Maternal Mortality In Utah

SCOTT JACOB, MD, LOIS BLOEBAUM RN, GULZAR SHAH, PhD
 AND MICHAEL W. VARNER, MD

Objective: To determine trends in maternal deaths in Utah, identify opportunities for preventive intervention, and analyze the mechanism of reporting maternal deaths.

Methods: A retrospective review was performed of maternal death certificates and medical records in Utah from January 1, 1982, through December 31, 1994.

Results: Sixty-two maternal deaths were identified. The risk of maternal death increased with maternal age and parity. The classic triad of hemorrhage (n=8), infection (n=5), and preeclampsia-eclampsia (n=3) remains an important contributor (16 of 62 or 25.8%). However, trauma (n=10), pulmonary embolism (n=10), and maternal cardiac disease (n=9) now account for 46.8% (29 of 62) of maternal deaths. A greater number of direct obstetric causes of maternal death (n=20) were deemed preventable than indirect obstetric causes (n=1) or nonobstetric causes (n=4).

Conclusion: Trauma, pulmonary embolism, and maternal cardiac disease have emerged as the most common identifiable causes of maternal death. Improvements in prevention, earlier diagnosis, and aggressive treatment of these conditions are necessary to achieve the Public Health Service year 2000 objective of a 50% reduction in maternal mortality ratios (using the 1987 ratio as a baseline). (Obstet Gynecol 1998;91:187-91).

The United States Public Health Service Year 2000 objective is a fifty percent reduction in the maternal mortality ratio (using 1987 ratio as a baseline).1 Despite marked decreases in the United States during the past five decades, maternal deaths remain devastating obstetric complications. Maternal mortality reviews have traditionally been used to identify and analyze trends in maternal deaths and to assess the quality of obstetric health care delivery. National maternal mortality surveillance studies compiled by the Centers for Disease Control were published in 1991 2 and 1996 3 to complement previously published state-wide reviews for the past five decades. 4-9

The goals of this study were to analyze trends in maternal deaths in Utah, to identify at-risk groups along with potential opportunities for preventative intervention and to analyze the mechanism of maternal death reporting to determine possible improvements for future reviews.

MATERIALS AND METHODS

During the years 1982 through 1994, 41 maternal deaths were identified by the Utah Department of Health (UDOH), Bureau of Vital Records. The interval of study was selected because computerized birth and death certificate data were available and because medical practices were similar to current practice. The Bureau of Vital Records definition for maternal death is the death of a woman while pregnant or within 42 days of termination of pregnancy, irrespective of the duration and the site of the pregnancy, from any cause related to or aggravated by the pregnancy or its management but not from accidental or incidental causes.

The Bureau of Vital Records currently identifies maternal deaths by the following methods: a) delivery mentioned on death certificate, b) pregnancy mentioned on death certificate, c) delivery related operation (cesarean section, etc.) mentioned on death certificate, d) post-partum listed on death certificate, or e) anything associated with pregnancy mentioned on death certificate. If any of the above are indicated by the physician who completes the death certificate, a vital records staff person checks the live birth and fetal death file to ascertain if the delivery has occurred within 42 days of the death. This method is known to miss a significant proportion of actual cases because some women are not identified as pregnant, or deceased as a result of pregnancy on the death certificate 10 . The Utah death certificate does not currently include a provision to specifically indicate whether the deceased was pregnant at the time of death. Because underreporting was suspected, we performed a further review of birth and death certificates by the Bureau of Vital Records using the CDC criteria and a search of computerized discharge summaries of the state's four tertiary referral centers; we identified an additional 21 cases. Data were then extracted from patient medical records. The Utah State Medical Examiner's Office also was used as a resource to supplement missing or incomplete medical records. Causes of death were then classified according to the 1991 Centers for Disease Control maternal mortality review guidelines 2 . All maternal deaths identified were deemed either preventable or non preventable by the authors on the basis of standards of care and medical technology currently available.

A maternal death was defined as any death occurring during pregnancy, or within one year following termination of pregnancy, resulting from complications of the pregnancy itself, by a chain of events initiated by the pregnancy or by the aggravation of an unrelated condition by the physiologic or pharmacologic effects of the pregnancy. Direct obstetric deaths were defined as maternal deaths resulting from obstetrical complications of the pregnancy state, labor or puerperium and from interventions, omissions, incorrect treatment or a chain of events due to any of these complications. Indirect obstetric deaths were defined as maternal deaths resulting from previously existing disease or diseases that developed during pregnancy, labor or the puerperium that were not directly due to obstetric causes but possibly aggravated by the physiologic effects of pregnancy. Non-obstetric deaths were defined as maternal deaths resulting from accidental or incidental causes not related to pregnancy or its management 11. Thus an additional 14 cases were identified using the expanded CDC guidelines, resulting in a total of 62 evaluable cases.

After the data were reviewed and coded, maternal death ratios (number of maternal deaths per 100,000 live births) were calculated. Correlations with maternal age, parity, level of education, urban versus rural residence, and marital status were determined using chi square statistics. The expected number of deaths for the chi square table were determined based on actual number of deaths in each category with the assumption that the deaths will be proportional to the number of live births in each category if the variable in question had no effect on maternal mortality. Multivariate analysis should have permitted the elimination of any spurious relationships by controlling for interrelationship between exogenous variables, but the number of cases under study were not large enough to permit sensible multivariate analysis.

RESULTS

During the thirteen years included in this review, 62 maternal deaths were identified. During this same interval, 484,789 live births were registered in the state of Utah, resulting in an overall maternal mortality ratio of 12.8 per 100,000 live births.
Table 1 shows the distribution and respective classifications of maternal deaths. Of all the maternal deaths, 56.5 % were related directly to obstetric causes, 21% were related indirectly to obstetric causes, 21% were related to non-obstetric causes, and 1.6 % were of unknown cause.

Maternal ages in the study group ranged from 15 to 39 (mean 27.7) years. Chi square goodness-of-fit testing demonstrated a positive correlation between the number of deaths and the maternal mortality ratio for progressive maternal age (chi square=8.35, p< .05), with ratios increasing from 8.5 for maternal ages 15-19 years to 27.1 for maternal ages 35-39 years.

The risk of maternal mortality increased with progressive parity, though the chi square test revealed that correlation was statistically non-significant. (chi square=1.40) The mean gravidity during the study period was 3.1.

Thirty-three (53.2%) of pregnancies resulted in live births. Eleven pregnancies (17.7%) ended in stillborn fetuses, and an additional ten (16.1%) died undelivered. The remaining eight cases included tubal ectopic pregnancies(n=3), miscarriage (n=2), unknown neonatal outcomes (n=2), and gestational trophoblastic disease (n=1).

Of the 62 cases reviewed, 85.5% (N = 53) were married; the maternal mortality ratio was higher for unmarried women than for married women (14.3 versus 12.4,), the difference being statistically non-significant. Fifty-seven women (91.9%) were white, four (6.5%) were Asian-Pacific Islander, and one (1.6%) was Native American. Maternal death ratios were highest among Asian-Pacific Islander patients (22.3), followed by white patients (12.6), Hispanic patients (12.1), and Native American patients (12.1). The average education was 12.4 years for patients in this review, with higher maternal mortality ratios among women with less than high school education (17.3) than those with high school diplomas (10.5) or 1 or more years or college (11.1). The average number of prenatal visits was 7.6, with the second month of pregnancy being the mode for the initiation of prenatal care. All deaths included in the review were to Utah residents; maternal death ratios were higher among rural residents (19.1) than among urban residents (12.4). Thirty-three deaths were investigated by autopsy (53.2%).

Thirty five of the deaths (56.5%) were due to direct obstetric causes (Table 2). The leading cause of direct obstetric death was pulmonary embolism. Of the ten patients who died as a result of a pulmonary embolism, two were primigravidas who died following uncomplicated spontaneous vaginal deliveries. Two other women died following operative vaginal deliveries and two after scheduled cesarean section. The remaining four deaths were antepartum.

Hemorrhage was the second most common cause of direct obstetric death. Ruptured tubal ectopic pregnancies accounted for five of these eight deaths. Two additional women died of hemorrhagic complications during home births and one died from uncontrollable intraoperative hemorrhage from complete placenta previa.

Five women died from puerperal infection. Two were caused by group B streptococcus, one by Bacteroides fragilis and one Klebsiella pneumoniae in association with adult respiratory distress syndrome. The fifth woman died at 5 weeks postpartum of clinical sepsis in association with immunosuppression for concurrent systemic lupus erythematosus.

Amniotic fluid embolism was determined to be the cause of four deaths. One of these women suffered the acute onset of shock and hemorrhage following a suction curettage for an 11-week missed abortion. Two others died following the acute onset of shock and respiratory decompensation immediately following operative vaginal deliveries. The fourth woman, who had a one-month history of chronic abruption, died undelivered at 16 weeks gestation after the sudden onset of seizures and shock.

Anesthetic complications accounted for three deaths. Two resulted from the intravascular injection of anesthetic agents during placement of regional anesthesia for cesarean section. The third woman died from aspiration pneumonitis following general anesthesia for cesarean delivery.

Complications of pre-eclampsia/eclampsia were the cause of three deaths, two as a result of intracranial hemorrhage and one from a cardiac arrhythmia from an atrioventricular nodal hemorrhage.

Fourteen maternal deaths (21%) were from indirect obstetric causes, primarily the result of cardiac disease (Table 3). Of the eleven deaths from cardiac causes, four were the result of maternal congenital heart disease. Three of these women had ventricular septal defects and Eisenmenger physiology, two unrecognized prior to autopsy. Cardiac arrhythmias accounted for three deaths. In addition, acute myocardial infarction and myocardiopathy identified preconceptionally each accounted for two additional deaths. Cerebrovascular accidents accounted for three deaths. All were embolic in nature and occurred during the puerperium.

Thirteen women (21%) died from non-obstetric causes (Table 4). The most common cause was trauma (N = 10). Non-obstetric infections accounted for two additional deaths and the final non-obstetric death was the result of a pancreatic malignancy.

An assessment of preventability was performed in this study (depicted in Tables 2, 3, & 4). The percentage of deaths determined to be preventable was greatest for direct obstetric causes and lowest for indirect obstetric causes. The only deaths deemed preventable in the non-obstetric category were those caused by motor vehicle accidents in which a single car was involved and where the patient's failure to wear a seatbelt was documented.

DISCUSSION

This review of maternal mortality in Utah found a ratio of 12.8 deaths per 100,000 live births over the entire 13-year study period. Although slightly higher than that reported in the recent CDC reviews 2,3 , this rate is comparable to other recently published state-wide reviews 4-9,11 . Despite the fact that Utah's population is largely Caucasian and relatively well-educated, 12 , maternal mortality remains a significant issue facing pregnant women, their families and health care professionals in Utah. Maternal mortality in Utah increases with progressive age and parity, confirming previous other states reviews 2,3,6,13 . This finding is of concern as data from the Utah Department of Health, Bureau of Vital Records indicates that the percent of births to women 35 years or older increased from 5.7% of total births in 1975 to 8.9% of total births in 1995. These observations support Grimes' 14 recent observation that, in spite of recent and impressive improvements, pregnancy remains a "risky business".

Pulmonary embolism is now the most common cause of direct obstetric death in Utah. Although low-dose heparin has yet to be proven effective for prophylaxis during pregnancy 15 , a recent review suggests that low-dose aspirin plus prophylactic heparin may be efficacious in at-risk women 16 . However, the majority of women in our series had no obvious clinical risk factors. Certainly, further studies aimed at improving medical management and prophylaxis for pregnant and puerperal women are critical in order to reduce this prominent cause of maternal death.

Hemorrhage was the second leading cause of direct obstetric death in this series. Of these eight maternal deaths, seven were considered preventable, five from ruptured tubal ectopic pregnancies and two from complications associated with home deliveries (uterine rupture and uterine atony). Although improved diagnostic techniques (transvaginal ultrasound, discriminatory serum HCG assays) 17 have decreased overall maternal morbidity and mortality associated with ectopic pregnancy, these deaths attest that continued clinical vigilance and suspicion are necessary.

Despite the availability of antibiotics and improvements in the management of labor, infection remains a significant cause of maternal mortality in this review. Physicians must appreciate the significance of fever in obstetric patients and should be aggressive in the evaluation of such patients, including appropriate cultures. Improved patient education regarding the warning signs and symptoms of postpartum infection seem increasingly important given the recent emphasis on shortened hospital stays following childbirth.

Cardiovascular disorders were the leading cause of indirect maternal death in this series. Particularly striking was the proportion of patients in this review who died of complications of pulmonary hypertension. A recent report 18 from Utah demonstrated the potential risks to women with structural heart defects, often repaired early in life, who then survive to reproductive age. Despite apparently normal cardiac function, thorough cardiac examinations, including echocardiography, should be considered in all patients with a known history of repaired congenital heart disease. Careful monitoring during labor, including multidisciplinary management, may also improve outcome.

Peripartum cardiomyopathy also accounted for two deaths in this series. This condition is nearly always unforeseen and thus not preventable. Cardiac transplantation is a more recent therapeutic option for such patients, with recent estimates showing that 11 - 17% of patients with peripartum cardiomyopathy eventually received cardiac transplants 19 . Although open heart surgery during pregnancy has been performed, exposure to cardiopulmonary bypass may have serious consequences for the fetus 20 .

Although some international, 21 national, 22 and statewide 5 reviews have chosen not to include non-obstetric deaths in their results, some have exclusively focused on them 23 . We have included these deaths because the impact of a maternal death on a family is devastating irrespective of cause. In addition, there exists the potential for improvement of maternal and fetal outcomes through education in these areas. Three of the eight deaths resulting from motor vehicle accidents were deemed preventable because of documented failure of the patient to wear her seatbelt. Seatbelt use during pregnancy reduces low birth weight and premature delivery rates related to automobile accidents. 24 Patients should be educated regarding the relative risks and benefits of seatbelt use during pregnancy 25.

Despite the dramatic decrease in maternal death ratios over the past half-century, recent studies 26-27 continue to estimate the degree of underreporting to be as high as 20 to 75%. In our study, 33.9% (21 of 62) of our eligible cases were missed by traditional death certificate searches, partially as a result of the continuing problem of inaccurate death certificate reporting. 10,28,29. Modification of death certificates to include pregnancy status enhances the reporting of pregnancy-related deaths . Furthermore, a recent study from New York City showed that a review of state medical examiner reports proved to be the most useful in the ascertainment of maternal deaths 31 . Perhaps the most cost efficient and systematic monitoring system is a computerized linkage of birth and fetal death certificates to death certificates of reproductive age female decedents. Because this system cannot identify pregnancy related deaths that do not generate a record of pregnancy outcome (eg, ectopic pregnancies, gestational trophoblastic disease and induced or spontaneous abortions) traditional methods need to be retained to capture these deaths 10 . The Utah Department of Health, Bureau of Vital Records has agreed to run a pilot program of this system in an attempt to more thoroughly identify maternal deaths in future.

TABLE 1

Number and Ratios of Maternal Deaths (per 100,000 live births)
in Utah by Classification and Year

RETURN TO TEXT OF REPORT

TABLE 2

Etiology of Direct Maternal Deaths in Utah

RETURN TO TEXT OF REPORT

TABLE 3

Etiology of Indirect Maternal Deaths in Utah

RETURN TO TEXT OF REPORT

TABLE 4

Etiology of Nonobstetric Maternal Deaths in Utah

RETURN TO TEXT OF REPORT

REFERENCES:

1. Healthy People 2000 National Health Promotion and Disease Prevention Objectives. U.S. Department of Health and Human Services, Public Health Service Publication 91-50212, 1991.

2. Koonin LM, Atrash HK, Lawson HW, Smith J Maternal mortality surveillance, United States, 1979-1986. In: Special Focus on Reproductive health Surveillance. CDC Surveillance Summaries 1991;40:1.

3. Berg C J, Atrash HK, Koonin LM, Tucker M Pregnancy-related mortality in the United States, 1987-1990. Obstet and Gynecol. 1996;88:161-167.

4. May WJ, Greiss FC. Maternal mortality in North Carolina: a forty-year experience. Am J Obstet Gynecol 1989;161:555-560.

5. Dye TD, Gordon H, Held B, Tolliver NJ, Holmes AP. Retrospective maternal mortality case ascertainment in West Virginia, 1985-1989. Am J Obstet Gynecol 1992;167:72-76.

6. Varner MW, Daly KD, Goplerud CP, Keettel WC. Maternal mortality in a major referral hospital, 1926-1980. Am J Obstet Gynecol 1982;143:325-329.

7. Gabel HD. Maternal mortality in South Carolina from 1970 to 1984: an analysis. Obstet Gynecol 1987;69:307-311.

8. Friede AM, Rochat RS. Maternal and perinatal mortality: an epidemiologic perspective. In: Sachs BP, Acker D (eds). Clinical obstetrics: a public health perspective. Little, Massachusetts, PSG, 1986, pp 35-64.

9. Moses V, DePersio ST, Lorenz D, Oberle MW, Rochat R, Fermo A. A thirty-year review of maternal mortality in Oklahoma, 1950 through 1979. Am J Obstet Gynecol 1987;157:1189-1193.

10. Jocums S, Mitchel EF, Entman S, and Piper JM Monitoring Maternal Mortality using vital records linkage. Am J Prev Med. 1995; 11: 75-78.

11. Moore J, Chauhan SP, Wiener WB, Morrison JC. Maternal mortality in Mississippi: 1987-1991. J Miss State Med Assoc 1993;34:35-39.

12. Utah Department of Health, Division of Community and Family Health Services. Report on maternal and infant health in Utah. Salt Lake City, Utah: Utah Department of Health, 1997.

13. O'Reilly-Green C, Cohen WR. Pregnancy in women aged 40 and older. Obstet Gynecol Clin N Am 1993;20:313-331.

14. Grimes DA. The morbidity and mortality of pregnancy: still a risky business. Am J Obstet Gynecol 1994;170:1489-1494.

15. Pearson J, Rees G. Technique of caesarean section. In: Chalmers I, Enkin MW, Keirse MJNC (eds), Effective Care in Pregnancy and Childbirth. Oxford, Oxford University Press, pp 1234-1245, 1989.

16. Greaves M. Anticoagulants in pregnancy. Pharmacol Ther 1993;59:311-327.

17. American College of Obstetricians and Gynecologists, Technical Bulletin 150, American College of Obstetricians and Gynecologists, Washington, DC, December 1990.

18. Jackson GM, Dildy GA, Varner MW, Clark SL. Severe pulmonary hypertension in pregnancy following successful repair of ventricular septal defects in childhood. Obstet Gynecol 1993;82:680-682.

19. Rickenbacher PR, Rizeq MN, Hunt SA, Billingham ME, Fowler MB. Long-term outcome after heart transplantation for peripartum cardiomyopathy. Am Heart J 1994;127:1318-1323.

20. Pomini F, Mercogliano D, Cavalletti C, Caruso A., Pomini P. Cardiopulmonary bypass in pregnancy. 1996;61: 259-68.

21. Tietze C. Maternal mortality. World Health Stat Q 1977;30:312-339.

22. Rochat RV. Maternal mortality. World Health Stat Q 1981;34:1-8.

23. Dannenberg AL, Carter DM, Lawson, HW, Ashton D, Dorfman, SF Graham EH. Homicide and other injuries as causes of maternal death in New York City, 1987-1991. Am J Obstet Gynecol. 1995; 172: 1557-1564.

24. Wolf ME, Alexander BH, Rivara FP, Hickok DE, Maier RV, Starzyk PM . A retrospective cohort study of seatbelt use and pregnancy outcome after a motor vehicle accident. J Trauma 1993;34:116-119.

25. Crosby WM, Costiloe JP. Safety of lap-belt restraint for pregnant victims of automobile collisions. N Engl J Med 1971;284:632-636.

26. Rochat RW, Koonin LM, Atrash HK, Jewett JF. Maternal mortality in the United States: report from the Maternal Mortality Collaborative. Obstet Gynecol 1988;72:91-97.

27. Atrash HK, Ellerbrock TV, Hogue C J R, Smith JC The need for national pregnancy mortality surveillance. Fam Plann Perspective. 1989; 21: 25-26.

28. Messite J, Stellman SD. Accuracy of death certificate completion: The need for formalized physician training. JAMA 1996;275:794-796.

29. Kirchner T, Anderson RE. Cause of death: Proper completion of the death certificate. JAMA 1987;258:349-352.

30. Comas A, Navarro A, Carrera A, Castellanos R, Perez N, Saliceti J, et al. Maternal mortality surveillance - Puerto Rico, 1989. MMWR 1991;40:521-522.