PREVENTING MATERNAL DEATHS

Preventing maternal deaths in America

By Suzanne M. Cox, MPH, Sarah J. Kilpatrick MD, PhD, and Stacie E. Geller, PhD

How can you help reduce pregnancy-linked deaths? Provide basic prenatal care and screen for ectopic pregnancy, STDs, eclampsia, spousal abuse, suicidal behavior, depression, and substance abuse.

One disturbing statistic has barely budged since 1982: The rate of pregnancy-related maternal mortality in the United States continues to be seven or eight deaths for every 100,000 live births—that is, when deaths are defined as occurring within 42 days of delivery or pregnancy termination.^1,2 The statistic may not be as staggering or compelling as the maternal mortality rates abroad, but given that the rate here at home hasn't declined in nearly 22 years, it's high time we examined maternal deaths in new ways—a process that researchers at the Centers for Disease Control and Prevention and elsewhere have already begun.

Taking the issue of maternal mortality one step further, many researchers feel that the only way to reduce the broader category of pregnancy-associated deaths is for clinicians to address the issue of homicide by intimate partners. Our goal is to propose practical strategies for how physicians and others can prevent maternal deaths, as we review the current state of maternal mortality research, including definitions, causes, surveillance and reporting, and disparities in the risk of death in the US.

Pregnancy-related versus pregnancy-associated deaths

The numbers could be even higher when you consider that the traditional way we tally maternal mortality may not identify the true number of deaths. For one thing, definitions of maternal death change over time and between states, with the timeframe during which a "postpartum" woman is considered a maternal death varying from 42 to 90 days to 1 year. For another, data are gathered from death certificates and coroners' reports that vary from state to state. The Pregnancy Related Mortality Ratio (PRMR), the most commonly reported measure of maternal mortality, is the number of pregnancy-related deaths per 100,000 live births. Currently, the CDC defines pregnancy-related death as a death that occurs during pregnancy or within 1 year after pregnancy and is caused by pregnancy-related complications.

Two examples of pregnancy-related deaths caused by "complications of the pregnancy itself" are hemorrhage and eclampsia. Postsurgical infection is an example of a death caused by a chain of events set in motion by the pregnancy. A third cause of pregnancy-related death is aggravation of an unrelated condition by the physiologic or pharmacologic effects of the pregnancy—cardiac problems, for example.

Until 1999, the National Center for Health Statistics' cutoff point for maternal death was 42 days postpartum. But when death certificate reporting changed from ICD-9 to ICD-10 codes, they added a new category for deaths from maternal causes that occur more than 42 days after delivery or pregnancy termination.

To expand the definition of maternal death, the CDC has now defined a pregnancy-associated death as any death from any cause during pregnancy or within 1 year of delivery or termination of pregnancy, regardless of the duration or anatomical site of the pregnancy. Therefore, pregnancy-associated deaths can be classified as unrelated to pregnancy, although the exact nature of the relationship between these deaths and pregnancy has not been fully determined. Two causes of pregnancy-associated but unrelated death are injury (motor vehicle accident or homicide) and unrelated medical conditions that are not aggravated by pregnancy (cancer, for example).

Clinicians can play a key role in improving surveillance

Accurate surveillance is the first step to coming to grips with the persistent problem of maternal mortality. The Pregnancy Mortality Surveillance System (PMSS), established in 1987 by the CDC's Division of Reproductive Health, uses improved methods to record maternal deaths. PMSS, which uses a 1-year postpartum cut-off point, shows mortality ratios actually increasing from 10.3 deaths per 100,000 live births in 1991 to 13.2 deaths per 100,000 live births in 1999.³ These ratios, which are higher than the often-cited 7.5 per 100,000 live births mentioned earlier, show that pregnancy-related deaths may be underreported.

Clinicians who complete death certificates and report deaths to state or local agencies are the first line of defense in gathering better information. For example, in states that tally postpartum deaths up to 1 year, physicians need to more fully investigate whether a woman had been pregnant during the past year.

A shortcoming of the PRMR is that it doesn't include pregnancy-associated deaths unrelated to pregnancy. Yet pregnancy in and of itself may increase the chances of injuries leading to death, such as homicide and motor vehicle crashes, some evidence suggests.⁴ Studies in four populous states find death due to injury to be the most common cause of maternal death, accounting for 33% to 46% of all maternal deaths.^5-8 While pregnancy-related deaths may result from the physiological changes accompanying pregnancy, pregnancy-associated but unrelated deaths may result from the social and cultural conditions of pregnancy—making it imperative that we raise awareness and improve reporting of pregnancy-associated but unrelated deaths.

Like pregnancy-related deaths, underreporting of pregnancy-associated but unrelated deaths is also likely because the person reporting deaths may not know pregnancy status nor consider it important.⁹ Furthermore, pregnancy-related deaths may be misclassified if the death certificate fails to adequately reflect the role of pregnancy in the death. For example, while pregnancy might have aggravated a woman's cardiac complications, the death certificate may only note the cardiac condition and not the pregnancy. This occurs more often in states where maternal deaths are captured using the range of maternal-related ICD-9 codes from the death certificates. Some 16 states and New York City have started putting a maternal death check-box on death certificates, which in one study captured 124 out of 425 pregnancy-related deaths (29%) that would otherwise have been missed.¹⁰ Additionally, a new option in ICD-10 codes for classifying indirect causes of death as pregnancy-related may uncover more pregnancy-related deaths.

Causes of pregnancy-related and pregnancy-associated deaths

Embolism is the leading cause of pregnancy-related deaths, while injury (chiefly homicide) causes up to 46% of pregnancy-associated deaths.

Pregnancy-related deaths. Hemorrhage is a cause of pregnancy-related deaths that's directly related to pregnancy, while a pre-existing health problem like a heart condition would be indirectly related. According to PMSS data for 4,200 deaths from 1991 to 1999 showing pregnancy-related causes for all pregnancy outcomes, the top three causes were embolism (19.6%), hemorrhage (17.2%), and pregnancy-induced hypertension (15.7%), which, when combined, accounted for more than 50% of those maternal deaths.³ The next most frequent causes were infection (12.6%), cardiomyopathy (8.3%), cerebrovascular accident (5.0%), anesthesia (1.6%), other causes (19.7%) (such as cardiovascular, pulmonary, and neurological), and finally, unknown causes (0.7%).

The pregnancy outcome (live birth vs. stillbirth, abortion, etc.) sometimes—but not always—affect the above percentages. Most women who die in childbirth do so after delivering a live infant. (In that case, hemorrhage causes only 2.7% of maternal deaths after live births versus 17.2% of deaths from hemorrhage overall.) In contrast, for women who've had a stillbirth or an ectopic pregnancy, hemorrhage was the leading cause of death (21.1% and 93.3%, respectively). Women who've had an abortion also had high rates of hemorrhage (21.8%) and infection (33.9%).³

Pregnancy-associated deaths. Those due to injury account for up to 46% of maternal deaths.^5-8 As many studies indicate, further investigation of nonobstetric causes of maternal death could potentially reduce maternal deaths dramatically. Although only a few studies have analyzed causes of deaths from injury, those that have list homicide as the leading cause (36%–63%), followed by motor vehicle crashes (12%–32%), suicide (8%–13%), and drug overdose (7%–20%).^5-8

We must dispel the notion that these "nonobstetric" deaths are isolated from the condition of pregnancy. Abuse during pregnancy occurs in 4% to 8% of all pregnancies and raises the risk of homicide.⁵ Moreover, adolescent homicide victims are 3.7 times more likely to be pregnant than adult homicide victims.¹¹ The risk of becoming an attempted homicide victim or being killed is three times higher in women abused during pregnancy, with 5.7% of attempted homicide victims and 4.8% of murder victims pregnant at the time.¹² A homicide related to domestic violence may be related to the stress of pregnancy, and a suicide following pregnancy may be related to postpartum depression or an unwanted pregnancy.¹³ Other researchers found recent substance use reported in 48% of injury deaths.⁷ In Chicago, 60% of deaths due to motor vehicle crashes were potentially preventable due to failure to use a seatbelt.

As you might suspect, pregnancy-related deaths are higher among African-Americans than white women and higher in women aged 35 and older.^13-16

How can we keep pregnant women alive?

There's a lot we can do. We can educate patients. Among other things, we can screen for and treat ectopic pregnancy, preeclampsia, and sexually transmitted diseases. One team of investigators found that 54% of all pregnancy-associated deaths in Massachusetts between 1990 and 1999 (which included 46% of injury deaths and 16% of deaths from medical causes) could have been prevented.⁸ Others in Chicago found that 36% of maternal deaths between 1992 and 1999 were preventable, which included 37% of pregnancy-related deaths and 30% of unrelated deaths.¹⁷ Sachs and colleagues found that physicians could have prevented 41% of maternal deaths, and patients might have prevented 15%.¹⁸

We must find ways to eliminate the incomplete management of cases, such as the failure to follow through on diagnostic tests, to refer to a tertiary care center, or to take action in a timely way. Finally, researchers must consider system factors such as communication, policies, and procedures, and staffing. What factors affect these preventable events?

Through patient education, we can encourage women to recognize and seek care for symptoms of illness, to undergo adequate prenatal care, and to adhere to a clinician's instructions. One study concluded, for instance, that prenatal care might have prevented five of the seven deaths, including three cases of preeclampsia or eclampsia, one of sepsis after prolonged rupture of membranes, and one from complications of hypertension. This study shows that the essential tools of prenatal care that may improve survival are "blood pressure measurement, urine protein determination, and an obstetric-focused review of systems."¹⁹ It further found that women receiving enhanced prenatal care coordination had no survival advantage over those who were not and that private prenatal care versus public prenatal care made no difference. This suggests that it's not necessarily the kind of prenatal care that makes a difference, but rather basic services provided in a timely fashion.

Strategies for change

These include developing more effective review committees and addressing deaths due to injury, such as homicide.

Quality assurance. Rather than viewing departmental quality assurance as punitive, we should welcome analysis of adverse outcomes as an opportunity to review institutional processes and structure.²⁰ One investigator points out that staff workload, night and weekend admission, and number of patients in an inpatient unit may increase mortality rates, perhaps due to increased staff fatigue.²⁰ Systems reviews may reveal opportunities to rectify unique policy and procedure issues that only occasionally cause adverse outcomes, such as adding needed language translation services or improving communication between laboratories and Labor and Delivery. The CDC and partner agencies have published Strategies to Reduce Pregnancy-Related Deaths: From Identification and Review to Action, which tells how to develop more effective review committees that address possible nonmedical and system-related factors that can contribute to a maternal death.^21,22

Addressing death by injury. Perhaps the only way to significantly reduce pregnancy-associated death is to address deaths due to injury that are currently defined as unrelated to pregnancy. Ob/gyns, as prenatal-care providers, need to address the important topic of homicide by intimate partners. Although ACOG guidelines suggest routine screening for domestic violence during prenatal care, just 17% of providers do so at the initial visit and only 10% screen at follow-up visits, according to a recent survey.⁴ Publications are available to help clinicians identify, assess, and intervene in domestic violence situations.^23,24 So are helpful Web sites (Table 1), like the one sponsored by the Massachusetts Public Health Department (http://www.state.ma.us/dph/fch/safemoms/preg02rg.htm), which highlights the clinician's role in ensuring safe motherhood by:

• Screening and re-screening all pregnant and postpartum women at regular intervals for domestic violence and other physical, sexual, and emotional abuse; suicidal behavior and/or ideation; depression and postpartum depression; alcohol, tobacco and drug use; and routine and proper seatbelt use.

• Educating all pregnant and postpartum women about domestic violence, postpartum depression, substance use, and routine and proper seatbelt use.

TABLE 1
Web sites of interest

National Center for Health Statistics has definitions of maternal mortality rates and how the change from ICD-9 to ICD-10 has affected surveillance. http://www.cdc.gov/nchs/datawh/nchsdefs/rates.htm.

The CDC and several partner agencies have published Strategies to Reduce Pregnancy-Related Deaths: From Identification and Review to Action, which provides guidance on reporting but also on developing more effective review committees that address non-medical and system-related factors that may have contributed to a maternal death. http://www.cdc.gov/nccdphp/drh/pdf/Strategies.pdf.

CDC Division of Reproductive Health includes reproductive health highlights from all CDC divisions. http://www.cdc.gov/nccdphp/drh.

Overview of Safe Motherhood in the US including the problem of maternal mortality and programs to address it. http://www.cdc.gov/nccdphp/aag/aag_drh.htm.

The Massachusetts Public Health Department has a useful Web site (http://www.state.ma.us/dph/fch/safemoms/preg02rg.htm), which highlights the role of the provider in ensuring safe motherhood.

Violence prevention and screening tools

Improving the Health Care Response to Domestic Violence: A Resource Manual for Health Care Providers. 2nd ed. San Francisco, Calif: Family Violence Prevention Fund, 1998. http://endabuse.org/programs/display.php3?DocID=238.

Safety behaviors of abused women after an intervention during pregnancy by McFarlane et al. http://jognn.awhonn.org/cgi/content/abstract/27/1/64.

Danger Assessment by Jacqueline C. Campbell, PhD, RN, helps women decide if they are in danger of becoming a homicide victim. http://www.nnvawi.org/Danger.doc.

Including near-miss maternal deaths. Despite researchers' efforts to further reduce maternal mortality, we've made no progress in the last two decades in reducing the number of pregnancy-related deaths in the US. However, broadening the research focus to include near-miss maternal morbidity—severe morbidity likely to lead to death—and other severe morbidities can strengthen the study of maternal death.²¹ By identifying a larger number of women with serious, life-threatening complications of pregnancy, the study of maternal morbidity can be incorporated into clinical case review, resulting in a win/win situation. We can then better identify opportunities to prevent both near-miss morbidity and maternal death. And this might just help us answer the question of why some women with life-threatening complications die while others do not. To this end, we've published two papers on the development and application of a scoring system to identify near-miss maternal morbidity (see "Scoring system for assessing near-miss maternal morbidity").^25,26

Even though maternal mortality is relatively rare in the US, when combined with the 1.7 million US women who experience maternal morbidity every year, illness and injury around the time of pregnancy represents a significant disease burden.²⁷ By addressing both, we have the opportunity to prevent not only unnecessary deaths but also complications of morbidity.

REFERENCES

1. Maternal mortality—United States, 1982-1996. MMWR Morb Mortal Wkly Rep. 1998;47:705-707. Available at: http://www.cdc.gov/mmwr/preview/mmwrhtml/00054602.htm. Accessed November 14, 2003.

2. National Center for Health Statistics. Healthy People 2000 Final Review. Hyattsville, Md: Public Health Service, 2001. Available at: http://www.cdc.gov/nchs/data/hp2000/hp2k01.pdf. Accessed November 14, 2003.

3. Chang J, Elam-Evans LD, Berg CJ, et al. Pregnancy-related mortality surveillance—United Sates, 1991-1999. MMWR Surveill Summ. 2003;52:1-8. Available at: http://www.cdc.gov/mmwr/preview/mmwrhtml/ss5202al.htm. Accessed November 14, 2003.

4. Fry V. Examining homicide's contribution to pregnancy-associated deaths. JAMA. 2001;285:1510-1511.

5. Harper M, Parsons L. Maternal deaths due to homicide and other injuries in North Carolina: 1992-1994. Obstet Gynecol. 1997;90:920-923.

6. Fildes J, Reed L, Jones N, et al. Trauma: the leading cause of maternal death. J Trauma. 1992;32:643-645.

7. Dannenberg AL, Carter DM, Lawson HW, et al. Obstetrics: homicide and other injuries as causes of maternal death in New York City, 1987 through 1991. Am J Obstet Gynecol. 1995;172:1557-1564.

8. Nannini A, Weiss J, Goldstein R, et al. Pregnancy-associated mortality at the end of the twentieth century: Massachusetts, 1990-1999. J Am Med Womens Assoc. 2002;57:140-143.

9. Krulewitch CJ, Pierre-Louis ML, de Leon-Gomez R, et al. Hidden from view: violent deaths among pregnant women in the District of Columbia, 1988-1996. J Midwifery Womens Health. 2001;46:4-10.

10. MacKay AP, Rochat R, Smith JC, et al. The check box: determining pregnancy status to improve maternal mortality surveillance. Am J Prev Med. 2000;19 (1 suppl):35-39.

11. Krulewitch CJ, Roberts DW, Thompson LS. Adolescent pregnancy and homicide: findings from the Maryland Office of the Chief Medical Examiner, 1994-1998. Child Maltreat. 2003;8:122-128.

12. McFarlane J, Campbell JC, Sharps P, et al. Abuse during pregnancy and femicide: urgent implications for women's health. Obstet Gynecol. 2002;100:27-36.

13. Horon IL, Cheng D. Enhanced surveillance for pregnancy-associated mortality—Maryland, 1993-1998. JAMA. 2001;285:1455-1459.

14. Pregnancy-related deaths among Hispanic, Asian/Pacific Islander, and American Indian/Alaska Native Women—United States, 1991-1997. MMWR Morb Mortal Wkly Rep. 2001;50(18):361-364. Available at: http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5018a3.htm. Accessed November 14, 2003.

15. Press Release: CDC reports pregnancy-related deaths still higher in black women than white women. Centers for Disease Control and Prevention, 2003. Available at: http://www.cdc.gov/od/oc/media/pressrel/r030220c.htm. Accessed November 14, 2003.

16. Safe motherhood: promoting health for women before, during and after pregnancy. Centers for Disease Control and Prevention, 2003. Available at: http://www.cdc.gov/nccdphp/aag/aag_drh.htm. Accessed November 14, 2003.

17. Panting-Kemp A, Geller SE, Nguyen T, et al. Maternal deaths in an urban perinatal network, 1992-1998. Am J Obstet Gynecol. 2000;183:1207-1212.

18. Sachs BP, Brown DA, Driscoll SG, et al. Maternal mortality in Massachusetts. Trends and prevention. N Engl J Med. 1987;316:667-672.

19. Harper MA, Byrington RP, Espeland MA, et al. Pregnancy-related death and health care services. Obstet Gynecol. 2003;102:273-278.

20. Minkoff H. Maternal mortality in America: lessons from the developing world. J Am Med Womens Assoc. 2002;57:171-172.

21. Callaghan WM, Berg CJ. Maternal mortality surveillance in the United States: moving into the twenty-first century. J Am Med Womens Assoc. 2002;57:131-139.

22. Berg C, Danel I, Atrash H, et al, eds. Strategies to reduce pregnancy-related deaths: from identification and review to action. Atlanta: Centers for Disease Control and Prevention, 2001.

23. Warshaw C, Ganley AL. Improving the health care response to domestic violence: a resource manual for health care providers. 2nd ed. San Francisco, Calif; Family Violence Prevention Fund, 1998. Available at: http://endabuse.org/programs/printable/display.php3t? DocID=238. Accessed November 17, 2003.

24. McFarlane J, Parker B, Soeken K, et al. Safety behaviors of abused women after an intervention during pregnancy. J Obstet Gynecol Neonatal Nurs. 1998; 27:64-69. Available at: http://jognn.awhonn.org/ cgi/content/abstract/27/1/64. Accessed November 17, 2003.

25. Geller SE, Rosenberg D, Cox S, Kilpatrick S. Defining a conceptual framework for near-miss maternal morbidity. J Am Med Womens Assoc. 2002;57:135-139.

26. Geller SE, Rosenberg D, Cox SM, et al. A scoring system for identifying near miss maternal morbidity. J Clin Epidemiol. In Press.

27. Danel I, Berg C, Johnson C, et al. Magnitude of maternal morbidity during labor and delivery: United States, 1993-1997. Am J Pub Health. 2003;93:631-634.

MS. COX is a Research Specialist in the Department of Obstetrics/Gynecology and a PhD candidate in the School of Public Health at the University of Illinois at Chicago, Ill.

DR. KILPATRICK is a Professor of Obstetrics and Gynecology and Head of the Department of Ob/Gyn at the University of Illinois at Chicago.

DR. GELLER is an Associate Professor in the Department of Ob/Gyn and the Director of the National Center of Excellence in Women's Health at the University of Illinois at Chicago.

Take-home messages

• It may sound obvious, but to improve a pregnant woman's odds of survival, measure blood pressure, determine urine protein, and do an obstetric-focused review of systems.

• Routinely screen for domestic violence and emotional abuse during prenatal care to gauge the likelihood of a potential homicide attempt by a woman's intimate partner. (Adolescent homicide victims are 3.7 times more likely to be pregnant than adult victims.)

• Screen for and treat ectopic pregnancy, preeclampsia, and STDs. Also screen every pregnant and postpartum patient for suicidal behavior, depression, and substance abuse.

• Encourage women to have adequate prenatal care. Educate them about domestic violence, postpartum depression, and routine and proper seatbelt use.

Scoring system for assessing "near-miss maternal morbidity"

These tables illustrate a scoring system to identify near-miss maternal morbidity. The system's development is fully explained in previous publications.^1,2 Table A summarizes the five-factor system. A total score was calculated as the weighted sum of the clinical factors present for each woman; with a score of 8 or higher considered a "near miss." Near-miss morbidities identified using the scoring system can be incorporated into clinical case review and epidemiologic studies to better monitor obstetric care and to better estimate the number of life-threatening complications in pregnancy.

TABLE A
Scoring system used to identify near-miss morbidities

TABLE B
Organ system failure criteria*

REFERENCES

1. Geller SE, Rosenberg D, Cox S, Kilpatrick S. Defining a conceptual framework for near-miss maternal morbidity. J Am Med Womens Assoc. 2002;57:135-139.

2. Geller SE, Rosenberg D, Cox SM, et al. A scoring system for identifying near miss maternal morbidity. J Clin Epidemiol. In Press.

Stacie Geller, Sarah Kilpatrick, Suzanne Cox. Preventing maternal deaths in America.

Contemporary Ob/Gyn

Sep. 1, 2004;49:78-88.