Abortion Surveillance -- United States, 1999

Persons using assistive technology might not be able to fully access information in this file. For assistance, please send e-mail to: mmwrq@cdc.gov. Type 508 Accommodation and the title of the report in the subject line of e-mail.

Laurie D. Elam-Evans, Ph.D.
Lilo T. Strauss, M.A.
Joy Herndon, M.S.
Wilda Y. Parker
Sara Whitehead, M.D.
Cynthia J. Berg, M.D.
Division of Reproductive Health
National Center for Chronic Disease Prevention and Health Promotion

Problem/Condition: CDC began abortion surveillance in 1969 to document the number and characteristics of women obtaining legal induced abortions and to monitor unintended pregnancy.

Reporting Period Covered: This report summarizes and describes data reported to CDC regarding legal induced abortions obtained in the United States in 1999.

Description of System: For each year since 1969, CDC has compiled abortion data by state or area of occurrence. From 1973 through 1997, data were received from or estimated for 52 reporting areas in the United States: 50 states, the District of Columbia, and New York City. Beginning in 1998, CDC compiled abortion data from 48 reporting areas. Alaska, California, New Hampshire, and Oklahoma did not report, and data for these areas were not estimated.

The availability of data regarding the characteristics of women who obtained an abortion in 1999 varied by state and by the number of states reporting each characteristic. The total number of legal induced abortions is reported by state of residence and also by state of occurrence for most areas; characteristics of women obtaining abortions in 1999 are reported by state of occurrence.

Results: A total of 861,789 legal induced abortions were reported to CDC for 1999, representing a 2.5% decrease from the 884,273 legal induced abortions reported by the same 48 reporting areas for 1998. The abortion ratio, defined as the number of abortions per 1,000 live births, was 256 in 1999, compared with 264 reported for 1998; the abortion rate for these 48 reporting areas was 17 per 1,000 women aged 15--44 years for 1999, the same as in 1997 and 1998.

The highest percentages of abortions were reported for women aged <25 years, women who were white, and unmarried women; slightly more than half were obtaining an abortion for the first time. Fifty-eight percent of all abortions for which gestational age was reported were performed at <8 weeks of gestation, and 88% were performed before 13 weeks. From 1992 (when these data were first collected) through 1999, increases have occurred in the percentage of abortions performed at <6 weeks of gestation. Few abortions were provided after 15 weeks of gestation; 4.3% were obtained at 16--20 weeks and 1.5% were obtained at >21 weeks. A total of 27 reporting areas submitted data stating that they performed medical (nonsurgical) procedures (two of these areas categorized medical abortions with "other" procedures), making up <1.0% of all procedures reported from all reporting areas.

In 1998 (for which data have not been published previously and the most recent year for which such data are available), nine women died as a result of complications from known legal induced abortion; no deaths were associated with known illegal abortion.

Interpretation: From 1990 through 1997, the number of legal induced abortions gradually declined. In 1998 and in 1999, the number of abortions continued to decrease when comparing the same 48 reporting areas. In 1998, as in previous years, deaths related to legal induced abortions occurred rarely.

Public Health Action: Abortion surveillance in the United States should continue so that trends and characteristics of women who obtain legal induced abortions can be examined and efforts to prevent unintended pregnancy can be enhanced. Introduction

CDC began conducting abortion surveillance in 1969 to document the number and characteristics of women obtaining legal induced abortions and to monitor unintended pregnancy. This report is based on abortion data for 1999 provided to CDC's National Center for Chronic Disease Prevention and Health Promotion (NCCDPHP), Division of Reproductive Health.

Methods

For 1999, CDC compiled data that were voluntarily provided from 48 reporting areas in the United States: 46 states (excluding Alaska, California, New Hampshire, and Oklahoma), the District of Columbia, and New York City. Legal induced abortion was defined as a procedure, performed by a licensed physician or someone acting under the supervision of a licensed physician, that was intended to terminate a suspected or known intrauterine pregnancy and to produce a nonviable fetus at any gestational age (1,2). The total number of legal induced abortions was available from all reporting areas; however, not all of these areas collected data regarding some or all of the characteristics of women who obtained abortions. Thus, the availability of these data varied by reporting area in 1999.

The majority of reporting areas (45 states, the District of Columbia, and New York City) collected and reported adequate abortion data (i.e., data categorized in accordance with surveillance variables and with <15% unknown values) by age of the woman, whereas only 24 states, the District of Columbia, and New York City collected and reported adequate abortion data by Hispanic ethnicity. Therefore, the findings in this report reflect characteristics of women only from reporting areas that submitted adequate data for those characteristics. The percentage data for most state tables include unknown values; unknowns have not been redistributed for the calculation of these percentages. However, percentages based only on known values are included for trend data, out-of-area residents, adolescent ages, and two-characteristics tables.

For the 48 reporting areas, data concerning the number of women obtaining legal induced abortions were provided by the central health agency.* These agencies provided data on numbers of abortions and characteristics of women by the state in which the abortions were performed (i.e., state of occurrence). For most states, only abortion totals were available by state of residence. However, two states, Delaware and Wisconsin, reported characteristics only for women who were residents and who obtained abortions in the state, but not for women from out of state; and one state (Iowa) provided both numbers and characteristics only for state residents. Two states (Florida and Louisiana) did not report abortion totals by resident status, and two states (Arizona and Massachusetts) provided only the total number of abortions for out-of-state residents without specifying their particular state or area of residence.

Women who obtained legal induced abortions were categorized by 5-year age groups and by single years of age for adolescents aged 15--19 years. Both abortion ratios (number of abortions per 1,000 live births in the same age group per year) and abortion rates (number of abortions per 1,000 women in the same age group per year) are presented by age group. Starting with 1996, ratios were calculated by using the number of live births to residents of each area from birth data reported to CDC's National Center for Health Statistics; numbers had previously been received from state health departments. Rates were calculated by using the number of women residents of each area from tabulations provided by the U.S. Census Bureau. Because nearly all (94%) abortions among women aged <15 years occurred among those aged 13--14 years in 1988 (the latest year for which this information is known) (3), the population of women aged 13--14 years was used as the denominator for calculating abortion rates for women <15 years. Rates for women aged >40 years were based on the number of women aged 40--44 years. Rates for all women who obtained abortions, however, were based on the population of women aged 15--44 years.

Race was categorized by three groups: white, black, and all other races. Other races included Asian/Pacific Islander, American Indian, Alaska Native, and women classified as "other" race. Ethnicity was categorized as Hispanic and non-Hispanic. As in previous reports, race and ethnicity were provided as separate variables and abortions were not cross-classified by race and ethnicity.

Marital status was reported as either married (including women who were married or separated) or unmarried (including those who were never married, divorced, or widowed).

Gestational age (in weeks) at the time of abortion was categorized as <6, 7, 8 and <8, 9--10, 11--12, 13--15, 16--20, and >21. Weeks of gestation were estimated in 20 reporting areas as the time elapsed since the woman's last menstrual period. For 18 other states, gestational age was reported on the basis of the physician's estimate (data from the clinical examination including ultrasound results). For the remaining five states, gestational age came from a combination of physician's estimates and the time elapsed since the woman's last menstrual period. Most areas (41 of 43) that reported adequate data on weeks of gestation at the time of abortion also reported abortions performed at <8 weeks separately for <6, 7, and 8 weeks of gestation.

CDC has periodically reported data on abortion-related deaths since these deaths were first included in the Abortion Surveillance Report for 1973 (4,5). An abortion-related death was defined as a death resulting from a) a direct complication of an abortion, b) an indirect complication caused by the chain of events initiated by abortion, or c) aggravation of a preexisting condition by the physiologic or psychologic effects of the abortion (1,2). Sources of data for abortion-related deaths included national and state vital records, maternal mortality review committees, surveys, private citizens and groups, media reports, health-care providers, medical examiners' reports, and computerized searches of full-text newspaper databases. All deaths associated with any type of abortion, induced or spontaneous, were investigated. For each death possibly related to an induced abortion or an abortion of unknown type, clinical records and autopsy reports were requested and reviewed by two clinically experienced medical epidemiologists to determine the cause of death and whether the death was abortion related. Each abortion-related death was then categorized as legal induced, illegal induced, spontaneous, or unknown (whether induced or spontaneous). Abortion-related deaths for 1972--1998 are provided in this report. The 1998 data have not been published previously and are the most recent data available. National case-fatality rates were calculated as the number of known legal induced abortion-related deaths per 100,000 reported legal induced abortions. Case-fatality rates for 1972--1997 are provided in this report. No case-fatality rate was calculated for 1998 because only 48 states reported abortions; thus, the denominator for the national case-fatality rate was unknown.

Results

Overall, the annual number of legal induced abortions in the United States increased gradually until it peaked in 1990, and it has generally declined thereafter (Figure 1). In 1999, a total of 861,789 legal induced abortions were reported to CDC by 48 reporting areas. This represents a 2.5% decrease from 1998, for which 884,273 legal induced abortions were reported from the same 48 reporting areas (5) (Table 1).

The national legal induced abortion ratio increased from 196 per 1,000 live births in 1973 (the first year that 52 areas reported) to 358 per 1,000 live births in 1979 and remained stable through 1981 (Figure 1) (Table 2). The ratio peaked at 364 per 1,000 live births in 1984 and since then has shown a generally steady decline. In 1999, the abortion ratio was 256 per 1,000 live births in the 48 reporting areas. This represents a 3.2% decrease from 1998 (264 per 1,000 live births) in the same 48 reporting areas (5) (Table 2). However, approximately 30--35 points of the apparent decrease in the 1998 and 1999 ratios compared with earlier years can be attributed to the current exclusion of the four nonreporting states, which had relatively high abortion ratios in earlier years.

The national legal induced abortion rate increased from 14 abortions per 1,000 women aged 15--44 years in 1973 (the first year that 52 areas reported) to 25 per 1,000 in 1980. In the 1980s and early 1990s, the rate remained stable at 23--24 abortions per 1,000 women, and from 1994 through 1997 it again stabilized at 20--21. The abortion rate has remained unchanged since 1997 at 17 per 1,000 women in the same 48 reporting areas. Again, the seeming decline after 1997 is due to the exclusion of the four nonreporting states with relatively high abortion rates.

The numbers, ratios, and rates of reported legal induced abortions are presented by area of residence as well as by area of occurrence (Table 3). In 1999, as in previous years, the highest number of reported legal induced abortions occurred in New York City (102,334), Florida (83,971), and Texas (80,739); the fewest occurred in Wyoming (110), South Dakota (740), and Idaho (867) (Table 3). The abortion ratios by state or area of occurrence ranged from 18 per 1,000 live births in Wyoming to 980 per 1,000 live births in the District of Columbia. The rates by occurrence ranged from 1 per 1,000 women aged 15--44 in Wyoming to 59 per 1,000 women aged 15--44 in the District of Columbia. These ratios and rates should be viewed with consideration of the sizable variation by state in the percentage of abortions obtained by out-of-state residents. In 1999, approximately 9% of reported abortions were obtained by out-of-state residents. The percentages ranged from 0.5% in Hawaii to 55% in the District of Columbia (Table 3). Data by state of residence are not complete because four states (Alaska, California, New Hampshire, and Oklahoma) did not report, and five states (Arizona, Florida, Iowa, Louisiana, and Massachusetts) could not provide any data concerning the residence status of all women obtaining abortions in their state.

Women with known age reported as 20--24 years obtained 32% of all abortions; women aged <15 years obtained <1.0% of all abortions in areas where age was reported (Table 4). Abortion ratios were highest for the youngest women (709 abortions per 1,000 live births for women aged <15 years) and lowest for women aged 30--34 years (152 per 1,000 live births) (Figure 2) (Table 4). In contrast to abortion ratios, among women for whom age was reported, abortion rates were highest for women aged 20--24 years (35 abortions per 1,000 women) and lowest for women at the extremes of reproductive age (2 abortions per 1,000 women aged 13--14 years and 2 per 1,000 women aged 40--44 years) (Table 4). Among adolescents (aged <20 years), the percentage of abortions obtained increased with increasing age. However, the abortion ratio was highest for those <15 years (708 abortions per 1,000 live births)^† and lowest for those aged 19 years (343 per 1,000 live births) (Table 5). Conversely, the rates of abortions were lowest for adolescents aged <15 years (2 per 1,000 women aged 13--14 years) and highest for women aged 19 years (30 per 1,000 women aged 19 years) (Table 5).

For women in the majority of age groups, the abortion ratio increased from 1974 through the early 1980s and declined thereafter, particularly for the younger (<19) and oldest reproductive-aged women (Figure 3). Abortion ratios for women <15 years have always been higher than those for the other age groups. The abortion ratio for women aged 20--34 years (the groups with the highest fertility rates) (6) has remained relatively stable since the mid-1980s.

In 1999, for women whose weeks of gestation at the time of abortion were adequately reported, 57% of reported legal induced abortions were known to have been obtained at <8 weeks of gestation, and 87% were reported at <13 weeks (Table 6). Overall, 22% of abortions were performed at <6 weeks of gestation, 17% at 7 weeks, and 18% at 8 weeks (Table 7). Few reported abortions were provided after 15 weeks of gestation; 4.3% were known to have been obtained at 16--20 weeks, and 1.5% at >21 weeks.

For women whose type of procedure was adequately reported, almost all (98%) abortions were known to have been performed by curettage and 0.2% by intrauterine instillation (Table 8). Hysterectomy and hysterotomy were included in the "other" procedure category and were used in <0.01% of all abortions. Twenty-seven reporting areas submitted data stating that they performed medical (nonsurgical) procedures^§ (two of these areas included medical abortions in the "other" category), making up approximately 1% of all procedures reported from all areas with adequate reporting. However, certain other reporting areas do not include medical abortions as a separate category in their abortion reporting form. For 1999, 25 reporting areas submitted data stating that they performed a total of 6,278 medical (nonsurgical) procedures. This reflects an increase of 28% from the 4,899 medical abortions reported by 22 reporting areas for 1998 (5). We do not know to what extent the 6,278 medical (nonsurgical) abortions reported to CDC for 1999 represent the use of this procedure in all reporting areas.

In the 37 areas for which race was adequately reported, approximately 55% of women who obtained legal induced abortions were known to be white, 36% were black and 6% were of other races. (Table 9). However, the abortion ratio for black women (529 per 1,000 live births) was 3.0 times the ratio for white women (177 per 1,000 live births). Additionally, the abortion ratio for women of other races (367 per 1,000 live births) was 2.1 times the ratio for white women. The abortion rate for black women (31 per 1,000 women) was 3.0 times the rate for white women (10 per 1,000 women). The abortion rate for women of other races (26 per 1,000 women) was 2.5 times the rate for white women.

Twenty-four states, the District of Columbia, and New York City reported adequate data^¶ concerning the ethnicity of women who obtained legal induced abortions (Table 10). The percentage of abortions known to have been obtained by Hispanic women in these reporting areas ranged from 0.2% in Kentucky to 46% in New Mexico. For Hispanic women in these reporting areas, the abortion ratio was 261 per 1,000 live births. The abortion rate for Hispanic women was 19 abortions per 1,000 women.

For women whose marital status was adequately reported, 78% of women who obtained abortions were known to be unmarried (Table 11). The abortion ratio for unmarried women was 8.6 times the ratio for married women (604 versus 70 abortions per 1,000 live births).

For women for whom data on previous live births were adequately reported, 39% of women who obtained legal induced abortions were known to have had no previous live births, and 86% had had <2 previous live births (Table 12). The abortion ratio was highest for women who had three previous live births (298 per 1,000 live births) and lowest for women who had one previous live birth (206 per 1,000 live births).

In 1999, of women who obtained an abortion and whose number of previous abortions was adequately reported, 52% were known to have obtained an abortion for the first time. Nineteen percent of women had >2 previous abortions (Table 13).

For women whose age and race were known, white women had a greater percentage of abortions at the youngest (<19 years) and oldest (>35 years) age groups compared with women of black or other races (20% versus 18%, and 12% versus 10%, respectively) (Table 14). For women whose marital status and race were both known, the percentage of reported abortions among black or other races that were obtained by unmarried women (84%) was higher than that obtained by unmarried white women (79%). Among women obtaining abortions whose age and Hispanic ethnicity were known (25 reporting areas), the percentage of abortions obtained by older women (>35 years) of non-Hispanic ethnicity (12%) was greater than that for older women of Hispanic ethnicity (9%) (Table 15). For women whose marital status and ethnicity were known (24 reporting areas), the percentage of reported abortions obtained by unmarried women was similar for non-Hispanic and Hispanic women (Table 15). Data were not available to cross-classify race by Hispanic ethnicity.

As in the past, approximately 88% of all abortions (for which gestational age at the time of abortion was reported and known) were obtained during the first 12 weeks of gestation (Table 1). The percentage of women who obtained an abortion at <8 weeks of gestation increased with age (Figure 4) (Table 16). This association is most evident for abortions obtained at <6 weeks' gestation (Table 17). The percentage of women who obtained an abortion at >21 weeks of gestation decreased with age for women through 25--29 years and remained stable for women in the older age groups (Table 16). Among women with known race and weeks of gestation, white women and women of other races were more likely than black women to obtain abortions at <6 or 7 weeks of gestation (Table 17). Among women with known ethnicity and weeks of gestation, 23% of Hispanic women obtained abortions at <6 weeks of gestation and 59% obtained abortions at <8 weeks' gestation (Table 17).

For women whose type of procedure and weeks of gestation were known, approximately 99% of reported abortions obtained at <15 weeks of gestation were performed using curettage (primarily suction procedures) (Table 18). Approximately 88% of reported medical (nonsurgical) abortions (n = 4,777) were performed at <8 weeks' gestation; these 4,777 represent 1.3% of all abortions that are performed at <8 weeks' gestation. At >16 weeks of gestation, medical (nonsurgical) abortions (n = 491) made up 1.3% of all abortions, whereas at all other gestational ages, medical (nonsurgical) abortions made up <0.2%. Intrauterine instillation involved the use of saline or prostaglandin and was used rarely (0.2% of all abortions), primarily at >16 weeks of gestation.

From the National Pregnancy Mortality Surveillance System, CDC identified 22 maternal deaths with some indication of abortion on the death certificate for 1998. Investigation of these cases indicated that nine of these 22 deaths were known to be related to legal induced abortion and none to illegal induced abortion (Table 19). Eleven deaths were due to spontaneous abortion, and one death was due to abortion of unknown type. One death was determined not to be abortion related. Numbers of deaths due to legal induced abortion were highest before the 1980s, with very few deaths occurring in 1998 (Table 19). Possible abortion-related deaths that occurred during 1999 are currently being investigated.

Discussion

A total of 861,789 legal induced abortions were reported for 1999 in the United States from 46 states, the District of Columbia, and New York City. This is a decline of 2.5% from the legal induced abortions reported for 1998 from the same 48 reporting areas (5). In previous years, a substantial number of legal induced abortions were estimated to have been performed in California, (e.g., >23% of the U.S. total in 1997) (7). The lack of data for California for 1999 explains most of the 27% decrease from the annual number of abortions reported in 1997 (7) as well as some of the decrease in the total ratio and rate. After 1997, data were no longer estimated for nonreporting states (5).

The abortion ratio for 1999 (256 per 1,000 live births) was a 3.2% decline from the previous year and is the lowest recorded level since 1975 (8). The abortion rate (17 per 1,000 women aged 15--44 years) was the same as in 1997 and 1998 (Table 2). The declines in the abortion ratio and rate over time may reflect multiple factors, including a decrease in the number of unintended pregnancies (9); a shift in the age distribution of women toward the older and less fertile ages (6); reduced access to abortion services, including the passage of abortion laws that affect adolescents (e.g., parental consent or notification laws and mandatory waiting periods) (10--13); and changes in contraceptive practices, including an increased use of contraception, particularly of condoms, and, among young women, of long-acting hormonal contraceptive methods that were introduced in the early 1990s (14--16).

The abortion rate reported here for the United States was higher than recent rates reported for Australia, Canada, and Western European countries and lower than rates reported for China, Cuba, most Eastern European countries, and several of the Newly Independent States of the former Soviet Union (17,18). The abortion rates reported for teenagers are higher in the United States than in most Western European countries and some Eastern European countries (19).

As in previous years, the abortion ratio in 1999 varied substantially by age (5). Although the ratio was highest for adolescents, the percentage of legal induced abortions obtained by women aged <19 years has decreased since the early 1970s (from 33% in 1973 to 22% in 1990 and to 20% in 1992) and has remained at that level (Table 1) (4,20--22). The abortion ratio has declined for women aged <15 and 15--19 years since the mid-1980s. Other studies also have indicated a decrease in birth rates for women aged 15--19 years from 1991 through 2000 and a decrease in adolescent pregnancy rates from 1991 to 1997 (6,23--29).

Since 1980, the percentage distribution of abortions by known weeks of gestation has been stable. From 1992 (when more detailed data for <8 weeks' gestation were first available) through 1999, data have shown some steady increases in procedures performed at <6 and 7 weeks' gestation, with decreases occurring in the percentage of abortions performed at 8, 9--10, and 11--12 weeks' gestation. The increase in the percentage of abortions known to have been performed at <6 and 7 weeks may be related to an increase in availability of early abortion services since 1992 as well as to an increase in medical (nonsurgical) and surgical procedures that can be performed early in gestation (30,31). Abortions performed early in pregnancy are associated with lower risks of mortality and morbidity (32). The proportions of abortions performed later in pregnancy (>13 weeks) have varied little since 1992. Several additional factors affect the timing of abortion. Age of the woman, race, marital status, level of education, availability and accessibility of services, timing of confirmation of pregnancy, personal decision-making, level of fear of discovery of pregnancy and denial of pregnancy all have an impact on timing (5,33,34).

Since the mid-1990s, two medical (nonsurgical) regimens -- methotrexate and mifepristone, each used in conjunction with misoprostol -- have been tested in clinical trials and used by clinical practitioners to perform early medical (nonsurgical) abortions (31,35). Surveillance data from CDC and other investigators indicate that >50% of all U.S. abortions are performed at <8 weeks of gestation, which is the timing of the regimen approved for both mifepristone and methotrexate (36). The medical (nonsurgical) procedures reported most often for abortions performed early in gestation (<7 weeks) are use of methotrexate with misoprostol and mifepristone with misoprostol (37,38). Mifepristone for medical abortion was approved by the Food and Drug Administration for use and distribution in the United States in September 2000. This approval might result in early medical (nonsurgical) induced abortions becoming more widespread (39,40).

In 1997, the U.S. Standard Report of Induced Termination of Pregnancy, published by the National Center for Health Statistics (NCHS) and used by providers for abortion reporting to state health departments, was revised to include a category for "medical (nonsurgical)" procedures (41). Medical (nonsurgical) legal induced abortion procedures have been included in this report since then as a separate category. CDC will continue to monitor early medical (nonsurgical) procedures.

The percentage of abortions known to be performed by curettage (which includes dilatation and evacuation [D&E]) increased from 89% in 1972 to 98% in 1999 (Table 1), while the percentage of abortions performed by intrauterine instillation declined sharply, from 10% to 0.2%. The increase in use of D&E is likely due to the lower risk for complications associated with the procedure (42,43). From 1974 (first year for which these data were available) through 1999, the percentage of second-trimester abortions performed by D&E (curettage) increased from 31% to 96%; the percentage of second-trimester abortions performed by intrauterine instillation decreased from 57% to 1.3% (Table 18) (44).

The differential between the abortion ratio for black women and that for white women has increased substantially since 1989 (45). In addition, the abortion rate for black women was 3.0 times the rate for white women. Census Bureau estimates and birth certificate data indicate that the majority of Hispanic women report themselves as white. NCHS vital statistics reports indicate that 97% of Hispanic women giving birth in 1999 were white (6). Data for some Hispanic women, therefore, are included in data for white women; in 1999, Hispanic women accounted for 24% of all births to white women (6). Race-specific and ethnicity-specific differences in legal induced abortion ratios and rates might reflect differences in factors such as socioeconomic status, access to family planning and contraceptive services, contraceptive use, and incidence of unintended pregnancies among groups.

In 1999, 37 states, the District of Columbia, and New York City reported Hispanic ethnicity of women who obtained abortions. Because of concerns regarding the completeness of such data (>15% unknown data) in some states, in 1999, data from only 24 states, the District of Columbia and New York City were used to determine the number and percentage of abortions obtained by women of Hispanic ethnicity. These geographic areas represent approximately 40% of all reproductive-age Hispanic women in the United States for 1999 and approximately 40% of U.S. Hispanic births (6,46). Thus, the number of Hispanic women who obtain abortions is underestimated, and the number, ratio, and rate of abortions for Hispanic women in this report are not generalizable to the overall Hispanic population in the United States.

The abortion ratio for Hispanic women in past years has either been slightly lower than or similar to that for non-Hispanic women; however, in 1999, the results shifted in that the ratio for Hispanic women (261) was substantially higher than for non-Hispanic women (252). As in the past, the abortion rate per 1,000 Hispanic women was higher than the rate per 1,000 non-Hispanic women, which is consistent with another study (47). NCHS vital statistics reports indicate that in 1999, fertility and live birth rates were substantially higher for Hispanic women as a whole than for non-Hispanic women for all age groups (6). However, because fertility and birth rates differ substantially among components of the Hispanic subgroups (Mexican, Puerto Rican, Cuban, and other Hispanic), and these differ substantially from rates among components of the non-Hispanic subgroups (white, black, other), comparisons between Hispanic and non-Hispanic groups are of limited value (6). The available abortion surveillance data do not permit cross-classification of race by Hispanic ethnicity.

Compared with 1972, the annual number of deaths associated with known legal induced abortion in 1998 has decreased nearly two thirds (Table 19). In 1972, 24 women died from causes known to be associated with legal abortions and 39 died as a result of known illegal abortions. In 1998, nine died as a result of legal induced abortion and none died as a result of illegal induced abortion. The national case-fatality rate for 1998 can not be appropriately calculated because a substantial number of abortions occurred in the four nonreporting states, making the denominator unknown.

Because these data are reported voluntarily, several limitations exist. First, abortion data are compiled and reported to CDC by reporting area where the abortion was performed rather than by where the woman resides. This inflates the numbers, ratios, and rates of abortions for areas where a high proportion of legal abortions are obtained by out-of-state residents and undercounts procedures for states with limited abortion services or more stringent legal requirements for obtaining an abortion (causing women to seek abortions elsewhere). Second, four states (Alaska, California, New Hampshire, and Oklahoma) did not collect or report abortion data in 1998 or 1999. Data for these four states were not estimated for 1998 or 1999. Data for California and Oklahoma had been estimated before 1998, and Alaska and New Hampshire reported for 1997. Third, data provided to state or area health departments by providers may be incomplete (48). Fourth, the overall number, ratio, and rate of abortions are conservative estimates because the total numbers of legal induced abortions provided by central health agencies and reported to CDC for 1999 were probably lower than the numbers actually performed. Additionally, totals provided by central health agencies are lower than those obtained by direct surveys of abortion providers (49). For example, the total number of abortions reported to CDC for 1997 was approximately 12% lower than that reported for 1997 by The Alan Guttmacher Institute, a private organization that contacts abortion providers directly to obtain the total number of abortions performed (27). Fifth, not all states collected or reported data on all characteristics (e.g., age, race, and weeks of gestation) of women obtaining a legal induced abortion in 1999. Thus, the numbers, rates, and ratios derived in this analysis may not be representative of all women who obtained abortions.

Despite these limitations, findings from ongoing national surveillance of legal induced abortion are useful for several purposes. First, public health agencies use data from abortion surveillance to identify characteristics of women who are at high risk for unintended pregnancy. Second, ongoing annual surveillance is used to monitor trends in the number, ratio, and rate of abortions in the United States. Third, statistics regarding the number of pregnancies ending in abortion are used in conjunction with birth data and fetal death computations to estimate pregnancy rates (e.g., pregnancy rates among adolescents) (23--27). Fourth, abortion and pregnancy rates can be used to evaluate the effectiveness of family planning programs and programs for preventing unintended pregnancy. Fifth, ongoing surveillance provides data for assessing changes in clinical practice patterns related to abortion (e.g., longitudinal changes in the types of procedures and trends in weeks of gestation at the time of abortion). Finally, numbers of abortions are used as the denominator in calculating abortion morbidity and mortality rates (32).

Welfare-reform legislation, specifically, the Personal Responsibility and Work Opportunity Reconciliation Act of 1996,** has increased interest in accurate state-based surveillance for induced abortion. In addition, certain states have instituted programs that emphasize the prevention of unintended pregnancy, particularly among adolescents. To help guide these efforts, an ongoing, accurate assessment of induced abortion is needed in all states (particularly abortion data by state of residence) to determine the number and characteristics of women who obtain these procedures.

Induced abortions usually result from unintended pregnancies, which often occur despite use of contraception (14,50,51). Data from the 1995 National Survey of Family Growth (NSFG), conducted by NCHS, indicate that approximately 49% of all pregnancies were unintended (9,52) and 31% of live births were associated with unintended pregnancy (i.e., either mistimed or unwanted at conception). Unintended pregnancy is a problem not just for adolescents, unmarried women, or poor women; it is a pervasive public health problem for all women of reproductive age (9,14,52).

A reduction in unintended pregnancy, and thus abortion, will require several complex strategies. In a study of abortion patients conducted during 1994--1995, 58% of patients reported that they "currently used" contraception during the month of their last menstrual period. However, their use of contraception might have been inconsistent or incorrect (47). In 1995, when the most recent NSFG was conducted, approximately 29% of sexually active U.S. women who used only oral contraceptives for birth control reported that they missed a birth-control pill one or more times during the 3 months before their NSFG interview. In addition, approximately 33% of U.S. women who were using only coitus- dependent contraceptive methods^†† during the 3 months before the interview used these methods inconsistently (9). Also, not all health insurance plans provide contraceptive benefits (53). Therefore, education regarding improved contraceptive use and practices as well as access to and education regarding safe, effective, and affordable contraception and family-planning services might help reduce the incidence of unintended pregnancy and, therefore, might reduce the number of legal induced abortions in the United States (54--56).

Copies of this and other MMWR reports containing statistical and epidemiologic data regarding abortions can be obtained through the CDC website at http://www.cdc.gov/nccdphp/drh/surv_abort.htm.

The authors thank Sonya Bowens, M.S. and Elizabeth Fitch, Division of Reproductive Health, NCCDPHP, CDC, for their critical work with computer programming and graphics for this surveillance project.

References

50. Fu H, Darroch JE, Haas T, Ranjit N. Contraceptive failure rates: new estimates from the 1995 National Survey of Family Growth. Fam Plann Persp 1999;31:56--63. Corrected data are available at http://www.agi-usa.org/pubs/journals/3105699.html.

51. Torres A, Forrest JD. Why do women have abortions? Fam Plann Perspect 1988;20:169--76.

52. Henshaw SK. Unintended pregnancy in the United States. Fam Plann Perspect 1998;30:24--9,46.

53. Gold RB, Darroch JE, Frost JJ. Mainstreaming contraceptive services in managed care -- five states' experiences. Fam Plann Perspect 1998;30:204--11.

54. Westoff CF. Contraceptive paths toward the reduction of unintended pregnancy and abortion. Fam Plann Perspect 1988;20:4--13.

55. Forrest JD. Epidemiology of unintended pregnancy and contraceptive use. Am J Obstet Gynecol 1994;170(suppl):1485--8.

56. Gonen JS. Value purchasing: investing in women's health. Washington DC: Jacobs Institute of Women's Health, 2000.

^†† Coitus-dependent contraceptive methods include male or female condoms, diaphragm, sponge, cream, jelly, or other methods that must be used at the time intercourse occurs.

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.