Sudden Death: Ectopic Pregnancy Mortality

Sudden Death: Ectopic Pregnancy Mortality Frank W. J. Anderson, MD, MPH, Joanne G. Hogan, PhD, and Rudi Ansbacher, MD, MS OBJECTIVE: To describe the trends in ectopic pregnancy mortality in Michigan from 1985 through 1999 and compare to those of previous time periods. METHODS: We reviewed all cases of maternal mortality from ectopic pregnancy in Michigan from 1985 through 1999. We extracted data from death certificates, hospital inpatient and records, medical examiner autopsy reports, and reviews by the Michigan Maternal Mortality Study. The Health Data Development Section of the Michigan Department of Community Health provided data on live births and maternal deaths RESULTS: Of the 268 pregnancy-related deaths, 16 (6%) were caused by complications of ectopic pregnancy. Mean age at death was 27 ( 6) years. Thirteen deaths were to African- women and 3 were to white women (P <.01). African- women had an ectopic mortality ratio 18 times higher than white women (3.25/100,000 live births, compared with 0.18/100,000) Three cases of pregnancy-related death due to complications of ectopic pregnancy were considered preventable, and 2 others were of unknown preventability. CONCLUSION: Ectopic pregnancy treatment has changed in the last 20 years coincident with a decrease in maternal mortality from ectopic pregnancy. Sudden death was the presenting scenario in 75% of nonpreventable ectopic deaths, an increase from previous analyses. A large racial disparity is apparent. Ideally, pregnancy care should start as soon as possible after the first missed menses; however, systemwide changes are needed to create a new norm promoting early access to pregnancy care and promoting education and testing to rule out pregnancy abnormalities. (Obstet Gynecol 2004;103:1218 23. 2004 by The College of Obstetricians and Gynecologists.) LEVEL OF EVIDENCE: II-2 Ectopic pregnancy is defined as a pregnancy that develops after implantation of the blastocyst anywhere other than the endometrium lining the uterine cavity. 1 It is the leading cause of pregnancy-related death in the first trimester, accounting for 9% of all pregnancy-related From the Department of Obstetrics and Gynecology, University of Michigan and the Michigan Department of Community Health, Ann Arbor, Michigan. The authors acknowledge the State of Michigan, Department of Community Health, for the continuous maintenance of the Michigan Maternal Mortality Study for the last 53 years and the Michigan Mortality Review Committee. deaths. 2 In the United States, the incidence of ectopic pregnancy increased from 4.5/1,000 pregnancies in 1970 to 19.7/1,000 pregnancies in 1992. 3 A Swedish hospitalbased study found that ectopic pregnancy rates increased from 7.7/1,000 pregnancies in 1970 1975 to a high in 1985 1989 of 16.6/1,000 pregnancies and then decreased by 30% in 1997. 4 In the United States, during the period 1970 1989, hospitalizations for ectopic pregnancy increased 4-fold, from 4.5/1,000 to 16.0/1,000 5 ; however, hospitalization data from 1990 to 1992 demonstrated a decrease to 1.14/1,000 pregnancies. 6 During the period 1970 1989 and despite the rise in the incidence of ectopic pregnancy, the risk of death associated with ectopic pregnancy in the United States decreased by 90%. 5 The case fatality rate declined from 35.5 deaths per 10,000 ectopic pregnancies in 1970 to 3.8/10,000 in 1989. 5 The contribution from ectopic pregnancy to maternal mortality in Michigan increased from 1950 through 1985. During the 1950 1969 period, mortality from ectopic pregnancy in Michigan constituted 6.8% of direct maternal deaths 7 ; during the 1975 1984 period, 15.3% of direct maternal deaths were due to ectopic pregnancy. 8 In a review of 21 deaths due to ectopic pregnancy in Michigan from 1975 to 1984, Ansbacher et al 8 found that 85.7% of the deaths resulted after some delay on the part of the patient in seeking medical attention and the physician in ordering appropriate diagnostic tests. Inappropriate diagnostic tests by physicians led to treatment delay in 15 patients. Three women had received no medical attention during the pregnancy and were found dead at home. Three others who had not previously sought medical care died in the hospital room. Of the 15 women who had received some medical attention, 3 were found dead at home within hours of the physician s office or visit. The remaining 12 died in the hospital, 5 after their second or third visit or hospital admission. The authors concluded that early diagnosis, aggressive surgical management, and adequate blood replacement could have been instrumental in preventing maternal death in most of the cases. The authors noted that highly sensitive testing for human chorionic gonadotropin had become available in 1980 but had not been used in any of VOL. 103, NO. 6, JUNE 2004 1218 2004 by The College of Obstetricians and Gynecologists. 0029-7844/04/$30.00 Published by Lippincott Williams & Wilkins. doi:10.1097/01.aog.0000127595.54974.0c

the reviewed cases. Only 2 patients underwent diagnostic laparoscopy, and none had ultrasonography. In the current study, maternal mortality from ectopic pregnancies that occurred from 1985 through 1999 in Michigan is reviewed. Our goal was to describe the trends in ectopic pregnancy mortality in Michigan from 1985 through 1999 and compare them with those of previous time periods. We examined epidemiologic patterns and details of the presentation and clinical course that have emerged as a result of changing clinical management patterns. MATERIALS AND METHODS We examined data for all deaths in Michigan due to ectopic pregnancy that were identified by the Michigan Maternal Mortality Study in the 1985 1999 period. Approval for this study was obtained from the University of Michigan institutional review board and the approval body at the Michigan Department of Community Health. Systematic surveillance of maternal mortality in Michigan began in 1950 as a collaborative effort among the Michigan Department of Community Health, the Michigan State Medical Society, and the medical schools in Michigan. The Michigan Maternal Mortality Study, which is administered by the Michigan Department of Community Health, is designed to document trends in the causes, characteristics, and avoidable and modifiable factors contributing to maternal mortality and to improve the teaching and practice of obstetrics in Michigan. 9 13 The Michigan Maternal Mortality Study follows the definition adopted by the College of Obstetricians and Gynecologists and the Centers for Disease Control and Prevention for classifying a maternal death. 14 A pregnancy-associated death is the death of a woman from any cause while she is pregnant or within 1 year after the end of the pregnancy. A pregnancy-related death is due to a complication of the pregnancy, a chain of events initiated by the pregnancy, or an aggravation of an unrelated condition by the physiologic or pharmacologic effects while pregnant or within 1 year after the end of the pregnancy. Pregnancy-related mortality ratios are defined as the number of pregnancy-related deaths per 100,000 live births. The Michigan Maternal Mortality Study uses an active multisource case ascertainment system, 15 which includes reporting by hospitals, medical examiners, and other health and social service providers, and newspaper obituaries. The Health Data Development Section of the Michigan Department of Community Health identifies deaths by electronic search and computerized match of each year s birth certificates to the death certificates of women of reproductive age for that year and the following year to identify late maternal deaths. 16 The major limitations to identifying every maternal death 17 are 1) the Michigan Maternal Mortality Study does not review hospital records and autopsy reports of all women of reproductive age; 2) by statute, death certificates are not matched to fetal death reports; and 3) the pregnancy check box 18 on the death certificate had not been implemented during the time of this study. An evaluation of the Michigan Maternal Mortality Study found that a multisource reporting system was superior to systems based on death certificate identification alone. 11 The Michigan Maternal Mortality Study is less likely to have missed deaths due to ectopic pregnancy because sudden deaths are referred to a medical examiner to establish cause of death on the death certificate, and the Health Data Development Section of the Michigan Department of Community Health sends a query to the certifying physician for deaths due to hemorrhage in women of reproductive age to determine if the woman was pregnant. The Michigan Maternal Mortality Study manager assembles available hospital charts, medical examiner investigation and autopsy reports, death certificate and, when applicable, birth certificate information. A senior obstetrician reviews the material and creates a standardized summary of the case, which is presented to the Michigan Maternal Mortality Study medical review committee. This committee includes obstetricians, maternal fetal medicine specialists, pediatricians, an anesthesiologist, a pathologist, a nurse midwife, and representatives from Michigan Department of Community Health. The committee reviews and categorizes each death as to cause, preventability, and responsible factors (patient, physician, hospital or community), and determines if the death was pregnancy associated or pregnancy related. 19 For this study, the 2 physician authors who are members of the Michigan Maternal Mortality Study reviewed death certificates, hospital inpatient and records, reports of medical examiner investigations and autopsies, reports from medical services, and Michigan Maternal Mortality Study summaries and findings for each death due to ectopic pregnancy. We abstracted the following information: age, race, marital status, gravidity, parity, abortions, previous ectopic pregnancy, rural/urban residence, surgical history, nonsurgical history, initial presentation, symptoms, contributing causes leading to death, reasons for failure or delay, cause of death, and findings from the Michigan Maternal Mortality Study review process. The Health Data Development Section of the Michigan Department of Community Health provided data on live births and maternal deaths. We used the t and 2 VOL. 103, NO. 6, JUNE 2004 Anderson et al Ectopic Pregnancy Mortality 1219

Table 1. Total Births, Pregnancy-Related Deaths, and Ectopic Pregnancies in Michigan, 1985 1999 Year Total births Pregnancyrelated deaths Ectopic deaths Ectopic morality ratio* Ectopic deaths as percentage of pregnancy-related deaths 1950 1954 885,413 557 25 2.8 4.5 1955 1959 1,011,841 492 23 2.3 4.7 1960 1964 924,645 417 21 2.3 5.0 1965 1969 819,832 269 15 1.8 5.6 1970 1974 759,729 190 16 2.1 8.4 1975 1979 686,979 121 13 1.9 10.7 1980 1984 692,499 76 8 1.2 10.5 1985 1989 703,943 81 6 0.9 7.4 1990 1994 723,789 107 6 0.8 5.6 1995 1999 668,027 80 4 0.6 5.0 * Ectopic deaths per 100,000 live births. tests within SPSS 11.5.0 for Windows (SPSS Inc, Cary, NC) for all analysis and considered a P value of less than or equal to.05 as significant. RESULTS In Michigan, from 1985 through 1999, there were 2,095,759 resident live births (Table 1). White live births comprised 78.8% (1,651,817), African- live births 19.0% (400,054), and the remaining 2% (43,888) were to Native- and Asian/Pacific-Island women. The median maternal age was 26 years. During the 15-year period, there were 704 pregnancy-associated deaths for a pregnancy-associated mortality ratio of 47/ 100,000 live births. After review of medical charts and other information, the Michigan Maternal Mortality Study medical review committee determined that 268 deaths were pregnancy related (12.8/100,000 live births). White and African- women each accounted for 49% of the pregnancy-related deaths. Of the 268 pregnancy-related deaths, 16 (6.0%) were caused by complications of ectopic pregnancy (0.8/ 100,000 live births). The age range at death from ectopic pregnancy was 14 36 years, with a mean of 27 ( 6) years, which was not significantly different from age at death for deaths not related to ectopic pregnancy (28 6 years). African- women had an ectopic pregnancy mortality ratio of 3.25/100,000 live births whereas among white women, that ratio was 0.182/100,000 live births. Thirteen of the ectopic deaths were in African- women and 3 in white women (P.01). African- women with an ectopic pregnancy were 18 times more likely to die of ectopic pregnancy related complications than white women. Table 2 presents case review information. In all cases, the ectopic gestation had implanted in the fallopian tube. There were no cases of abdominal, cervical, or ovarian implantations. Eleven cases were determined to be in the first trimester. In the remaining 5 cases, the gestational age was unknown, and there were no indications that a more advanced gestational age contributed to the death. Of the 16 cases, 2 misdiagnosed cases and 1 case of an medical services delay were judged preventable deaths by the Medical Review Committee. The preventable case scenarios are: Case 2. In 1986, a 22-year-old woman of unknown parity with a history of sickle cell anemia presented to the room with abdominal pain and nausea. She was treated for what was termed flu and sickle cell crisis for approximately 20 hours until she collapsed and died. The autopsy revealed an 11-week ruptured ectopic pregnancy. Case 3. In 1987, a 34-year-old, gravida 4, para 0 woman presented with abdominal pain and weakness. A negative culdocentesis steered the diagnosis toward septic shock, intracranial bleeding and myocardial infarction. The patient arrested less than 90 minutes after arrival. Case 16. In 1998, a 35-year-old, gravida 2, para 1 woman experienced a 58-minute lapse between the medical services call from the patient and her arrival in the room. At her home, medical services personnel noted an alert female with a blood pressure of 110/70 mm Hg. At the room, her heart rate was 34 beats per minute with an agonal rhythm. She died after 54 minutes of resuscitative efforts. She had not sought medical care because she had some abdominal pain with a previous pregnancy. In 2 cases, the committee was not able to determine preventability. These case scenarios are: Case 4. In 1988, a 24-year-old woman had contact with an /urgent care center 2 days before her death and was diagnosed with influenza and unknown parity. She presented dead on arrival to the. 1220 Anderson et al Ectopic Pregnancy Mortality OBSTETRICS & GYNECOLOGY

Table 2. Description of Ectopic Deaths in Michigan, 1985 1999 Case Location Age (y) Race Gravity Parity Abortion 1 33 African 4 First trimester Gestational age Year Clinical scenario Error 1986 Full arrest on arrival in Medical contact this pregnancy 2 22 African 11 wk* 1986 Missed diagnosis Yes 3 Urban 34 African 4 0 3 4 wk 1987 Missed diagnosis Yes 4 Urban 29 Caucasian 7 9 wk* 1988 Dead on arrival in Flu diagnosed 2 days prior 5 Urban 24 African 6 2 3 First trimester 1988 Dead on arrival in 6 Urban 31 African 7 Urban 27 African 2 1 0 12 wk 1989 Dead on arrival in 6 7 wk* 1991 Died in, full arrest 8 Rural 25 Caucasian 2 1 0 1991 Died in, full arrest 9 Urban 19 African 6 8 wk* 1991 Died in, full arrest 10 Urban 36 Caucasian 2 1 0 6 7 wk 1991 Died in, full arrest, heterotopic 11 Urban 32 African 4 6 wk* 1992 Found dead at home 12 Urban 22 African 1994 Found dead at home 13 Urban 22 African 14 Urban 14 African 15 Urban 22 African 16 Urban 35 African * Reported in the autopsy results. 2 0 1 1995 Resuscitated in, died in operating 1 0 0 1997 Dead on arrival in 5 6 wk* 1997 Found dead at home 2 1 0 6 8 wk* 1998 Died in full arrest, delay medical services Positive home pregnancy test result Yes IUP Yes Patient knew she was pregnant, unknown prenatal visit Patient knew she was pregnant 3 weeks before she died Case 10. In 1991, a 36-year-old, gravida 2, para 1 woman with a known intrauterine pregnancy developed acute onset of lower abdominal pain. Her physician was waiting for her in the of 1 hospital while medical services transported her to the closest facility, where she died during resuscitative efforts. Heterotopic pregnancy was confirmed at autopsy. In the remaining 11 cases, 3 were found dead at home, 3 were dead on arrival in the, 4 died in the, and 1 was resuscitated in the room but died in the operating room. Two of the 11 women knew that they were pregnant, but no prenatal care was documented. prenatal care was sought during the index pregnancy in 4 of these cases, and prenatal care could not be deter- VOL. 103, NO. 6, JUNE 2004 Anderson et al Ectopic Pregnancy Mortality 1221

mined in the other 7. Based on the scenarios of those cases, it is suspected that most had not sought medical care. DISCUSSION The diagnosis and treatment of ectopic pregnancy has changed radically in the last 20 years. 20 24 In the past, it was common to diagnose an ectopic pregnancy in the after rupture. Massive hemoperitoneum was the presenting sign, treated by surgery and fluid and blood replacement. Currently, it is more common to treat the small and unruptured tubal pregnancy through minimally invasive surgery, or by nonsurgical means. The availability of sensitive and specific radioimmunoassays for human chorionic gonadotropin and serum progesterone levels and the development of high-resolution transvaginal ultrasonography have allowed for a more complete understanding of the natural history of early ectopic pregnancies and are effective screening tools for the detection of ectopic pregnancy. A number of novel surgical and nonsurgical approaches for treatment have also been developed that allow for the prerupture treatment that decreases the risk of catastrophic rupture and improves the chance of a future patent fallopian tube. 21 24 It is apparent from our analysis that improvements in the care of women with known ectopic pregnancy have resulted in decreased risk and mortality. In this study, the risk of death from ectopic pregnancy in Michigan and the percentage of maternal deaths due to ectopic pregnancy have declined in association with improvements in treatment. The ectopic mortality ratio has declined from 2.8/100,000 live births in 1950 1954 6 to 0.6 for 1995 1999 (Table 1). During the same period, the percentage of deaths from ectopic pregnancy in Michigan has decreased. After a high of 15.3% in 1975 1984, the 5.9% ectopic deaths as a percentage of total pregnancy-related deaths in the 1985 1999 period is comparable with the rate in the 1950 1969 period of 6.8%. The clinical scenario around which an ectopic pregnancy occurs in Michigan has also changed, with a decrease in errors and an increase in sudden death. Whereas 85.7% of cases of ectopic death during the 1975 1984 period were due to some medical error, 8 in the 1985 1999 period, 18.7% were from medical errors. Sudden death as the presenting sign is now the most common clinical scenario. In the 1975 1984 period, 28.5% of ectopic deaths involved women who were found dead at home or who died in the, 8 whereas in the 1985 1999 period, 75% of nonpreventable or unknown-if-preventable ectopic deaths involved women who were found dead at home, were dead on arrival, or died in the during resuscitative efforts. The majority of women who died in the most recent analysis had unknown ectopic pregnancies and died a sudden death at home, in the ambulance, or in the room during resuscitative efforts. These women most likely did not recognize their early pregnancy and dismissed symptoms they may have had, resulting in tubal rupture and its consequences. Racial disparities are especially high in this series. African- women were 18 times more likely to die of complications of ectopic pregnancy than white women. Racial disparities in the overall maternal mortality ratio have been persistent. In 1990, the maternal mortality ratio for African- women was 3.3 times greater than for white women. 25 Further study of this relationship, especially the contribution to this disparity from ectopic pregnancy, is warranted. The establishment of a fertilized ovum into tissue, whether it be uterine or other, defines a period of increased risk. Ectopic pregnancies are easily identifiable through simple screening tests and can be effectively treated before rupture. It is important for all women of reproductive age, especially African- women, to have an awareness of ectopic pregnancy and its symptoms and consequences to aid in early diagnosis. The large racial disparity identified in this study underscores the need to educate both providers and patients about risk factors for ectopic pregnancy, the early signs and symptoms, and the actions to take when they occur. Any abnormality in menstrual flow combined with abdominal discomfort or pain should initiate contact by the woman with a health care provider. It is equally important that providers caring for women of reproductive age include menstrual patterns in their history taking and liberally utilize the quantitative test for -hcg and vaginal probe ultrasonography to rule out ectopic pregnancy. Ideally, pregnancy care begins during the preconceptual period. 14 In prenatal care research, early prenatal care is variably defined as that occurring before 12 or up to 20 weeks gestation. 26,27 Delayed prenatal care is defined as that which occurs after 12 weeks of gestation. 28 In practice, enrollment into formal prenatal care is often not initiated until the establishment of a viable intrauterine pregnancy. A change should be made whereby access to care at the first sign of pregnancy (ie, after the first missed period or any suspicion of pregnancy) becomes the norm so that education about all early pregnancy complications can be delivered and screening tests for ectopic pregnancy can be used as indicated. This should reduce the unattended sudden death from ectopic pregnancy. 1222 Anderson et al Ectopic Pregnancy Mortality OBSTETRICS & GYNECOLOGY

REFERENCES 1. Mishell DR. Ectopic pregnancy etiology, pathology, diagnosis, management, fertility prognosis. In: Stenchever MA, Droegmueller W, Herbst AL, editors. Comprehensive gynecology. 4th ed. St. Louis (MO): Mosby; 2001. 2. Centers for Disease Control and Prevention. Ectopic pregnancy surveillance, United States, 1970 1987. MMWR CDC Surveill Summ 1990;39(SS-4):9 17. 3. Ectopic pregnancy: United States, 1990 1992. JAMA 1995;273:533. 4. Kamwendo F, Forslin L, Bodin L, Danielsson D. Epidemiology of ectopic pregnancy during a 28-year period and the role of pelvic inflammatory disease. Sex Transm Infect 2000;76:28 32. 5. Centers for Disease Control and Prevention. Surveillance for ectopic pregnancy: United States, 1970 1989. MMWR CDC Surveill Summ 1993;42:73 85. 6. Centers for Disease Control and Prevention. Current trends ectopic pregnancy: United States, 1990 1992. MMWR Morb Mortal Wkly Rep 1995;44:46 8. 7. Schneider J, Berger CJ, Cattell C. Maternal mortality due to ectopic pregnancy: a review of 102 deaths. Obstet Gynecol 1977;49:557 61. 8. Ansbacher R, Mills EM, Thrush JC, Stevenson LD. Ectopic pregnancy and maternal mortality in Michigan. Am J Gynecol Health 1989;3:118 23. 9. Mills EM. Maternal Mortality Study. Spectrum 1988;1: 1 2. 10. Mills EM. The Michigan Maternal Mortality Study: review of activities and findings from the 1980 preliminary report. Lansing (MI): Michigan Department of Public Health; 1990. 11. Endler G, Mariona F, Sokol R. Maternal death as a result of anesthesia: the Michigan experience. Mich Med 1988; 88:28 9. 12. Lockett C. An evaluation of the Maternal Mortality Surveillance System in Michigan. Lansing (MI): Michigan Department of Community Health; 2000. 13. Hogan JG. The Michigan Maternal Mortality Study: EPI- Insight. Lansing (MI): Michigan Department of Community Health Bureau of Epidemiology; 2001. 14. College of Obstetricians and Gynecologists, Academy of Pediatrics. Guidelines for perinatal care. 5th ed. Washington (DC): College of Obstetricians and Gynecologists, Academy of Pediatrics; 2002. 15. 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 5. 16. Jocums S, Mitchel EF, Entman SS, Piper JM. Monitoring maternal mortality using vital records linkage. Am J Prev Med 1995;11:75 8. 17. Horon IL, Cheng D. Enhanced surveillance for pregnancy-associated mortality: Maryland, 1993 1998. JAMA 2001;285:1455 9. 18. Mackay AP, Rochat R, Smith JC, Berg CJ. The check box: determining pregnancy status to improve maternal mortality surveillance. Am J Prev Med 2000;19(1S):35 9. 19. Hogan JG. Live births and maternal morality, Michigan, 1990 1999. Lansing (MI): Michigan Department of Community Health; 2003. 20. Pisarska MD, Carson SA, Buster JE. Ectopic pregnancy. Lancet 1998;351:1115 20. 21. Lipscomb GH, Stovall TG, Ling FW. nsurgical treatment of ectopic pregnancy. N Engl J Med 2000;343: 1325 9. 22. Lemus JF. Ectopic pregnancy: an update. Curr Opin Obstet Gynecol 2000;12:369 75. 23. Tulandi T, Semour A. Evidence-based management of ectopic pregnancy. Curr Opin Obstet Gynecol 2000;12: 289 92. 24. Buster JE, Heard MJ. Current issues in medical management of ectopic pregnancy. Curr Opin Obstet Gynecol 2000;12:525 52. 25. Centers for Disease Control and Prevention. Differences in maternal mortality among black and white women: United States, 1990. MMWR Morb Mortal Wkly Rep 1995;44:6 7, 13 14. 26. Centers for Disease Control and Prevention. Adequacy of prenatal-care utilization California, 1984 1994. MMWR Morb Mortal Wkly Rep 1996;45:653 6. 27. Johnson AA, El-Khorazaty MN, Hatcher BJ, Wingrove BK, Milligan R, Harris C, et al. Determinants of late prenatal care initiation by African women in Washington DC. Maternal Child Health J 2003;7:103 14. 28. Centers for Disease Control and Prevention. Entry into prenatal care: United States, 1989 1997. MMWR Morb Mortal Wkly Rep 2000;49:393 8. Address reprint requests to: Frank W. J. Anderson, 1500 East Medical Center Drive L4000 WH, Ann Arbor, MI 48109; e-mail: fwja@umich.edu. Received December 10, 2003. Received in revised form February 19, 2004. Accepted March 3, 2004. VOL. 103, NO. 6, JUNE 2004 Anderson et al Ectopic Pregnancy Mortality 1223