The Right Start Online >
Definitions, Data Sources, and Reporting Issues for Cities
| Right Start Home |
 |
Introduction Prior to 2003, all states used the 1989 Revision of the U.S. Standard Certificate of Live Birth. However, beginning in 2003, states began to adopt the 2003 Revision of the U.S. Standard Certificate of Live Birth.1 In 2003, Pennsylvania and Washington adopted the 2003 Certificate, and in January 2004, Idaho, Kentucky, New York (excluding New York City), South Carolina, and Tennessee adopted the revised certificate. Two additional states, Florida and New Hampshire, implemented the revised birth certificate in 2004, but after January 1. Data for educational attainment, prenatal care, and tobacco use, although collected on both the revised and unrevised certificates, are not considered comparable between revisions, The discussion of individual measures in the next section provides details on the lack of comparability. More specifically, for cities in the states that have adopted the revised certificate, it is not appropriate to construct trends for maternal education, prenatal care, and tobacco use if the trend period includes the year in which the state adopted the revised certificate. Similarly, it is not appropriate to compare data on maternal education, prenatal care, and tobacco use for cities in states using the 1989 certificate with data on these three measures for cities in states using the 2003 certificate. All percentages in the city tables are rounded to the nearest tenth of a percent. Thus, 9.46 percent is rounded to 9.5 percent in the tables. Definitions and data sources Sources: 1990-2004 Natality Data Set CD Series 21, numbers 2-9, 11-12, 14-16 (SETS versions), and 16H and 17Ha (ASCII version), National Center for Health Statistics. Late or no prenatal care (percent of total births to mothers receiving late or no prenatal care) is the percentage of births that occurred to mothers who reported receiving prenatal care only in the third trimester of their pregnancy, or reported receiving no prenatal care. Birth certificates that did not report information about prenatal care were not included in this calculation. Beginning in 2003, the adoption of the revised birth certificate in several states produced “substantive changes in both question wording and the sources for [prenatal care] information [that] have resulted in data that are not comparable” with data for previous years.2 Prior to the revision there were other reporting issues for a few states in some years. See the next section for more details. See the next section for more details. Sources: 1990-2004 Natality Data Set CD Series 21, numbers 2-9, 11-12, 14-16 (SETS versions), and 16H and 17Ha (ASCII version), National Center for Health Statistics. Low-birthweight births (percent low-birthweight births) is the percentage of live births weighing less than 2,500 grams (5.5 pounds). Births of unknown weight were not included in these calculations. Sources: 1990-2004 Natality Data Set CD Series 21, numbers 2-9, 11-12, 14-16 (SETS versions), and 16H and 17Ha (ASCII version), National Center for Health Statistics. Low maternal education (percent of total births to mothers with less than 12 years of education) is the percentage of women who had completed fewer than 12 years of education at the time of the birth. Birth certificates on which maternal education was not reported were not included in this calculation. Data on maternal education were not available for Seattle in the early 1990s. Beginning in 2003, the adoption of the revised birth certificate in several states produced substantive changes in the wording of the questions on maternal education that have resulted in data that are not comparable with data for previous years. See the next section for more details. Sources: 1990-2004 Natality Data Set CD Series 21, numbers 2-9, 11-12, 14-16 (SETS versions), and 16H and 17Ha (ASCII version), National Center for Health Statistics. Number of births by race and ethnicity of mother contains information about race/Hispanic origin of the mother. On birth certificates, as on most federal data collection forms, the question regarding whether a person is Hispanic is separate from the question asking whether a person is white, black, Asian or Pacific Islander, or American Indian. Thus, people are asked to select a racial group and to indicate whether they are of Hispanic origin. A birth to a woman who reported that she was Hispanic and white would usually be included in figures for both of these groups. In order to create mutually exclusive categories, Hispanics were removed from the black and white racial categories in the tabulations presented here. This allows more meaningful comparisons between minorities and the group people typically think of when we say “white.” It should also be noted that these figures represent the race of the mother, not the race of the child. This is important because increasing numbers of children are born to parents of different races. Births for which Hispanic origin was not reported are included in the “Other” category. In 1990, data on Hispanic origin were not collected for Oklahoma City and Tulsa, OK. Data on race/ethnicity were less reliable for Hartford, CT, in the early 1990s than in the late 1990s. In El Paso, TX, non-Hispanic white births may have been underreported in 1997 and earlier years. Starting in 2003, multiple race reporting was allowed by cities in several states. See the next section for more details on each city. Sources: 1990-2004 Natality Data Set CD Series 21, numbers 2-9, 11-12, 14-16 (SETS versions), and 16H and 17Ha (ASCII version), National Center for Health Statistics. Preterm births (percent preterm births) is the percentage of babies born with a gestational age of less than 37 completed weeks. Birth certificates that did not report gestational age were not included in this tabulation. Sources: 1990-2004 Natality Data Set CD Series 21, numbers 2-9, 11-12, 14-16 (SETS versions), and 16H and 17Ha (ASCII version), National Center for Health Statistics. Repeat teen births (percent of teen births to women who were already mothers) is the percentage of births that were second or higher order births to mothers who were under the age of 20 at the time of the birth. Birth certificates that did not contain information on birth order were not included in this calculation. Sources: 1990-2004 Natality Data Set CD Series 21, numbers 2-9, 11-12, 14-16 (SETS versions), and 16H and 17Ha (ASCII version), National Center for Health Statistics. Smoking during pregnancy (percent of total births to mothers who smoked during pregnancy) is the percentage of women who smoked during pregnancy. Data for smoking were not collected in NCHS’ standard format in California. Birth certificates on which information on smoking during pregnancy was not reported were not included in this calculation. Beginning in 2003, the adoption of the revised birth certificate in several states produced substantive changes in the wording of the questions on tobacco use that have resulted in data that are not comparable with data for previous years. See the next section for more details. Sources: 1990-2004 Natality Data Set CD Series 21, numbers 2-9, 11-12, 14-16 (SETS versions), and 16H and 17Ha (ASCII version), National Center for Health Statistics. Teen births (percent of total births to teens) was calculated by dividing the number of births to women under age 20 by the total number of births to women of all ages. It should be noted that this is not the same as a teen birth rate, which measures the risk that a teenager will give birth. The percentage of total births to women under age 20 is affected by the fertility of women over age 20, as well as the fertility of teenagers. Sources: 1990-2004 Natality Data Set CD Series 21, numbers 2-9, 11-12, 14-16 (SETS versions), and 16H and 17Ha (ASCII version), National Center for Health Statistics. The 50-City Average is the unweighted mean of the data for each of the 50 largest cities in 1997 (based on Census Bureau population estimates).3 In other words, this figure is the average of the figures for each of the 50 cities without taking into consideration the size of each city. The average for each year is calculated using only those cities with reported and reliable data. For measures which have changed for some cities due to their adoption of the revised birth certificate, the Multi-city average only includes cities in states that are still using the 1989 version of the birth certificate. In 2003-2004, two different multi-city averages have been used for comparison purposes. For cities in states using the 1989 certificate, a multi-city average restricted to cities in states using the 1989 certificate is shown. For the three cities using the 2003 certificate in 2003 (Philadelphia, Pittsburgh, and Seattle), no multi-city average is displayed. For cities in states using the 2003 certificate in 2004, a multi-city average restricted to those cities in states that adopted the 2003 certificate no later than January 1, 2004, is shown. Reporting issues Births to unmarried women.4 Since 1998, births to unmarried women have been “identified by a question on the birth certificates of all but two States. . . . In the two States (Michigan and New York) that use inferential procedures to compile birth statistics by marital status . . . , a birth is inferred as nonmarital if either of these factors is present: a paternity acknowledgment5 was received or the father’s name is missing.” However, in 1990, the first year included in this volume, six states used inferential procedures.6 Thus, during the 1990-1998 period covered by this volume, four states—California (1997), Connecticut (1998), Nevada (1997), and Texas (1994)—switched from using inferential procedures to using a direct question to identify births to unmarried women.7 At each of these transition points, there is the potential for the percentage of births to unmarried women to change because of the change in the method of identifying births to unmarried women. In addition, among those states using inferential procedures, in some states the inferential procedures changed during the 1990-2002 period. For example, in the counts of births to unmarried women submitted by Michigan to NCHS, births with paternity acknowledgments were counted as births to unmarried women beginning in 1994. Thus, there is an undercount of about 25 percent for births to unmarried women in Michigan between 1990 and 1993. Data for Texas cities are also incomplete during 1990-1993. Data for Las Vegas, NV, for 1995 and 1996 were affected by a computer processing error and are therefore not presented. In 2003, mother’s marital status was not reported on 0.04 percent of birth records in the 48 states and the District of Columbia, where information was obtained via a direct question. For the missing records, marital status was imputed as married if the father’s age was known and unmarried if the father’s age was not known. This change in imputation had essentially no impact on percentages and rates of nonmarital births.8 Except in 2003, all cases of unreported marital status were imputed as married. Table A2.1 summarizes the reporting transitions made by each of the affected states by providing the date of each change and the likely effect of the change in reporting as estimated by NCHS. Cities included in this volume are listed under their respective states. In some cases, reporting changes have substantial effects on the percentage of births to unmarried women that are included in the tables in this volume. In such cases, trend analysis is either impossible or must be limited to an abbreviated period. All such cases are indicated with footnotes in the table. City of residence. Many people have a mailing address that uses the name of an incorporated city even though the address is not located within the corporate limits of that city. When asked if their address falls within the corporate limits of the city, some of these people answer (incorrectly) that it does. When this happens, births are attributed to the city incorrectly. Scattered evidence suggests that this misreporting problem can be serious—especially in built-up but unincorporated suburban areas that share the name of an incorporated city or town. In such cases, the number of births attributed to a city or town can be substantially inflated, which, in turn, leads to an overstated birth rate.9 This problem has been documented for Georgia but may also affect data for other cities in this volume, especially cities whose names are used in mailing addresses that are located outside the corporate limits of the city. Late or no prenatal care. Beginning in 2003, for states adopting the 2003 birth certificate, “substantive changes in both question wording and the sources for this information have resulted in data that are not comparable . . .” with data from prior years. “The wording of the prenatal care item was modified to ‘Date of first prenatal visit’ from ‘Month prenatal care began.’ In addition, the 2003 revision process resulted in recommendations that the prenatal care information be gathered from the prenatal care or medical records, whereas the 1989 revision did not recommend a source for these data.”10 Although we have reported data both pre- and post-revision years for cities in those states adopting the revised certificate, these data are not comparable. Data for cities in all other reporting states were based on the 1989 revision of the U.S. Standard Certificate of Live Birth. In earlier years, there were several cities in which the percentage of women who received late or no prenatal care changed substantially between two consecutive years. In these situations, we consulted with both NCHS staff and staff of the appropriate state department of health to check into the possibility that reporting problems were responsible, at least in part, for the changes. Table A2.2 summarizes the instances in which reporting problems were documented. Low maternal education. In 1990 and 1991, Washington and New York State (exclusive of New York City) did not require reporting of educational attainment. Beginning in 2003, for states adopting the 2003 birth certificate, substantive changes in question wording have resulted in data from prior years. The 2003 revisions to the U.S. Standard Certificate of Live Birth ask for the highest degree or level of school completed while the 1989 standard certificate asks the highest grade completed. These questions are too dissimilar to yield comparable results across years. Although we have reported data both pre- and post-revision years for those cities in states adopting the revised certificate, these data are not comparable. Data for cities in all other reporting states were based on the 1989 revision of the U.S. Standard Certificate of Live Birth. Table A2.3 summarizes all instances in which reporting issues were documented. Race and ethnicity. The number of El Paso births with Hispanic origin not reported was 411 in 1997, compared with 17 in 1998. Generally, when Hispanic origin is not reported, the mother is non-Hispanic, so it is likely that non-Hispanic white births were underreported in 1997 and earlier years. Data on ethnicity for Hartford are less reliable in the early 1990s than in later years. In 1990, nearly 13 percent of Hartford births were coded as unknown/not stated for Hispanic origin of the mother. In 1990, Oklahoma did not ask about Hispanic origin. Consequently, data on the race/ethnicity of births in Oklahoma City and Tulsa are not available in 1990. In Providence, Hispanic origin was not reported or unknown for 10 percent of total births in 1998—up from only 6 percent in 1990. Consequently, the distribution of births by Hispanic origin is less reliable during the late 1990s than in the early 1990s. Nearly 80 percent of births with unknown Hispanic origin in 1998 were to white women. The percentage of births to women whose Hispanic origin was unknown or not reported fell to 3 percent in 2000, returning to 6 percent in 2001. In 2003, multiple race was reported by cities in Pennsylvania and Washington, which used the 2003 revision of the U.S. Standard Certificate of Live Birth, as well as cities in California, Hawaii, Ohio (for births occurring in December 2003 only),and Utah. Data from the remaining 44 States and the District of Columbia reported only the four races stipulated in the 1977 OMB standards: White, Black, American Indian or Alaskan Native, and Asian or Pacific Islander. In 2004, multiple race was reported by Florida (for births occurring from March 1, 2004, only), Idaho, Kentucky, New Hampshire (for births occurring from July 19, 2004, only), New York State (excluding New York City), Pennsylvania, South Carolina, Tennessee, and Washington, which used the 2003 revision of the U.S. Standard Certificate of Live Birth, as well as California, Hawaii, Michigan (for births at selected facilities only), Minnesota, Ohio, and Utah, which used the 1989 revision of the U.S. Standard Certificate of Live Birth. To maintain uniform and comparable data between cities in states that report multiple race and those that do not, multiple race is mapped to one of the four races stipulated in the 1977 OMB standards. This report provides the number of births for four groups defined by race and Hispanic origin: non-Hispanic Whites, non-Hispanic Blacks, Hispanics, and Other, which includes American Indians, Asian-Pacific Islanders, and all those births where Hispanic origin is unknown or not stated. Repeat teen births. In 1997 and 1998, the number of teen births in which birth order was unknown or not reported was extremely high in Oklahoma. Consequently, the statistics on the percentage of Oklahoma City and Tulsa teen births to women who were already mothers are not reliable in those years and are not reported in this volume. Generally, most births with birth order unknown are first births. If it is assumed that most of the births with birth order unknown are actually first-order births, the resulting percentage of repeat teen births for Oklahoma City and Tulsa is comparable to that in most other cities. In 1999, Oklahoma’s reporting of live birth order improved considerably. Consequently, statistics on Oklahoma repeat teen births are reported in this volume starting with 1999. Smoking during pregnancy. Indiana did not report smoking during pregnancy on the birth certificate in a manner consistent with other states during 1990-1998. Consequently, 1990-1998 data on smoking during pregnancy are not available from NCHS for Indianapolis. However, subsequent data are consistent with other states and, therefore, are included. California asked its question on smoking in a form that was not compatible with the standard recommended by NCHS. Thus, data on smoking during pregnancy are not available for any cities in California. Oklahoma began reporting smoking data on the birth certificate in 1991. Consequently, data on smoking during pregnancy are not available for Oklahoma City and Tulsa in 1990. New York City began reporting of smoking during pregnancy in 1994.11 Data for New York State became available in 1999. Beginning in 2003, for states adopting the 2003 birth certificate, substantive changes in question wording have resulted in data from prior years. The tobacco-use-during-pregnancy question that was used in the 1989 revision had a yes or no question, while the 2003 revision asks for the number of cigarettes smoked at different intervals before and during pregnancy. These questions are too dissimilar to yield comparable results across years. Although we have reported data both pre- and post-revision years for those cities states adopting the revised certificate, these data are not comparable. Data for cities in all other reporting states were based on the 1989 revision of the U.S. Standard Certificate of Live Birth. Table A2.4 summarizes all instances in which reporting issues were documented. TABLE A2.1. Summary of reporting changes for births to unmarried women, 1990-2004
Sources: Ventura, S.J., Martin, J.A., Curtin S.C., Mathews T.J., and Park M.M. (2000), “Births: Final data for 1998,” National Vital Statistics Reports; Vol. 48, no. 3. Hyattsville, Maryland: National Center for Health Statistics. Ventura, S.J., Bachrach, C.A. (2000), “Nonmarital childbearing in the United States, 1940-1999,” National Vital Statistics Reports, Vol. 48, No. 16. Hyattsville, MD: National Center for Health Statistics. Ventura, S.J., Martin, J.A., Curtin S.C., Menacker,F., and Hamilton, B.E. (2001), “Births: Final data for 1999,” National Vital Statistics Reports; Vol. 49, no. 1. Hyattsville, Maryland: National Center for Health Statistics.
TABLE A2.3 Summary of reporting issues for the percentage of women with low maternal education
ABLE A2.4 Summary of reporting issues for the percentage of women who smoked during pregnancy
1 All information on the revision to the U.S. Certificate of Live Birth are based on the discussion in the Technical Notes of Martin JA, Hamilton BE, Sutton PD, et al. Births: Final data for 2004. National vital statistics reports; vol 55 no 1. Hyattsville, MD: National Center for Health Statistics. 2006. In some cases, to preserve the precision of the description of the changes, we have quoted directly from the report. 2 Martin, et al., 2006. 3 We are using 1997 population to determine the 50 largest cities to be consistent with the original Right Start data book. 4 Unless otherwise indicated, this discussion and all direct quotations are drawn from Ventura, S.J., and Bachrach, C.A. (2000), “Nonmarital Childbearing in the United States, 1940-99,” National Vital Statistics Reports; vol. 48, no. 16. Hyattsville, Maryland: National Center for Health Statistics. 5 A paternity acknowledgment is an acknowledgment from a man who is not married to a child’s mother that he is the child’s father. 6 Ventura, S.J., and Martin, J.A. (1993), “Advance Report of Final Natality Statistics, 1990,” Monthly Vital Statistics Reports; vol. 41, no. 9 (supplement). Hyattsville, Maryland: National Center for Health Statistics. 7 With the exception of Connecticut, these changes took place at the beginning of the year cited. The date of Connecticut’s change was June 15. 8 Unless otherwise indicated, all 2003 revisions and reporting issues are based on Martin, J.A., Hamilton, B.E., Sutton, P.D., Ventura, S.J., Menacker, F., Munson, M.L. (2005). “Births: Final data for 2003,” National Vital Statistics Reports; Vol. 54, no. 2. Hyattsville, Maryland: National Center for Health Statistics. 9 MacDorman, M.F., and Gay, G.A. (1999), “State Initiatives in Geocoding Vital Statistics Data.” Journal of Public Health Management and Practice, Vol. 5 No. 2, 91-93, 1999. 10 Martin, et al. 2006. 11 See Mathews, T.J. (1998), “Smoking During Pregnancy, 1990-96,” National Vital Statistics Reports; vol. 47, no. 10, Hyattsville, Maryland: National Center for Health Statistics and Mathews, T.J. (2001), “Smoking During Pregnancy in the 1990s,” National Vital Statistics Reports; vol. 49, no. 7, Hyattsville, Maryland: National Center for Health Statistics. 12 All changes occurred at the beginning of the year unless otherwise indicated. 13 Source: unpublished tabulation provided by Stephanie Ventura, National Center for Health Statistics. 14 All changes occurred at the beginning of the year unless otherwise indicated. 15 All changes occurred at the beginning of the year unless otherwise indicated.
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||