The content, links, and pdfs are no longer maintained and might be outdated.
Mortality Data from the National Vital Statistics System
Mortality data from the National Vital Statistics System are a primary source of information for identifying and monitoring chronic diseases and other public health problems. This article describes the sources of mortality data, the distinction between provisional and final data, the roles of CDC's National Center for Health Statistics (NCHS) and the World Health Organization (WHO) in compiling these data, the methods used to tabulate and rank leading causes of death, the distinction between underlying and multiple causes of death, and the completeness and quality of mortality information from death certificates.
The vital statistics system (including births, deaths, and other reported vital events) is the principal standardized source of health-related data in the United States. Mortality statistics derived from information reported on death certificates are among the most widely used sources of health data at the national, state, and local levels. These data have several important strengths (1,2): 1) coverage is universal because state laws require death certificates for disposition of bodies and because the certificates are often needed for legal purposes, including estate settlement; 2) considerable uniformity in content and format is achieved among the states through federal-state cooperation in the design of the death certificate; and 3) stand- ardization in processing and data presentation is promoted through cooperation with states, professional societies, and WHO.
Mortality data from the vital statistics system are used to identify health problems and monitor health programs because these data are unique as a means for measuring and comparing mortality at the national, state, and local levels. Therefore, many state and national initiatives in disease prevention and health promotion are predicated on and evaluated with mortality data (3). SOURCES OF DATA
Mortality data from the National Vital Statistics System are cooperatively produced by NCHS and the state vital statistics offices. U.S. death registration is based on state law; death certificates are filed and maintained in state vital statistics offices. In addition, an increasing share of the data processing activities have shifted from the national to the state level througha collaborative arrangement. This arrangement and WHO recommendations have enabled implementation of procedures and practices for uniform collection, processing, and dissemination of mortality statistics. This approach ensures a high level of comparability in mortality statistics not only among the states but also between the United States and other countries.
The basic source of information about mortality is the death certificate. The U.S. Standard Certificate of Death, recommended for use by the states, is revised approximately once every 10 years with collaboration by states, NCHS, other federal agencies, and subject-matter experts (4). The current revision, effective for 1989, has been adopted with minor variations by the states. The death certificate is used for all deaths regardless of the decedent's age. Information on fetal deaths or spontaneous abortions is collected using a different form, the U.S. Standard Report of Fetal Death.
The information on the death certificate is provided by two groups of persons: 1) the certifying physician, medical examiner, or coroner and 2) the funeral director. The certifying physician, medical examiner, or coroner (5) certifies the causes of death. Instructions for completing these items are available in the Physicians' Handbook on Medical Certification of Death (6,7). The funeral director provides the demographic information, (e.g., age, race, and sex) and files the certificate with the state vital registration office. Instructions for completing these items are available in the Funeral Directors' Handbook on 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 filed annually in the United States. The more timely provisional mortality data are based in part on a systematic sample of death certificates.
Final mortality information is processed principally in state vital statistics offices. Information from the death certificate is coded from copies of the original certificates using uniform specifications developed under rigorous quality-control procedures by NCHS (9,10). In 1986, all states and the District of Columbia submitted to NCHS precoded demographic data on computer tapes for all deaths; i addition, 22 states submitted precoded medical data, and the remaining 28 states, New York City, and the District of Columbia submitted copies of the original certificates from which NCHS coded the medical data. These final mortality data are disseminated in the annual volumes of Vital Statistics of the United States, Volume II, Mortality (11), and on public-use computer tapes (12). They are summarized in Advance Report of Final Mortality Statistics (13). Final data are most recently available for 1986. The interval between close of a data (calendar) year and publication of data from the final mortality file is approximately 18-24 months.
Provisional mortality data are published 3-4 months after the death certificates are filed in the state vital statistics office and comprise 1) counts of the number of death certificates (based on the number of deaths) filed during the month in the state vital statistics offices and 2) a 10% systematic sample (called the Current Mortality Sample) of death certificates filed in the state offices and coded by NCHS. Estimates of the total numbers of deaths and the total death rate for the United States are available for October 1988; sample numbers of deaths 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 Statistics Report and in the Annual Summary of Births, Marriages, Divorces, and Deaths: United States (15). ROLE OF WHO
WHO plays a major role in collecting, classifying, and tabulating mortality statistics for the United States and other countries. The United States is a signatory to an international agreement coordinated by WHO that promotes standardization of mortality statistics through the International Classification of Diseases (ICD) (16). The ICD specifies the detailed title for each of more than 5000 categories to which medical entities and circumstances of death may be assigned. ICD chapters are organized principally by anatomical system (e.g., circulatory system, respiratory system); a few chapters are organized by disease (e.g., neoplasms, infectious and parasitic diseases, and mental disorders). The external causes of injuries and poisoning are covered in a supplementary chapter.
The ICD also provides recommendations for the broad categories used for tabulating and ranking mortality data, as well as standard definitions for such concepts as maternal mortality, underlying cause of death*, and fetal death. WHO also provides rules for selecting one underlying cause of death from among the many medical conditions that physicians may indicate contributed to the death. These rules are especially useful for guiding medical coders when ambiguous diagnoses or illogical or implausible sequences are recorded on the death certificate. WHO prescribes in the ICD how cause-of-death information should be collected and indicates how the death certificate should be completed. An expansion of the ICD, the International Classification of Diseases, Clinical Modification, is used to classify morbidity statistics in the United States (17).
The ICD has been revised approximately once each decade since the beginning of this century. The last revision, ICD-9, was implemented in 1979; however, the next revision--the 10th--is planned for implementation in 1993. Interim changes in the classification system have been made infrequently between major revisions; these have included the introduction of a special category for sudden infant death syndrome in 1973 (18) and for human immunodeficiency virus (HIV) infection in 1987 (19). These changes are documented in the annual volumes 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 one underlying cause for each death. However, underlying-cause data can be augmented with additional information on the other conditions that the medical certifier reported as contributing to death (20). Because several chronic conditions are often reported, multiple-cause data may be 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 mortality data; several of these lists combine detailed cause-of-death categories into broader groups (9). Those most commonly used for presentation of mortality data are the list of 72 selected causes of death for general mortality and a list of 61 categories for infant deaths (9). The categories in these lists are exhaustive and, when summed, account for all causes of death.
The ranking of causes of death is important to differentiate the magnitude of various health problems. A standard approach that facilitates uniform presentation of mortality data has been developed and adopted by the states and NCHS. The ranking of leading causes of death is based solely on the list of 72 selected causes of death for persons of all ages and on the list of 61 causes for infant deaths (9). Effective with 1987 final mortality data, HIV infection will become a rankable cause of death. MEASUREMENT OF MORTALITY
NCHS and the states use measures of mortality--such as crude death rates, age-specific death rates, age-standardized death rates, and life table indices--that have been developed and standardized by practices of WHO, health statisticians, and public health agencies. Other measures, such as potential years of life lost and standardized mortality ratios, are used principally in detailed analyses of mortality data. Standardization of mortality rates is generally done using the direct method, with the 1940 U.S. population distribution as the standard. This procedure is widely used by NCHS, state vital statistics offices, and the research community. Use of the 1940 population as a standard has the advantage of historic continuity; however, other standard populations (e.g., 1970 or 1980) are sometimes used. Although age- standardized death rates based on alternative standards are usually similar, they cannot be directly compared (22). COMPLETENESS AND QUALITY OF INFORMATION
Reliance on mortality data as a primary basis for public health measurement requires understanding the completeness and validity of information reported on death certificates. All states have adopted laws that require the registration of deaths and the reporting of fetal deaths. More than 99% of the deaths in the United States are thought to be registered. In contrast, fetal deaths at greater than or equal to28 weeks' gestation may be reported less completely than other deaths.
Quality assurance of NCHS mortality data is promoted during each phase of data collection and data processing. During data collection, states are encouraged to scrutinize records with questionable entries, using guidelines specified in instruction manuals for demographic (23) and medical (24) items. During processing, quality is maintained through:
The validity of the medical certification of cause of death reflects both the ability of the medical certifier to make the proper diagnosis and the correctness with which he/she records this information on the death certificate. Efforts used by NCHS and the states to promote accurate reporting include dissemination of video and audio cassettes and handbooks that describe proper completion of the death certificates. NCHS is also encouraging states to evaluate death certificates for potential errors as an integral aspect of their vital statistics programs using a manual developed by NCHS (24). The current version of the U.S. Standard Certificate of Death includes examples of properly completed cause-of-death certifications. Efforts are also being directed at educating physicians during medical school and residency and through continuing education about proper completion of death certificates.
One index of the quality of reporting causes of death is the proportion of death certificates coded to the ICD-9, Chapter XVI, "Symptoms, Signs, and Ill-Defined Conditions" (rubrics 780-799). This proportion generally indicates the care and consideration given to the certification by the medical certifier and may be used as an approximate measure of the specificity of the medical diagnoses made by the 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-death information is important for interpreting mortality data (27). Reported by: Mortality Statistics Br, Div of Vital Statistics, National Center for Health Statistics, CDC.
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.**Questions or messages regarding errors in formatting should be addressed to email@example.com.
Page converted: 08/05/98
This page last reviewed 5/2/01