Pregnancy-related morbidity and mortality due to hemorrhage continue to be a serious public health burden. Hemorrhage has been identified as the leading cause of maternal death in the United States, accounting for 30.2% of maternal deaths in 1979–1986 and 28.7% in 1987–1990.1,2 During 1986 and 1987, there were 1.7 million nondelivery, pregnancy-related hospitalizations in the United States (22.2 per 100 births); 65.8% were antenatal hospitalizations, and 34.2% were pregnancy loss hospitalizations (ectopic pregnancies; molar pregnancies; induced, spontaneous, missed, and unspecified abortions); 102,000 (9.1%) of the antenatal hospitalizations were due to early pregnancy hemorrhage, not including pregnancies with abortive outcomes (ectopic pregnancies, spontaneous, missed, and unspecified abortions).3 For 1991–1992, the ratio of pregnancy-related hospitalizations dropped to 16.7 per 100 births, and the proportion of antenatal hospitalizations due to hemorrhage dropped to 7.3%.4
We analyzed data from the Pregnancy Mortality Surveillance System of the Centers for Disease Control and Prevention (CDC) to examine the incidence, trends, characteristics, and risk factors of pregnancy-related deaths caused by hemorrhage for 1979–1992.
Materials and Methods
The 52 reporting areas (health departments in the 50 states, the District of Columbia, and New York City) provided CDC with copies of death certificates, and when available, matched live birth certificates or fetal death records for identified pregnancy-related deaths from 1979–1992. Deaths were classified as pregnancy-related if they occurred during pregnancy or within 1 year after pregnancy termination, and resulted from complications of pregnancy itself, a chain of events initiated by pregnancy, or aggravation of an unrelated condition by the physiologic effects of pregnancy.1,2 Deaths were considered potentially pregnancy-related if pregnancy status boxes on death certificates were marked, pregnancy status was indicated by language on death certificates (ie, one of the pregnancy key words was used to describe the woman or cause of death), a matched birth certificate or a fetal death record indicated pregnancy within the previous year, or other supplementary materials indicated pregnancy (such as notes on or added to the death certificate, autopsy reports, or maternal mortality committee reports).1,2 Medical epidemiologists reviewed the information available on all deaths and determined their status accordingly (whether they were pregnancy-related or not), and coded each of them appropriately. Matched outcome records (birth certificates and fetal death records) were available for 95% of deaths after live births, and 86% of deaths after stillbirths. For the 763 deaths due to hemorrhage (study population), matched records were available for 94% of deaths after live births and 90% of deaths after stillbirths.
We limited our analysis to deaths from hemorrhage associated with intrauterine pregnancies, excluding 511 deaths due to ruptured ectopic pregnancies because different risk factors might precede the onset of that kind of pregnancy and because ectopic pregnancy has its own management issues.
Pregnancy-related mortality ratios were calculated as pregnancy-related deaths per 100,000 live births.1,2 The live birth data used to calculate pregnancy-related mortality ratios were obtained from the 1979–1992 national natality files compiled by the CDC's National Center for Health Statistics.5
For analysis by race, women's race was defined as reported on death certificates for the numerator (deaths). For the denominator (live births), race was defined as that of the mother. Women were classified as white, black, or other (Asians or Pacific Islanders, American Indians, Alaska natives, and women reported as “other”). Women whose ethnicity was reported as Hispanic were classified as black, white, or other based on their reported race. Maternal age was grouped into 5-year intervals, from 10–14 to 45–49, but the 10–14 and 15–19 age groups and the 40–44 and 45–49 age groups were combined because of small numbers.
A total of 4915 pregnancy-related deaths were reported to the CDC from 1979–1992. The leading causes of pregnancy-related deaths were hemorrhage (29.0%), embolism (22.3%), and pregnancy-induced hypertension (17.0%). Deaths due to hemorrhage included 511 deaths associated with ectopic pregnancies and 763 deaths associated with intrauterine pregnancies. Our analysis was limited to the 763 deaths associated with intrauterine pregnancies.
The pregnancy-related mortality ratio due to hemorrhage from intrauterine pregnancy was 1.4 deaths per 100,000 live births. Black women and women of other races were at greater risk of pregnancy-related death due to hemorrhage than white women. Among black women, the risk of death from hemorrhage was three times that of white women; for women of other races, the risk was 2.8 times that of white women. Black women had a higher pregnancy-related mortality ratio due to hemorrhage than white women for every year of the study (Table 1). In black women, 13.4% of all pregnancy-related deaths were due to hemorrhage; in white women, 16.0% of all pregnancy-related deaths were due to hemorrhage; and for women of other races, 30.3% of pregnancy-related deaths were due to hemorrhage.
The risk of pregnancy-related death due to hemorrhage increased with maternal age (Table 2). The risk of death for women 40–49 years of age was nearly 13 times that of women 24 years of age and younger. The proportion of all pregnancy-related deaths attributable to hemorrhage increased with maternal age. In women younger than 20 years of age, 10.4% of pregnancy-related deaths were due to hemorrhage. The proportion increased consistently to 21.0% for women 40–49 years of age.
For women of all races, the pregnancy-related mortality ratio due to hemorrhage increased with age. Age-specific pregnancy-related mortality ratios were higher for black women than white at all ages (Table 3). The disparity in the risk of death between black and white women became more pronounced for women 25 years of age and older.
Fifty-six percent of all pregnancy-related deaths due to hemorrhage resulted in a live birth. Other common pregnancy outcomes were stillbirth (15.5%), abortion (6.9%), and undelivered pregnancy (6.3%) (Table 4). The most frequent specific cause of death due to hemorrhage in the study population was abruptio placentae (18.5%), followed by laceration and uterine rupture (16.4%), and uterine atony and postpartum bleeding (15.1%). The specific causes of pregnancy-related deaths due to hemorrhage differed by pregnancy outcome. For women who died after live births, the leading cause was atony and postpartum bleeding. For those whose pregnancies ended in stillbirth, it was abruptio placentae. Laceration and uterine rupture were the leading causes of death among women whose pregnancies ended in abortion, and among women who were still pregnant at the time of death (undelivered pregnancies).
The specific cause of pregnancy-related death due to hemorrhage varied by age group and race (data not shown). Uterine atony and postpartum bleeding were the leading causes of death for women younger than 20 years of age and women 40–49 years of age. Abruptio placentae was the leading cause of death for women age 20–24 and 30–34 years of age. Laceration and uterine rupture were the leading causes of death for the remaining age groups (25–29 years and 35–39 years). Among white women, abruptio placentae was the leading cause of pregnancy-related death, but laceration and uterine rupture ranked first among black women, and uterine atony and postpartum bleeding were first for women of other races.
Eighty-nine percent of the 763 pregnancy-related deaths due to hemorrhage occurred in hospitals (678 cases), and 3.4% of deaths occurred at home; 1.4% of cases were dead on arrival at hospitals. The place of death was unknown for 5.4% of cases. Abruptio placentae was the leading cause of deaths in hospitals (18.9%), and laceration and uterine rupture were the leading causes of deaths at home (30.8%).
The Pregnancy Mortality Surveillance System provides a unique data set for studying maternal deaths in the United States. The only other national source of information on pregnancy-related mortality for the United States are routinely collected vital (mortality) statistics. Matching of birth and fetal death records with maternal death certificates provides the system with useful information about pregnancy-related deaths. For pregnancies resulting in live births, such information includes initiation of prenatal care, mother's education, gravidity, parity, method of delivery, and complications of pregnancy, labor, and delivery. Unfortunately, that information is often missing. For example, in more than 25% of cases in our study, the mother's level of education and the trimester in which prenatal care began could not be determined.
The United States Public Health Service identified pregnancy-related mortality as an important public health problem and set specific goals for its reduction. Those goals call for an overall maternal mortality ratio of no more than 3.3 deaths per 100,000 live births and a ratio no greater than five deaths per 100,000 live births among black women.6 However, as of 1995, vital statistics records showed that the goals are still far from being met. The overall maternal mortality ratio from vital statistics in 1995 was 7.1 deaths per 100,000 live births. The maternal mortality ratio for black women (22.1 deaths/100,000 live births) was five times greater than for white women (4.2 deaths/100,000 live births).7 Several reports indicated that pregnancy-related mortality is underreported both in developing and developed countries, including the United States. However, we believe that underreporting should not affect our conclusions because there was no reason to believe that deaths due to hemorrhage are more or less likely to be reported if they occurred among specific groups of women (differential underreporting).8
Our analysis indicated that hemorrhage continues to be the leading cause of pregnancy-related deaths. Although many risk factors have been identified for hemorrhage, it remains hard to predict. Many women in whom postpartum hemorrhage develops, for example, have no known preexisting risk factors.9 In some cases, hemorrhage might occur intra-abdominally, or in other cases, a slow persistent blood loss might go on unnoticed until it develops into unexpected severe complications.9 Reviews of the literature indicate that most of the pregnancy-related deaths due to hemorrhage are preventable with early diagnosis, effective management, and proper preventive measures in place.9,10
Our findings are consistent with previous reports of increased risk of pregnancy-related death due to hemorrhage with age. For example, Jouppila11 found that advanced maternal age was associated strongly with an increasing risk of hemorrhage, and Iyasu et al12 reported an increasing risk of placenta previa with advanced age. Saftlas et al13 identified women 30 years of age or older and those younger than 20 years of age as at increased risk of abruptio placentae.
Our study provided more evidence of the racial gap in pregnancy-related mortality. Narrowing that gap will continue to be a challenge. Other studies found that the black-white gap in risk of overall maternal mortality persists after controlling for parity, education, and marital status.14 In addition, studies on causes of obstetric hemorrhage (such as placenta previa and abruptio placentae) have indicated a greater prevalence among black women than white women. From 1979–1987, Saftlas et al13 reported an increase in incidence ratio of abruptio placentae among women of all racial groups, with a 50% increase among black women compared with a 20% increase among white women. The available data did not allow an in-depth investigation of risk factors, so the reasons for racial disparity in pregnancy-related mortality ratios could not be determined. More in-depth studies on this subject should be conducted. Many researchers believe that race acts as an indicator of other risk factors, such as access to health care, quality of care, and financial and social status.14 Enhancing and modifying existing data sources (eg, death certificates, live birth certificates, and fetal death records) might help identify risk factors that have gone unnoticed.
1. Atrash HK, Koonin LM, Lawson HW, Franks AL, Smith JC. Maternal mortality in the United States, 1979–1986. Obstet Gynecol 1990;76:1055–60.
2. Berg CJ, Atrash HK, Koonin LM, Tucker M. Pregnancy-related mortality in the United States, 1987–1990. Obstet Gynecol 1996;88:161–7.
3. Franks AI, Kendrick JS, Olson DR, Atrash HK, Saftlas AF, Moien M. Hospitalization for pregnancy complications: United States, 1986 and 1987. Am J Obstet Gynecol 1992;166:1339–44.
4. Bennett TA, Kotelchuck M, Cox CE, Tucker MJ, Nadeau DA. Pregnancy-associated hospitalization in the United States in 1991 and 1992: A comprehensive view of maternal morbidity. Am J Obstet Gynecol 1998;178:346–54.
5. National Center for Health Statistics. Vital statistics of the United States, 1992, vol. 1, natality. Washington, DC: United States Public Health Service, 1995.
6. United States Department of Health and Human Services, Public Health Service. Health people 2000: National health promotion and disease prevention objectives. Washington, DC: DHHS Publication No. (PHS) 91-50212, 1991.
7. Anderson RN, Kochanek KD, Murphy SL. Report of final mortality statistics, 1995. Monthly Vital Statistics Report; vol. 45 no. 11, suppl 2. Hyattsville, Maryland: National Center for Health Statistics, 1997.
8. Atrash HK, Alexander S, Berg CJ. Maternal mortality in developed countries: Not just a concern of the past. Obstet Gynecol 1995;86:700–5.
9. Norris TC. Management of postpartum hemorrhage. Am Fam Physician 1997;55:635–40.
10. World Health Organization. Postpartum hemorrhage module: Education material for teachers of midwifery. Geneva: World Health Organization, 1996.
11. Jouppila P. Postpartum hemorrhage. Curr Opin Obstet Gynecol 1995;7:446–50.
12. Iyasu S, Saftlas AK, Rowley DL, Koonin LM, Lawson HW, Atrash HK. The epidemiology of placenta previa in the United States, 1979 through 1987. Am J Obstet Gynecol 1993;168:1424–9.
13. Saftlas AF, Atrash HK, Rowley D. National trends in the incidence of abruptio placentae, 1979–1987. Obstet Gynecol 1991;78:1081–6.
14. Atrash HK, Rowley D, Hogue C. Maternal and perinatal mortality. Curr Opin Obstet Gynecol 1992;4:61–71.