Mortality Data from the National Vital Statistics System (2024)

Mortality data from the National Vital Statistics System are a primarysource of information for identifying and monitoring chronic diseasesand other public health problems. This article describes the sources ofmortality data, the distinction between provisional and final data, theroles of CDC's National Center for Health Statistics (NCHS) and theWorld Health Organization (WHO) in compiling these data, the methodsused to tabulate and rank leading causes of death, the distinctionbetween underlying and multiple causes of death, and the completenessand quality of mortality information from death certificates.

The vital statistics system (including births, deaths, and otherreported vital events) is the principal standardized source ofhealth-related data in the United States. Mortality statistics derivedfrom information reported on death certificates are among the mostwidely used sources of health data at the national, state, and locallevels. These data have several important strengths (1,2): 1) coverageis universal because state laws require death certificates fordisposition of bodies and because the certificates are often needed forlegal purposes, including estate settlement; 2) considerable uniformityin content and format is achieved among the states throughfederal-state cooperation in the design of the death certificate; and3) stand- ardization in processing and data presentation is promotedthrough cooperation with states, professional societies, and WHO.

Mortality data from the vital statistics system are used to identifyhealth problems and monitor health programs because these data areunique as a means for measuring and comparing mortality at thenational, state, and local levels. Therefore, many state and nationalinitiatives in disease prevention and health promotion are predicatedon and evaluated with mortality data (3).SOURCES OF DATA

Mortality data from the National Vital Statistics System arecooperatively produced by NCHS and the state vital statistics offices.U.S. death registration is based on state law; death certificates arefiled and maintained in state vital statistics offices. In addition, anincreasing share of the data processing activities have shifted fromthe national to the state level througha collaborative arrangement.This arrangement and WHO recommendations have enabled implementation ofprocedures and practices for uniform collection, processing, anddissemination of mortality statistics. This approach ensures a highlevel of comparability in mortality statistics not only among thestates but also between the United States and other countries.

The basic source of information about mortality is the deathcertificate. The U.S. Standard Certificate of Death, recommended foruse by the states, is revised approximately once every 10 years withcollaboration by states, NCHS, other federal agencies, andsubject-matter experts (4). The current revision, effective for 1989,has been adopted with minor variations by the states. The deathcertificate is used for all deaths regardless of the decedent's age.Information on fetal deaths or spontaneous abortions is collected usinga different form, the U.S. Standard Report of Fetal Death.

The information on the death certificate is provided by two groups ofpersons: 1) the certifying physician, medical examiner, or coroner and2) the funeral director. The certifying physician, medical examiner, orcoroner (5) certifies the causes of death. Instructions for completingthese items are available in the Physicians' Handbook on MedicalCertification of Death (6,7). The funeral director provides thedemographic information, (e.g., age, race, and sex) and files thecertificate with the state vital registration office. Instructions forcompleting these items are available in the Funeral Directors' Handbookon Death Registration and Fetal Death Reporting (8).FINAL AND PROVISIONAL MORTALITY DATA

A distinction is made between final and provisional mortality data.Final data are based on processing all 2 million death records filedannually in the United States. The more timely provisional mortalitydata are based in part on a systematic sample of death certificates.

Final mortality information is processed principally in state vitalstatistics offices. Information from the death certificate is codedfrom copies of the original certificates using uniform specificationsdeveloped under rigorous quality-control procedures by NCHS (9,10). In1986, all states and the District of Columbia submitted to NCHSprecoded demographic data on computer tapes for all deaths; iaddition, 22 states submitted precoded medical data, and the remaining28 states, New York City, and the District of Columbia submitted copiesof the original certificates from which NCHS coded the medical data. These final mortality data are disseminated in the annual volumes ofVital Statistics of the United States, Volume II, Mortality (11), andon public-use computer tapes (12). They are summarized in AdvanceReport of Final Mortality Statistics (13). Final data are most recentlyavailable for 1986. The interval between close of a data (calendar)year and publication of data from the final mortality file isapproximately 18-24 months.

Provisional mortality data are published 3-4 months after the deathcertificates are filed in the state vital statistics office andcomprise 1) counts of the number of death certificates (based on thenumber of deaths) filed during the month in the state vital statisticsoffices and 2) a 10% systematic sample (called the Current MortalitySample) of death certificates filed in the state offices and coded byNCHS. Estimates of the total numbers of deaths and the total death ratefor the United States are available for October 1988; sample numbers ofdeaths and estimated death rates by age, race, sex, and cause of death,based on the Current Mortality Sample, are available for September 1988(14). Provisional data are published in the Monthly Vital StatisticsReport and in the Annual Summary of Births, Marriages, Divorces, andDeaths: United States (15).ROLE OF WHO

WHO plays a major role in collecting, classifying, and tabulatingmortality statistics for the United States and other countries. TheUnited States is a signatory to an international agreement coordinatedby WHO that promotes standardization of mortality statistics throughthe International Classification of Diseases (ICD) (16). The ICDspecifies the detailed title for each of more than 5000 categories towhich medical entities and circ*mstances of death may be assigned. ICDchapters are organized principally by anatomical system (e.g.,circulatory system, respiratory system); a few chapters are organizedby disease (e.g., neoplasms, infectious and parasitic diseases, andmental disorders). The external causes of injuries and poisoning arecovered in a supplementary chapter.

The ICD also provides recommendations for the broad categories usedfor tabulating and ranking mortality data, as well as standarddefinitions for such concepts as maternal mortality, underlying causeof death*, and fetal death. WHO also provides rules for selecting oneunderlying cause of death from among the many medical conditions thatphysicians may indicate contributed to the death. These rules areespecially useful for guiding medical coders when ambiguous diagnosesor illogical or implausible sequences are recorded on the deathcertificate. WHO prescribes in the ICD how cause-of-death informationshould be collected and indicates how the death certificate should becompleted. An expansion of the ICD, the International Classification ofDiseases, Clinical Modification, is used to classify morbiditystatistics in the United States (17).

The ICD has been revised approximately once each decade since thebeginning of this century. The last revision, ICD-9, was implemented in1979; however, the next revision--the 10th--is planned forimplementation in 1993. Interim changes in the classification systemhave been made infrequently between major revisions; these haveincluded the introduction of a special category for sudden infant deathsyndrome in 1973 (18) and for human immunodeficiency virus (HIV)infection in 1987 (19). These changes are documented in the annualvolumes of Vital Statistics of the United States, Volume II, Mortality.UNDERLYING AND MULTIPLE CAUSES OF DEATH

Cause-of-death data are traditionally presented in terms of oneunderlying cause for each death. However, underlying-cause data can beaugmented with additional information on the other conditions that themedical certifier reported as contributing to death (20). Becauseseveral chronic conditions are often reported, multiple-cause data maybe important in chronic disease surveillance (21). The NCHS multiple-cause data base is produced annually on public-use tapes (12).TABULATING AND RANKING CAUSE-OF-DEATH INFORMATION

NCHS uses lists of cause-of-death categories to tabulate mortalitydata; several of these lists combine detailed cause-of-death categoriesinto broader groups (9). Those most commonly used for presentation ofmortality data are the list of 72 selected causes of death for generalmortality and a list of 61 categories for infant deaths (9). Thecategories in these lists are exhaustive and, when summed, account forall causes of death.

The ranking of causes of death is important to differentiate themagnitude of various health problems. A standard approach thatfacilitates uniform presentation of mortality data has been developedand adopted by the states and NCHS. The ranking of leading causes ofdeath is based solely on the list of 72 selected causes of death forpersons of all ages and on the list of 61 causes for infant deaths (9).Effective with 1987 final mortality data, HIV infection will become arankable cause of death.MEASUREMENT OF MORTALITY

NCHS and the states use measures of mortality--such as crude deathrates, age-specific death rates, age-standardized death rates, and lifetable indices--that have been developed and standardized by practicesof WHO, health statisticians, and public health agencies. Othermeasures, such as potential years of life lost and standardizedmortality ratios, are used principally in detailed analyses ofmortality data. Standardization of mortality rates is generally doneusing the direct method, with the 1940 U.S. population distribution asthe standard. This procedure is widely used by NCHS, state vitalstatistics offices, and the research community. Use of the 1940population as a standard has the advantage of historic continuity;however, other standard populations (e.g., 1970 or 1980) are sometimesused. Although age- standardized death rates based on alternativestandards are usually similar, they cannot be directly compared (22).COMPLETENESS AND QUALITY OF INFORMATION

Reliance on mortality data as a primary basis for public healthmeasurement requires understanding the completeness and validity ofinformation reported on death certificates. All states have adoptedlaws that require the registration of deaths and the reporting of fetaldeaths. More than 99% of the deaths in the United States are thought tobe registered. In contrast, fetal deaths at greater than or equal to28weeks' gestation may be reported less completely than other deaths.

Quality assurance of NCHS mortality data is promoted during each phaseof data collection and data processing. During data collection, statesare encouraged to scrutinize records with questionable entries, usingguidelines specified in instruction manuals for demographic (23) andmedical (24) items. During processing, quality is maintained through:

1. follow-up to the states to verify those records of deaths, including reported diseases of public health concern (e.g., cholera) (25), 2) computer edits to ensure consistency between demographic characteristics--such as age and sex--and reported causes of death (26), and 3) independent coding and verification by NCHS of a monthly sample of state records. For 1986, the estimated average error rates for coding the demographic and medical items were 0.3% and 3%-4%, respectively (9).

The validity of the medical certification of cause of death reflectsboth the ability of the medical certifier to make the proper diagnosisand the correctness with which he/she records this information on thedeath certificate. Efforts used by NCHS and the states to promoteaccurate reporting include dissemination of video and audio cassettesand handbooks that describe proper completion of the deathcertificates. NCHS is also encouraging states to evaluate deathcertificates for potential errors as an integral aspect of their vitalstatistics programs using a manual developed by NCHS (24). The currentversion of the U.S. Standard Certificate of Death includes examples ofproperly completed cause-of-death certifications. Efforts are alsobeing directed at educating physicians during medical school andresidency and through continuing education about proper completion ofdeath certificates.

One index of the quality of reporting causes of death is theproportion of death certificates coded to the ICD-9, Chapter XVI,"Symptoms, Signs, and Ill-Defined Conditions" (rubrics 780-799). Thisproportion generally indicates the care and consideration given to thecertification by the medical certifier and may be used as anapproximate measure of the specificity of the medical diagnoses made bythe certifier in various areas. In 1986, 1.5% of all reported U.S.deaths were assigned to the rubric for ill-defined or unknown causes.However, this percentage varied among the states, from 0.3% to 4.0%.Awareness of geographic differences in the quality of cause- of-deathinformation is important for interpreting mortality data (27).Reported by: Mortality Statistics Br, Div of Vital Statistics, NationalCenter for Health Statistics, CDC.

References

1. Zemach R. What the vital statistics system can and cannot do. Am J Public Health 1984;74:756-8. 2.Glasser JH. The quality and utility of death certificate data. Am J Public Health 1981; 71:231-3. 3.National Center for Health Statistics. Health, United States, 1986 and Prevention Profile. Hyattsville, Maryland: US Department of Health and Human Services, Public Health Service, 1986; DHHS publication no. (PHS)87-1232. 4.Freedman MA, Gay GA, Brockert JE, Potrzebowski PW, Rothwell CJ. The 1989 revisions of the US standard certificates of live birth and death and the US standard report of fetal death. Am J Public Health 1988;78:168-72. 5.CDC. Death investigations--United States, 1987. MMWR 1989;38:1-4. 6.National Center for Health Statistics. Physicians' handbook on medical certification of death. Hyattsville, Maryland: US Department of Health and Human Services, Public Health Service, 1987; DHHS publication no. (PHS)87-1108. 7.National Center for Health Statistics. Medical examiners' and coroners' handbook on death registration and fetal death reporting. Hyattsville, Maryland: US Department of Health and Human Services, Public Health Service, 1987: DHHS publication no. (PHS)87-1110. 8.National Center for Health Statistics. Funeral directors' handbook on death registration and fetal death reporting. Hyattsville, Maryland: US Department of Healt and Human Services, Public Health Service, 1987; DHHS publication no. (PHS)87-1109. 9.National Center for Health Statistics. Vital statistics of the United States, 1986. Vol II: Mortality, part A. Technical appendix. Hyattsville, Maryland: US Department of Health and Human Services, Public Health Service, 1988; DHHS publication no. (PHS)88-1122. 10.National Center for Health Statistics, Harris KW. A methodological study of quality control procedures for mortality medical coding. Hyattsville, Maryland: US Department of Health, Education, and Welfare, Public Health Service, 1980; DHEW publication no. (PHS)80-1355. (Vital and health statistics; series 2--Data evaluation and methods research; no. 81). 11.National Center for Health Statistics. Vital statistics of the United States, 1986. Vol II: Mortality, part A. Hyattsville, Maryland: US Department of Health and Human Services, Public Health Service, 1988; DHHS publication no. (PHS)88-1122. 12.National Center for Health Statistics. Catalog of public use data tapes. Hyattsville, Maryland: US Department of Health and Human Services, Public Health Service, 1988; DHHS publication no. (PHS)88-1213. 13.National Center for Health Statistics. Advance report of final mortality statistics, 1986. Hyattsville, Maryland: US Department of Health and Human Services, Public Health Service, 1988; DHHS publication no. (PHS)88-1120. (Monthly vital statistics report; vol 37, no. 6, suppl). 14.National Center for Health Statistics. Births, marriages, divorces, and deaths for October 1988. Hyattsville, Maryland: US Department of Health and Human Services, Public Health Service, 1989; DHHS publication no. (PHS)89-1120. (Monthly vital statistics report; vol 37, no. 10). 15.National Center for Health Statistics. Annual summary of births, marriages, divorces, and deaths: United States, 1987. Hyattsville, Maryland: US Department of Health and Human Services, Public Health Service, 1988; DHHS publication no. (PHS)88-1120. (Monthly vital statistics report; vol 36, no. 13). 16.World Health Organization. Manual of the international statistical classification of diseases, injuries, and causes of death--based on the recommendations of the Ninth Revision Conference, 1975. Geneva: World Health Organization, 1977. 17.Commission on Professional and Hospital Activities. The international classification of diseases, 9th revision, clinical modification. Ann Arbor, Michigan: US Department of Health and Human Services, Public Health Service, Health Care Financing Administration, 1981; DHHS publication no. (PHS)80-1260. 18.National Center for Health Statistics. Vital statistics of the United States, 1978. Vol II: Mortality, part A. Technical appendix. Hyattsville, Maryland: US Department of Health and Human Services, Public Health Service, 1982; DHHS publication no. (PHS)83-1101. 19.CDC. Human immunodeficiency virus (HIV) infection codes, official authorized addendum, ICD-9-CM (revision no. 1): effective January 1, 1988. MMWR 1987;36(suppl S-7). 20.National Center for Health Statistics, Chamblee RF, Evans MC. TRANSAX, the NCHS system for producing multiple cause-of-death data statistics, 1968-78. Hyattsville, Maryland: US Department of Health and Human Services, Public Health Service, 1986; DHHS publication no. (PHS)86-1322. (Vital and health statistics, series 1, no. 20). 21.Israel RA, Rosenberg HM, Curtin LR. Analytical potential for multiple cause-of-death data. Am J Epidemiol 1986;124:161-79. 22.Curtin LR, Maurer JD, Rosenberg HM. On the selection of a standard population for computing age-adjusted death rates. In: 1980 Proceedings of the Social Statistics Section, American Statistical Association. Washington, DC: American Statistical Association, 1980: 218-23. 23.National Center for Health Statistics. Instruction manual: part 18--Guidelines for implementing field and query programs for registration of births and deaths, 1983. Hyattsville, Maryland: US Department of Health and Human Services, Public Health Service, 1982. 24.National Center for Health Statistics. Instruction manual: part 20--Cause of death querying, 1985. Hyattsville, Maryland: US Department of Health and Human Services, Public Health Service, 1985. 25.National Center for Health Statistics. Instruction manual: part 2a--Instructions for classifying the underlying cause of death, 1984. Hyattsville, Maryland: US Department of Health and Human Services, Public Health Service, 1983. 26.National Center for Health Statistics. Instruction manual: part 11--Vital statistics computer edits for mortality data, effective 1979. Hyattsville, Maryland: US Department of Health and Human Services, Public Health Service, 1979. 27.Sorlie TD, Gold EB. The effect of physician terminology preference on coronary heart disease mortality: an artifact uncovered by the 9th revision, ICD. Am J Public Health 1987;77:148-52. *Defined as "underlying disease or injury which initiates the train of morbid events leading directly to death, or the circ*mstances of the accident or violence which produced the fatal injury" (16).

Disclaimer All MMWR HTML documents published before January 1993 are electronic conversions from ASCII text into HTML. This conversion may have resulted in character translation or format errors in the HTML version. Users should not rely on this HTML document, but are referred to the original MMWR paper copy for the official text, figures, and tables. An original paper copy of this issue can be obtained from the Superintendent of Documents, U.S. Government Printing Office (GPO), Washington, DC 20402-9371; telephone: (202) 512-1800. Contact GPO for current prices.

Page converted: 08/05/98