Appendix

Technical Notes

National Health Interview Survey

Sampling Frame: The National Health Interview Survey (NHIS) sampling frame covers the civilian noninstitutionalized population in all 50 states and the District of Columbia.

The survey is administered by in-person interview in the sample household. During 1979--1996, a knowledgeable adult provided information on their own health and that of others in the household. A one-sixth subsample (approximately 20,000 of 120,000 persons) that is representative of the whole sample was asked about 17 respiratory conditions, including asthma. Asthma prevalence was determined by an affirmative response to the question "During the past 12 months has anyone in the family had asthma?" As a result, during 1980--1996, prevalence estimates were referred to as 12-month asthma prevalence. The survey methodology changed in 1997. One of the changes eliminated proxy respondents for adult medical information. Since 1997, information collected on the whole family, including sociodemographic data and use of health-care services information, is collected from the primary household respondent. However, medical information (including asthma-related questions) is collected from a single sampled household adult and by adult proxy for a sampled child in households with children. In addition, the question wording was changed to "Has a doctor or other health professional ever told you that you had asthma?" which measured lifetime asthma prevalence. Asthma attack prevalence was measured by responses to the question, "During the past 12 months, have you had an episode of asthma or an asthma attack?"

Beginning in 2001, all persons answering yes to the question, "Has a doctor or other health professional ever told you that you had asthma?" were then asked if they still had asthma. Analysis of the responses to these questions indicates that approximately 5% of persons who stated that they no longer had asthma also reported that they had had an asthma attack during the preceding 12 months. In estimating the number of asthma attacks and the attack prevalence percent with the U.S. population as the denominator, all persons who stated they had an attack during the preceding 12 months were included, regardless of whether they stated they still had asthma. This is consistent with procedures used in other National Center for Health Statistics (NCHS) publications to produce national asthma attack estimates from NHIS. When estimating the percent of asthma attacks per person with asthma, persons with attacks who stated they no longer had asthma were excluded. This exclusion maintains consistency in determining attack rates only in persons who self-report that they still have asthma and is also consistent with the Council of State and Territorial Epidemiologists (CSTE) case definition for asthma prevalence.

Race categorization and current asthma and asthma attack statistics. In the presentation of NHIS race data in the current asthma and asthma attack tables (Tables 1, 2, 3, 4, 5, 6, 7, 8, 9, 10), data before 1996 is presented using the NHIS race variable that coded race into three categories---white, black, and other. Starting with 1997 data, estimates for race categories are presented using both the new OMB classifications and a race classification that follows the 1977 standards. For 1997 and 1998, 1977-comparable categories were created from the "main race" variable found on the NHIS file. After 1998, the data files already included a 1977-comparable variable. The new OMB race classifications were grouped into categories of single race white, single race black, and all other (which includes respondents indicating multiple races [approximately 1.5% of respondents]).

Poverty Status: Because income information is missing for a large proportion of respondents (23%--26% of sample adults and children, depending on year), estimates should be interpreted with caution.

National Ambulatory Medical Care Survey and National Hospital Ambulatory Medical Care Survey

Sampling Frame: The National Ambulatory Medical Care Survey (NAMCS) samples office-based physicians, primarily in patient care, excluding radiologists, anesthesiologists, and pathologists, from the master files of the American Medical Association and the American Osteopathic Association. Approximately 2,000 randomly selected physicians participate each year by reporting data on a sample of approximately 30,000 patient encounters. The clinical modification to the International Classification of Diseases Ninth revision (ICD9-CM) code 493 has been used to identify asthma office visits since 1979 (http://www.cdc.gov/nchs/data/ahcd/NAMCS-DATA-ITEMS.pdf). Before 1989, Alaska and Hawaii were not included in NAMCS. Since 1989, all 50 states, including the District of Columbia, were included in NAMCS.

The National Hospital Ambulatory Medical Care Survey (NHAMCS) sampling frame is nonfederal, general, and short-stay hospitals (including children's hospitals) in the 50 states and the District of Columbia that have an emergency department or an outpatient department (http://www.cdc.gov/nchs/data/ahcd/NHAMCS-DATA-ITEMS.pdf). This survey has been administered annually since 1992. Approximately 400 hospitals report data on a sample totaling approximately 30,000 outpatient visits and 30,000 emergency department visits each year. The ICD9-CM code 493 has been used to identify asthma hospital outpatient and emergency department visits since 1979. Estimates with asthma coded as the first listed diagnosis are presented for outpatient visits and for emergency department visits. Because the sampling frame did not control for region in 1980, regional differences for that year should be interpreted with caution.

Race: Before 1999, only one race could be reported per patient record. Beginning with 1999, data reflect standards mandated in 1997 by the (OMB) to promote comparability of data among federal data sources. One of the mandates requires the option to report more than one race per patient, where appropriate. In 1999, the race item on the survey instrument was modified to include a "multiple race" category. In these surveys, race is collected from medical record rather than by personal reporting; as a result, "multiple race" is underreported when compared with census data. Because of small numbers, records with a single category of "Asian" or "Native Hawaiian/other Pacific Islander" or "American Indian" were combined with "multiple race" records and are presented in the table category labeled "other races."

Population figures and rate calculation: The denominators used in calculating physician office visit, hospital outpatient, and emergency department visit rates for sex, race, ethnicity, age, and region are from special tabulations that were developed by the U.S. Bureau of the Census to estimate the civilian, noninstitutionalized population covered in these surveys. The denominator counts used in this report were provided by NCHS. Starting in 2001, both numerator and denominator estimates reflect the transition to multirace reporting. In 1999 and 2000, the numerator counts included "multiple race" in the "other races" category, and the denominator data did not. Before 1999, neither numerator nor denominator allowed for the "multiple race" option. Because "multiple race" is underreported in these surveys, visit counts and rates for the "other race" category (which includes the multiple race group) presented for 2001--2003 are considered underestimates. Because the transition to the 1997 OMB standards was not clearly demarcated as it was in the other data sources, and because the size of the "other races" group is small, the tables for office visits, outpatient visits, and emergency department visits do not distinguish between the years before and after the transition to "multiple race" coding.

National Hospital Discharge Survey: Hospital inpatient visits

Sampling Frame: The National Hospital Discharge Survey (NHDS) consists of short-stay (average length of stay: <30 days for all patients), general, and children's general hospitals. The survey includes patients discharged from approximately 500 nonfederal general and short-stay specialty hospitals excluding federal, military, and the Department of Veterans Affairs hospitals in the 50 States and the District of Columbia. These hospitals must have six or more beds for patient use. Sampled hospitals produce approximately 275,000 discharge records each year. Before 1988, hospitals with an average length of stay >30 days were excluded from the survey, regardless of specialty. A hospitalization for asthma is defined as a first-listed discharge diagnosis of asthma (ICD9-CM code 493). Before 1988, a two-stage sample design was used. In 1988, NHDS was redesigned using a three-stage design.

Standard Errors: Before 1988, standard error estimates from NHDS were produced using a computerized routine based on a rigorously unbiased algebraic estimator of the variance. Using approximations, standard errors were prepared that were applicable to a wide variety of statistics. To calculate standard errors for 1980 and 1985 estimates, survey year-specific approximate standard errors of estimated numbers of discharges for first-listed diagnoses by selected patient and hospital characteristics were used. For simplicity, linear interpolation was used to obtain specific relative standard errors (RSE) for each estimate, although the RSE curves are not strictly linear functions.

Underreporting Race: Underreporting of race in NHDS is attributed to the number of hospitals in NHDS samples that do not report race for any of their patients and the number of patients identified as Hispanic with missing race data. Race information is missing for a large proportion of hospital discharge records (5%--30% each year). Hospitals that did not report race probably had a higher proportion of white discharges than hospitals that reported race. Because most discharges with missing race data represent white persons, the differences between racial groups probably are smaller than these estimates indicate.

Race Categorization: Starting in 2000, a "multiple race" category was added to the NHDS data set. The "multiple race" category consists of respondents who identified with two or more races. Because of small numbers, records with a single category of American Indian or Alaska Native, Asian, Native Hawaiian or Other Pacific Islander, Some Other Race and "multiple race" were grouped together and are presented in the table category labeled "Other Races".

Population figures for rate calculations: Population-based rates were computed using July 1 estimates of the civilian population of the United States. Starting in 2000, both numerator and denominator estimates include "multiple race" in the "other races" category. The Census population tables contained race categories for three primary racial groups: white, black, and all other races. The denominator for the at-risk--based rates was derived from NHIS, which covers the civilian, noninstitutionalized population rather than the civilian population covered by NHDS. The civilian, noninstitutionalized population represents 99% of the civilian population. Consequently, the at-risk--based hospitalization rates are slight overestimates of the actual rates.

Exclusions: Data for newborns are excluded from this report.

National Vital Statistics System

Race Categorization: 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. In 1997 OMB standards were introduced to collect data on ethnicity and five single race categories (white, black, American Indian, and various Asian and Pacific Islander categories) and a multiple race category. However, as of 2003, this format had been introduced in only a few states.

Hispanic ethnicity: Hispanic mortality data were collected from 17 reporting states and published for the first time in 1984. The number of reporting states increased over time as did the quality of the data. Data year 1997 was the first year that mortality data for the Hispanic population were available for all 50 states and the District of Columbia.

Death rates for Hispanic, American Indian or Alaska Native, and Asian or Pacific Islander should be interpreted with caution because of inconsistencies in reporting Hispanic origin or race on the death certificate as compared with race or Hispanic origin on the censuses, surveys, and birth certificates. Studies have documented underreporting on the death certificates of these decedents and undercounts of these groups on the censuses.

Population figures and rate calculation: The denominators used in calculating population-based mortality rates for sex, race, ethnicity, age, and region are prepared by the U.S. Bureau of the Census. Death rates for the United States and individual states are based on the total resident populations for the respective years. These populations exclude members of the Armed Forces stationed abroad but include the members of the Armed Forces stationed domestically, unless noted. The denominator for the at-risk--based rates was derived from NHIS, which covers the civilian, noninstitutionalized population rather than the resident population covered by the National Vital Statistics System. The civilian, noninstitutionalized population represents approximately 98% of the resident population. Consequently, the at-risk--based death rates are slight overestimates of the actual rates.

Population data through the 1990s also were 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 five single-race categories (American Indian or Alaska Native, Asian, black, Hawaiian or Other Pacific Islander, and white) and permit the reporting of more than one race. As a result, 31 racial groups were included on the 2000 census (five 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. These Census population tables contained race categories for three primary racial groups: white, black, and all other races.

Use of trade names and commercial sources is for identification only and does not imply endorsement by the U.S. Department of Health and Human Services. References to non-CDC sites on the Internet are provided as a service to MMWR readers and do not constitute or imply endorsement of these organizations or their programs by CDC or the U.S. Department of Health and Human Services. CDC is not responsible for the content of pages found at these sites. URL addresses listed in MMWR were current as of the date of publication.

Disclaimer   All MMWR HTML versions of articles 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 electronic PDF version and/or 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.

Date last reviewed: 10/5/2007