Clinical Articles Fatal pulmonary embolism during legal induced abortion in the United States from 1972 to 1985 Herschel W. Lawson, MD, Hani K. Atrash, MD, MPH, and Adele L. Franks, MD Atlanta, Georgia To determine the risk factors for abortion-related deaths caused by pulmonary embolism. we investigated all deaths from legal abortions in the United States from 1972 through 1985. Of 213 deaths. 45 (21%) were due to air. blood clot. or amniotic fluid embolism. The risk of embolism death was higher among minority women and older women (34 to 44 years). Our analysis revealed that curettage at :521 weeks and abortions at :512 weeks. regardless of method. were both associated with the least risk of embolism death. In comparing 1972 to 1978 and 1979 to 1985. we found that the embolism mortality rate decreased 79%. During 1979 to 1985. the number of abortions performed by noncurettage methods decreased 58%. possibly as a result of earlier abortion morbidity studies. which showed that these methods carried a greater risk of complications. Although a decrease in mortality rates may be partially attributable to the declining use of these methods. our analysis suggests that changes in methods over time have not been universally applied to all racial groups. (AM J OBSTET GVNECOL 1990;162:986-90.)
Key words: Abortion. mortality, pulmonary embolism
Recent maternal mortality studies have shown that pulmonary embolism is the leading cause of maternal death in the United States." 2 Likewise. a recent study of mortality from legal abortions reported that pulmonary embolism is the leading cause of death associated with legal induced abortion.' Because scant information exists on abortion-related deaths from embolism, little can be suggested to prevent those deaths. Therefore we reviewed all legal abortion-related deaths from pulmonary embolism reported to the Centers for Disease Control (CDC) from 1972 through 1985 to determine the incidence. trends, and risk factors for such deaths. We present the results of that review and update previously published information on amniotic fluid embolism associated with legally induced abortion. Throughout this article. the terms "embolism" and "pulmonary embolism" are used interchangeably. Material and methods
All deaths were identified through CDC's nationwide surveillance of abortion-related deaths. This surveil-
From the Pregnancy and Infant Health Branch, Division of Reproductive Health, Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control. Received for publication September 5, 1989; revised November 29, 1989; accepted December 29,1989. Reprint requests: Herschel W. Lawson, MD, Centers for Disease Control, 1600 Clifton Rd., Mailstop C06, Atlanta, GA 30333. 611119143 QA~
lance system, which was begun in 1972, identifies deaths from multiple sources, including (1) vital records maintained by the National Center for Health Statistics; (2) state health departments; (3) state maternal mortality committees; (4) the Commission on Professional and Hospital Activities; (5) published cases; and (6) private sources, such as physicians, private citizens, and newspaper reports. Induced abortion is defined as a procedure intended to terminate a known or suspected intrauterine pregnancy at any gestational age to produce a nonviable fetus. A legal abortion is defined as an induced abortion performed by a licensed physician or a person under the supervision of a licensed physician. A case of fatal pulmonary embolism is defined as either one documented by imaging methods before death or during autopsy or one in which the patient had symptoms and signs often associated with pulmonary embolism (i.e., sudden respiratory distress, cardiovascular collapse, or coma) with no other cause of death found during autopsy.' For those deaths in which an autopsy or imaging studies were not performed, the diagnosis was made on clinical features alone. All reported deaths from legal induced abortion from thrombotic, air, and amniotic fluid embolism were included in the study. Denominators for calculating embolism mortality rates consisted of all reported legal abortions derived from CDC's abortion surveillance system for 1972 through 1985. The rates are reported as legal abortion-related embolism deaths per million abortions.
Fatal embolism during legal induced abortion
Volume 162 Number 4
987
Table I. Cases of embolism death related to legal abortion in the United States from 1972 to 1985 Yr
72 72 72 72 72 72 73 73 73 73 73 73 73 74 74 74 74 74 74 74 75 75 75 75 75 75 75 75 75
76 76 77 78 78 79 79 79
80
81 82 83 83
84 84 85
When
Method
16
23 25 21
18
29 19
35
20 20 44 37 20 31
22
35
19
17 20
22 31
20 23 20 18 18
35
23 20
34 26 36
21 37
29
24
32 22
21
29
16 16
09
23 II
20 15
10
17
19
08 13
18
10 34
18 20 10
20
10 08
16
08
12
09
25 12
16
22
18 13
17 20
08
20 23 20 16
14
21 28
23
30
II
16
21
19
16
AFE AFE Air AFE Air AFE PE PE AFE AFE PE PE PE PE AFE PE AFE PE AFE PE PE AFE PE PE Air AFE PE AFE AFE PE PE PE AFE PE AFE AFE AFE AFE PE PE AFE AFE PE PE Air
Saline Saline Suction Saline Unknown Saline Vaginal prostaglandin Suction Saline Saline Hysterectomy Hysterotomy Saline Suction Saline Saline Prostaglandin Suction Saline Suction Suction Hysterotomy Suction Suction Suction Suction Suction Saline U rea instillation Hysterectomy Suction Prostaglandin Saline Suction Saline Saline Saline D&E D&E Unknown Saline D&E Suction Suction D&E
y
y
y
y y
y
y
y
y y y y
y y
y
y
y
y
y y y y
y
y
y
y
y
y y
y
y
y y
Unknown During After After Unknown After After After During After After After After After During After After After Before After After During After After Before After After During After After After Unknown During After During During During After After After During During After After During
Preexisting condition
Hypertension Asthma
Marfan's syndrome Leiomyoma Leiomyoma
Heart disease Obesity Hodgkin's disease Heart disease
Heart disease Collagen vascular disease Hypertension Seizures Previous CVA Pancreatitis Seizures
Wilms' tumor Obesity
GA, Gestational age; DX, diagnosis; eVA, cerebrovascular accident; AFE, amniotic fluid embolism; PE, pulmonary embolism; D & E, dilation and evacuation.
For our analysis, we categorized a woman's age at the time of abortion as 15 to 19 years, 20 to 24 years, 25 to 29 years, 30 to 34 years, or 35 to 44 years. Gestational age was divided into three groups: (1) ~12 weeks, (2) 13 to 15 weeks, and (3) ~ 16 weeks. We divided abortion methods into three groups: curettage, instillation procedures, and hysterectomy/hysterotomy. Curettage performed at 13 weeks and beyond was considered dilation and evacuation. Methods classified as other were excluded from the study. To compare 1972 through 1978 and 1979 through 1985, we categorized the women's ages into two groups: 15 to 24 years and 25 to 44 years. We divided gestational age at the time of abortion into two groups: ~ 12 weeks and"" 13 weeks, and we divided abortion methods into two groups: curettage and instillation and hysterectomy /hysterotomy.
To analyze gestational age and method by race when no appropriate denominators were available for 1972 through 1985, we used a Poisson distribution to compare the observed number of embolism deaths among women of black and other races with the expected number based on the distribution of deaths among white women. Statistical significance of this comparison was tested with Fisher's exact method.' Because gestational age and method are somewhat interrelated and yet each may independently affect the risk of embolism death, we stratified our analysis by gestational age and method and calculated 95% confidence intervals around risk ratios by the Woolf method." For this analysis we excluded all abortions conducted after 21 weeks' gestation, because these procedures were unlikely to be performed by curettage.
988
Lawson, Atrash, and Franks
April 1990 Am J Obstet Gynecol
Table II. Numbers of embolism deaths, abortions, embolism mortality rates, and risk ratios, with 95% confidence intervals, by race, age group, gestational age, and method in the United States from 1972 to 1985 Variable
Race White Black and other Age group (yr) 15-19 20-24 25-29 30-34 35-44 Gestational age (wk)* $12 13-15 ;::16 Methodt Curettage Instillation H/H
EMR
RR (95% Cl)
10,374,147 4,757,591
2.2 4.6
1.0 Referent 2.1 (1.5-2.8)
8 20 5 6 6
4,263,429 5,204,472 2,948,923 1,556,692 986,433
1.9 3.8 1.7 3.9 6.1
1.0 2.0 0.9 2.1 3.2
14 5 24
13,484,539 791,469 833,395
1.0 6.3 28.8
1.0 Referent 6.1 (2.9-12.9) 27.7 (21.8-35.4)
19 21 4
14,302,277 623,819 24,641
1.3 33.7 162.3
1.0 Referent 25.3 (18.9-34.0) 122.2 (50.1-297.8)
Deaths
Abortions
23 22
Referent (1.6-2.6) (0.5-1.8) (1.1-3.8) (1.8-5.9)
EMR, Embolism mortality rate; RR, risk ratio; C/, confidence interval; H/H, hysterectomy/hysterotomy. *Gestational age unknown for two deaths. tMethod unknown for one death.
Risk ratios are calculated by dividing the rate of one group by the rate of a comparison group. Where the confidence interval does not include 1.0, the risk ratio is considered statistically significant at the p < 0.05 level. Results
Demographic and clinical characteristics. Overall, 45 cases of fatal embolism associated with legal induced abortion were reported for 1972 through 1985 (Table I). Amniotic fluid embolism accounted for 22 cases, thrombotic pulmonary embolism for 19, and air embolism for four. Twenty of these cases were confirmed by imaging studies or at autopsy. The women who died were primarily young (mean age, 25 years), were white (51 %), were unmarried (67%), and had their abortions during the second trimester of pregnancy (mean gestational age, 16 weeks). One death occurred after a saline solution instillation for delivery of a 35-week anencephalic fetus. This death was included in the total because it met the CDC definition of induced abortion. Eighteen women had preexisting conditions. Nineteen women underwent curettage for termination of their pregnancies, 21 underwent instillation procedures, 2 women underwent hysterotomy, and 2 hysterectomy (hysterectomy I hysterotomy). The type of anesthesia used was known in 24 cases: 15 women had general anesthesia, 6 had local, and 2 had other forms, including one spinal block and one intravenous sedation for placental removal. The initial presenting symptoms and signs included tachycardia (3), cardiopulmonary collapse (27), hypotension (12), pain (3), seizures (2), and vaginal bleeding (5). Symptoms occurred before labor
or the onset of the procedure in 2 women, during the procedure or labor in 10, and after the completion of the procedure in 22. Disseminated intravascular coagulation (DIC) was diagnosed in 12 patients. In 11 of those 12 patients, DIC occurred during the second trimester; eight were associated with instillation procedures and four with curettage procedures. Nine cases of DIC were associated with amniotic fluid embolism and three with pulmonary embolism. Epidemiologic characteristics. From 1972 through 1985, the 45 deaths from embolism represented 21 % of all legal abortion-related deaths, with an embolism mortality rate of 3.0 deaths per million abortions. The embolism mortality rate for amniotic fluid embolism was 1.5, the rate for thrombotic pulmonary embolism was 1.3, and the rate for air embolism was 0.3. Women in the older age groups (~30 years) had the highest rates; these were significantly elevated compared with rates among women in the youngest group (Table II). The embolism mortality rate for women of black and other minority races was more than twice that for white women (4.6 vs 2.2) (Table II). The 19 embolism deaths among women of black and other minority races during the second trimester were a significantly higher number (p < 0.01) than the expected number of 9.5. Likewise, the 16 embolism deaths attributable to noncurettage abortions among women of black and other minority races were higher than the expected number of 9.4; however, the probability of this difference occurring by chance alone was 0.06. For the 16-year period, the highest embolism mortality rates were reported among women undergoing
Fatal embolism during legal induced abortion
Volume 162 Number 4
989
Table III. Numbers of embolism deaths, abortions, embolism mortality rates, and risk ratios with 95% confidence intervals around risk ratios for gestational age by method in the United States from 1972 to 1985 Gestational age (wh)
$12 Curettage N oncurettage* 13-21 Curettage Noncurettage*
Deathst
Abortionst
EMR
RR (95% Cl)
12 I
13,600,455 60,290
0.9 16.6
1.0 Referent 18.8 (2.9-123.6)
5 18
884,335 573,151
5.7 31.4
6.4 (3.1-13.4) 35.6 (26.6-47.7)
EMR, Embolism mortality rate; RR, risk ratio; C/, confidence interval.
procedures at or after 16 weeks' gestation (embolism mortality rate = 28.8, Table II). When analyzed by method, the rate associated with instillation procedures was 33.7 times that of curettage and the rate for death resulting from hysterectomy/hysterotomy was 162.3 times that of curettage (Table II). In our stratified analysis of method by gestational age, we compared curettage and noncurettage (instillation and hysterectomy/hysterotomy procedures and used curettage at $12 weeks' gestation as the referent. This analysis revealed that noncurettage procedures at $12 weeks' gestation, with an embolism mortality rate of 16.6 deaths per million legal abortions, were associated with 18.8 times greater risk of embolism death than curettage procedures at $12 weeks (Table III). Curettage procedures at 13 to 21 weeks' gestation, with an embolism mortality rate of 5.7 per million, had 6.4 times the risk of embolism death compared with curettage procedures at $12 weeks, whereas noncurettage procedures at 13 to 21 weeks gestation with an embolism mortality rate of 31.4 deaths per million abortions had 35.6 times the risk of curettage at $12 weeks. From 1972 through 1978, 34 (24%) deaths from legal abortions were attributable to embolism, whereas from 1979 through 1985, 11 (15%) were attributable to embolism. The embolism mortality rate fell 79%, from 5.6 embolism deaths per million abortions from 1972 through 1978 to 1.2 deaths per million from 1979 through 1985. Between the comparison periods, the proportion of deaths from amniotic fluid embolism increased from 38% to 45%, and the proportion of deaths from thrombotic embolism decreased from 44% to 36%. The embolism mortality rates decreased in an absolute sense for both race groups over time. However, during the comparison periods from 1972 through 1978 and 1979 through 1985, the risk of embolism death among white women decreased 90%, whereas the risk among women of black and other minority races decreased only 62%. As a result, the risk ratio of embolism death among women of black and other minority races rose almost fourfold, from risk ratio = 1.5
to 5.8. Embolism mortality rates decreased for all age groups, and no changes in relative risk among age groups were noted over time. When risks of death by gestational age were compared over time, the 96% decrease in the risk of embolism death associated with abortions at $12 weeks versus the 71 % decrease in the risk of embolism death associated with abortions at ~ 13 weeks resulted in an almost sixfold increase in the relative risk of death associated with abortions occurring at ~ 13 weeks' gestation, from risk ratio = 12 to 71. Comment
Pulmonary embolism manifests during legal induced abortion in a fashion similar to the way it does during term delivery: sudden cardiovascular collapse, respiratory distress, cyanosis, and coma. The symptoms that often herald this dramatic event include chills, sweating, anxiety, coughing, convulsions, and pulmonary edema. 7 Of those patients diagnosed as having amniotic fluid embolisms, if the initial insult does not immediately result in death, approximately 40% are at risk of DIC." In several abortion-related deaths not included in this article, in which intraamniotic instillation of saline solution was used, amniotic fluid embolism may have been the event that began the chain reaction, resulting in DIC as the ultimate cause of death. Thus our measures of risk may be underestimates. The volume of amniotic fluid or amniotic fluid debris necessary to cause a fatal event has not been established!' 9. IU Since the CDC began to collect data on abortion mortality in 1972, the number of deaths and the case fatality rates associated with legal induced abortion have decreased significantly. Although embolism is reported as the leading cause of legal abortion-related death, our analysis showed that over time, the absolute risk of embolism death and the percentage of women dying of this condition declined. Contributing to this decrease in part is the large body of epidemiologic data on abortion morbidity and mortality collected in the United States during the 1970s. These studies found that abortions performed during the second trimes-
990
Lawson, Atrash, and Franks
ter of pregnancy carried a higher risk of morbidity and mortality than did those performed during the first trimester and also that instillation and hysterectomy/hysterotomy methods increased the risk of morbidity and mortality over those performed by curettage or dilation and evacuation. 7 • 11-13 After these reports, physicians' practices changed, and the numbers of instillation and hysterectomy/hysterotomy procedures declined. Thus the decrease in absolute risk of embolism death shown by our analysis may be attributed, at least in part, to the 58% decrease in the use of instillation and hysterectomy/hysterotomy methods that occurred during the comparison periods (1972 through 1978 and 1979 through 1985).14 In addition, we found that at all reported gestational ages, curettage carries the least risk of embolism death, and that regardless of method, abortions at :s 12 weeks carry the least risk of embolism death. Consistent with studies of other types of abortion morbidity and mortality, the risk of embolism death is the lowest for curettage abortions through 12 weeks' gestational age. 7 Although the risk of embolism death decreased for all races over time, the risk decreased more for white women than for women of black and other minority races. This resulted in an almost fourfold increase in the relative risk of embolism death among women of black and other minority races. Our results are suggestive that timing of abortion contributed to the observed increase in embolism mortality risk among women of black and other minority races, because a higher than expected proportion of embolism deaths for these women occurred in the second trimester. Although women of black and other races are reported to obtain less and later pregnancy-related care, further studies are necessary to better understand if other reasons for the racial disparity in the risk of embolism death exist so that effective interventions can be devised. 15• 16 REFERENCES I. Kaunitz A, Hughes ], Grimes D, Smith ], Rochat R, Kafrissen M. Causes of maternal mortality in the United States. Obstet Gynecol 1985;65:605-11.
April 1990 Am J Obstet Gynecol
2. Rochat R, Koonin L, Atrash H, et al. Maternal mortality in the United States: report from the maternal mortality collaborative. Obstet Gynecol 1988;72:91-7. 3. Atrash H, MacKay H, Binkin N, Hogue C. Legal abortion mortality in the United States: 1972 to 1982. AM] OBSTET GYNECOL 1987;156:605-11. 4. Guidotti R, Grimes D, Cates W. Fatal amniotic fluid embolism during legally induced abortion, United States, 1972-1978. AM] OBSTET GYNECOL 1981;156:257-61. 5. Rothman K, Boice J. Epidemiologic analysis with a programmable calculator. Boston: Epidemiologic Resources Inc, 1982:29-30. 6. Kahn H. An introduction to epidemiologic methods. New York: Oxford University Press, 1983:46-8. 7. Grimes D, Cates W. Complications of legally-induced abortion: a review. Obstet Gynecol Surv 1979;34:177-91. 8. Courtney L. Amniotic fluid embolism. Obstet Gynecol Surv 1974;29:169-77. 9. Lee W, Ginsburg K, Cotton D, Kaufman R. Squamous and trophoblastic cells in the maternal pulmonary circulation identified by invasive hemodynamic monitoring during the peripartum period. AM ] OBSTET GYNECOL 1986; 155:999-1001. 10. Kuhlman K, Hidvegi D, Tamura R, Depp R. Is amniotic fluid material in the central circulation of peripartum patients pathologic? Am] Perinatol 1985;2:295-9. II. Grimes D, Schulz K. Morbidity and mortality from second-trimester abortions.] Reprod Med 1985;30:50514. 12. Buehler ], Schulz K, Grimes D, Hogue C. The risk of serious complications from induced abortion: do personal characteristics make a difference? AM] OBSTET GYNECOL 1985;153:14-9. 13. Cates W, Schulz K, Gold ], Tyler C. Complications of surgical evacuation procedures for abortions after 12 weeks' gestation. In: Zatuchni G, Sciarra ], Speidel J. Pregnancy termination, procedures, safety, and new developments. Hagerstown, Maryland: Harper & Row, 1979:206-17. 14. Lawson H, Atrash H, Saftlas A, Koonin L, Ramick M, SmithJ. Abortion surveillance, 1984-1985, United States. CDC surveillance summaries. MMWR 1989;38(SS-2):1145. 15. National Center for Health Statistics. Health, United States, 1984. Hyattsville, Maryland: National Center for Health Statistics, 1984: 10-1; DHHS publication no. (PHS) 85-1232. 16. National Center for Health Statistics. Health, United States, 1986. Hyattsville, Maryland: National Center for Health Statistics, 1986:25; DHHS publication no. (PHS) 87-1232.
Key words: Abortion. mortality, pulmonary embolism
Recent maternal mortality studies have shown that pulmonary embolism is the leading cause of maternal death in the United States." 2 Likewise. a recent study of mortality from legal abortions reported that pulmonary embolism is the leading cause of death associated with legal induced abortion.' Because scant information exists on abortion-related deaths from embolism, little can be suggested to prevent those deaths. Therefore we reviewed all legal abortion-related deaths from pulmonary embolism reported to the Centers for Disease Control (CDC) from 1972 through 1985 to determine the incidence. trends, and risk factors for such deaths. We present the results of that review and update previously published information on amniotic fluid embolism associated with legally induced abortion. Throughout this article. the terms "embolism" and "pulmonary embolism" are used interchangeably. Material and methods
All deaths were identified through CDC's nationwide surveillance of abortion-related deaths. This surveil-
From the Pregnancy and Infant Health Branch, Division of Reproductive Health, Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control. Received for publication September 5, 1989; revised November 29, 1989; accepted December 29,1989. Reprint requests: Herschel W. Lawson, MD, Centers for Disease Control, 1600 Clifton Rd., Mailstop C06, Atlanta, GA 30333. 611119143 QA~
lance system, which was begun in 1972, identifies deaths from multiple sources, including (1) vital records maintained by the National Center for Health Statistics; (2) state health departments; (3) state maternal mortality committees; (4) the Commission on Professional and Hospital Activities; (5) published cases; and (6) private sources, such as physicians, private citizens, and newspaper reports. Induced abortion is defined as a procedure intended to terminate a known or suspected intrauterine pregnancy at any gestational age to produce a nonviable fetus. A legal abortion is defined as an induced abortion performed by a licensed physician or a person under the supervision of a licensed physician. A case of fatal pulmonary embolism is defined as either one documented by imaging methods before death or during autopsy or one in which the patient had symptoms and signs often associated with pulmonary embolism (i.e., sudden respiratory distress, cardiovascular collapse, or coma) with no other cause of death found during autopsy.' For those deaths in which an autopsy or imaging studies were not performed, the diagnosis was made on clinical features alone. All reported deaths from legal induced abortion from thrombotic, air, and amniotic fluid embolism were included in the study. Denominators for calculating embolism mortality rates consisted of all reported legal abortions derived from CDC's abortion surveillance system for 1972 through 1985. The rates are reported as legal abortion-related embolism deaths per million abortions.
Fatal embolism during legal induced abortion
Volume 162 Number 4
987
Table I. Cases of embolism death related to legal abortion in the United States from 1972 to 1985 Yr
72 72 72 72 72 72 73 73 73 73 73 73 73 74 74 74 74 74 74 74 75 75 75 75 75 75 75 75 75
76 76 77 78 78 79 79 79
80
81 82 83 83
84 84 85
When
Method
16
23 25 21
18
29 19
35
20 20 44 37 20 31
22
35
19
17 20
22 31
20 23 20 18 18
35
23 20
34 26 36
21 37
29
24
32 22
21
29
16 16
09
23 II
20 15
10
17
19
08 13
18
10 34
18 20 10
20
10 08
16
08
12
09
25 12
16
22
18 13
17 20
08
20 23 20 16
14
21 28
23
30
II
16
21
19
16
AFE AFE Air AFE Air AFE PE PE AFE AFE PE PE PE PE AFE PE AFE PE AFE PE PE AFE PE PE Air AFE PE AFE AFE PE PE PE AFE PE AFE AFE AFE AFE PE PE AFE AFE PE PE Air
Saline Saline Suction Saline Unknown Saline Vaginal prostaglandin Suction Saline Saline Hysterectomy Hysterotomy Saline Suction Saline Saline Prostaglandin Suction Saline Suction Suction Hysterotomy Suction Suction Suction Suction Suction Saline U rea instillation Hysterectomy Suction Prostaglandin Saline Suction Saline Saline Saline D&E D&E Unknown Saline D&E Suction Suction D&E
y
y
y
y y
y
y
y
y y y y
y y
y
y
y
y
y y y y
y
y
y
y
y
y y
y
y
y y
Unknown During After After Unknown After After After During After After After After After During After After After Before After After During After After Before After After During After After After Unknown During After During During During After After After During During After After During
Preexisting condition
Hypertension Asthma
Marfan's syndrome Leiomyoma Leiomyoma
Heart disease Obesity Hodgkin's disease Heart disease
Heart disease Collagen vascular disease Hypertension Seizures Previous CVA Pancreatitis Seizures
Wilms' tumor Obesity
GA, Gestational age; DX, diagnosis; eVA, cerebrovascular accident; AFE, amniotic fluid embolism; PE, pulmonary embolism; D & E, dilation and evacuation.
For our analysis, we categorized a woman's age at the time of abortion as 15 to 19 years, 20 to 24 years, 25 to 29 years, 30 to 34 years, or 35 to 44 years. Gestational age was divided into three groups: (1) ~12 weeks, (2) 13 to 15 weeks, and (3) ~ 16 weeks. We divided abortion methods into three groups: curettage, instillation procedures, and hysterectomy/hysterotomy. Curettage performed at 13 weeks and beyond was considered dilation and evacuation. Methods classified as other were excluded from the study. To compare 1972 through 1978 and 1979 through 1985, we categorized the women's ages into two groups: 15 to 24 years and 25 to 44 years. We divided gestational age at the time of abortion into two groups: ~ 12 weeks and"" 13 weeks, and we divided abortion methods into two groups: curettage and instillation and hysterectomy /hysterotomy.
To analyze gestational age and method by race when no appropriate denominators were available for 1972 through 1985, we used a Poisson distribution to compare the observed number of embolism deaths among women of black and other races with the expected number based on the distribution of deaths among white women. Statistical significance of this comparison was tested with Fisher's exact method.' Because gestational age and method are somewhat interrelated and yet each may independently affect the risk of embolism death, we stratified our analysis by gestational age and method and calculated 95% confidence intervals around risk ratios by the Woolf method." For this analysis we excluded all abortions conducted after 21 weeks' gestation, because these procedures were unlikely to be performed by curettage.
988
Lawson, Atrash, and Franks
April 1990 Am J Obstet Gynecol
Table II. Numbers of embolism deaths, abortions, embolism mortality rates, and risk ratios, with 95% confidence intervals, by race, age group, gestational age, and method in the United States from 1972 to 1985 Variable
Race White Black and other Age group (yr) 15-19 20-24 25-29 30-34 35-44 Gestational age (wk)* $12 13-15 ;::16 Methodt Curettage Instillation H/H
EMR
RR (95% Cl)
10,374,147 4,757,591
2.2 4.6
1.0 Referent 2.1 (1.5-2.8)
8 20 5 6 6
4,263,429 5,204,472 2,948,923 1,556,692 986,433
1.9 3.8 1.7 3.9 6.1
1.0 2.0 0.9 2.1 3.2
14 5 24
13,484,539 791,469 833,395
1.0 6.3 28.8
1.0 Referent 6.1 (2.9-12.9) 27.7 (21.8-35.4)
19 21 4
14,302,277 623,819 24,641
1.3 33.7 162.3
1.0 Referent 25.3 (18.9-34.0) 122.2 (50.1-297.8)
Deaths
Abortions
23 22
Referent (1.6-2.6) (0.5-1.8) (1.1-3.8) (1.8-5.9)
EMR, Embolism mortality rate; RR, risk ratio; C/, confidence interval; H/H, hysterectomy/hysterotomy. *Gestational age unknown for two deaths. tMethod unknown for one death.
Risk ratios are calculated by dividing the rate of one group by the rate of a comparison group. Where the confidence interval does not include 1.0, the risk ratio is considered statistically significant at the p < 0.05 level. Results
Demographic and clinical characteristics. Overall, 45 cases of fatal embolism associated with legal induced abortion were reported for 1972 through 1985 (Table I). Amniotic fluid embolism accounted for 22 cases, thrombotic pulmonary embolism for 19, and air embolism for four. Twenty of these cases were confirmed by imaging studies or at autopsy. The women who died were primarily young (mean age, 25 years), were white (51 %), were unmarried (67%), and had their abortions during the second trimester of pregnancy (mean gestational age, 16 weeks). One death occurred after a saline solution instillation for delivery of a 35-week anencephalic fetus. This death was included in the total because it met the CDC definition of induced abortion. Eighteen women had preexisting conditions. Nineteen women underwent curettage for termination of their pregnancies, 21 underwent instillation procedures, 2 women underwent hysterotomy, and 2 hysterectomy (hysterectomy I hysterotomy). The type of anesthesia used was known in 24 cases: 15 women had general anesthesia, 6 had local, and 2 had other forms, including one spinal block and one intravenous sedation for placental removal. The initial presenting symptoms and signs included tachycardia (3), cardiopulmonary collapse (27), hypotension (12), pain (3), seizures (2), and vaginal bleeding (5). Symptoms occurred before labor
or the onset of the procedure in 2 women, during the procedure or labor in 10, and after the completion of the procedure in 22. Disseminated intravascular coagulation (DIC) was diagnosed in 12 patients. In 11 of those 12 patients, DIC occurred during the second trimester; eight were associated with instillation procedures and four with curettage procedures. Nine cases of DIC were associated with amniotic fluid embolism and three with pulmonary embolism. Epidemiologic characteristics. From 1972 through 1985, the 45 deaths from embolism represented 21 % of all legal abortion-related deaths, with an embolism mortality rate of 3.0 deaths per million abortions. The embolism mortality rate for amniotic fluid embolism was 1.5, the rate for thrombotic pulmonary embolism was 1.3, and the rate for air embolism was 0.3. Women in the older age groups (~30 years) had the highest rates; these were significantly elevated compared with rates among women in the youngest group (Table II). The embolism mortality rate for women of black and other minority races was more than twice that for white women (4.6 vs 2.2) (Table II). The 19 embolism deaths among women of black and other minority races during the second trimester were a significantly higher number (p < 0.01) than the expected number of 9.5. Likewise, the 16 embolism deaths attributable to noncurettage abortions among women of black and other minority races were higher than the expected number of 9.4; however, the probability of this difference occurring by chance alone was 0.06. For the 16-year period, the highest embolism mortality rates were reported among women undergoing
Fatal embolism during legal induced abortion
Volume 162 Number 4
989
Table III. Numbers of embolism deaths, abortions, embolism mortality rates, and risk ratios with 95% confidence intervals around risk ratios for gestational age by method in the United States from 1972 to 1985 Gestational age (wh)
$12 Curettage N oncurettage* 13-21 Curettage Noncurettage*
Deathst
Abortionst
EMR
RR (95% Cl)
12 I
13,600,455 60,290
0.9 16.6
1.0 Referent 18.8 (2.9-123.6)
5 18
884,335 573,151
5.7 31.4
6.4 (3.1-13.4) 35.6 (26.6-47.7)
EMR, Embolism mortality rate; RR, risk ratio; C/, confidence interval.
procedures at or after 16 weeks' gestation (embolism mortality rate = 28.8, Table II). When analyzed by method, the rate associated with instillation procedures was 33.7 times that of curettage and the rate for death resulting from hysterectomy/hysterotomy was 162.3 times that of curettage (Table II). In our stratified analysis of method by gestational age, we compared curettage and noncurettage (instillation and hysterectomy/hysterotomy procedures and used curettage at $12 weeks' gestation as the referent. This analysis revealed that noncurettage procedures at $12 weeks' gestation, with an embolism mortality rate of 16.6 deaths per million legal abortions, were associated with 18.8 times greater risk of embolism death than curettage procedures at $12 weeks (Table III). Curettage procedures at 13 to 21 weeks' gestation, with an embolism mortality rate of 5.7 per million, had 6.4 times the risk of embolism death compared with curettage procedures at $12 weeks, whereas noncurettage procedures at 13 to 21 weeks gestation with an embolism mortality rate of 31.4 deaths per million abortions had 35.6 times the risk of curettage at $12 weeks. From 1972 through 1978, 34 (24%) deaths from legal abortions were attributable to embolism, whereas from 1979 through 1985, 11 (15%) were attributable to embolism. The embolism mortality rate fell 79%, from 5.6 embolism deaths per million abortions from 1972 through 1978 to 1.2 deaths per million from 1979 through 1985. Between the comparison periods, the proportion of deaths from amniotic fluid embolism increased from 38% to 45%, and the proportion of deaths from thrombotic embolism decreased from 44% to 36%. The embolism mortality rates decreased in an absolute sense for both race groups over time. However, during the comparison periods from 1972 through 1978 and 1979 through 1985, the risk of embolism death among white women decreased 90%, whereas the risk among women of black and other minority races decreased only 62%. As a result, the risk ratio of embolism death among women of black and other minority races rose almost fourfold, from risk ratio = 1.5
to 5.8. Embolism mortality rates decreased for all age groups, and no changes in relative risk among age groups were noted over time. When risks of death by gestational age were compared over time, the 96% decrease in the risk of embolism death associated with abortions at $12 weeks versus the 71 % decrease in the risk of embolism death associated with abortions at ~ 13 weeks resulted in an almost sixfold increase in the relative risk of death associated with abortions occurring at ~ 13 weeks' gestation, from risk ratio = 12 to 71. Comment
Pulmonary embolism manifests during legal induced abortion in a fashion similar to the way it does during term delivery: sudden cardiovascular collapse, respiratory distress, cyanosis, and coma. The symptoms that often herald this dramatic event include chills, sweating, anxiety, coughing, convulsions, and pulmonary edema. 7 Of those patients diagnosed as having amniotic fluid embolisms, if the initial insult does not immediately result in death, approximately 40% are at risk of DIC." In several abortion-related deaths not included in this article, in which intraamniotic instillation of saline solution was used, amniotic fluid embolism may have been the event that began the chain reaction, resulting in DIC as the ultimate cause of death. Thus our measures of risk may be underestimates. The volume of amniotic fluid or amniotic fluid debris necessary to cause a fatal event has not been established!' 9. IU Since the CDC began to collect data on abortion mortality in 1972, the number of deaths and the case fatality rates associated with legal induced abortion have decreased significantly. Although embolism is reported as the leading cause of legal abortion-related death, our analysis showed that over time, the absolute risk of embolism death and the percentage of women dying of this condition declined. Contributing to this decrease in part is the large body of epidemiologic data on abortion morbidity and mortality collected in the United States during the 1970s. These studies found that abortions performed during the second trimes-
990
Lawson, Atrash, and Franks
ter of pregnancy carried a higher risk of morbidity and mortality than did those performed during the first trimester and also that instillation and hysterectomy/hysterotomy methods increased the risk of morbidity and mortality over those performed by curettage or dilation and evacuation. 7 • 11-13 After these reports, physicians' practices changed, and the numbers of instillation and hysterectomy/hysterotomy procedures declined. Thus the decrease in absolute risk of embolism death shown by our analysis may be attributed, at least in part, to the 58% decrease in the use of instillation and hysterectomy/hysterotomy methods that occurred during the comparison periods (1972 through 1978 and 1979 through 1985).14 In addition, we found that at all reported gestational ages, curettage carries the least risk of embolism death, and that regardless of method, abortions at :s 12 weeks carry the least risk of embolism death. Consistent with studies of other types of abortion morbidity and mortality, the risk of embolism death is the lowest for curettage abortions through 12 weeks' gestational age. 7 Although the risk of embolism death decreased for all races over time, the risk decreased more for white women than for women of black and other minority races. This resulted in an almost fourfold increase in the relative risk of embolism death among women of black and other minority races. Our results are suggestive that timing of abortion contributed to the observed increase in embolism mortality risk among women of black and other minority races, because a higher than expected proportion of embolism deaths for these women occurred in the second trimester. Although women of black and other races are reported to obtain less and later pregnancy-related care, further studies are necessary to better understand if other reasons for the racial disparity in the risk of embolism death exist so that effective interventions can be devised. 15• 16 REFERENCES I. Kaunitz A, Hughes ], Grimes D, Smith ], Rochat R, Kafrissen M. Causes of maternal mortality in the United States. Obstet Gynecol 1985;65:605-11.
April 1990 Am J Obstet Gynecol
2. Rochat R, Koonin L, Atrash H, et al. Maternal mortality in the United States: report from the maternal mortality collaborative. Obstet Gynecol 1988;72:91-7. 3. Atrash H, MacKay H, Binkin N, Hogue C. Legal abortion mortality in the United States: 1972 to 1982. AM] OBSTET GYNECOL 1987;156:605-11. 4. Guidotti R, Grimes D, Cates W. Fatal amniotic fluid embolism during legally induced abortion, United States, 1972-1978. AM] OBSTET GYNECOL 1981;156:257-61. 5. Rothman K, Boice J. Epidemiologic analysis with a programmable calculator. Boston: Epidemiologic Resources Inc, 1982:29-30. 6. Kahn H. An introduction to epidemiologic methods. New York: Oxford University Press, 1983:46-8. 7. Grimes D, Cates W. Complications of legally-induced abortion: a review. Obstet Gynecol Surv 1979;34:177-91. 8. Courtney L. Amniotic fluid embolism. Obstet Gynecol Surv 1974;29:169-77. 9. Lee W, Ginsburg K, Cotton D, Kaufman R. Squamous and trophoblastic cells in the maternal pulmonary circulation identified by invasive hemodynamic monitoring during the peripartum period. AM ] OBSTET GYNECOL 1986; 155:999-1001. 10. Kuhlman K, Hidvegi D, Tamura R, Depp R. Is amniotic fluid material in the central circulation of peripartum patients pathologic? Am] Perinatol 1985;2:295-9. II. Grimes D, Schulz K. Morbidity and mortality from second-trimester abortions.] Reprod Med 1985;30:50514. 12. Buehler ], Schulz K, Grimes D, Hogue C. The risk of serious complications from induced abortion: do personal characteristics make a difference? AM] OBSTET GYNECOL 1985;153:14-9. 13. Cates W, Schulz K, Gold ], Tyler C. Complications of surgical evacuation procedures for abortions after 12 weeks' gestation. In: Zatuchni G, Sciarra ], Speidel J. Pregnancy termination, procedures, safety, and new developments. Hagerstown, Maryland: Harper & Row, 1979:206-17. 14. Lawson H, Atrash H, Saftlas A, Koonin L, Ramick M, SmithJ. Abortion surveillance, 1984-1985, United States. CDC surveillance summaries. MMWR 1989;38(SS-2):1145. 15. National Center for Health Statistics. Health, United States, 1984. Hyattsville, Maryland: National Center for Health Statistics, 1984: 10-1; DHHS publication no. (PHS) 85-1232. 16. National Center for Health Statistics. Health, United States, 1986. Hyattsville, Maryland: National Center for Health Statistics, 1986:25; DHHS publication no. (PHS) 87-1232.
