Search NIOSH  |  NIOSH Home  |  NIOSH Topics  |  Site Index  |  Databases and Information Resources  |  NIOSH Products  |  Contact Us

National Occupational Respiratory Mortality System — Glossary

A selected list of terms used on NORMS Occupational Respiratory Disease Surveillance pages.

Terms followed by this question-mark icon
are linked to other definitions within this Glossary.

A B C  D E F  G H I  J K L  M N O  P Q R  S T U  V W X  Y Z 

Age-adjusted Death Rates

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 or the U.S. Year 2000 Standard Population may be used as the standard.  The specific age intervals used are 0-14, 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 are 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).  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-grouped, mid-year population in the comparable geopolitical unit.  Age-adjusted rates then are computed as described above.   For more information, see Age Adjustment.

Age-specific Death Rates

Same as crude death rates except that an age-specific rate typically is calculated for a particular age group, whereas a crude death rate typically is calculated for all decedents regardless of age.

Code Fields

For analytical purposes, each cause of death that is written on a death certificate is converted into an International Classification of Diseases code.   There can be as many as 20 causes of death listed on a death certificate.   These causes of death are collectively referred to as a field of codes (i.e., a code field).   A "Multiple Cause" query in NORMS is directly comparable to the entity axis multiple cause-of-death code-field data published annually by NCHS.   An "Underlying Cause" query in NORMS is directly comparable to the underlying cause-of-death code-field data published annually by NCHS.   A "Contributing Cause" code-field query in NORMS generates data for the number of decedents where the selected respiratory conditions are a contributing cause-of-death on the entity axis .   See “Detail Record Layout” at www.cdc.gov/nchs/about/major/dvs/mcd/1998mcd.htm for a discussion of the appropriate uses, advantages, and disadvantages of the code fields.   For an example of the code-field data published annually by NCHS, see 1998 Control Total Table 1 (PDF only; 90KB; 43 pgs).   Please note that NCHS record axis code-field data are not included in NORMS.

Comparability Ratios

A comparability ratio estimates the difference between a sample of death certificates for which the causes of death have been coded according to the standards of two International Classification of Diseases revisions.   NORMS data users should first read the following selected definition and caveats and methods for determining and applying   (PDF only; 338KB; 32 pgs) comparability ratios when interpreting or analyzing NORMS trend data that span one or more ICD revisions.   The preceding sources contain underlying-cause comparability ratios estimated for some of the respiratory conditions included in the NORMS database.

Contributing Cause-of-Death

A condition listed on a decedent's entity axis code field that is not also coded as the underlying cause-of-death for that decedent.

Crude Death Rates

To compute annual cause-specific crude death rates, the total number of decedents with a specified condition in a given year is divided by the population of the same geopolitical unit in the same year.  Race-specific, sex-specific, and/or Hispanic origin-specific rates are computed from the appropriate subsets of the data.  For any multi-year period, the average annual number of decedents with a specified condition is divided by the mid-year population of the same geopolitical unit.

Entity Axis

This code field includes all of the diseases, injuries, or medical complications, as well as the location (part, line, and sequence) of the information recorded on each death certificate.   See Parts 2b and 2f of the Instruction Manuals for classifying multiple cause-of-death data.

Hispanic Origin

NORMS does not include any data for Hispanic origin prior to 1990.  Most states began reporting this information on death certificates in 1990, however, it was not until 1997 that all states adopted the practice.  The following states did not report Hispanic origin (in the following years):

• Louisiana, New Hampshire, and Oklahoma (1990)
• New Hampshire and Oklahoma (1991-1992)
• Oklahoma (1993-1996)

International Classification of Diseases (ICD)

The most recent ICD was published by the World Health Organization (WHO) [ External Link: http://www.who.int/en/ ].   It is used to categorize diseases and other health problems.  Depending on the condition/disease of interest, occupational respiratory disease mortality data may span one to three revisions of the ICD: ICDA-8 (1968-1978); ICD-9 (1979-1998); and ICD-10 (1999-present).  See the table of the occupational respiratory disease ICD codes that are included in NORMS.

Industry/Occupation

Between 1993 and 1999, the 1990 U.S. Bureau of Census (BoC) Index of Industries and Occupations classification system was used for coding death certificate information on the NCHS multiple cause-of-death data files.  Most codes and titles in the 1990 system do not differ from the 1980 BoC system used for coding death certificates between 1985 and 1992.   All NORMS tables reporting BoC industry (CIC) and occupation (COC) codes and titles, except those listed in Appendix D of the 2002 WoRLD Surveillance Report, follow the 1980 BoC classification system.   Refer to the "Technical Appendix for 1995" at www.cdc.gov/nchs/about/major/dvs/mcd/1998mcd.htm for additional information.   For a comparison of the 1990 CIC and COC codes with more recent industrial and occupational coding systems, refer to the "Industry and Occupation Crosswalks" on the U.S Census Bureau Industry and Occupation 2000 [ External Link: http://www.census.gov/hhes/www/ioindex/ ioindex.html ] Web page.   Use the Name Finder to search the lists of CIC and COC titles, or to obtain a complete listing.

Multiple Cause-of-Death

NCHS has made available annual public-use multiple cause-of-death data files since 1968.   These files contain records of all deaths in the United States (approximately two million annually) that are reported to state vital statistics offices.   Each death record includes codes for up to 20 (14 for ICDA-8 data) conditions derived from the "Cause of Death" section of the death certificate form, in two fields: the entity axis , which preserves diagnostic detail for all listed conditions and their placement on the death certificate; and the record axis , which reorders the codes, removes redundancies, and (infrequently) combines some associated conditions.   For additional information concerning multiple cause-of-death, see “Detail Record Layout” at www.cdc.gov/nchs/about/major/dvs/mcd/1998mcd.htm).   Please note that record axis data are not included in NORMS.

Number of Deaths

The number of deaths for each occupational respiratory condition is the number of individual decedents for which the condition was coded.  Since 1999, deaths with underlying-cause International Classification of Diseases code J65 (pneumoconiosis associated with tuberculosis) are included in the underlying cause-of-death tabulations of each specific type of pneumoconiosis (except byssinosis), and for each type of tuberculosis.  Similarly, deaths with underlying-cause code J92.0 (pleural plaque with presence of asbestos) are included in asbestosis underlying cause-of-death tabulations.  Data are restricted to United States residents, based on state of residence at death.  Race is classified as white, black, and all others.

Population Estimates

National population estimates are based on national, state-level, and county-level data from the United States Bureau of the Census (BoC).   All population estimates used to compute rates for 1968-1999 have been those obtainable through the CDC computer system.   Estimates obtained from unmodified intercensal Demo-Detail files were used for 1970-1979 and for 1980-1989.   The unmodified 1970 intercensal population estimates were used for 1968-1969 because no other county-level population estimates were available.   Postcensal Demo-Detail estimates were used for 1990-1995.   Comparable postcensal estimates prepared by the BoC were used for 1996-1999.   Since 2000, NCHS bridged-race estimates have been used:   for data years 2000-2002, we used the bridged-race vintage 2002 postcensal estimates of the resident population of the United States as of July 1, 2000, July 1, 2001, and July 1, 2002 (file name "pcenv2002.txt");   for data year 2003, we used the bridged-race vintage 2003 postcensal estimates of the resident population of the United States as of July 1, 2003 (file name "pcen_v2003_y03.txt");   for data year 2004, we used the bridged-race vintage 2003 postcensal estimates of the resident population of the United States as of July 1, 2004 (file name "pcen_v2004_y04.txt");   for data year 2005, we used the bridged-race vintage 2003 postcensal estimates of the resident population of the United States as of July 1, 2005 (file name "pcen_v2005_y05.txt");   and for data year 2006, we used the bridged-race vintage 2004 postcensal estimates of the resident population of the United States as of July 1, 2006 (file name "pcen_v2006_y06.txt").   [Note: Comparing population statistics from Demo-Detail and BoC postcensal estimates for each year from 1990 through 1995, we observed a maximum annual difference of less than 0.05 percent, and a difference of 0.01 percent or less in a majority of years.  State-specific differences for the same years were less than one percent for all states, with very rare exceptions.]   Hispanic origin population estimates were not originally coded in the 1990-1995 Demo-Detail file, so an algorithm was used to compute the ratios of Hispanic and non-Hispanic population estimates in the 1990-1995 BoC postcensal files, by age, race, and sex, which was used to apportion the Demo-Detail population estimates by Hispanic origin.   Because of the modifications that were made to the original Demo-Detail estimates, the population estimates used in NORMS for calculating death rates may be difficult to ascertain.   Therefore, the Census Database query is available for generating the population estimates used in NORMS.

Proportionate Mortality Ratio (PMR)

The data used for PMR analyses are a subset of the NCHS multiple cause-of-death files for which usual industry and occupation codes are available and meet quality criteria set by NCHS.  The PMR is defined as the observed number of deaths with the condition of interest in a specified industry/occupation, divided by the expected number of deaths with that condition.  The expected number of deaths is the total number of deaths in BoC 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.  The PMRs generated by NORMS are internally adjusted by five-year age groups (i.e., 15-19, 20-24, ...), sex, and race (i.e., white, black, and all other).  Confidence intervals are calculated assuming Poisson distribution of the data.  A PMR greater than 1.0 indicates that there were more deaths associated with the condition in a specified occupation or industry than expected.

Record Axis

This code field is generated by a computer algorithm that interprets and translates the information on the entity axis by removing duplicate codes and contradictory diagnoses, and reorders them alphanumerically.   See Part 2f of the Instruction Manuals for classifying multiple cause-of-death data, and Series Report Number 20.   The record axis data are not included in NORMS.

Standard Population

The Standard Population is the distribution of U.S. residents, by age group, for a given year. NORMS weights age-adjusted death rates by the percentage of the census population within each of the following age groups, for either the 1940 or the 2000 Census:

The standard-million population distribution is used when the selected lower age limit is less than 15

Age Group	1940 Census		2000 Census
	Population	Proportion (%)	Population	Proportion (%)
0-14	250,416	25.04	214,700	21.47
15-24	181,677	18.17	138,646	13.86
25-34	162,066	16.21	135,573	13.56
35-44	139,237	13.92	162,613	16.26
45-54	117,811	11.78	134,834	13.48
55-64	80,294	8.03	87,247	8.72
65-74	48,426	4.84	66,037	6.60
75-84	17,303	1.73	44,842	4.48
85+	2,770	0.27	15,508	1.55
Total	1,000,000	100.00	1,000,000	100.00

The working-age population distribution is used when the selected lower age limit is 15 or older

Age Group	1940 Census		2000 Census
	Population	Proportion (%)	Population	Proportion (%)
15-24	181,677	24.24	138,646	17.66
25-34	162,066	21.62	135,573	17.26
35-44	139,237	18.58	162,613	20.71
45-54	117,811	15.72	134,834	17.17
55-64	80,294	10.71	87,247	11.11
65-74	48,426	6.46	66,037	8.41
75-84	17,303	2.31	44,842	5.71
85+	2,770	0.37	15,508	1.97
Total	749,584	100.00	785,300	100.00

For more information, see Age Standardization of Death Rates: Implementation of the Year 2000 Standard (PDF only; 260KB; 20 pgs), which is also accessible from http://www.cdc.gov/nchs/datawh/nchsdefs/ageadjustment.htm.

Statistical comparisons of rates

Methods for statistical comparison of rates are described on pages 111-118 of http://www.cdc.gov/nchs/data/nvsr/nvsr55/nvsr55_19.pdf   (PDF only; 3.36MB; 120 pgs).   NORMS generates standard errors (pages 113-115), when rates are based on 100 or more deaths, for making statistical comparisons between two rates using the normal approximation method (page 116).   NORMS also provides 95% confidence intervals based on the normal approximation (page 116) or the gamma method (pages 116-117) for statistical comparisons when one or both rates are based on fewer than 100 deaths.

Underlying Cause-of-Death

This code field is represented by a single International Classification of Diseases code that is assigned by a computer algorithm based on the information found in the "Cause of Death" section of the death certificate provided by the certifier of the death.  The World Health Organization (WHO) [ External Link: http://www.who.int/en/ ] defines the underlying cause-of-death as (a) the disease or injury which initiated the train of events leading directly to death, or (b) the circumstances of the accident or violence which produced the fatal injury.   See Parts 2a and 2c of the Instruction Manuals for classifying underlying cause-of-death data.

U.S. Census Regions

The Northeast (CT, ME, MA, NJ, NY, PA, RI, VT) includes the New England (CT, ME, MA, RI, VT) and Middle Atlantic (NJ, NY, PA) divisions.   New York City (Bronx, Kings, New York, Queens, and Richmond counties) is within the Middle Atlantic Division.   The South (AL, AR, DC, DE, FL, GA, KY, LA, MD, MS, NC, OK, SC, TN, TX, VA, WV) includes the South Atlantic (DC, DE, FL, GA, MD, NC, SC, VA, WV), East South Central (AL, KY, MS, TN), and West South Central (AR, LA, OK, TX) divisions.   The Midwest (IL, IN, IA, KS, MI, MN, MO, NE, ND, OH, SD, WI) includes the East North Central (IL, IN, MI, OH, WI) and West North Central (IA, KS, MN, MO, NE, ND, SD) divisions.   The West (AK, AZ, CA, CO, HI, ID, MT, NV, NM, OR, UT, WA, WY) includes the Mountain (AZ, CO, ID, MT, NV, NM, UT, WY) and Pacific (AK, CA, HI, OR, WA) divisions, however, NORMS does not include the non-contiguous states (AK and HI) in the Pacific Division.

Years of Potential Life Lost (YPLL)

YPLL are calculated using the method described by the Centers for Disease Control (CDC) (MMWR Surveill Summ 1986/35(2S)).  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.   Except for 2005 life-table data, which were obtained from http://www.cdc.gov/nchs/data/nvsr/nvsr56/nvsr56_10.pdf   (PDF only; 2.28MB; 121 pgs), life tables published annually by NCHS were used to determine race/sex life expectancies for the categories white/male, white/female, black/male, and black/female.   The overall U.S. population life expectancy is used for 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.

Note to Users

ICD Codes