Number of Deaths

The number of deaths for each occupational respiratory condition is the number of decedents for which the condition was coded as a cause of death on the entity axis in the National Center for Health Statistics (NCHS) multiple cause-of-death data files. [Note: In eWoRLD updates prior to June 2008, the numbers of deaths were generally based on the numbers of decedents for which each condition was coded as a cause of death on the record axis. Changing to the entity axis multiple cause-of-death permits more complete ascertainment of the diseases of interest.] See Detail Record Layout at http://www.cdc.gov/nchs/about/major/dvs/mcd/1998mcd.htm for a discussion of the appropriate uses, advantages, and disadvantages of the multiple cause-of-death code fields (entity axis and record axis).

Percentages of deaths for the pneumoconioses, malignant mesothelioma, and hypersensitivity pneumonitis that NCHS coded as the underlying cause of death are listed by year. Since 1999, deaths with underlying cause-of-death International Classification of Diseases (ICD)-10 code J65 (pneumoconiosis associated with tuberculosis) are included in the underlying cause-of-death tabulations of each specified type of pneumoconiosis (except byssinosis), and for each type of tuberculosis. Deaths with J65 coded as a cause of death on the entity axis are included in the entity axis multiple cause-of-death tabulations for code J64 (unspecified pneumoconiosis) and codes A15, A16, A19, or B90.9 (collectively defined as respiratory tuberculosis) in updates later than August, 2008. Similarly, deaths with underlying cause-of-death code J92.0 (pleural plaque with presence of asbestos) are included in asbestosis underlying cause-of-death tabulations. Deaths with malignant mesothelioma of the pericardium (ICD-10 code C45.2) are grouped into malignant mesothelioma of other sites (ICD-10 code C45.7).

The sum of individual pneumoconiosis (i.e., asbestosis, coal workers' pneumoconiosis, silicosis, byssinosis, and unspecified/other pneumoconiosis) death totals may be greater than the total number of deaths for all pneumoconioses because some decedents have more than one type of pneumoconiosis listed on their death certificate. The sum of individual malignant mesothelioma (i.e., pleura, peritoneum, other sites, and unspecified) death totals may be greater than the total number of deaths for all sites because some decedents have more than one type of malignant mesothelioma listed on their death certificate.

To avoid inadvertent disclosure of confidential data, county-level statistics are computed for time periods that overlap with similar statistics presented in eWoRLD by at least 5 years, and omit data from counties having less than 10 deaths with the condition of interest. Since 2005, any state-level mortality statistics for less than 10 decedents are also omitted from tables and figures.

Reported deaths are restricted to U.S. residents, 15 years or older, based on state and county of residence at death. Race is classified as white, black, and all others. Industry and occupation classifications are described later in the Industry/Occupation Codes and Titles section below.

Crude Death Rates

To compute annual cause-specific crude death rates, the total number of decedents, 15 years and older, with a specified condition coded as a cause of death in the NCHS data files in a given year is divided by the population, 15 years and older, of the same geopolitical unit in the same year. Race-specific and sex-specific, and/or Hispanic origin-specific rates are computed from the appropriate subsets of the data. Rates are calculated annually for each specified condition, as well as for selected periods. Crude death rates for all conditions except malignant mesothelioma, which has only been classified as a separate condition since 1999 (see International Classification of Diseases (ICD) Codes), are computed at the national, state, and county level for multi-year periods. Malignant mesothelioma rates are computed from 1999 at the national and state levels and from 2000 at the county level. For any multi-year period, the average annual number of decedents, 15 years and older, with a specified condition is divided by the mid-year population, 15 years and older, of the same geopolitical unit.

Comparability ratios are not applied to rates. No adjustments are made to account for any potential variation in the classification of respiratory diseases and conditions across ICD revisions. eWoRLD users should read the following selected guidelines and methods for determining and applying comparability ratios when interpreting trend data that span one or more ICD revisions. (See Comparability of Cause-of-death Between ICD Revisions at http://www.cdc.gov/nchs/nvss/mortality/comparability_icd.htm .)

Age-Adjusted Death Rates

Age-adjusted death rates are based on deaths with the condition of interest coded as a cause of death in the NCHS data files. Rates are calculated annually for each specified condition, as well as for selected periods. For a given year, the age-adjusted rate represents the rate that would have been observed if the age-specific rates for specified age groups had occurred in a population with the same age distribution as that of the standard population. Either the U.S. Year 1940 Standard Population (updates prior to December 2005) or the U.S. Year 2000 Standard Population (updates since December 2005) may be used as the standard. The specific age intervals used are 15–24, 25–34, 35–44, 45–54, 55–64, 65–74, 75–84, and 85 years and older. Rates for the entire United States population and for each sex/race group are age-adjusted separately, using the same standard population.

Age-adjusted rates are computed by the direct method. First, the annual age-specific rates for the population of interest are calculated. The product of the age-specific rates and the number in the comparable age-specific group in the standard population equals the expected number of deaths per million population for each age group. The total expected numbers of deaths are then obtained by summing over all age groups. The total expected number of deaths is divided by the sum of the standard population and the resulting quotient is multiplied by 1,000,000 to produce the age-adjusted rate (per million).

Age-adjusted rates are computed at the national and state level for multi-year periods. Rates are computed at the county level either for periods of 10 or more years, except for Malignant Mesothelioma where rates are calculated for periods of less than 10 years. For any time period, age-specific rates first are computed by dividing the average annual number of deaths in each age group by the corresponding age-group, mid-year population in the comparable geopolitical unit. Age-adjusted rates then are computed as described above. For more information, see Healthy People 2010, Statistical Note No. 20 at http://www.cdc.gov/nchs/products/hp_pubs.htm.

Years of Potential Life Lost (YPLL)

Years of Potential Life Lost (YPLL) are based on deaths with the condition of interest coded as a cause of death on the NCHS data files. They are calculated using the method described by the Centers for Disease Control and Prevention (MMWR December 19, 1986/35(2S);1s–11s at http://www.cdc.gov/mmwr/preview/mmwrhtml/00001773.htm ). YPLL are calculated both to age 65 and to life expectancy. YPLL to age 65 may be considered as a loss of years from a traditional working life, while YPLL to life expectancy may be considered as a loss of years from the overall life span. To compute YPLL to life expectancy, the number of deaths in each race/sex age group (the same age intervals used for computing age-adjusted rates) first is multiplied by the difference between the mid-point of the age group and life expectancy for that race/sex age group. Life tables published annually by NCHS are used to determine race/sex life expectancies for categories male, female, white, black, white/male, white/female, black/male, and black/female. The overall U.S. population life expectancy is used for other, other/male, and other/female. To compute YPLL to age 65, the number of deaths in age groups up to and including 55–64 is multiplied by the difference between 65 years and the mid-point of each age group (e.g., 65 minus 20 years for the 15–24 age group). The age-specific YPLLs then are summed over all specified age groups to obtain the total YPLL.

State-specific YPLLs (to life expectancy) per death also are calculated for a given time period. To calculate this index, the total number of all race/sex deaths in each age group is multiplied by the corresponding U.S. population life expectancy, then summed over all age groups to obtain the total YPLL, and then divided by the total number of deaths for each state during the given time period. To avoid inadvertent disclosure of confidential data, any state-level deaths for which there are less than 10, the YPLL is not presented in tables from 2005 onwards.

Due to a revision of the methodology used to determine life expectancy, 2005 YPLL to life-expectancy estimates were revised. This revision is described in the NCHS National Vital Statistics Report Volume 58, No. 10 at http://www.cdc.gov/nchs/products/nvsr.htm. In previous updates, these YPLLs were based on preliminary 2005 life-expectancy data obtained from the National Vital Statistics Report Volume 56, No. 10. For any given year, life-table data presented are obtained from the National Vital Statistics Reports for Preliminary or Final Death Data for that year found at http://www.cdc.gov/nchs/products/nvsr.htm .

Rank Order

For each state, a rank order is presented for each of several mortality measures. Depending on the specific mortality measures, a rank order of "1" indicates the greatest number of deaths, highest death rate, or highest years of potential life lost among all states in the United States.

Industry/Occupation Codes and Titles

NCHS provided specific industry and occupation information in the multiple cause-of-death data files from 1985 to 1999. Industry and occupation was obtained from the following items that appeared on the U.S. Standard death certificate: Decedent's usual occupation and Kind of business/industry, respectively. Since 2000, funding for states to code industry ended and NCHS no longer provided industry and occupation information with the multiple cause-of-death data.

Between 1993 and 1999, the 1990 Bureau of Census (BoC) Index of Industries and Occupations classification system was used for coding industry and occupation. Most codes and titles in the 1990 system do not differ from the 1980 BoC system, which was used prior to 1993. All tables reporting BoC industry (CIC) and occupation (COC) codes and titles, except those listed in Changes in Bureau of Census Industry and Occupation Codes and Titles, follow the 1980 BoC classification system. For more information see, for example, Technical Appendix for 1995 or Technical Appendix for 1999 at http://www.cdc.gov/nchs/products/vsus.htm#appendices , and pages XI–XIV in U.S. Dept. Commerce 1992, Alphabetical Index of Industries and Occupations, Bureau of the Census, 1990 CPH-R-3.

NIOSH established the National Occupational Research Agenda (NORA) as a partnership program to stimulate innovative research and improved workplace practices (see http://www.cdc.gov/niosh/nora/default.html). To assign CIC codes to NORA Sector Groups (see http://www.cdc.gov/niosh/nora/sector.html ), the CIC codes were merged using the Census 2000 Industrial Classification System and Crosswalk to North American Industry Classification System (NAICS) available at http://www.census.gov/people/io/ and the 1997 NAICS-US Matched to 2002 NAICS-US Excel spreadsheet available at http://www.census.gov/eos/www/naics/concordances/concordances.html . The CIC codes are listed by NORA Sector Group in Bureau of Census Industry Codes by NORA Sector Group.

Most Frequently Recorded Industries/Occupations

The ten most frequently recorded Bureau of Census industries and occupations with at least two decedents are listed for specified causes of death from selected states and years (see States (and Years) for which Industry and Occupation Codes from Death Certificates Met NCHS Quality Criteria, 1985-1999). Where more than one industry/occupation is tied for tenth place, all those that are tied are listed.

Proportionate Mortality Ratio

The data used for proportionate mortality ratio (PMR) analyses are a subset of the NCHS multiple cause-of-death files for which usual industry and occupation codes are available and met quality criteria set by NCHS (see States (and Years) for which Industry and Occupation Codes from Death Certificates Met NCHS Quality Criteria, 1985-1999 for a list of states and years for which data qualified).

The PMR is defined as the observed number of deaths with the condition of interest coded as a cause of death on the entity axis of the NCHS data files in a specified industry/occupation from selected states and years (see States (and Years) for which Industry and Occupation Codes from Death Certificates Met NCHS Quality Criteria, 1985-1999), divided by the expected number of deaths with that condition. The expected number of deaths is the total number of deaths in the Bureau of Census industry (CIC) or occupation (COC) of interest multiplied by a proportion defined as the number of cause-specific deaths for the condition of interest in all industries and/or occupations, divided by the total number of deaths in all industries/occupations; this includes deaths in retired, non-paid, homemaker, and unreported industries/occupations. Current PMRs and those in eWoRLD updates dated December 2005 and June 2008 are internally adjusted by five-year age groups (i.e., 15–19, 20–24, . . ., and 85 years and older), sex, and race (i.e., white, black, and all other). Confidence intervals are calculated assuming a Poisson distribution of the data.

A PMR greater than 1.0 indicates that there are more deaths associated with the condition in a specified occupation or industry than expected. Industries/occupations with five or more decedents with the condition of interest and a lower 95% confidence limit exceeding 1.0 (in some instances the lower 95% confidence limit may equal 1.0 due to rounding) are only presented.

Industries with Elevated PMRs and Most Frequently Recorded on Death Certificates

The number of samples, geometric mean exposures, and percent of samples exceeding the permissible exposure limit (PEL) or recommended exposure limit (REL) by selected industries for exposure agents related to elevated occupational lung disease mortality are presented. For further information about related exposures see Exposure Methods.

For asbestosis, coal workers' pneumoconiosis, silicosis, byssinosis, and all pneumoconiosis, separate tables present data for the ten most frequently recorded industries with five or more decedents and significantly elevated PMRs. [Note: see Industry/Occupation Codes and Titles section above.]

State and County Designations

The "number of states" displayed on maps sums to 51 because the District of Columbia is included. Counties are coded according to the 1990 Federal Information Processing Standards (FIPS) Codes system. A small number of counties or county equivalents have split, merged with, or separated from surrounding or adjacent subdivisions (see Split, Merged, or Renamed Counties and County Equivalents). Readers should be cautious in assessing geographic patterns and temporal trends for subdivisions that have split or merged.