDC). Mortality information is collected by state registries and provided to the National Vital Statistics System. Underlying cause of death and demographic descriptors are indicated on the death certificates.

To learn more about the methods and source of these data please reference:

Population Denominator Data Sources The population estimates on the CMF are based on Bureau of the Census estimates of total U.S., State, and county resident populations. The 1968 and 1969 State and county population estimates were calculated by NCHS using linear extrapolation. The 1970, 1980, 1990, 2000, and 2010 population estimates are April 1 modified census counts. The estimates for 1971-79, 1981-89, 1991-99, and 2001-2009 are intercensal estimates of July 1 resident populations. Note that the estimates for 1991 and later are based on bridged-race categories. The population estimates on the CMF are by geographic unit (total United States, State, and county), year, race (white, black, other races), sex, and age group (13 age groups). To permit the calculation of infant mortality rates, NCHS live-birth data are included on the file.

For more information on the population estimates, see:



Population Estimates on the Compressed Mortality File

Reference the following topics to learn more about population denominators for rate calculation:

Population Information for the Current Release

1968-1969 Population Estimates
1970 Population Estimates
1971-1979 Population Estimates
1980 Population Estimates
1981- 1989 Population Estimates
1990 Population Estimates
1991-1999 Population Estimates
2000 Population Estimates
2001- 2009 Population Estimates
2010 Population Estimates
Compressed Mortality File Archives: Population Revisions
Compressed Mortality File Archives:  About the 1990-based Postcensal Population Data



Population Information for the Current Release

The population data on the CMF are derived from U.S. Census Bureau files. The population estimates for the Census years:  1970, 1980, 1990, 2000, and 2010 are April 1, modified census counts. The population estimates for the non-Census years:  1968-1969, 1971-79, 1981-89, 1991-99, and 2001-2009 are intercensal estimates of the July 1, resident population. The population estimates for the years 2001- 2009 are from the revised intercensal estimates of the July 1 resident population (released by NCHS on 10/26/2012).

The following modifications of the Census population estimates were made by NCHS for CMF:

  1. To permit the calculation of infant mortality rates, NCHS live-birth data are included for "Infant age Groups." The race code for these records is derived from "race of mother".
  2. When the age group 1-4 years did not appear on the Census file, the age group 0-4 years was multiplied by 0.8 to obtain an estimate of the population 1-4 years.
  3. For the years 1988 through 1991, there was an additional county in Georgia with a "missing" county code of "999". This county was assigned a population count of zero.
  4. Geographic changes that create new counties or delete old counties often are implemented in the population files before they are implemented in the mortality files. As a result, county FIPS codes for some counties on the Census population files have been modified so that the FIPS codes on the CMF population and mortality file match. For more information, see County Geography Changes.

Specific Details

  1. 1968-1969 Population Estimates

    National population estimates are U.S. Bureau of the Census intercensal estimates of the July 1 resident population. State and county population estimates were calculated by NCHS using linear extrapolation from the corresponding July 1, 1970 and July 1, 1971 estimates. Source: Current Population Reports, Series P-2, Number 519), rounded to the nearest 1,000, of the July 1, 1968 and July 1, 1969 resident population of the U.S. by 5-year age group (under 1, 1-4, 5-9, ..., 85+ years), sex, and race (White, Black, Other races). NCHS live-birth counts were substituted for the estimates of the population under 1 year of age; the live-birth counts were not rounded.

  2. 1970 Population Estimates

    National, state, and county population estimates are from a modified version of the April 1, 1970 census. Live birth counts were substituted for the estimates of the population under 1 year of age. The original census counts were modified by the U.S. Bureau of the Census to produce a special census file of the April 1, 1970 resident population by 5-year age group (0-4, 5-9, ..., 85+), sex, and race (White, Black, Other races) and to correct errors in the data:

    1. Numerous small changes at the county and sub-county level were made to the original census file to correct errors discovered after publication of the original data. These changes resulted in an increase of 93,494 persons in the total U.S. population.
    2. The race classifications for the 1970 and 1980 census data were adjusted to be consistent with each other and with vital statistics. Some 327,000 persons of Hispanic origin who reported their race as "Other" (race not specified) were transferred to "White". The Black population was not affected by the race adjustment.
    3. Some 103,000 persons who reported their age as over 100 years had their age recoded to between 85 and 100 years. This adjustment did not affect this file because the oldest age group on the file is 85 years and over.

  3. 1971-1979 Population Estimates

    National and county estimates are U.S. Census Bureau intercensal estimates of the July 1 resident population. The Census Bureau did not produce state population estimates by age, race, and sex for the 1970s. Therefore, the state population estimates for 1979 are simply the sum of population estimates of the counties in each state.

    1. For 1979, the Census Bureau produced a national series of estimates and a county series of estimates. While these series are consistent with each other, the sum of the population estimates for all counties may not equal the national population estimates. This is due to rounding error that results because fractional estimates are not allowed. The Census Bureau recommends that national population estimates be used to calculate national death rates and county population estimates be used to calculate county death rates. CDC WONDER does this automatically.
    2. For 1979, NCHS used national population estimates rounded to the nearest thousand to calculate published death rates. The population estimate used by NCHS to calculate the death rate of an aggregate group (e.g. aggregated across race, gender, or age) was obtained by summing unrounded population estimates for each group in the aggregate and then rounded to the nearest 1,000. Unfortunately, the national population estimates available for use in the CMF for 1979 are all prerounded to the nearest 1,0000. This means that the population estimates for aggregate groups are obtained by summing rounded population estimates, rather than by summing unrounded population estimates and then rounding. As a result, for 1979, some rounding error may result when aggregating across race, gender, or age groups as summing rounded numbers may result in a different total than summing unrounded numbers and then rounding.
    3. Three Virginia independent cities (Manassas, Manassas Park, and Poquoson) did not appear on the Census file. Therefore, the 1979 populations for these three cities were estimated from the July 1, 1980 and July 1, 1981 estimates of these cities. The 1979 population estimates for the counties containing the cities were reduced by the estimated city populations.

  4. 1980 Population Estimates

    National, state, and county population estimates are from a modified version of the April 1, 1980 census The national and state estimates are the sum of the county population counts. The original census counts were modified by the U.S. Census Bureau so that persons who reported their race as "other" (the majority of these persons being of Hispanic ethnicity) were reassigned to one of the four single-race groups specified in the Office of Management and Budget (OMB) 1977 Standards on Race and Ethnicity:  White, Black, American Indian or Alaska Native, and Asian or Pacific Islander.

    1. For 1980, NCHS used unrounded population estimates to calculate national death rates. Therefore, on the CMF, the national population estimates used in rate calculation and displayed are unrounded.
    2. April 1, 1980 population estimates for three Virginia independent cities, (Manassas, Manassas Park, and Poquoson) had to be extrapolated from July 1, 1980 estimates. The April 1 populations for the three cities were calculated as a proportion of the April 1 county population, where the proportion was calculated from the July 1, 1980 estimates. The April 1 population estimates for the counties containing the three cities were reduced by the cities estimated April 1 populations.

  5. 1981-89 Population Estimates

    National, state, and county estimates are U.S. Census Bureau intercensal estimates of the July 1 resident population.

    1. For 1981-89, the U.S. Census Bureau has separate series of estimates for each geographic level of estimates (i.e. national, state, and county). While these series are consistent with each other, the sum of the population estimates of counties within a state may not equal the state population estimate, and the sum of all state population estimates or all county population estimates may not equal the national population estimates. This is due to rounding error that results because fractional estimates are not allowed. The Census Bureau recommends that national population estimates be used to calculate national death rates, state population estimates be used to calculate state death rates, and county population estimates be used to calculate county death rates. CDC WONDER does this automatically.
    2. For 1981-89, NCHS used national population estimates rounded to the nearest thousand to calculate published death rates. The population estimate used by NCHS to calculate the death rate of an aggregate group (e.g. aggregated across race, gender, or age) was obtained by summing unrounded population estimates for each group in the aggregate and then rounding to the nearest 1,000. This is also done on the CMF in CDC WONDER, and thus, for these years, national death rates calculated on CDC WONDER agree with those published by NCHS.

  6. 1990 Population Estimates

    National, state, and county population estimates are from the April 1, 1990 age-race-sex modified census counts (national and state estimates are the sum of the county census counts). The original census counts were modified by the U.S. Census Bureau: 

    1. To correct the bias in reported age -- about 10 percent of persons were actually a year younger as of April 1 than reported;
    2. To assign persons who reported their race as "other" to one of the four single-race groups specified in the Office of Management and Budget (OMB) 1977 Standards on Race and Ethnicity:  White, Black, American Indian or Alaska Native, and Asian or Pacific Islander.
    For 1990, NCHS used unrounded population estimates to calculate national death rates. Therefore, on the CMF, the national population estimates used in rate calculation and displayed are unrounded.

  7. 1991- 1999 Population Estimates

    As of September 2003, National, state, and county population estimates are U.S. Census Bureau bridged-race intercensal estimates of the July 1, resident population, based on the 1990 census and the bridged-race 2000 census. Derivation of the race-specific intercensal population estimates for the 1990s was complicated by the incomparability of the race data on the 1990 and 2000 censuses. Before the intercensal estimates for the 1990s could be derived, the race groups on the 2000 census had to be made consistent with ("bridged to") the race groups on the 1990 census. Race data on the 2000 Census were collected in accordance with the 1997 Office of Management and Budget's standards on race and ethnicity. The 1997 standards specify 5 single-race categories (American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or Other Pacific Islander, and White) and permit the reporting of more than one race. As a result, there were 31 race groups on the 2000 census (5 single-race groups and 26 multiple-race groups). NCHS, in collaboration with the Census Bureau, developed methodology for bridging the multiple-race groups to single-race categories (American Indian or Alaska Native, Asian or Pacific Islander, Black, and White). The 1990 census race groups are White, Black, American Indian or Alaska Native, and Asian or Pacific Islander.

    1. The national and state estimates are the sum of the bridged-race county estimates.
    2. Prior to September 2003, the population estimates on the CMF for the years 1991-1999 were postcensal estimates based on the 1990 census.
    3. Refer to Population Revisions and Description of the 1990-based Postcensal Population Data for more information.

  8. 2000 Population Estimates

    National, state, and county population estimates are from the U.S. Census Bureau April 1, bridged modified race 2000 Census counts. The original census counts were modified by the U.S. Census Bureau to assign persons who reported their race as "other " to one of the 31 single or multiple-race groups specified in the Office of Management and Budget (OMB) 1997 Standards on Race and Ethnicity. The resulting counts were then bridged to (made consistent with) the four single-race categories (White, Black, American Indian or Alaska Native, and Asian or Pacific Islander).

  9. 2001 - 2009 Population Estimates

    The population estimates for 2001-2009 are July 1 resident population estimates from the U.S. Census Bureau's revised bridged-race intercensal series, based on the year 2000 and the year 2010 census counts (released by NCHS on 10/26/2012).

    The bridged-race population files have estimates for the four single-race categories (White, Black, American Indian or Alaska Native, and Asian or Pacific Islander).

    For the current release of the 1999-2010 data, the national, region, division, and state estimates were obtained by summing the county estimates of the revised intercensal 2001-2009 series, so the national, region, division, state and county estimates are consistent with each other.

    Notes on 2001-2009 populations estimates in the archives:
    • On the archive CMF 1999-2009 on-line database (published 2012):
      • The region, division, state and county 2001-2008 populations estimates are from the 2009 series of the U.S. Census Bureau's bridged-race postcensal series.
      • Region, division, and state estimates were obtained by summing the county estimates of the Vintage 2009 series.
      • The national population figures for 2001 and beyond are bridged-race postcensal estimates of the July 1 resident population from the corresponding postcensal series: 2001 from the Vintage 2001 series, 2002 from the Vintage 2002 series, 2003 from the Vintage 2003 series, 2004 from the Vintage 2004 series, 2005 from the Vintage 2005 series, 2006 from the Vintage 2006 series, 2007 from the Vintage 2007 series, 2008 from the Vintage 2008 series, and 2009 from the 2009 series.
      • Population figures for 2000 are bridged-race April 1 census counts.
      • Population figures for 1999 are bridged-race intercensal estimates of the July 1 resident population.
    • On the archive CMF 1999-2008 on-line database (published 2011):
      • The region, division, state and county 2001-2008 populations estimates are from the 2009 series of the U.S. Census Bureau's bridged-race postcensal series.
      • Region, division, and state estimates were obtained by summing the county estimates of the Vintage 2009 series.
      • The national population figures for 2001 and beyond are bridged-race postcensal estimates of the July 1 resident population from the corresponding postcensal series: 2001 from the Vintage 2001 series, 2002 from the Vintage 2002 series, 2003 from the Vintage 2003 series, 2004 from the Vintage 2004 series, 2005 from the Vintage 2005 series, 2006 from the Vintage 2006 series, 2007 from the Vintage 2007 series, and 2008 from the Vintage 2008 series.
      • Population figures for 2000 are bridged-race April 1 census counts.
      • Population figures for 1999 are bridged-race intercensal estimates of the July 1 resident population.
    • On the archive CMF 1999-2007 on-line database (published 2010):
      • The region, division, state and county 2001-2007 populations estimates are from the 2008 series of the U.S. Census Bureau's bridged-race postcensal series.
      • Region, division, and state estimates were obtained by summing the county estimates of the Vintage 2008 series.
      • The national population figures for 2001 and beyond are bridged-race postcensal estimates of the July 1 resident population from the corresponding postcensal series: 2001 from the Vintage 2001 series, 2002 from the Vintage 2002 series, 2003 from the Vintage 2003 series, 2004 from the Vintage 2004 series, 2005 from the Vintage 2005 series, 2006 from the Vintage 2006 series, and 2007 from the Vintage 2007 series.
      • Population figures for 2000 are bridged-race April 1 census counts.
      • Population figures for 1999 are bridged-race intercensal estimates of the July 1 resident population.
    • On the archive CMF 1999-2006 on-line database (published 2009):
      • The region, division, state and county 2001-2006 populations estimates are from the 2007 series of the U.S. Census Bureau's bridged-race postcensal series.
      • Region, division, and state estimates were obtained by summing the county estimates of the Vintage 2007 series.
      • The national population figures for 2001 and beyond are bridged-race postcensal estimates of the July 1 resident population from the corresponding postcensal series: 2001 from the Vintage 2001 series, 2002 from the Vintage 2002 series, 2003 from the Vintage 2003 series, 2004 from the Vintage 2004 series, 2005 from the Vintage 2005 series, and 2006 from the Vintage 2006 series.
      • Population figures for 2000 are bridged-race April 1 census counts.
      • Population figures for 1999 are bridged-race intercensal estimates of the July 1 resident population.
    • On the archive CMF 1999-2005 on-line database (published 2008):
      • The region, division, state and county 2001-2005 populations estimates are from the 2006 series of the U.S. Census Bureau's bridged-race postcensal series.
      • Region, division, and state estimates were obtained by summing the county estimates of the Vintage 2006 series.
      • The national population figures for 2001 and beyond are bridged-race postcensal estimates of the July 1 resident population from the corresponding postcensal series: 2001 from the Vintage 2001 series, 2002 from the Vintage 2002 series, 2003 from the Vintage 2003 series, 2004 from the Vintage 2004 series and 2005 from the Vintage 2005 series.
      • Population figures for 2000 are bridged-race April 1 census counts.
      • Population figures for 1999 are bridged-race intercensal estimates of the July 1 resident population.
    • On the archive CMF 1999-2004 on-line database (published 2007):
      • The region, division, state and county 2001-2004 populations estimates are from the 2005 series of the U.S. Census Bureau's bridged-race postcensal series.
      • Region, division, and state estimates were obtained by summing the county estimates of the Vintage 2005 series.
      • The national population figures for 2001 and beyond are bridged-race postcensal estimates of the July 1 resident population from the corresponding postcensal series: 2001 from the Vintage 2001 series, 2002 from the Vintage 2002 series, 2003 from the Vintage 2003 series, and 2004 from the Vintage 2004 series.
      • Population figures for 2000 are bridged-race April 1 census counts.
      • Population figures for 1999 are bridged-race intercensal estimates of the July 1 resident population.
    • On the archive CMF 1999-2003 on-line database (published 2006):
      • The region, division, state and county 2001-2003 populations estimates are from the 2004 series of the U.S. Census Bureau's bridged-race postcensal series.
      • Region, division, and state estimates were obtained by summing the county estimates of the Vintage 2004 series.
      • The national population figures for 2001 and beyond are bridged-race postcensal estimates of the July 1 resident population from the corresponding postcensal series: 2001 from the Vintage 2001 series, 2002 from the Vintage 2002 series, and 2003 from the Vintage 2003 series.
      • Population figures for 2000 are bridged-race April 1 census counts.
      • Population figures for 1999 are bridged-race intercensal estimates of the July 1 resident population.
    • See CMF Archives for more information.

  10. 2010 Population Estimates

    National, state, and county population estimates are from the U.S. Census Bureau April 1, bridged modified race 2010 Census counts. The original census counts were modified by the U.S. Census Bureau to assign persons who reported their race as "other" to one of the 31 single or multiple-race groups specified in the Office of Management and Budget (OMB) 1997 Standards on Race and Ethnicity. The resulting counts were then bridged to (made consistent with) the four single-race categories (White, Black, American Indian or Alaska Native, and Asian or Pacific Islander).

If you have additional questions about the population estimates, please see U.S. Census Populations With Bridged Race Categories (http://www.cdc.gov/nchs/about/major/dvs/popbridge/popbridge.htm) or contact PopEst@cdc.gov.


Revised Population Estimates

In September 2003, all of the population estimates for years 1991-1999 were revised; the postcensal estimates for this period were replaced with intercensal estimates. For the CMF 1999-2002 through CMF 1999-2009, state and county population data for 2001 and later changed annually to reflect updated postcensal estimates. In February 2013, all of the population estimates for years 2001-2009 were revised; the postcensal estimates for this period were replaced with intercensal estimates. Previous releases are available as "archive" data for a period of time so that researchers can replicate previously published rates. The query pages labeled "archive" calculate rates with the previous population estimates.

Notes:

Description of the 1990-based Postcensal Population Data in Archived CMF

In the archive data, the national, state, and county estimates for 1991-99 are postcensal estimates developed by the U.S. Census Bureau. Postcensal population estimates are estimates made for the years following a decennial census before the next census has been taken. The Census Bureau annually produces postcensal estimates of the population for various levels of geography and demographic detail. The estimates produced in this annual production cycle are collectively referred to as a series of estimates, and the current series of estimates is referred to by the last year in the series. Each series contains estimates for each year from the census year forward, so each new series contains estimates for all of the years in the previous series. The estimates in a given series may not be the same as the estimates in the previous series. For example, the 1998 series of state and county estimates contains estimates for 1997 which in some cases differ substantially from the estimates for 1997 released in the 1997 series of estimates.

  1. National estimates for 1991-1999 in the archive data:
    The postcensal national estimates on the archived CMF 1979-98,CMF 1999, and CMF 1999-2001 are the same as those used by NCHS to calculate published death rates. Each year when the CMF was updated, national estimates for the new year were added to the file and the estimates for previous years were retained. As a result, the postcensal national estimates for the archive 1991-1999 CMF are from different series of estimates, as outlined below.
    • The national estimates of the July 1, 1991 resident population of the U.S. are from the 792I series of intercensal estimates developed by the U.S. Census Bureau. They are consistent with the postcensal estimates for that year produced shortly thereafter.
    • The national estimates for July 1, 1992 are from the 1990-93 annual time series described in Population Division Product PPL-8.
    • The national estimates for July 1, 1993 are from the 1990-94 annual time series.
    • The July 1, 1994 and July 1, 1995 national estimates are from the 1990-95 annual time series described in Population Division Product PPL-41.
    • The July 1, 1996 national estimates are from the 1990-96 annual time series described in Population Division Product PPL-57.
    • The July 1, 1997 national estimates are from the 1990-97 annual time series described in Population Division Product PPL-91R.
    • The July 1, 1998 national estimates are from the 1990-98 annual time series.
    • The July 1, 1999 national estimates are from the 1990-99 annual time series.
  2. State estimates for 1991-1999 in the archive data:
    The state estimates are from the 1990-1999 annual time series of postcensal estimates of the July 1 resident population of the 50 States and the District of Columbia prepared by the U.S. Census Bureau.
  3. County estimates for 1991-1999 in the archive data:
    The county estimates are from the U.S. Census Bureau 1990-99 annual time series of postcensal estimates of the July 1 resident population of the 3,141 counties in the U.S. (as defined in 1990).

Frequently Asked Questions about Compressed Mortality

The questions are in three sections:

Questions about Death Rates
Data Release Questions
Race and Ethnicity Questions

  1. Questions about Death Rates

    1. How are crude death rates calculated in WONDER?

      The "crude death rate" is the number of deaths divided by the population, multiplied by 100,000.

      Crude Death Rate = (number of deaths / population) * 100,000

      Note: 100,000 is the default multiplier, other multipliers can be specified in the query.

    2. How are age-adjusted death rates calculated in WONDER?

      The age-adjusted rate is calculated by multiplying the age-specific death rate for each age group by the corresponding weight from the specified standard population, summing across all age groups, and then multiplying this result by 100,000 (or whatever multiplier is specified in the query).

      Age-Adjusted Death Rate = Sum of (Age Specific Death Rate * Standard Population weight) * 100,000

      The age-specific death rate is the number of deaths for a given age group divided by the population of that age group.

      Age Specific Death Rate = (number of deaths in age group / population of age group)

      The "standard population weight " for an age group is calculated by dividing the population for the age group by the sum of the populations for all of the age groups in the query. Please see the question below on "children under 1 year" age categories.

      Standard Population Weight = population for age group
      / sum of age group populations for all age groups in query

      See http://seer.cancer.gov/seerstat/tutorials/aarates/definition.html for a step-by-step tutorial with an example of the calculations.

    3. What are the "Standard" and "Non-Standard" populations?

      WONDER allows the user to select the population distribution used for calculating age-adjusted rates. Three "Standard" populations are offered: the year 2000 standard population (the default), the 1970 standard population, and the 1940 standard population. Alternatively, the user can specify a "Non-Standard" population for use as the population distribution in the age-adjustment.

      • The 1940 and 2000 standard populations were obtained from the National Center for Health Statistics. Beginning with the 1999 data year, NCHS adopted the year 2000 projected population of the U.S. as the standard population for use in age adjusting death rates. The year 2000 standard replaced the 1940 standard population that had been used for over 50 years. The new population standard affects levels of mortality, and to some extent, trends and group comparisons.

      • The 1970 standard population is the one used by the National Cancer Institute.

      • When the user requests that a "non-standard " population is used in the calculation of age-adjusted rates, WONDER uses the Census population estimates/counts included in the Compressed Mortality File to determine the weights used in the age-adjustment. See Population Data Description for more information.

    4. What age categories are used for age-adjusted rates?

      Only age groups that fall within the age range specified in the query are used to calculate an age-adjusted rate. The "total population" for a query is the sum of the populations of each age group included in that query. For example, if an age-adjusted rate is requested for 45-74 year olds, then the total population is the sum of the 45-54 year olds, 55-64 year olds, and 65-74 year old populations. For the 1940, 1970 and 2000 standard populations, the possible age groups are: 

      less than 1 year, 1-4, 5-14, 15-24, 25-34, ....85 years and over.

      If the user specifies a "non-standard" population for use in age-adjustment, the possible age groups are:

      less than 1 year, 1-4, 5-9, 10-14, 15-19, 20-24, 25-34, ....85 years and over.

      Note that age groups differ from the age groups used for "standard" years, affecting the classification of deaths in the age range from 5 years to 24 years of age.

      See Age Adjustment of Death Rates for more information.

    5. What about children under 1 year and rate calculation?

      When calculating mortality rates for "Infant Age Groups" (under 1 day, 1-6 days, 7-27 days, 28-264 days), the population is the number of live births in the given time period.

      Notes:

      • Age-adjusted rates cannot be calculated for infant age groups, because the denominator population is the number of live births in the specified years.
      • In the CMF 1999-2010 and the archive CMF 1999-2009 online databases, rates for infant age groups are not shown for fewer than twenty deaths, nor are race and Hispanic Origin data available, due to privacy constraints; see Assurance of Confidentiality for more information. However, rates for race and Hispanic Origin are available for the age group "under 1 year of age." The "under 1 year of age" age group represents the population estimates for the given time period.
      • In the CMF 1999-2003 online database, the "under 1 year of age" age group represents the population estimates for the given time period. (Prior to August 2006, in all previous CMF online databases, the number of live births in the given time period had been substituted for the "under 1 year of age" population estimates.)
      • For more information, see Mortality for Infants.

    6. Why are death rates sometimes flagged as "Unreliable" or "Suppressed"?

      Death rates based on counts less than twenty (death count <=20) are flagged as "Unreliable". A death rate based on fewer than 20 deaths has a relative standard error (RSE(R))of 23 percent or more. A RES(R ) of 23 percent is considered statistically unreliable.

      Death counts and death rates are "Suppressed" when the data meets the criteria for confidentiality constraints. See Assurance of Confidentiality for more information.

    7. Why do the rates change for past years?

      The rates change because the population estimates used the denominators in the rate calculations change. The population estimates used as denominators change in order to provide the most recently available best estimate. In the years following the decennial census, the Census Bureau annually produces a set of estimates containing estimates of the current year population and revised estimates of the population for previous years. Each set of estimates is referred to as a postcensal series. reference Description of the 1990-based Postcensal Population Estimates. See also Population Revisions.

      National death rates for 1979, 1981-89 and 1991-1999 computed using the CMF may not agree with death rates published previously by NCHS. The population estimates for these years on the CMF are intercensal estimates, which have replaced the postcensal population estimates originally used by NCHS to calculate published death rates. NCHS has published some revised death rates using the intercensal population estimates found on the current data release. To reproduce the earlier rates, please use the appropriate archive data.

  2. Data Release Questions

    1. What are the Assurance of Confidentiality constraints?

      Data reports for years 1989 and later meet the NCHS data use restrictions. Vital statistics data are suppressed due to confidentiality constraints, in order to protect personal privacy. For data from year 1989 and later years, the term "Suppressed" replaces sub-national death counts, births counts, death rates and associated confidence intervals and standard errors, as well as corresponding population figures, when the figure represents zero to nine (0-9) persons.

      With the release of CMF 1999-2008 on February 15, 2012, additional privacy constraints apply to infant mortality statistics representing infant age groups and live births as the denominator population. When an infant mortality measure represents fewer than ten (0-9) infant deaths, all corresponding live birth population denominator figures are suppressed. When the infant mortality measure represents ten to nineteen (10-19) infant deaths, the number of deaths and live births are shown, but rates and associated measures are not shown. Race and Hispanic origin data are not available in this online database for infant age groups. However, race and Hispanic origin data are available for persons under one year of age in the other age groups, which use population estimates as population denominator data. Race and Hispanic origin detail for infant mortality statistics are available in the Linked Birth / Infant Death Records data collections.

      Prior to May 23, 2011, data cells in tables for year 1989 and later years were suppressed only for single county-level data, when the data represented five or fewer (1-5) deaths for a time period less than three years, and the county's total population in the April 1st, 2000 Census was fewer than one hundred thousand (100,000) persons.

      Totals and sub-totals are suppressed when the value falls within scope of the suppression criteria, or when the summary value includes a single suppressed figure, in order to prevent the inadvertent disclosure of suppressed values.

      The confidentiality constraints and use of the "Unreliable" flag are established by the original data providers. For more information, please contact the data providers.

    2. What are my responsibilities in accessing this data?

      See Data Use Restrictions to review the policies affecting access to the Compressed Mortality data. Note that use of the data implies consent or agreement to abide by the policies.

    3. Why separate query pages for CMF 1968-1978, CMF 1979-1998 and CMF 1999 and later?

      The International Classification of Diseases (ICD) is used to classify underlying cause of death. The ICD Eighth Revision is used from 1968-1978, the ICD Ninth Revision is used from 1979-1998, and the ICD Tenth Revision is used since 1999. The classification systems differ significantly. As a result of these changes, the two classification schemes are different enough to make direct comparisons of cause-of-death difficult. NCHS has conducted a comparability study to measure the discontinuities between the Ninth and Tenth Revisions of the ICD for selected causes of death.

      In order to avoid delivering potentially misleading information, separate Mortality query pages are available on CDC WONDER, one for the Eighth Revision data years (1969-1978), one for the Ninth Revision data years (1979-1998), and one for the Tenth Revision data years (1999 and later). Deaths cannot be tabulated across the 1998-1999 boundary in a single query. We fully recognize that this approach creates limitations for users of these data. We are hopeful that better approaches can be implemented once a method for addressing the data discontinuity has been developed.

    4. Why are there "Archive" query pages for previous releases?

      CDC's current policy on data release requires that previously published data remain available for researchers for comparison purposes. WONDER provides the previously published data designated as "archive" data, in contrast to the currently available published release. Each year the Compressed Mortality File is updated with a new year of data. Other revisions are made to the preceding years of data, such as changes in denominator population estimates and changes in county location codes. For more information, please see

      Population Revisions
      County Geography Changes
      CMF Archives

  3. Race and Ethnicity Questions

    1. What about differences in race reporting on death certificates and self-reporting in the Census?

      See Race Reporting for more information about the issues of race categories and ethnicity in the data.

    2. What racial categories are included in the "Other" classification for CMF Vintage 2005 and earlier?

      The racial categories for all years of the CMF are white, black, and other races. Other races includes the American Indian or Alaskan Native race category and the Asian or Pacific Islander race category.

    3. How are multi-racial persons classified in CMF?

      Race data are collected on death certificates in accordance with the 1977 OMB standards on race and ethnicity. The 1977 standards specified four race categories (white, black, American Indian or Alaska Native, and Asian or Pacific Islander) and did not permit more than one racial category to be identified for an individual. Population data through the 1990s were also obtained in accordance with the 1977 standards. Race data on the 2000 census were collected in accordance with the 1997 OMB standards on race and ethnicity. The 1997 standards specify 5 single-race categories (American Indian or Alaskan Native, Asian, black, Hawaiian or Other Pacific Islander, and white) and permit the reporting of more than one race. As a result, there were 31 racial groups on the 2000 census (5 single-race groups and 26 multiple-race groups). NCHS, in collaboration with the Census Bureau, developed methodology for bridging the multiple-race groups to single-race categories, so that the race categories in the population data would match the race categories in the mortality data. Please see U.S. Census Populations With Bridged Race Categories.

    4. How are Hispanic persons who reported their race as "Other" on the census assigned to a race group?

      The Census Bureau has assigned all persons (including Hispanic persons) who specified their race as "other" on the census (1980, 1990, and 2000) to one of the OMB specified racial categories. The algorithm used by the Census Bureau to make these assignments has differed for the 1980, 1990, and 2000 censuses, and is described in the Population Data section. See also Population Revisions.


CMF Archives

Archives of previously published CMF versions are provided so that researchers can validate statistics and references from previous years. Each year the Compressed Mortality File is updated with a new year of data. Other revisions are made to the preceding years of data, such as changes in denominator population estimates and changes in county location codes. Changes in population estimates are reflected in changes in mortality rates. Death counts are not revised in later releases of Compressed Mortality File, other than dividing the deaths in previously combined county or county equivalents.

CMF 1999 - 2009 with ICD-10 codes Archive

  • This database was the active CMF data release for 1999-2009 (Series 20, No. 2O, 2012) from July 2012 to February 2013, when it was replaced by the CMF 1999-2010 ( Series 20 No. 2P, 2013).
  • 2001-2009 population estimates: the region, division, state and county population estimates for years 2001 and later, on the archive CMF 1999-2009 are estimates of the July 1 resident population from the county-level bridged-race Vintage 2009 postcensal series. National population estimates for 2001-2009 are from the corresponding postcensal vintage (e.g., Vintage 2001 for 2001, Vintage 2002 for 2002).
  • 2000 population estimates: The population estimates for 2000 are bridged-race modified April 1 census counts.
  • 1999 population estimates: The population estimates for 1999 are estimates of the July 1 resident population from the bridged-race intercensal estimates.
  • Data are available for Broomfield county, Colorado [08014] and Denali, Alaska [02068], only for years 2003 and later. Before 2003, data for these areas are aggregated with neighboring areas. Data for Clifton Forge City, Virginia (51560) are available only for the years 1999-2000. In the year 2001, the area merged with neighboring Alleghany county (51005).
CMF 1999 - 2008 with ICD-10 codes Archive

  • This database was the active CMF data release for 1999-2008 (Series 20, No. 2N, 2011) from February 2012 to July 2012, when it was replaced by the CMF 1999-2009 ( Series 20 No. 2O, 2012).
  • 2001-2008 population estimates: The region, division, state and county population estimates for years 2001 and later, on the archive CMF 1999-2008 are estimates of the July 1 resident population from the county-level bridged-race Vintage 2009 postcensal series. National population estimates for 2001-2008 are from the corresponding postcensal vintage (e.g., Vintage 2001 for 2001, Vintage 2002 for 2002).
  • 2000 population estimates: The population estimates for 2000 are bridged-race modified April 1 census counts.
  • 1999 population estimates: The population estimates for 1999 are estimates of the July 1 resident population from the bridged-race intercensal estimates.
  • Data are available for Broomfield county, Colorado [08014] and Denali, Alaska [02068], only for years 2003 and later. Before 2003, data for these areas are aggregated with neighboring areas. Data for Clifton Forge City, Virginia (51560) are available only for the years 1999-2000. In the year 2001, the area merged with neighboring Alleghany county (51005).
CMF 1999 - 2007 with ICD-10 codes Archive

  • This database was the active CMF data release for 1999-2007 (Series 20, No. 2M, 2010) from October 2010 to January 2012, when it was replaced by the CMF 1999-2008 ( Series 20 No. 2N, 2011).
  • 2001-2007 population estimates: The region, division, state and county population estimates for years 2001 and later, on the archive CMF 1999-2007 are estimates of the July 1 resident population from the county-level bridged-race Vintage 2008 postcensal series. National population estimates for 2001-2007 are from the corresponding postcensal vintage (e.g., Vintage 2001 for 2001, Vintage 2002 for 2002).
  • 2000 population estimates: The population estimates for 2000 are bridged-race modified April 1 census counts.
  • 1999 population estimates: The population estimates for 1999 are estimates of the July 1 resident population from the bridged-race intercensal estimates.
  • Data are available for Broomfield county, Colorado [08014] and Denali, Alaska [02068], only for years 2003 and later. Before 2003, data for these areas are aggregated with neighboring areas. Data for Clifton Forge City, Virginia (51560) are available only for the years 1999-2000. In the year 2001, the area merged with merged with neighboring Alleghany county (51005).
CMF 1999 - 2006 with ICD-10 codes Archive

  • This database was the active CMF data release for 1999-2006 (Series 20, No. 2L, 2010) from August 2009 to October 2010, when it was replaced by the CMF 1999-2007 ( Series 20 No. 2M, 2010).
  • 2001-2006 population estimates: The region, division, state and county population estimates for years 2001 and later, on the archive CMF 1999-2006 are estimates of the July 1 resident population from the county-level bridged-race Vintage 2007 postcensal series. National population estimates for 2001-2006 are from the corresponding postcensal vintage (e.g., Vintage 2001 for 2001, Vintage 2002 for 2002).
  • 2000 population estimates: The population estimates for 2000 are bridged-race modified April 1 census counts.
  • 1999 population estimates: The population estimates for 1999 are estimates of the July 1 resident population from the bridged-race intercensal estimates.
  • Data are available for Broomfield county, Colorado [08014] and Denali, Alaska [02068], only for years 2003 and later. Before 2003, data for these areas are aggregated with neighboring areas. Data for Clifton Forge City, Virginia (51560) are available only for the years 1999-2000. In the year 2001, the area merged with merged with neighboring Alleghany county (51005).
  • New measures are available beginning with the CMF 1999-2006 database: 95% confidence intervals and standard errors.
CMF 1999 - 2005 with ICD-10 codes Archive

  • This database was the active CMF data release for 1999-2005 (Series 20, No. 2K, 2008) from March 2008- August 2009, when it was replaced by the CMF 1999-2006 ( Series 20 No. 2L, 2009).
  • 2001-2005 population estimates: The region, division, state and county populations estimates for years 2001 and later on the archive CMF 1999-2005 are estimates of the July 1 resident population from the county-level bridged-race Vintage 2006 postcensal series.
  • 2000 population estimates: The population estimates for 2000 are bridged-race modified April 1 census counts.
  • 1999 population estimates: The population estimates for 1999 are estimates of the July 1 resident population from the bridged-race intercensal estimates.
  • Data are available for Broomfield county, Colorado [08014] and Denali, Alaska [02068], only for years 2003 and later. Before 2003, data for these areas are aggregated with neighboring areas. Data for Clifton Forge City, Virginia (51560) are available only for the years 1999-2000. In the year 2001, the area merged with merged with neighboring Alleghany county (51005).
  • The CMF 1999-2005 has not adjusted state and county level populations for migration due to Hurricanes Katrina and Rita in 2005, see About Population Migration due to Hurricanes in 2005 for more information.
  • The CMF 1999-2005 does not provide data for Hispanic Origin, nor data for these race groups: American Indian or Alaskan Native; Asian or Pacific Islander.
CMF 1999 - 2004 with ICD-10 codes Archive

  • This database was the active CMF data release for 1999-2004 (Series 20, No. 2J, 2007) from April 2007 - March 2008, when it was replaced by the CMF 1999-2005 ( Series 20 No. 2K, 2008).
  • 2001-2004 population estimates: The region, division, state and county populations estimates for years 2001 and later, on the archive CMF 1999-2004 are estimates of the July 1 resident population from the county-level bridged-race Vintage 2005 postcensal series.
  • 2000 population estimates: The population estimates for 2000 are bridged-race modified April 1 census counts.
  • 1999 population estimates: The population estimates for 1999 are estimates of the July 1 resident population from the bridged-race intercensal estimates.
  • Data are available for Broomfield county, Colorado [08014] and Denali, Alaska [02068], only for 2003 and later. Before 2003, data for these areas are aggregated with neighboring areas. Data for Clifton Forge City, Virginia (51560) are available only for the years 1999-2000. In the year 2001, the area merged with merged with neighboring Alleghany county (51005).
  • The CMF 1999-2004 does not provide data for Hispanic Origin, nor data for these race groups: American Indian or Alaskan Native; Asian or Pacific Islander.
CMF 1999 - 2003 with ICD-10 codes Archive

  • This database was the active CMF data release for 1999-2003 (Series 20, No. 2I, 2006) from August 2006 to April 2007, when it was replaced by the CMF 1999-2004 ( Series 20 No. 2J, 2007).
  • 2001-2003 population estimates: The region, division, state and county populations estimates on the archive CMF 1999-2003 are estimates of the July 1 resident population from the county-level bridged-race Vintage 2004 postcensal series.
  • 2000 population estimates: The population estimates for 2000 are bridged-race modified April 1 census counts.
  • 1999 population estimates: The population estimates for 1999 are estimates of the July 1 resident population from the bridged-race intercensal estimates.
  • Data are available for Broomfield county, Colorado [08014] and Denali, Alaska [02068], only for 2003. Before 2003, data for these areas are aggregated with neighboring areas. Data for Clifton Forge City, Virginia (51560) are available only for the years 1999-2000. In the year 2001, the area merged with merged with neighboring Alleghany county (51005).
  • The CMF 1999-2005 does not provide data for Hispanic Origin, nor data for these race groups: American Indian or Alaskan Native; Asian or Pacific Islander.

CMF 1999 with ICD-10 codes Archive

  • This database was the active CMF data release (CMF 1989-2000, Series 20, No. 2C, 2001) from January 2002 to November 2003, when it was replaced by CMF 1999-2001 (Series 20, No. 2G, 2004).
  • 1999 population estimates: The population figures for 1999 are 1990-based postcensal population estimates, whereas the current CMF release has intercensal population estimates (based on both the 1990 and 2000 censuses) for 1999.
  • This query only provides data for 1999, due to differences in ICD-9 and ICD-10 disease classifications.
  • The CMF 1999 does not provide data for Urbanization, Hispanic Origin, nor data for these race groups: American Indian or Alaskan Native; Asian or Pacific Islander.

CMF 1979-1998 with ICD-9 codes Archive

  • This database was the active CMF data release (CMF 1989-2000, Series 20, No. 2C, 2001) from January 2002 through November 2003, when it was replaced by CMF 1989-1998 (Series 20, No. 2E, 2003).
  • 1991-1998 population estimates: The population figures for 1991-1998 are 1990-based postcensal population estimates, whereas the current CMF release has bridged-race intercensal population estimates (based on both the 1990 and 2000 censuses) for the years 1991-1998.
  • This query only provides data for 1979 through 1998, due to differences in disease classifications.
  • County-level data for the state of Alaska are not available in this database. Dade county, Florida represents Miami-Dade, Florida for all years 1979-1998. The database was updated in May 2003 to include death count and populations for La Paz county, Arizona since 1989, and for Cibola county, New Mexico since 1994. Location exceptions for the Archive CMF 1979-1998 are summarized in the table below:
    Location  Available   Unavailable  Notes
    Alaskan boroughs and census areas --*-- 1979-1998 Only Alaska state level available
    Cibola, New Mexico (35006) 1989-1998 1979-1988 Aggregated with Valencia (35061) before1989
    Dade, Florida (12025) 1979-1998 --*-- Represents Miami-Dade [12086)
    Kalawao, Hawaii [15005) --*-- 1979-1998 Aggregated with neighboring areas
    La Paz, Arizona (04012) 1994-1998 1979-1993 Aggregated with Yuma (04027) before1994
    Miami-Dade, Florida (12086) --*-- 1979-1998 Represented by Dade (12025)
    South Boston, Virginia (51780) 1979-1994 1995-1998 Aggregated with Halifax (51083) in 1995
    Unknown, Georgia (13999) 1988-1991 1979-1987,
    1992-1998
    AIDS deaths only
  • The CMF 1979- 1998 does not provide data for Urbanization, Hispanic Origin, nor data for these race groups: American Indian or Alaskan Native; Asian or Pacific Islander.

For more information, please see

Population Revisions
Locations: About County Geography Changes
Archive Mortality Data Requests
Data Source Citations for CMF On-line Databases



Mortality for Infants (under 1 Year of Age)

Causes of death among persons less than one year of age vary greatly during the first year of life, and therefore special "rates" (actually, ratios) have long been used in public health to provide meaningful indicators of infant mortality. Infant mortality rates are typically calculated as the number of deaths per 1,000 live births. Infant Age Groups" for rates calculated using the number live births as the population denominator. The default multiplier for Infant Age Groups (live births population) is 1,000 births. However, the default multiplier for death rates calculated with the population estimate for persons under 1 year of age is 100,000 persons.

Three commonly used indicators of infant mortality that can be calculated in WONDER are: 

  1. Infant Mortality Rate
    Number of deaths of infants (less than 1 year of age of death)
    divided by the number of live births during a given period,
    then multiplied by 1,000;

    (Deaths of persons under 1 year of age) / Live Births) * 1000

  2. Neonatal Mortality Rate
    Number of deaths of infants less than 28 days of age
    divided by the number of live births during a given period,
    then multiplied by 1,000;

    (Deaths of persons under 28 days of age) / Live Births) * 1000

  3. Postneonatal Mortality Rate
    Number of deaths of infants 28 days to 1 year of age
    divided by the number of live births during a given period,
    then multiplied by 1,000.

    (Deaths of persons age 28 days to 1 year of age) / Live Births) * 1000

Note that all three indicators use the same denominator:   number of live births during a given period.

To support these and other infant mortality indicators, the Compressed Mortality File provides first-year mortality data as follows: 

Infant Age Groups
less than one day old;
1 to 6 days old;
7 to 27 days old;
and 28 to 364 days old.

Hints:
  • Select "Infant Age Groups" for rates calculated using the number live births as the population denominator. The default multiplier for Infant Age Groups (live births population) is 1,000 births.
  • The single age group labeled "under 1 of age year" in the standard "Age Groups" list represents the population estimates for this age group. (Prior to August 2006, the number of live births in the given time period had been substituted for the "under 1 year of age" population estimates. Since August 2006, the number of live births for the specified year are available separately.) Death rates based on the population estimates are calculated per 100,000 persons, unless another multiplier is specified.
  • Change the default multiplier for the rate per number of persons in the Rate Options section on the Request screen.
Notes:
  • Age-adjusted rates are not available for the live births population, because each infant age group uses the same population, the number of live births, to produce the age-specific rates.
  • Age-adjusted rates are available for the "under 1 year of age" group, when this group is combined with other age groups. Age-adjusted rates are not calculated for any single age group because the ratio is effectively "1" in this case.
  • The number of live births are not summed together for the population total when the data are grouped by infant age groups, because the number of live births is used as the population denominator for each infant age group.
  • Note that rates for infant age groups are not shown for fewer than twenty deaths, nor are race and Hispanic Origin data available, due to privacy constraints; see Assurance of Confidentiality for more information.


County Geography Changes:  notes about specific county-level changes in boundaries and codes

Changes in county geography occur from time to time. For example, one county may absorb another, two counties may be merged resulting in the deletion of both of them and the creation of a new county, or part of a county may be split off to form a new county or be merged with an existing county. Creation and deletion of counties results in missing data for those counties for some years. County boundary changes can result in substantial increases or decreases in the population of the affected counties and hence impact birth and death counts, population estimates, and birth and death rates for those counties. Creation and deletion of counties often is implemented later in the vital statistics system than in the Census population files. Thus, for the CMF, county codes on the population data have been modified so that they match those in the corresponding vital statistics files. County geography changes that involve creation or deletion of counties in the CMF data are listed below. They are organized by state and county name.

  1. Alaska boroughs and census areas:
    The boroughs and census areas of Alaska undergo frequent changes making it difficult to work with individual Alaskan areas. Prior to 1989, the CMF did not include data for the individual Alaska areas. The FIPS codes used for vital statistics coding for Alaska changed in 1994 and in 2003 and therefore changed on the CMF. Specific area reporting details for Alaska are given below and summarized in the table.
    1. Alaska (Entire State)
      No data are available on the CMF for individual Alaska areas for the years 1965-1988. For these years, only state-level records (all individual areas combined) with a FIPS code of 02900 appear on the CMF. Alaska data are available for all areas combined for years 1965-1988. Data are reported as "Entire State, AK" (FIPS code 02900) for years 1968-1978, and as "Alaska (all counties, 1979-1988)" (FIPS code 02900) for years 1979-1988. Death counts and population estimates are available for individual areas within Alaska for years 1989-present.
    2. Aleutians East borough, Alaska
      Aleutians East borough, Alaska (FIPS code 02013) deaths and populations are on the CMF for 1994-present. For years 1989-1993, deaths and populations are reported for Aleutians Islands Census Area. Aleutians East borough (FIPS code 02013) was created from part of Aleutian Islands Census Area (FIPS code=02010) in 1987 from part of Aleutian Islands census area (FIPS code 02010) and an unpopulated part of Dillingham (FIPS code 02070).
    3. Aleutian Islands Census Area, Alaska
      Aleutian Islands Census Area, Alaska (FIPS code 02010) deaths and populations are on the CMF only for the years 1989-1993. For years 1994 - present, deaths and populations are reported for two areas, Aleutians East Borough (FIPS code=02013) and Aleutians West Census Area (FIPS code=02016).
    4. Aleutians West Census Area, Alaska
      Aleutians West Census Area, Alaska (FIPS code 02016) deaths and populations for Aleutians West are on the CMF for 1994 - present. For years 1989-1993, deaths and populations are reported for Aleutians Islands Census Area. Aleutians West Census Area, Alaska (FIPS code 02016) was created in 1990 from that part of Aleutian Islands (FIPS code 02010) that did not become part of Aleutians East (FIPS code 02013).
    5. Denali borough, Alaska
      Denali Borough, Alaska (FIPS code 02068) deaths and populations are on the CMF for years 2003 - present. For years 1989-2003, deaths and populations are reported for Yukon-Koyukuk (FIPS code 02290). Denali Borough, Alaska (FIPS code 02068) was created from the Yukon-Koyukuk Census Area (FIPS code 02290) and a small part of Southeast Fairbanks (FIPS code 02240) on December 7, 1990.
    6. Dillingham census area, Alaska
      Dillingham, Alaska (FIPS code 02070) deaths and populations are available on the CMF for years 1989 - present. There may be a discontinuity between 1993 and 1994 in the data reported for Dillingham because in 1989 part of Dillingham was removed to form Lake and Peninsula Borough (FIPS code 02164). This change was implemented in the vital records files and the CMF beginning 1994.
    7. Hoonah-Angoon Census Area, Alaska
      Hoonah-Angoon Census Area, Alaska (FIPS code 02105) deaths and populations are not available on the CMF. In 2008 Skagway-Hoonah-Angoon Census Area (FIPS code 02232) was divided into Hoonah-Angoon Census Area (FIPS code 02105) and Skagway Municipality (FIPS code 02230). This change has been implemented in the Census Bureau’s population files, but not in the mortality files. Therefore, the 2001-2009 revised intercensal population estimates for Hoonah-Angoon and Skagway Municipality are recoded to Skagway-Hoonah-Angoon Census Area for the CMF population file.
    8. Kobuck
      Kobuck, Alaska (FIPS code 02140) became Northwest Arctic borough (FIPS code 02188). There are no records for Kobuck on the CMF.
    9. Lake and Peninsula Borough, Alaska
      Lake and Peninsula Borough, Alaska (FIPS code 02164) deaths and populations are available on the CMF for years 1994 - present. For years 1989-1993, deaths and populations for this area are reported for Dillingham. Lake and Peninsula Borough, Alaska (FIPS code 02164) was created from part of Dillingham, Census Area, Alaska (FIPS code 02070) in 1989.
    10. Northwest Arctic borough, Alaska
      Northwest Arctic Borough, Alaska (FIPS code 02185) deaths and populations are available on the CMF for years 1989 - present. This area was formed in 1982 when an unpopulated part of North Slope Borough (FIPS code=02185) was combined with Kobuck (FIPS code=02140).
    11. Petersburg Census Area, Alaska
      Petersburg Census Area, Alaska (FIPS code 02195) deaths and populations are not available on the CMF. Petersburg Census Area (FIPS code 02195) was created in 2009 from the Petersburg part of Wrangell-Petersburg Census Area (FIPS code 02280). This change has been implemented in the Census Bureau’s population files, but not in the mortality files. Therefore, the 2001-2009 revised intercensal population estimates for Petersburg Census Area (FIPS code 02195) have been aggregated with those for the new Wrangell Borough (FIPS code 02275) and recoded to the former Wrangell-Petersburg (FIPS code 02280).
    12. Prince of Wales-Hyder Census Area, Alaska
      Prince of Wales-Hyder Census Area, Alaska (FIPS code 02198) deaths and populations are not available on the CMF. Prince of Wales-Hyder Census Area was formed in 2008 from part of Prince of Wales-Outer Ketchikan Census Area (FIPS code 02201). This change was implemented in the Census Bureau’s population files but has not been implemented in the vital records files. Therefore, the 2001-2009 revised intercensal population estimates for Prince of Wales-Hyder have been recoded to Prince of Wales-Outer Ketchikan Census Area. Population estimates for years 2001-2009 for Outer Ketchikan, which were aggregated into the Ketchikan Gateway estimates population files, cannot be recovered, but are estimated to be small.
    13. Prince of Wales-Outer Ketchikan Census Area, Alaska
      Prince of Wales-Outer Ketchikan Census Area, Alaska (FIPS code 02201) deaths and populations are available on the CMF for years 1989 - present. In 2009, Prince of Wales-Outer Ketchikan was disaggregated. The Outer Ketchikan portion was annexed by Ketchikan Gateway Borough, the Meyers Church part of Prince of Wales was annexed by Wrangell City and Borough, and the remainder became Prince of Wales-Hyder Census Area. This change was implemented in the Census Bureau’s population files but has not yet been implemented in the vital records files. Therefore, the 2001-2009 revised intercensal population estimates for Prince of Wales-Outer Ketchikan were obtained by recoding the population estimates for Prince of Wales-Hyder to Prince of Wales-Outer Ketchikan Census Area. There may be a discontinuity between the 2000 and 2001 estimates for this area because the populations of Outer Ketchikan and Meyers Church cannot be recovered.
    14. Skagway Municipality, Alaska
      Skagway Municipality, Alaska (FIPS code 2230) deaths and populations are not available on the CMF. Skagway-Hoonah-Angoon Census Area (FIPS code 02232) was divided into Skagway Municipality (FIPS code 02230) and Hoonah-Angoon Census Area (FIPS code 02105) in 2008. This change was implemented in the Census Bureau's population files but has not yet been implemented in the vital records files. Therefore, the 2001-2009 revised intercensal population estimates for Skagway Municipality have been aggregated with those for Hoonah-Angoon Census Area and recoded to Skagway-Hoonah-Angoon Census Area for the CMF population file.
    15. Skagway-Hoonah-Angoon Census Area, Alaska
      Skagway-Hoonah-Angoon Census Area, Alaska (FIPS code 02232) deaths and populations are available on the CMF for years 1994 - present. For years 1989-1993, data for Skagway-Hoonah-Angoon are reported for Skagway-Yakutat-Angoon (FIPS code 02231). Skagway-Hoonah-Angoon was formed in September 22, 1992 from that part of Skagway-Yakutat-Angoon not incorporated into Yakutat Borough (FIPS code=02282). In 2008, Skagway-Hoonah-Angoon was split into Hoonah-Angoon Census Area (FIPS code 02105) and Skagway Municipality (FIPS code 02230). This change was implemented in the Census Bureau’s population files but has not yet been implemented in the vital records files. Therefore, the 2001-2009 revised intercensal population estimates for Hoonah-Angoon Census Area and Skagway Municipality have been aggregated and recoded to Skagway-Hoonah-Angoon Census Area.
    16. Skagway-Yakutat-Angoon Census Area, Alaska
      Skagway-Yakutat-Angoon, Alaska (FIPS code 02231) deaths and populations are on the CMF for years 1989-1993. For years 1994 - present, deaths and populations for this area are reported for Yakutat (FIPS code 02282) and Skagway-Hoonah-Angoon (FIPS code 02232. Effective September 22, 1992, Skagway-Yakutat-Angoon Census Area was deleted after Yakutat (FIPS code 02282) and Skagway-Hoonah-Angoon (FIPS code 02232) were formed. This change was implemented in the vital records files, and hence in the CMF, beginning with the 1994 data year.
    17. Wrangell City and Borough, Alaska
      Wrangell City and Borough, Alaska (FIPS code 02275) deaths and populations are not available on the CMF. Wrangell City and Borough (FIPS code 02275) was created in 2009 from part of Wrangell-Petersburg Census Area (FIPS code 02280) and the Meyers Church part of Prince of Wales-Outer Ketchikan Census Area (FIPS code 02201). This change was implemented in the Census Bureau’s population estimates but has not yet been implemented in the vital records files. Therefore, the 2001-209 revised intercensal population estimates for Wrangell City and Borough have been aggregated with those for Petersburg (FIPS code 02195) and recoded to Wrangell-Petersburg Census Area for the CMF population file. The recombined area includes the additional population for Myers Church.
    18. Wrangell-Petersburg Census Area, Alaska
      Wrangell-Petersburg Census Area, Alaska (FIPS code 02280) deaths and populations are available on the CMF for years 1989 - present. In 2009, Wrangell-Petersburg (FIPS code 02280) split to form part of Wrangell City and Borough (FIPS code 02275) and all of Petersburg Census Area (FIPS code 02195). This change was implemented in the Census Bureau’s population estimates but has not yet been implemented in the vital records files. Therefore, the 2001-2009 revised intercensal population estimates for this area on the CMF were obtained by combining the estimates for Wrangell City and Borough with those for Petersburg Census Area. Because the new Wrangell City and Borough includes part of the former Prince of Wales-Outer Ketchikan (Meyers Church), the population of the recombined area is augmented somewhat.
    19. Yakutat Borough, Alaska
      Yakutat Borough, Alaska (FIPS code 02282) deaths and populations are available on the CMF for years 1994 - present. For years 1991-1993, deaths and populations for Yakutat are aggregated with those for Skagway-Hoonah-Angoon (FIPS code 02231). Yakutat Borough was created September 22, 1992 from part of Skagway-Yakutat-Angoon (FIPS code 02231). This change was implemented in the vital records files for the 1994 data year.
    20. Yukon-Koyukuk, Alaska
      Yukon-Koyukuk Census Area, Alaska (FIPS code 02290) deaths and populations are available on the CMF for years 1989 - present. However, there are discontinuities between 2002 and 2003 in the mortality and population data for Yukon-Koyukuk, because part of Yukon-Koyukuk Census Area was removed to form Denali Borough, Alaska (FIPS code 02068). This change was implemented in the vital records files and the CMF in the 2003 data year
    21. Table of Alaska boroughs and census areas in the CMF from 1989 - 1998

      					 FIPS  Code
      					______________________
      
      Borough / Census Area Name               1989-93        1994-98
      ______________________________________________________________
      
      Aleutian Islands Census Area		02010		--*--
      Aleutians East Borough 			--*--		02113
      Aleutians West Census Area		--*--		02016
      Anchorage Borough			02020		02020
      Bethel Census Area			02050		02050
      Bristol Bay Borough			02060		02060
      Denali Borough (02068)			--*--		--*--
      Dillingham Census Area			02070		02070
      Fairbanks North Star Borough		02090		02090
      Haines Borough				02100		02100
      Juneau Borough				02110		02110
      Kenai Peninsula Borough			02122		02122
      Ketchikan Gateway Borough		02130		02130
      Kobuck(-2140)				--*--		--*--
      Kodiak Island Borough			02150		02150
      Lake and Peninsula Borough 		--*--		02164
      Matanuska-Susitna Borough		02170		02170
      Nome Census Area			02180		02180
      North Slope Borough			02185		02185
      Northwest Arctic Borough		02188		02188
      Prince of Wales-Outer Ketchikan C.A	02201		02201
      Sitka Borough				02220		02220
      Skagway-Yakutat-Angoon Census Area	02231		--*--
      Skagway-Hoonah-Angoon Census Area	--*--		02232
      Southeast Fairbanks Census Area		02240		02240
      Valdez-Cordova Census Area		02261		02261
      Wade Hampton Census Area		02270		02270
      Wrangell-Petersburg Census Area		02280		02280
      Yakutat Borough				--*--		02282
      Yukon-Koyukuk Census Area		02290		02290
      
         --*-- Code does not appear on file for these years.
      

  2. Arizona: 
    1. La Paz, Arizona
      La Paz county, Arizona (FIPS code 04012) deaths and populations are available for years 1994 - present. For years 1965 - 1993, deaths and populations for this area are aggregated with those for Yuma county, Arizona (FIPS code 04027). In January, 1983, La Paz county, Arizona (FIPS code 04012) was formed from the northern portion of Yuma county (FIPS code 04027). This change was implemented in the vital records files and in the CMF beginning in 1994.
    2. Yuma, Arizona
      Yuma county, Arizona (FIPS code 02027) deaths and populations are available on the CMF for years 1968 - present. However, there are discontinuities between 1993 and 1994 in the data reported for Yuma County. These discontinuities occur because part of Yuma County was removed to form La Paz county (FIPS code 04012) in January 1983. This change was implemented in the vital records files and in the CMF beginning in 1994.
  3. Colorado: 
    1. Adams, Colorado
      Adams county, Colorado (FIPS code 08001) shows a discontinuity in the mortality and population data between 2002 and 2003. This discontinuity occurs because territory in each of these four counties has been combined to form a new county, Broomfield, Colorado (FIPS code 08014). Beginning in 2003, deaths and population counts for this territory are reported for Broomfield county and are no longer included with data for Adams county.
    2. Boulder, Colorado
      Boulder county, Colorado (FIPS code 08003) shows a discontinuity in the mortality and population data between 2002 and 2003. This discontinuity occurs because territory in each of these four counties has been combined to form a new county, Broomfield, Colorado (FIPS code 08014). Beginning in 2003, deaths and population counts for this territory are reported for Broomfield county and are no longer included with data for Boulder county.
    3. Broomfield county
      Broomfield county, Colorado (FIPS code 08014) was created effective November 15, 2001 from parts of four counties: Adams, Boulder, Jefferson, and Weld. Deaths and population estimates for Broomfield county appear on the CMF beginning in the year 2003. Deaths and population estimates before 2003 are coded to the original locations.
    4. Jefferson, Colorado
      Jefferson county, Colorado (FIPS code 08059) shows a discontinuity in the mortality and population data between 2002 and 2003. This discontinuity occurs because territory in each of these four counties has been combined to form a new county, Broomfield, Colorado (FIPS code 08014). Beginning in 2003, deaths and population counts for this territory are reported for Broomfield county and are no longer included with data for Jefferson county.
    5. Weld, Colorado
      Weld county, Colorado (FIPS code 080123) shows a discontinuity in the mortality and population data between 2002 and 2003. This discontinuity occurs because territory in each of these four counties has been combined to form a new county, Broomfield, Colorado (FIPS code 08014). Beginning in 2003, deaths and population counts for this territory are reported for Broomfield county and are no longer included with data for Weld county.
  4. Florida:  Dade county and Miami city
    Dade county, Florida (FIPS code 12025) was renamed Miami-Dade County and its FIPS code changed to 12086, effective November 13, 1997. The CMF 1999-2009 and CMF 1979-1998 have the area coded to Miami-Dade (12086) for all years. The archive CMF 1999-2003 and later archives CMF releases have the area coded to Miami-Dade (12086) for all years. The CMF 1968-1978, the archive CMF 1999, and the archive CMF 1979-1998 have the area coded to Dade County, FL (12025) for all years.
  5. Georgia:
    1. Columbus city and Muscogee county, Georgia
      The independent city Columbus, Georgia does not appear on the CMF. Death counts and population estimates for Columbus city (FIPS code 13510) have been aggregated with those for Muscogee county (FIPS code 13215).
    2. Georgia, unknown county
      For years 1988-1991, an additional county code (FIPS code 13999) was created for an “unknown” county in Georgia. Deaths occurring in Georgia in years 1988-1991 with HIV infection mentioned as the underlying cause of death or one of the multiple causes of death are assigned to this fictitious county when fewer than 3 occurred in the actual county of residence of the decedent. No population estimates are available for FIPS code 13999. The deaths assigned to this FIPS code are included in Georgia state totals.
  6. Hawaii: Kalawao, Hawaii
    Kalawao, Hawaii [FIPS code 15005) has data on the CMF beginning in the year 1989. There are no data for the years for years 1979-1988 in the CMF 1979-1998 online database, and no data for the full span of years in the archive 1979-1998 online database. There are no data for the full span of years in the 1969-1978 online database. Death counts and population estimates for Kalawao, Hawaii (FIPS code = 15005) are aggregated with those for Maui county, Hawaii (FIPS code=15009), when data for Kalawao, Hawaii are not available.
  7. Louisiana and Orleans and East Baton Rouge parishes after Hurricane Katrina in 2005:
    The July 1st 2005 population estimates for Louisiana do not reflect population changes that occurred after Hurricane Katrina on August 29, 2005. The population of Orleans parish in Louisiana was reduced following Hurricane Katrina; East Baton Rouge parish in Louisiana had an increased population. As a result, 2005 death rates for Orleans parish calculated using the population estimates will be too low; the death rates calculated for East Baton Rouge will be somewhat too high. Death rates for Orleans parish are especially affected.
  8. Maryland:  Baltimore city and Baltimore county
    The independent city of Baltimore, Maryland has been treated as a county. Death counts and population estimates are reported separately for Baltimore city (FIPS code 24510) and Baltimore county (FIPS code 24005).
  9. Missouri:
    1. St. Genevieve county, Missouri
      In order to achieve alphabetical consistency, the FIPS code for St. Genevieve, Missouri was changed in 1979 from 29193 to 29186. The new code (29186) has been used throughout the CMF.
    2. St. Louis city and St. Louis county, Missouri
      The independent city of St. Louis, Missouri has been treated as a county. Death counts and population estimates are reported separately for St. Louis city (FIPS code 29510) and St. Louis county (FIPS code 29189).
  10. Montana:  Yellowstone National Park
    Until November 7, 1997, the Montana portion of Yellowstone Park was not in any county and therefore was treated as a county equivalent (FIPS code 30113). On that date, the Montana portion of Yellowstone Park became part of Gallatin, Montana (FIPS code 30031) and Park, Montana (FIPS code 30067). The number of deaths in Yellowstone Park was so small that this should not create a discontinuity. Data for Yellowstone Park (30113) does not appear after data year 1988 on the CMF 1979-1998. However, Yellowstone Park (30113) data are available for the years 1979-1998 on the Archive CMF 1978-1998. and Yellowstone Park (30113) data are available for the years 1969-1978 in the 1968-1978 CMF. Deaths and population coded as Yellowstone National Park, Montana (FIPS code 30113) in the year 1968 are recoded to Park County, Wyoming (FIPS code 56029).
  11. Nevada:  Carson City
    The independent city of Carson City, Nevada (FIPS code 32510) has been treated as a county. Death counts and population estimates are reported for Carson City.
  12. New Mexico:  Cibola county and Valencia county
    Valencia county, New Mexico (FIPS code 35061) data are available on the CMF for years 1968 - present. Data for Cibola county, New Mexico (FIPS code 35006) are available in the CMF for years 1989 - present. However, there are discontinuities between 1988 and 1989 in the data reported for Valencia County. These discontinuities occur because part of Valencia County was removed to form Yuma County (in 1981) and this change was implemented in the vital records files and in the CMF beginning in 1989. Hence Valencia shows an abrupt drop in deaths and populations in 1989.
  13. New York:  New York City boroughs
    The five boroughs of New York City have been treated as counties and maintained as separate entities on this file.

    Borough County FIPS Code
    Bronx Bronx 36005
    Brooklyn Kings 36047
    Manhattan New York 36061
    Queens Queens 36081
    Staten Island Richmond 36085

  14. South Dakota:  Jackson county and Washabaugh county
    In 1979, Washabaugh county, South Dakota (FIPS code 46131) merged with Jackson county, South Dakota (FIPS code 46071). For all years, death counts and population estimates for Washabaugh county have been aggregated with those for Jackson county.
  15. Virginia independent cities: 
    1. Alleghany, Virginia
      Alleghany county, Virginia (FIPS code 51005) deaths and populations are available on the CMF for years 1968 - present. Clifton Forge city, Virginia (FIPS code 51560) data are available for years 1968-2000. Beginning with the 2001 data year, Clifton Forge deaths and population estimates are aggregated with those of Alleghany County because this independent city merged with the county in 2001. As a result of this change, there is a discontinuity between 2000 and 2001 in the data reported for Alleghany County.
    2. Clifton Forge city, Virginia
      Clifton Forge city, Virginia (FIPS code 51560) deaths and populations are available on the CMF for years 1968-2000. On July 1, 2001, Clifton Forge city, formerly an independent city, merged with Alleghany county (FIPS code 51005).
    3. Nansemond city, Virginia
      Nansemond city, Virginia (FIPS code 51123) has been part of the independent city of Suffolk, VA (FIPS code 51800) since 1979. For all years, death counts and population estimates for Nansemond have been aggregated with those for Suffolk city.
    4. Halifax, Virginia
      For CMF 1979-1998, there is a discontinuity in the mortality and population data for Halifax, Virginia (FIPS code 51083) between 1988 and 1989. The discontinuity occurs because South Boston city, Virginia (FIPS code 51780) merged with Halifax county. For CMF 1979-1998, beginning in data year 1989, deaths and population counts for South Boston city are reported with those of Halifax county. However, for CMF 1979-1998 Archive data, death counts and population estimates for South Boston city are aggregated with those of Halifax county since 1995.
    5. South Boston city, Virginia
      South Boston city, Virginia (FIPS code 51780) deaths and populations are available on the CMF for years 1968-1988, and available through year 1994 in the archive CMF 1979-1998. Death counts and population estimates for South Boston city are aggregated with those of Halifax county since 1989 in the current release; however South Boston city data are aggregated with Halifax county since 1995 for the archive CMF 1979-1998. On June 30, 1995, South Boston city, Virginia (FIPS code 51780), formerly an independent city, merged with Halifax county (FIPS code 51083).
    6. Table of Virginia independent cities and counties
      The Virginia independent cities are treated as counties and appear on the CMF with the following FIPS codes:

         Independent City	       County
      

      Name	    FIPS code	Name	    FIPS code	City Data available * 
      

      Alexandria 	51510	Arlington	51013	1968-current
      Bedford		51515	Bedford		51019	1979-current			
      Bristol		51520	Washington	51191	1979-current
      Buena Vista	51530	Rockbridge	51163	1979-current
      Charlottesville	51540	Albemarle 	51003	1979-current
      Chesapeake	51550				1968-current 
      Clifton Forge	51560	Alleghany 	51005 	1979-2000 
      Colonial Heights51570	Chesterfield 	51041	1979-current
      Covington	51580	Alleghany 	51005	1979-current
      Danville	51590	Pittsylvania 	51143	1979-current
      Emporia		51595	Greensville	51081	1979-current
      Fairfax		51600	Fairfax	 	51059	1979-current
      Falls Church	51610	Fairfax 	51059	1979-current
      Franklin	51620	Southampton	51175	1979-current
      Fredericksburg	51630	Spotsylvania	51177	1979-current
      Galax		51640	Grayson 	51077   1979-current     
      Hampton		51650				1968-current	
      Harrisonburg	51660	Rockingham	51165	1979-current
      Hopewell	51670	Prince George	51149	1979-current
      Lexington	51678	Rockbridge	51163	1979-current
      Lynchburg	51680	Campbell 	51031	1979-current
      Manassas	51683	Prince William	51153	1979-current
      Manassas Park	51685	Prince William	51153	1979-current
      Martinsville	51690	Henry	 	51089	1979-current
      Newport News	51700				1968-current
      Norfolk		51710				1968-current
      Norton		51720	Wise	 	51195	1979-current
      Petersburg	51730	Dinwiddie	51053	1979-current
      Poquoson	51735	York	 	51199	1979-current
      Portsmouth	51740	Norfolk city	51710	1979-current
      Radford		51750	Montgomery	51121	1979-current
      Richmond	51760	Henrico	 	51087	1979-current
      Roanoke		51770	Roanoke	 	51161	1979-current
      Salem		51775	Roanoke	 	51161	1979-current
      South Boston *  51780	Halifax 	51083	1979-1988, 1979-1994 in Archive
      Staunton	51790	Augusta	 	51015	1979-current
      Suffolk		51800				1968-current
      Virginia Beach	51810				1968-current
      Waynesboro	51820	Augusta	 	51015	1979-current
      Williamsburg	51830	James City	51095	1979-current
      Winchester	51840	Frederick	51069	1979-current
      
      * Note that data discontinuities occur when data are not available for all years.




This page last reviewed: Monday, April 28, 2014
This information is provided as technical reference material. Please contact us at cwus@cdc.gov to request a simple text version of this document.