Materials and Methods
To monitor trends in pregnancy-related deaths, the Centers for Disease Control and Prevention's (CDC's) Division of Reproductive Health, in collaboration with ACOG and state health departments, conducts pregnancy-related mortality surveillance. Health departments in the 50 states, the District of Columbia, and New York City provided CDC with copies of death certificates and, when available, the corresponding matched pregnancy-outcome records (birth or fetal death certificates) for all identified pregnancy-related deaths for 1979–1992, the most recent available data. Classification of deaths as pregnancy-related by this system has been described.10
Deaths were considered pregnancy-related if they occurred during pregnancy or within 1 year of pregnancy termination and resulted from complications of the pregnancy itself, a chain of events initiated by pregnancy, or aggravation of an unrelated event by the physiologic effects of pregnancy. We reviewed all information about each death to ascertain whether the death was pregnancy-related.
Pregnancy-related mortality ratios were defined as the number of pregnancy-related deaths per 100,000 live births. To determine Hispanic pregnancy-related mortality ratios, we limited our analysis to states that reported information on Hispanic origin on live birth certificates in a given year. Beginning in 1978, states began to identify ethnicity or Hispanic origin of mothers on birth certificates, including information on Hispanic subgroups (eg, Mexican, Puerto Rican, Cuban).11 In 1979, 19 states collected information on Hispanic origin on birth certificates.11 These states accounted for 60% of Hispanic live births in that year. The next year, 22 states (90% of Hispanic live births) collected information on Hispanic origin for births,12 and in 1992, 49 states and the District of Columbia (99% of Hispanic live births) did so.7 The number of live births in the states that collected data on Hispanic origin was obtained from public-use tapes maintained by the CDC's National Center for Health Statistics. We excluded deaths from Puerto Rico because published natality data did not include births there.
Race and ethnicity were classified separately. Race was defined as the race of the mother. Hispanic origin was assigned if the maternal death certificate, matched live birth certificate, or fetal death certificate indicated maternal Hispanic origin. All personal identifiers were removed from copies of death certificates submitted by the states; therefore, Hispanic surnames could not be used to assign Hispanic origin. Because 98% of the Hispanic women who died from pregnancy-related causes during the study period were white, race-specific mortality was not calculated for Hispanic women. Non-Hispanic white women were reported as “non-Hispanic white,” and non-Hispanic black women were reported as “black.” Mortality rates in women of other races, including Asian-Pacific Islanders and American Indian–Alaskan Natives, were not analyzed.
Hispanic women were further categorized by ethnic subgroup, including Mexican, Puerto Rican, Cuban, Central or South American, other Hispanic, and unknown Hispanic. Nativity status (ie, U.S.-born or foreign-born) was obtained from the maternal place of birth as listed on the maternal death certificates or matched live birth or fetal death certificates, and was only available beginning in 1987. Potential risk factors (eg, age, live-birth order) could not be calculated by Hispanic subgroups because of the small size of some groups.
Maternal age was grouped in 5-year intervals. Because of small numbers, women aged 10–19 years were grouped together, as were women aged 40–49 years. Because parity was not reported on vital records, we examined live-birth order as a proxy for parity. Information on live-birth order (reported on birth certificates) was available only for deaths associated with live births. Information on prenatal care also was available only for women who died after live births. Prenatal care was assessed using a modification of the Adequacy of Prenatal Care Index developed by Kotelchuck.13 Our modified index combined “intermediate” care with “inadequate” care and added a category for women who received no prenatal care. Deaths from states not reporting one or more of the three variables needed to calculate the Adequacy of Prenatal Care Index were excluded from the analysis of prenatal care adequacy for those years during which they did not report (Texas, 1979; New Mexico, 1979–1985; California, 1979–1988). Because of the large percentage of death certificates missing information needed for computing prenatal care adequacy (22.5%), the results of this analysis must be interpreted with caution.
Causes of death and outcomes of pregnancy were classified independently and determined for each case using all available data, including death certificates and, when available, birth or fetal death certificates, maternal mortality review committee reports, and medical records. Deaths resulting from abortion included all abortion deaths reported as spontaneous, induced, or of unknown type. Except for cause of death and pregnancy outcome, all unknown, not stated, or missing information was proportionally redistributed in known categories for each variable.
In the states that reported Hispanic origin during all or part of the study period, there were 3777 pregnancy-related deaths; 623 deaths were of women of Hispanic origin. There were 6.06 million live births to Hispanic women for the reporting states in the years 1979–1992. The pregnancy-related mortality ratio was 10.3 deaths per 100,000 live births for Hispanic women, 6.0 for non-Hispanic white women, and 25.1 for black women (Figure 1). Pregnancy-related mortality varied by Hispanic ethnic subgroup (Figure 1). Mexican women, the largest Hispanic subgroup (67% of Hispanic live births in 1992), had a mortality ratio of 9.7. Puerto Rican women (9% of Hispanic live births in 1992) had the highest ratio at 13.4; Cuban women, the smallest group (1.8% of Hispanic live births in 1992), had the lowest ratio of 7.8.
From 1987 to 1992, foreign-born Hispanic women had a higher risk of pregnancy-related death than Hispanic women born in the United States (Table 1). Foreign-born women whose Hispanic origin was “other or unknown” had the highest pregnancy-related mortality ratio, and U.S.-born women of Cuban ancestry had the lowest. During 1987–1992, 40% of all Hispanic live births were to women born in the United States and 60% were to immigrant mothers.
Mortality ratios for Hispanic women in every age group were higher than for non-Hispanic white women, but markedly lower than for black women (Table 2). In each of these groups, women aged 30 years and older had a higher risk of pregnancy-related death than younger women. For Hispanic women, as for non-Hispanic white and black women, the risk of pregnancy-related death after a live birth rose with increasing live-birth order after the second live birth (Table 2). For both Hispanic women and non-Hispanic white women, the fifth or later birth had nearly three times the risk of the second live birth. For all women who died after a live birth, the risk of death due to pregnancy was greater for those who received no prenatal care than for those who received some care (Table 2). The increased risk of death associated with no prenatal care was smaller for Hispanic women than for non-Hispanic white or black women.
Pregnancy outcomes for women who died from pregnancy-related causes differed by race and Hispanic origin. In Hispanic women, 71% of these deaths occurred after live births or stillbirths, versus 64% in non-Hispanic white women and 60% in black women. Ectopic pregnancy was associated with 6% of the deaths in Hispanic women, 10% in non-Hispanic white women, and 15% in black women. Four percent of the deaths of Hispanic women and non-Hispanic white women occurred after abortions (spontaneous or induced); 6% did so in black women.
The leading cause of pregnancy-related death after a live birth or stillbirth for Hispanic women was complications of pregnancy-induced hypertension (Table 3). Hispanic women were approximately three times as likely to die from complications of pregnancy-induced hypertension as non-Hispanic white women. The risk of death from all other causes after a live birth or stillbirth was up to twice that of non-Hispanic white women. Hemorrhage was the primary cause of death for all women who had ectopic pregnancies (data not shown). Women whose pregnancies ended with abortions died primarily from infection, bleeding, embolism, and anesthesia complications. Among women who died while pregnant (undelivered), embolism was the leading cause of death.
We found that the risk of pregnancy-related death for Hispanic women is higher than that for non-Hispanic white women and markedly lower than that for black women. In contrast, Liao et al14 found that overall mortality rates for Hispanic women aged 18–44 years were similar to those for non-Hispanic white women and significantly lower than those for black women.
Because minority status and lower socioeconomic resources contribute to poor health outcomes,3,15 we expected Hispanic women to fare worse than non-Hispanic white women, but we found the disparity in the risk of pregnancy-related death between black and Hispanic women striking. The socioeconomic disadvantage of both Hispanics and blacks in the United States has been well documented.4,5 In 1996, 60% of Hispanics, 55% of blacks, and 26% of whites lived in families classified as poor or nearly poor.5 Other researchers, however, have found that socioeconomic status and health outcomes correlate less well for Hispanics than for blacks, and have labeled this the “Hispanic paradox.”16
Our analysis found that the risk of pregnancy-related death varied among Hispanic ethnic subgroups; other researchers have also found important differences by subgroup. For example, socioeconomic profiles and health insurance status differed among Mexican-Americans, Cuban-Americans, and Puerto Ricans.8,17 In the physician assessment portion of the Hispanic Health and Nutrition Survey, greater proportions of Mexican-American and Cuban-American women than Puerto Rican women of reproductive age were reported to be in excellent health.8 Mortality rates among Hispanic infants were highest for Puerto Ricans and lowest for Cuban-Americans.6
The increased risk of death for Hispanic women associated with increasing maternal age and the pattern of mortality associated with live-birth order shown here are similar to findings in other studies of pregnancy mortality.10 An increased risk of pregnancy-related mortality for women with three or more live births is important for Hispanic women because of their higher fertility rates.7 The 1995 National Survey of Family Growth reported that 50% of Hispanic women, 30% of white women, and 34% of black women expect to have three or more births.18
Prenatal care use by Hispanic women is lower than that for non-Hispanic white women and similar to that of black women. Access to the health care system is limited for many Hispanic women, particularly those who are younger, have lower incomes, are less acculturated, are uninsured, and have poorer perceived health status.4,19 The markedly lower pregnancy-related mortality ratio for Hispanic women who had received no prenatal care compared with non-Hispanic white and black women who received no care was unexpected. However, these results may have been influenced by the large proportion of women for whom no prenatal care information was available.
Our finding that pregnancy-induced hypertension was responsible for approximately one-third of the pregnancy-related deaths of Hispanic women after live births or stillbirths fits well with a previous study in which pregnancy-induced hypertension was the leading risk factor for pregnancy-related deaths of Hispanic women in Puerto Rico.20 In contrast, a study of hospitalizations for severe complications of pregnancy that used population-based California hospital discharge data from 1987 to 1992 found rates of hospitalization for pregnancy-induced hypertension to be similar for Hispanic women and white women.21 Pregnancy-related deaths from pregnancy-induced hypertension are preventable. Early prenatal care, subsequent detection of pregnancy-induced hypertension, and careful monitoring and treatment during pregnancy are essential to prevent serious complications from this problem.22
Several limitations of this analysis should be considered. Pregnancy-related mortality encompasses a complex array of causes and pregnancy outcomes, and the underlying risk factors associated with death may vary with the cause of death and pregnancy outcome. Although matched live birth and fetal death certificates provide more and better information than death certificates alone, detailed clinical information was not available.
Pregnancy-related mortality ratios are underestimated for all groups. Despite improved ascertainment methods initiated during the study period, we estimate that more than half of pregnancy-related deaths are not identified through routine surveillance methods.23,24 Misclassification of race and ethnicity may also have influenced the results of this study. If Hispanic origin was not indicated on death certificates or matched live birth or fetal death certificates, pregnancy-related mortality among Hispanic women may have been underreported.
Further studies are needed to assess the rates of morbidity and subsequent mortality due to pregnancy-induced hypertension in Hispanic women. Because Hispanic women are similar socioeconomically to black women but have a lower risk of death from pregnancy than black women, further study is needed of the factors that contribute to this Hispanic paradox in pregnancy-related mortality.
Prevention of pregnancy-related deaths from all causes among Hispanic women should take into account medical and nonmedical factors and the heterogeneity of the Hispanic population. Additional sources of data, such as clinical data, access to and content of prenatal care, and the effects of socioeconomic indices, social environment, and lifestyle, may help to explain the differences in pregnancy-related death between Hispanic and non-Hispanic white and black women. Clinical and public health interventions targeting Hispanic women are needed to reduce pregnancy-related mortality among this growing part of our population.
1. U.S. Bureau of the Census. Statistical abstract of the United States: 1997. 117th ed. Washington, DC: US Government Printing Office, 1997.
2. Ventura SJ, Peters KD, Martin JA, Maurer JD. Births and deaths: United States, 1996. Mon Vital Stat Rep 1997;46(suppl 2):10.
3. Klitsch M. Hispanic ethnic groups face variety of serious health, social problems. Fam Plann Perspect 1991;23:186–8.
4. Council on Scientific Affairs. Hispanic health in the United States. JAMA 1991;265:248–52.
5. Pamuk E, Makuc D, Heck K, Reuben C, Lochner K. Socioeconomic status and health chartbook. Health, United States, 1998. Hyattsville, Maryland: National Center for Health Statistics, 1998.
6. Becerra JE, Hogue CJR, Atrash HK, Perez N. Infant mortality among Hispanics: A portrait of heterogeneity. JAMA 1991;265:217–21.
7. Ventura SJ, Martin JA, Taffel SM, Mathews MS, Clarke SC. Advance report of final natality statistics, 1992. Mon Vital Stat Rep 1994;43(suppl):39–46.
8. Stroup-Benham CA, Trevino FM. Reproductive characteristics of Mexican-American, mainland Puerto-Rican, and Cuban-American women. JAMA 1991;265:222–6.
9. U.S. Department of Health and Human Services. Healthy people 2000 review: National health promotion and disease prevention objectives. PHS no. 91-50212. Hyattsville, Maryland: US Department of Health and Human Services, 1990:373–5.
10. Koonin LM, MacKay AP, Berg CJ, Atrash HK, Smith JC. Pregnancy-related mortality surveillance—United States, 1987–1990. MMWR CDC Surveill Summ 1997;46:17–36.
11. Ventura SJ. Births of Hispanic parentage, 1979. Mon Vital Stat Rep 1982;31(suppl):1–6.
12. Ventura SJ. Births of Hispanic parentage, 1980. Mon Vital Stat Rep 1983;32(suppl):1–8.
13. Kotelchuck M. An evaluation of the Kessner Adequacy of Prenatal Care Index and a proposed adequacy of prenatal care utilization index. Am J Public Health 1994;84:1414–20.
14. Liao Y, Cooper RS, Cao G, Durazo-Arvizu R, Kaufman JS, Luke A, et al. Mortality patterns among adult Hispanics: Findings from the NHIS, 1986 to 1990. Am J Public Health 1998;88:227–32.
15. Cooper R. Health and the social status of blacks in the United States. Ann Epidemiol 1993;3:137–44.
16. Markides K, Coreil J. The health of Hispanics in the southwestern United States: An epidemiologic paradox. Public Health Rep 1986;101:253–65.
17. Valdeviesco R, Davis C. US Hispanics: Challenging issues for the 1990's. Washington, DC: Population Reference Bureau, 1988.
18. Abma J, Chandra A, Mosher W, Peterson L. Fertility, family planning and women's health: New data from the 1995 National Survey of Family Growth. Vital Health Stat 1997;23(19):21.
19. Estrada AL, Trevino FM, Ray LA. Health care utilization barriers among Mexican Americans: Evidence from HHANES 1982–84. Am J Public Health 1990;80(suppl):27–31.
20. Speckhard ME, Comas-Urrutia AC, Rigau-Perez JG, Adamson K. Risk factors associated with pregnancy-related deaths in Puerto Rico. Bol Asoc Med P R 1986;78:2–6.
21. Scott CL, Chavez GF, Atrash HK, Taylor DJ, Shah RS, Rowley D. Hospitalizations for severe complications of pregnancy, 1987–1992. Obstet Gynecol 1997;90:225–9.
22. Cunningham FG, MacDonald PC, Gant NF, McDonald PC, Leveno KJ, Gilstrap LC, et al. Williams obstetrics. 20th ed. Norwalk, Connecticut: Appleton & Lange, 1997.
23. Centers for Disease Control. Enhanced maternal mortality surveillance—North Carolina, 1988 and 1989. MMWR Morb Mortal Wkly Rep 1991;40:469–71.
24. Centers for Disease Control. Misclassification of maternal deaths—Washington state. MMWR Morb Mortal Wkly Rep 1986;35:621–3.