wiliam obstetri

SECTION IX - REPRODUCTIVE SUCESS AND FAILURE

33. Abortion

The definition of abortion is the termination of pregnancy by any means before the fetus is sufficiently developed to survive. In the United States this definition is confined to the termination of pregnancy before 20 weeks based upon the date of the first day of the last normal menses. Another commonly used definition is the delivery of a fetus-neonate that weighs less than 500 g.

SPONTANEOUS ABORTION

When abortion occurs without medical or mechanical means to empty the uterus, it is referred to as spontaneous. Another widely used term is miscarriage.

PATHOLOGY

Hemorrhage into the decidua basalis and necrotic changes in the tissues adjacent to the bleeding usually accompany abortion. The ovum becomes detached, and this stimulates uterine contractions that result in expulsion. When the sac is opened, fluid is commonly found surrounding a small macerated fetus, or alternatively there may be no visible fetus in the sac, the so-called blighted ovum.

Blood or carneous mole is an ovum that is surrounded by a capsule of clotted blood. The capsule is of varying thickness, with degenerated chorionic villi scattered through it. The small, fluid-containing cavity within appears compressed and distorted by thick walls of old blood clot.

In later abortions, several outcomes are possible. The retained fetus may undergo maceration. The bones of the skull collapse and the abdomen becomes distended with blood-stained fluid. The skin softens and peels off in utero or at the slightest touch, leaving behind the corium. Internal organs degenerate and undergo necrosis. Amnionic fluid may be absorbed when the fetus becomes compressed upon itself and desiccated to form a fetus compressus. Occasionally, the fetus eventually becomes so dry and compressed that it resembles parchment, so-called fetus papyraceous.

RESUMPTION OF OVULATION. Ovulation may resume as early as 2 weeks after an abortion. Lahteenmaki and Luukkainen (1978) detected a surge of luteinizing hormone (LH) 16 to 22 days after abortion in 15 of 18 women studied. Moreover, plasma progesterone level, which had plummeted after the abortion, increased soon after the LH1 surge. These hormonal events are in temporal agreement with histological changes observed in endometrial biopsies as described by Boyd and Holmstrom (1972). Therefore, it is important that effective contraception be initiated soon after abortion.

ETIOLOGY. More than 80 percent of abortions occur in the first 12 weeks, and the rate decreases rapidly thereafter ( Harlap and Shiono, 1980). Chromosomal anomalies cause at least half of these early abortions, and their incidence likewise decreases thereafter ( Fig. 33-1). The risk of spontaneous abortion increases with parity as well as with maternal and paternal age (Warburton and Fraser, 1964; Wilson and associates, 1986 ). The frequency of clinically recognized abortion increases from 12 percent in women less than 20 years old to 26 percent in those over age 40. The effect of advancing maternal age is illustrated in Figure 33-2. For the same paternal ages, the increase is from 12 to 20 percent. Finally, the incidence of abortion is increased if a woman conceives within 3 months of a term birth ( Harlap and Shiono, 1980).

The exact mechanisms responsible for abortion are not always apparent, but in the very early months of pregnancy, spontaneous expulsion of the ovum is nearly always preceded by death of the embryo or fetus. For this reason, etiological considerations of early abortion involve ascertaining whenever possible the cause of fetal death. In the subsequent months, the fetus frequently does not die in utero before expulsion, and other explanations for its expulsion must be invoked.

FETAL FACTORS ABNORMAL ZYGOTE DEVELOPMENT. The most common morphological finding in early spontaneous abortions is an abnormality of development of the zygote, embryo, early fetus, or at times the placenta. In an analysis of 1000 spontaneous abortions, Hertig and Sheldon (1943) observed pathological ("blighted") ova in which the embryo was degenerated or absent in half. Such an abnormal ovum can be seen in Figure 33-3.

Poland and co-workers (1981) identified morphological disorganization of growth in 40 percent of abortuses that were expelled spontaneously before 20 weeks. Among embryos of less than 30 mm crown-rump length, the frequency of abnormal morphological development was 70 percent. Of the embryos on which tissue culture and chromosomal analyses were performed, 60 percent had chromosomal abnormalities. For fetuses of 30 to 180 mm crown-rump length, the frequency of chromosomal abnormalities was 25 percent.

ANEUPLOID ABORTION. Chromosomal abnormalities are common among embryos and early fetuses that are aborted spontaneously, and account for much or most of early pregnancy wastage. Approximately 50 to 60 percent of early spontaneous abortions are associated with a chromosomal anomaly of the conceptus ( Table 33-1). Jacobs and Hassold (1980) reported that approximately one fourth of chromosomal abnormalities were due to maternal gametogenesis errors and 5 percent to paternal errors. In a study of fetuses and newborns with trisomy 13, Robinson and colleagues (1996) reported that in 21 of 23 cases, the extra chromosome was of

maternal origin.

Autosomal trisomy is the most frequently identified chromosomal anomaly associated with first-trimester abortions ( Table 33-1). As discussed in Chapter 36, trisomies can be the result of an isolated nondisjunction, maternal or paternal balanced translocation, or balanced chromosomal inversion. Balanced structural chromosomal rearrangements are present in 2 to 3 percent of couples with a history of recurrent abortions ( American College of Obstetricians and Gynecologists, 1995). Translocations may be identified in either parent. Balanced chromosomal inversions may also be identified in couples with recurrent abortions. Trisomies for all autosomes except chromosome number 1 have been identified in abortuses, but autosomes 13, 16, 18, 21, and 22 are most common.

Monosomy X (45,X) is the next most common chromosomal abnormality and is compatible with live-born females (Turner syndrome). Triploidy is often associated with hydropic placental degeneration. Incomplete hydatidiform moles may have fetal development that is triploid or trisomic for chromosome number 16. Fetuses associated with these frequently abort early, and the few carried longer are all grossly malformed. Advanced maternal and paternal age are not associated with this abnormality. Tetraploid abortuses are rarely live born and are most often aborted very early in gestation.

Chromosomal structural abnormalities are unusual causes of abortion and have been identified only since the development of banding techniques. Some of these infants are live born with balanced translocations and can be normal. Autosomal monosomy is extremely rare and is incompatible with life. Sex chromosomal polysomy (47,XXX or 47,XXY) is unusual in abortus material but is commonly seen in live births.

EUPLOID ABORTION. Kajii and co-workers (1980) reported that three fourths of aneuploid abortions were before 8 weeks, while euploid abortions peaked at about 13 weeks. Stein and associates (1980) presented evidence that the incidence of euploid abortions increases dramatically after the maternal age of 35 years. The reasons for euploid abortions are generally unknown, but the following are possibilities:

1.      A genetic abnormality such as an isolated mutation or polygenic factors.

2.      Various maternal factors.

3.      Possibly some paternal factors.

Simpson (1980) observed that approximately 0.5 percent of live births have chromosomal abnormalities, while at least 2 percent of live births have diseases associated with a single-gene mutation or a polygenic mechanism of inheritance.

MATERNAL FACTORS. A variety of medical disorders, environmental conditions, and developmental abnormalities have been implicated in euploidic abortion.

INFECTIONS. Some chronic infections have been implicated in causing abortion. Brucella abortus and Campylobacter fetus are well-known causes of chronic abortion in cattle, but they are not a significant cause in humans ( Sauerwein and associates, 1993). Evidence that Toxoplasma gondii causes abortion in humans is inconclusive. There is no evidence in humans that either Listeria monocytogenes or Chlamydia trachomatis produce abortions ( Feist and associates, 1999; Osser and Persson, 1996; Paukku and associates, 1999). Herpes simplex, however, has been associated with an increased incidence of abortion following genital infection in early pregnancy. Temmerman and colleagues (1992) reported that spontaneous abortion was independently associated with maternal human immunodeficiency virus-1 (HIV-1) antibody, with maternal syphilis seroreactivity, and with vaginal colonization with group B streptococci. Serological evidence supportive of a role for Mycoplasma hominis and Ureaplasma urealyticum in abortion was provided by Quinn and co-workers (1983). Conversely, Temmerman and associates (1992) found no association between genital mycoplasma and spontaneous abortion. Interestingly, Berg and associates (1999) reported that erythromycin treatment for women with mycoplasma culture-positive amnionic fluid undergoing genetic amniocentesis resulted in a significant decrease in midtrimester losses (11 versus 44 percent).

CHRONIC DEBILITATING DISEASES. In early pregnancy, chronic wasting diseases such as tuberculosis or carcinomatosis have seldom caused abortion. Hypertension is seldom associated with abortion before 20 weeks, but rather may lead to fetal death and preterm delivery. Celiac sprue has been reported to cause both male and female infertility and recurrent abortions ( Sher and colleagues, 1994).

ENDOCRINE ABNORMALITIES

HYPOTHYROIDISM There does not appear to be an increased incidence of abortion associated with clinical hypothyroidism (Montoro and associates, 1981). Thyroid autoantibodies were associated with an increased incidence of abortion despite the lack of overt hypothyroidism ( Dayan and Daniels, 1996; Stagnaro-Green and associates, 1990). Conversely, others have found no increase in the incidence of antithyroid antibodies in women who have recurrent abortions when compared with normal controls (Esplin and colleagues, 1998; Pratt and associates, 1994).

DIABETES MELLITUS. As reviewed recently by Greene (1999), spontaneous abortion and major congenital malformations are both increased in women with insulin-dependent diabetes. The risk is related to the degree of metabolic control in the first trimester. In a prospective study, Mills and associates (1988) reported that early glucose control (within 21 days of conception) resulted in a similar spontaneous abortion rate compared with nondiabetic controls. Lack of glucose control, however, resulted in a marked increase in the abortion rate. In a study from the Children's Hospital of Pittsburgh, Dorman and associates (1999) reported a significantly higher rate of spontaneous abortion (27 versus 8 percent) for diabetic women compared with nondiabetic partners of type I diabetic men. There was a temporal decline in the spontaneous abortion rate in these diabetic women from 26 percent before 1969 to 5.7 percent from 1980 to 1989. These authors postulated that this decrease may be secondary to changes in medical care such as glucose self-monitoring.

PROGESTERONE DEFICIENCY. Insufficient progesterone secretion by the corpus luteum or placenta has been associated with an increased incidence of abortion. It has been suggested that abnormal levels of one or more hormones might help to forecast abortion. Unfortunately, reduced levels of these hormones are usually the consequence rather than the cause ( Salem and co-workers, 1984). There are well-documented cases of luteal phase defects, but they are uncommon.

NUTRITION. There is no conclusive evidence that dietary deficiency of any one nutrient or moderate deficiency of all nutrients is an important cause of abortion. The nausea and vomiting that develop rather commonly during early pregnancy, and any inanition so induced, are rarely followed by spontaneous abortion.

DRUG USE AND ENVIRONMENTAL FACTORS. A variety of different agents has been reported, but not confirmed, to be associated with an increased incidence of abortion.

TOBACCO. Smoking has been associated with an increased risk for euploidic abortion ( Harlap and Shiono, 1980). For women who smoked more than 14 cigarettes a day, the risk was approximately twofold compared with controls ( Kline and associates, 1980). Armstrong and associates (1992) calculated that the abortion risk increased in a linear fashion by a factor of 1.2 for each 10 cigarettes smoked per day.

ALCOHOL. Both spontaneous abortion and fetal anomalies may result from frequent alcohol use during the first 8 weeks of pregnancy ( Floyd and associates, 1999). Spontaneous abortion was increased even when alcohol was consumed "in moderation." Kline and co-workers (1980) reported that the abortion rate was doubled in women drinking twice weekly and trebled in women who consumed alcohol daily compared with nondrinkers. Armstrong and colleagues (1992) computed that abortion risk increased by an average of 1.3 for each drink per day. In contrast, Cavallo and associates (1995) , in a prospective study of 546 women, reported that a low level of alcohol consumption during pregnancy was not associated with a significant risk for abortion. Somewhat worrisome is the fact that in one cross-sectional study from the Centers for Disease Control and Prevention, Floyd and associates (1999) found that half of all pregnant women in the study drank alcohol during the 3 months preceding pregnancy recognition and 5 percent drank moderate to heavy levels!

CAFFEINE. Coffee consumption at greater than four cups per day appears to slightly increase the risk of abortion ( Armstrong and associates, 1992). The risk appears to increase with increasing amounts. In a study by Klebanoff and associates (1999), maternal paraxanthine (a caffeine metabolite) levels were associated with a significant twofold risk of spontaneous abortion only if extremely high. These authors concluded that moderate consumption of caffeine was unlikely to be associated with spontaneous abortion.

RADIATION. In sufficient doses, radiation is a recognized abortifacient. As discussed in Chapter 42, the human dose is not precisely known.

CONTRACEPTIVES. There is no evidence to support that oral contraceptives or spermicidal agents used in contraceptive creams and jellies are associated with an increased incidence of abortion. Intrauterine devices, however, are associated with an increased incidence of septic abortion after contraceptive failure ( Chap. 58, p. 1538).

ENVIRONMENTAL TOXINS. In some studies abortion rates in exposed women were not increased ( Axelsson and Rylander, 1982). Rowland and associates (1995) reported an increased risk for spontaneous abortion among dental assistants exposed to 3 or more hours of nitrous oxide in offices without scavenging equipment, but not in offices using such equipment. In a meta-analysis, Boivin (1997) concluded that there was an increased risk of spontaneous abortion for women occupationally exposed to anesthetic gases based on data from the prescavenging era.

In most instances, there is little information to indict any specific environmental agent; however, there is evidence that arsenic, lead, formaldehyde, benzene, and ethylene oxide may cause abortion (Barlow and Sullivan, 1982). Video display terminals and exposure to the accompanying electromagnetic fields do not increase the risk of abortion (Schnorr and co-workers, 1991). Short-waves and ultrasound also do not increase the risk ( Taskinen and colleagues, 1990 ).

IMMUNOLOGICAL FACTORS. Much attention has focused on the immune system as important in recurrent pregnancy loss. Two primary pathophysiological models that have evolved are the autoimmune theory (immunity against self) and the alloimmune theory (immunity against another person).

AUTOIMMUNE FACTORS. It has been determined from compiled studies that approximately 15 percent of over 1000 recurrent pregnancy loss patients have recognized autoimmune factors (Kutteh and Pasquarette, 1995). The most significant antibodies have specificity against negatively charged phospholipids and are most commonly detected by testing for lupus anticoagulant (LAC) and anticardiolipin antibody (ACA). Women with both a history of early fetal loss and high levels of antibodies may suffer a 70 percent miscarriage recurrence ( Dudley and Branch, 1991). Pooling studies totaling 1500 women with recurrent loss yields an average incidence of 17 percent for anticardiolipin antibody and 7 percent for the lupus anticoagulant. In contrast, only 1 to 3 percent of normal obstetrical patients are found to have either of these (Harris and Spinnato, 1991; Lockwood and colleagues, 1989). In a prospective study of 860 women screened for anticardiolipin antibody in the first trimester, Yasuda and colleagues (1995) reported that 7 percent were positive. Spontaneous abortion occurred in 25 percent of the antibody-positive group compared with 10 percent of the negative group. In another recent study, however, Simpson and associates (1998) found no association between early pregnancy loss and the presence of either anticardiolipin antibody or lupus anticoagulant. Despite these controversies with early abortion, there is a consensus regarding increased midtrimester pregnancy losses and the antiphospholipid antibody syndrome ( Blumenfeld and Brenner, 1999; Cowchock, 1997; Simpson and associates, 1998).

The lupus anticoagulant is an immunoglobulin (IgG, IgM, or both) that interferes with one or more of the phospholipid-dependent tests of in vitro coagulation. The term is a misnomer because it is associated with clinically important increases in thromboembolic events. Importantly, the lupus anticoagulant is most often diagnosed in patients who do not meet the diagnostic criteria for lupus.

Antiphospholipid antibodies are acquired antibodies targeted against a phospholipid. They can be of the IgG, IgA, or IgM isotope. The mechanism of pregnancy loss in these women is thought to involve placental thrombosis and infarction. One mechanism may involve the inhibition of prostacyclin release ( Fig. 33-4). This product of endothelial cells is a potent vasodilator and inhibitor of platelet aggregation. On the other hand, platelets produce thromboxane A 2, a vasoconstrictor and platelet aggregator. These antibodies therefore may reduce prostacyclin production, facilitating a thromboxane dominant milieu that leads to thrombosis. In addition, they have been shown to inhibit protein C activation.

Investigators have proposed various treatments for the antiphospholipid antibody syndrome, including low-dose aspirin, prednisone, heparin, and intravenous immunoglobulin (Coulam, 1995). These treatments are thought to counteract the adverse action of antibodies by affecting both the immune and coagulation systems. Cowchock and colleagues (1992) performed a randomized trial comparing prednisone to low-dose heparin therapy in 20 women with antibodies and recurrent pregnancy loss. Live-birth rates were equal (75 percent) for both groups. However, those women receiving a glucocorticoid demonstrated a significantly greater incidence of maternal and fetal morbidity. Kutteh (1996) described 50 such women who were treated with either heparin and low-dose aspirin or aspirin alone. Heparin was initiated at 5000 units subcutaneously twice daily with a positive pregnancy test and titrated according to the partial thromboplastin time and platelet count. Although 76 percent of women in the heparin plus aspirin group delivered viable infants, only 44 percent of those treated with aspirin alone had a live birth. Maternal and obstetrical complications were low in both groups. Recent data indicate that antibodies bind directly to heparin in vitro and function in a similar way in vivo, thereby decreasing the adverse effects of the antibodies ( Ermel and associates, 1995).

Branch and associates (2000) conducted a placebo-controlled pilot study of immune globulin for the treatment of the antiphospholipid antibody syndrome during pregnancy and found that intravenous immune globulin did not improve pregnancy outcomes above that achieved with heparin and low-dose aspirin.

ALLOIMMUNE FACTORS. A number of women with recurrent pregnancy loss have been diagnosed as having an alloimmune cause. They have received a variety of therapies targeted at stimulating maternal immune tolerance of fetal material. Diagnosis of an alloimmune factor has centered on several tests:

1.      Maternal and paternal HLA comparison.

2.      Assessment of maternal serum for the presence of cytotoxic antibodies to paternal leukocytes.

3.      Maternal serum testing for blocking factors for maternal-paternal mixed lymphocyte reactions.

In essence, those couples determined to have significant HLA-type homology, or in which the women were found to have minimal antipaternal antibodies, were judged to represent an alloimmune disorder.

The validity of this model remains doubtful. For example, human HLA sharing clearly does not preclude successful pregnancies ( Ober and co-workers, 1983). Most importantly, other studies have compared HLA sharing frequency in couples with recurrent losses with those with reproductive success and observed no differences (Bellingard, 1995; Houwert-de Jong, 1989; Sargent, 1988; and their associates). Although some investigators have found the presence of lymphocytotoxic antibodies and mixed lymphocyte culture inhibitors to be associated with women with successful pregnancies, Coulam (1992) has conclusively shown these to be a function of the duration and number of pregnancies rather than a reason for pregnancy loss. Furthermore, the results of these three tests were found to have no predictive value in assessing risks for subsequent pregnancy outcome in a population of women with recurrent losses ( Cowchock and Smith, 1992).

Notwithstanding the uncertainties surrounding the most prevalent hypotheses of alloimmune causes of recurrent pregnancy loss, a number of studies have described outcomes following therapy to improve the maternal immune milieu. The majority of these women received either paternal leukocytes or pooled human immunoglobulin. Fraser and associates (1993) performed meta-analysis of 19 case series and concluded that immunotherapy does not significantly improve pregnancy outcome. A retrospective worldwide observational study and meta-analysis on allogenic leukocyte immunization therapy for recurrent pregnancy loss in over 400 cases demonstrated marginal improvement for immunized women (Coulam and colleagues, 1994 ). The considerable expense and potential morbidity associated with immunization therapy make full disclosure of relevant information and informed consent important ( Bux and co-workers, 1992; Katz and associates, 1992).

Some physicians have infused pooled human immunoglobulin as an alternative to paternal lymphocyte therapy. A prospective, double-blind, placebo-controlled trial using intravenous gamma globulin to treat women with recurrent loss demonstrated an improvement in women receiving this treatment versus placebo ( Coulam, 1995). In another prospective study of 47 women with a history of three or more unexplained pregnancy losses, Stricker and associates (2000) reported that low-dose intravenous immunoglobulin (IVIG) was beneficial in improving pregnancy outcome. Of the 24 women receiving IVIG therapy who subsequently became pregnant, 22 had a term pregnancy. There were 11 women who refused IVIG therapy and seven subsequently became pregnant but all had first-trimester spontaneous abortions.

INHERITED THROMBOPHILIA. There have been numerous reports of an association of spontaneous abortions and inherited thrombophilias ( Blumenfeld, 1999; Girling, 1998; Nelen, 1997; Ridker; 1998; Souza, 1999; Younis, 2000 and all their associates). As discussed in Chapter 49 (p. 1330), other pregnancy complications have also been associated with these thrombophilias ( Table 33-2). In a study of 78 consecutive women with two or more first- or second-trimester losses, Younis and associates (2000) reported that 38 percent versus 8 percent of controls had activated protein C resistance and 19 percent versus 6 percent of controls had factor V Leiden mutation. Nelen and colleagues (2000) reported that elevated serum homocysteine levels were also a risk factor. Blumenfeld and Brenner (1999) recently reviewed thrombophilia-associated pregnancy wastage.

The optimal treatment for the various thrombophilias during pregnancy is unclear, but heparin (including low molecular weight heparin) appears to be efficacious for the treatment of antithrombin III deficiency as well as protein C and S deficiency. Aspirin plus heparin seems to be efficacious for treatment of factor V Leiden mutation and antiphospholipid syndrome ( Blumenfeld and Brenner, 1999).

AGING GAMETES.Guerrero and Rojas (1975) found an increased incidence of abortion relative to successful pregnancies when insemination occurred 4 days before or 3 days after the time of shift in basal body temperature. They concluded, therefore, that aging of the gametes within the female genital tract before fertilization increased the chance of abortion. Dickey and colleagues (1992) reported that infertility patients over 35 had a higher incidence of small amnionic sac syndrome and an increased incidence of euploidic abortion. Whether ovulation induction or in vitro fertilization result in aging of gametes prior to implantation is not known.

LAPAROTOMY. There is no evidence that surgery performed during early pregnancy causes increased abortions ( Chap. 42, p. 1144). For example, ovarian tumors

and pedunculated myomas are generally removed without interfering with pregnancy. Peritonitis increases the likelihood of abortion.

PHYSICAL TRAUMA. Trauma that failed to interrupt the pregnancy is often forgotten. Only the particular event apparently related temporally to abortion is

remembered. Most spontaneous abortions, however, occur some time after death of the embryo or fetus.

UTERINE DEFECTS ACQUIRED UTERINE DEFECTS. Even large and multiple uterine leiomyomas usually do not cause abortion. When associated with abortion,

their location is apparently more important than their size ( Chap. 35, p. 927). Uterine synechiae (Asherman syndrome) are caused by destruction of large areas of

endometrium by curettage. This in turn results in amenorrhea and recurrent abortions believed to be due to insufficient endometrium to support implantation. The

diagnosis can be made by a hysterosalpingogram that shows characteristic multiple filling defects, but the most accurate and direct diagnosis is made by

hysteroscopy (Raziel and colleagues, 1994). Romer (1994) reported that the incidence of intrauterine adhesions diagnosed by hysteroscopy was about the same after

the first incomplete or missed abortion (20 percent), but was significantly higher in women with recurrent abortions (approximately 50 percent). Recommended

treatment is lysis of the adhesions via hysteroscopy and placement of an intrauterine contraceptive device to prevent recurrence. Continous high-dose estrogen

therapy is also recommended by some practitioners for 60 to 90 days. March and Israel (1981) reported that abortions decreased from 80 to 15 percent with such

therapy.

DEVELOPMENTAL UTERINE DEFECTS. These defects are the consequence of abnormal mullerian duct formation or fusion; or they may occur spontaneously or be

induced by in utero exposure to diethylstilbestrol ( Chap. 35). Some types, such as uterine septa, may be associated with abortions. Porcu and associates (2000)

described pregnancy outcomes in 63 women with a septate uterus. They all underwent hysteroscopic resection of the septa because of repeat pregnancy loss or

abnormal fetal presentation. There were 26 term live births following this procedure. In a recent review, Homer and associates (2000) reported that hysteroscopic

septoplasty resulted in improved pregnancy outcome in women with repeated pregnancy loss.

INCOMPETENT CERVIX. The term incompetent cervix is applied to a discrete obstetrical entity. It is characterized by painless cervical dilatation in the second

trimester or perhaps early in the third trimester, with prolapse and ballooning of membranes into the vagina, followed by rupture of membranes and expulsion of an

immature fetus. Unless effectively treated, this sequence tends to repeat in each pregnancy.

Numerous methods have been described in nonpregnant women to make the diagnosis, usually by documenting a more widely dilated internal cervical os than is

normal. Methods have included hysterography, pull-through techniques of inflated catheter balloons, and acceptance without resistance at the internal os of

specifically sized cervical dilators ( Ansari and Reynolds, 1987). During pregnancy, attempts have been made with moderate success to predict premature cervical

dilation using ultrasonic techniques ( Michaels and associates, 1989). Iams and co-workers (1995) performed a cross-sectional study of cervical length measured by

transvaginal ultrasonography in women with a prior preterm delivery, those with cervical incompetence, and normal controls delivered at term. Gestational age at the

first preterm delivery was significantly correlated with cervical length in the pregnancy evaluated at each gestational age between 20 and 30 weeks. Andrews and

associates (2000), in a study of 53 women with ultrasound evaluation prior to 20 weeks, reported an association of short cervical lengths or funneling of the internal

cervical os and early spontaneous preterm births. Several authors have reported on the use of transfundal pressure during transvaginal ultrasound evaluation of the

cervix as an aid in the detection of asymptomatic incompetent cervix ( Guzman and colleagues, 1997a, 1997b, 1998; Rocco and Garrone, 1999).

Kurup and Goldkrand (1999), in a study comparing elective, emergent, or urgent cerclage, concluded that ultrasound was useful in identifying women with subtle

changes in the cervix who would benefit from urgent cerclage. Ultrasound has also been utilized to demonstrate an increase in cervical length after prophylactic or

therapeutic cerclage (Althuisius and co-workers, 1999; Funai and colleagues, 1999 ). The use of magnetic resonance imaging in the diagnosis of incompetent cervix

was recently reported by Maldjian and associates (1999).

There is little doubt that ultrasound, especially transvaginal, is a useful adjunct for the diagnosis of cervical shortening or funneling of the internal os and in the early

detection of cervical incompetence. The diagnosis, however, remains difficult in most women and is still often based on clinical examination and history.

ETIOLOGY. Although the cause of cervical incompetence is obscure, previous trauma to the cervix—especially in the course of dilatation and curettage, conization,

cauterization, or amputation—appears to be a factor in many cases. In other instances, abnormal cervical development, including that following exposure to

diethylstilbestrol in utero, plays a role ( Chap. 35, p. 918).

TREATMENT. The treatment of cervical incompetence is surgical, consisting of reinforcement of the weak cervix by some type of purse-string suture. Bleeding,

uterine contractions, or ruptured membranes are usually contraindications to surgery.

PREOPERATIVE EVALUATION. Cerclage should generally be delayed until after 14 weeks so that early abortions due to other factors will be completed. There is no

consensus as to how late in pregnancy the procedure should be performed. The more advanced the pregnancy, the more likely surgical intervention will stimulate

preterm labor or membrane rupture. For these reasons, some clinicians prefer bed rest rather than cerclage some time after midpregnancy. We usually do not perform

cerclage after 24 to 26 weeks.

Aarts and associates (1995) provided a review of late second-trimester cerclage, commonly known as an emergency cerclage. They concluded that emergency

cerclage can be of benefit in some women, but that the incidence of complications, especially infection, is high. According to Schorr and Morales (1996) , bulging

membranes are associated with significantly increased failure rates. Caruso and associates (2000) reported their experience with emergency cerclage in 23 women

with a dilated cervix and protruding membranes (gestational age 17 to 27 weeks). There were 11 live-born infants and they concluded that the success of the

procedure was unpredictable. In a 10-year review of 75 emergency cerclages, Chasen and Silverman (1998) reported that 65 percent of the women delivered at 28

weeks or later, and half delivered at 37 weeks or greater. Only 44 percent of those with bulging membranes at the time of cerclage reached 28 weeks. Amnioreduction

at the time of emergency cerclage may improve pregnancy prolongation (Locatelli and associates, 1999 ).

Sonography to confirm a living fetus and to exclude major fetal anomalies is done prior to cerclage. Obvious cervical infection should be treated, and cultures for

gonorrhea, chlamydia, and group B streptococci are recommended. For at least a week before and after surgery, there should be no sexual intercourse.

If there is a question as to whether cerclage should be performed, the woman is placed at decreased physical activity. Proscription of intercourse is essential, and

frequent cervical examinations should be conducted to assess cervical effacement and dilatation. Weekly ultrasonic surveillance of the lower uterine segment

between 14 and 27 weeks may prove useful in some women (Guzman and associates, 1998; Michaels and colleagues, 1989). Unfortunately, rapid effacement and

dilatation develop even with such precautions ( Witter, 1984).

CERCLAGE PROCEDURES. Three types of operations are commonly used during pregnancy. One is a simple procedure recommended by McDonald (1963) and

illustrated in Figure 33-5. The second is the more complicated Shirodkar operation (1955). The third is the modified Shirodkar procedure shown in Figure 33-6. There

is less trauma and blood loss with both the McDonald and modified Shirodkar procedures than with the original Shirodkar procedure.

In many cases, filling the bladder with 600 mL of saline through an indwelling Foley catheter will serve to push the fetus and membranes upward from the ballooning

lower segment. Some clinicians recommended placement of a 30-mL balloon Foley catheter through the cervix and inflating the balloon with saline to deflect the

amnionic sac cephalad.

Success rates approaching 85 to 90 percent are achieved with both McDonald and modified Shirodkar techniques ( Caspi and associates, 1990; Kuhn and Pepperell,

1977). Thus, there appears to be little reason for performing the more complicated original Shirodkar procedure. The modified Shirodkar procedure is often reserved

for previous McDonald cerclage failures and structural cervical abnormalities. Success rates are higher when cervical dilatation was minimal and membrane prolapse

was absent.

Transabdominal cerclage placed at the level of the uterine isthmus has been recommended in some instances, especially in cases of anatomical defects of the cervix

or failed transvaginal cerclage ( Cammarano and colleagues, 1995; Gibb and Salaria, 1995; Herron and Parer, 1988). The procedure requires laparotomy for

placement of the suture and another laparotomy for its removal, for delivery of the fetus, or both. The potential for trauma and other complications initially and

subsequently is much greater with this procedure than with the vaginal procedures. Turnquest and colleagues (1999) recently described fetal salvage in 9 of 11

women with abdominal cerclage.

COMPLICATIONS. Charles and Edward (1981) identified complications, especially infection, to be much less frequent when cerclage was performed by 18 weeks.

When performed much after 20 weeks, there was a high incidence of membrane rupture, chorioamnionitis, and intrauterine infection. With clinical infection, the suture

should be cut, and labor induced.

There is no evidence that prophylactic antibiotics prevent infection, or that progestational agents or ß-mimetic drugs are of any adjunctive value ( Thomason and

co-workers, 1982). In the event that the operation fails and signs of imminent abortion or delivery develop, it is urgent that the suture be released at once; failure to do

so may result in grave sequelae. Rupture of the uterus or cervix may be the consequence of vigorous uterine contractions with the ligature in place. Membrane

rupture during suture placement or within the first 48 hours of surgery is considered by some to be an indication to remove the cerclage. Kuhn and Pepperell (1977)

reported that when the membranes rupture in the absence of labor, the likelihood of serious fetal or maternal infection is increased appreciably if the suture is left in

situ and delivery is delayed. Still, the range of management options spans from observation, to removal of the cerclage with observation, to removal of the cerclage

and labor induction (Barth, 1995). There are insufficient data upon which to base any firm recommendation, and the optimal management of such patients remains

controversial (O'Connor and associates, 1999).

Following the Shirodkar operation, the suture can be left in place if it remains covered by mucosa, and cesarean delivery performed near term. Conversely, the

Shirodkar suture may be removed and vaginal delivery permitted.

PATERNAL FACTORS. Little is known about paternal factors in the genesis of spontaneous abortion. Certainly, chromosomal translocations in sperm can lead to

abortion. Kulcsar and associates (1991) found adenovirus or herpes simplex virus in nearly 40 percent of semen samples obtained from sterile men. The viruses were

detected in latent form in 60 percent of the cells, and the same viruses were found in abortuses.

CATEGORIES AND TREATMENT OF SPONTANEOUS ABORTION

It is convenient to consider the clinical aspects of spontaneous abortion under five subgroups: threatened, inevitable, incomplete, missed, and recurrent abortion.

THREATENED ABORTION. The clinical diagnosis of threatened abortion is presumed when any bloody vaginal discharge or bleeding appears during the first half of

pregnancy. It is an extremely commonplace occurrence, and one out of four or five women has vaginal spotting or heavier bleeding during early gestation. Of those

women who bleed in early pregnancy, approximately half will abort. The bleeding of threatened abortion frequently is slight, but it may persist for days or weeks.

Unfortunately, an increased risk of suboptimal pregnancy outcome in the form of preterm delivery, low birthweight, and perinatal death persists ( Batzofin and

associates, 1984; Funderburk and colleagues, 1980). Importantly, the risk of a malformed infant does not appear to be increased.

Some bleeding about the time of expected menses may be physiological. Cervical lesions are likely to bleed in early pregnancy, especially after intercourse. Polyps

presenting at the external cervical os as well as decidual reaction in the cervix tend to bleed in early gestation. A significant clinical point is that lower abdominal pain

and persistent low backache do not accompany bleeding from these benign causes.

Because most physicians consider any bleeding in early pregnancy to be a sign of threatened abortion, any treatment of so-called threatened abortion has a

considerable likelihood of success. Most women who actually are threatening to abort will ultimately do so no matter what is done. If, however, bleeding is attributable

to one of the unrelated causes mentioned previously, it is likely to disappear, regardless of treatment.

If an intrauterine device is still present and the "string" is visible, the device should be removed for reasons cited in Chapter 58 (p. 1539).

Bleeding usually begins first, and cramping abdominal pain follows a few hours to several days later. The pain of abortion may be anterior and clearly rhythmic; it may

be a persistent low backache, associated with a feeling of pelvic pressure; or it may be a dull, midline, suprapublic discomfort. Whichever form the pain takes,

prognosis for pregnancy continuation in the presence of bleeding and pain is poor. Higher perinatal mortality rates are observed in women whose pregnancies were

complicated early by threatened abortion.

Each woman should be examined for there is always the possibility that the cervix already is dilated and abortion is inevitable, or there is a serious complication such

as extrauterine pregnancy or torsion of an unsuspected ovarian cyst. She may be kept at home in bed with analgesia given to help relieve the pain. If the bleeding

persists, she should be reexamined and the hematocrit rechecked. If blood loss is sufficient to cause anemia or hypovolemia, evacuation of the pregnancy is generally

indicated.

Women with threatened abortion have been treated with progesterone intramuscularly or with a wide variety of synthetic progestational agents orally or

intramuscularly. Unfortunately, there evidence of effectiveness is lacking. "Success" from the use of these agents often results in no more than a missed abortion.

Occasionally, slight bleeding may persist for weeks. It then becomes essential to decide whether there is any possibility of continuation of the pregnancy. Vaginal

sonography, serial serum quantitative chorionic gonadotropin (hCG) levels, and serum progesterone values, measured alone or in various combinations, have proven

helpful in ascertaining if a live intrauterine pregnancy is present. Fossum and associates (1988) reported that a fetal sac can usually be seen using vaginal

sonography between 33 and 35 days from the last menstrual period ( Table 33-3). This was associated with a chorionic gonadotropin level of about 1000 mIU/mL.

Thus, if a gestational sac can be seen and serum hCG2 is less than 1000 mIU/mL, the gestation is not likely to survive. If any doubt exists, however, serial

gonadotropin levels should be measured.

Al-Sebai and associates (1995) reported that a single progesterone measurement had an 88 percent sensitivity and specificity in predicting a live versus dead

intrauterine fetus or a tubal pregnancy. Stovall and associates (1992) reported that only about 1 percent of abnormal pregnancies (spontaneous incomplete abortions

and ectopic pregnancies) have a serum progesterone level of 25 ng/mL or greater. A serum progesterone value of less than 5 ng/mL was associated with a dead

conceptus, but this did not localize the pregnancy as uterine or extrauterine. Hahlin and colleagues (1990) reported that no live intrauterine pregnancy had a

progesterone level less than 10 ng/mL; and 88 percent of ectopic pregnancies and 83 percent of spontaneous abortions had lower values. Therefore, with a fetal sac

clearly visible, a gonadotropin level of less than 1000 mIU/mL, and a serum progesterone value of less than 5 ng/mL, there almost certainly is not an intrauterine

pregnancy.

Sonographic demonstration of a distinct, well-formed gestational ring with central echoes from the embryo implies that the products of conception are reasonably

healthy (Table 33-3). A gestational sac with no central echoes from an embryo or fetus implies strongly, but does not prove, death of the conceptus. When abortion is

inevitable, mean gestational sac diameter is frequently smaller than appropriate for gestational age. All viable intrauterine pregnancies are visible by transvaginal

ultrasonography by 41 days' gestation ( Lipscomb and colleagues, 2000). Moreover, by approximately 45 days after the last menses and thereafter, fetal heart action

should be discernible using real-time ultrasound. Emerson and associates (1992) and Pellerito and colleagues (1992) reported excellent results in identifying live

intrauterine gestations using vaginal color and pulsed Doppler flow imaging techniques.

After death of the conceptus, the uterus should be emptied. All tissue passed should be examined to determine whether the abortion is complete. Unless all of the

fetus and placenta can be positively identified, curettage may be required. Vaginal probe or abdominal ultrasound may assist in this decision-making process. If

significant amounts of material are retained within the uterine cavity, curettage is recommended by most clinicians. Ectopic pregnancy should always be considered in

the differential diagnosis of threatened abortion. This is especially so if the gestational sac or fetus are not identified. Frozen section of the curettings may further

assist with diagnosis.

Women who are D negative with a threatened abortion probably should receive anti-D immunoglobulin ( Chap. 39, p. 1067). Von Stein and associates (1992) reported

that more than 10 percent of such women had significant fetomaternal hemorrhage.

INEVITABLE ABORTION. Inevitability of abortion is signaled by gross rupture of the membranes in the presence of cervical dilatation. Under these conditions,

abortion is almost certain. Rarely, a gush of fluid from the uterus during the first half of pregnancy is without serious consequence. The fluid may have collected

previously between the amnion and chorion. Most often, however, either uterine contractions begin promptly, resulting in expulsion of the pregnancy, or infection

develops.

With obvious membrane rupture during the first half of pregnancy, the possibility of salvaging the pregnancy is very unlikely. If in early pregnancy the sudden

discharge of fluid, suggesting ruptured membranes, occurs before any pain or bleeding, the woman may be put to bed and observed for further leakage of fluid,

bleeding, cramping, or fever. If after 48 hours there has been no further escape of amnionic fluid, no bleeding or pain, and no fever, she may get up and, except for

any form of vaginal penetration, continue her usual activities. If, however, the gush of fluid is accompanied or followed by bleeding and pain, or if fever ensues,

abortion should be considered inevitable and the uterus emptied.

INCOMPLETE ABORTION. The fetus and placenta are likely to be expelled together in abortions occurring before 10 weeks, but separately thereafter. When the

placenta, in whole or in part, is retained in the uterus, bleeding ensues sooner or later, to produce the main sign of incomplete abortion. With abortions that are more

advanced, bleeding occasionally may be massive to the point of producing profound hypovolemia.

In instances of incomplete abortion, it is often unnecessary to dilate the cervix before curettage. In many cases, the retained placental tissue simply lies loose in the

cervical canal and can be lifted from an exposed external os with ovum or ring forceps. Suction curettage, as described later, is effective for evacuating the uterus. A

woman with a more advanced pregnancy, or a woman who is bleeding heavily, should be hospitalized and the retained tissue removed without delay. Hemorrhage

from incomplete abortion is occasionally severe but rarely fatal. Fever is not a contraindication to curettage once appropriate antibiotic treatment has been started (p.

877).

Nielsen and Hahlin (1995) performed a randomized study comparing expectant management with curettage for spontaneous abortions less than 13 weeks.

Spontaneous resolution of pregnancy occurred within 3 days in 80 percent of women treated conservatively, although vaginal bleeding lasted on the average one day

longer. Complications were similar between the groups.

MISSED ABORTION. This is defined as retention of dead products of conception in utero for several weeks. The rationale for an exact time period is not clear, and it

serves no useful clinical purpose. In the typical instance, early pregnancy appears to be normal, with amenorrhea, nausea and vomiting, breast changes, and growth

of the uterus. After fetal death, there may or may not be vaginal bleeding or other symptoms denoting a threatened abortion. For a time, the uterus seems to remain

stationary in size, but mammary changes usually regress. The woman is likely to lose a few pounds. Thereafter, it becomes apparent that the uterus not only has

ceased to enlarge but also has become smaller. Many women have no symptoms during this period except persistent amenorrhea. If the missed abortion terminates

spontaneously, and most do, the process of expulsion is the same as in any abortion. If retained several weeks after death, it becomes a shriveled sac containing a

macerated fetus (Fig. 33-7). Egarter and associates (1995) reported that gemeprost vaginal suppositories (prostaglandin E 1) were effective in the termination of

first-trimester missed abortions in 77 percent of women.

Occasionally, after prolonged retention of the dead fetus, serious coagulation defects develop. This is more likely when the gestation reached the second trimester

before fetal death. The woman may note troublesome bleeding from the nose or gums and especially from sites of slight trauma. The pathogenesis and treatment of

coagulation defects and any attendant hemorrhage in instances of prolonged retention of a dead fetus are considered in Chapter 25 (p. 659).

The reason some abortions do not terminate after fetal death is not clear. The use of potent progestational compounds to treat threatened abortion, however, may

contribute to this. Smith and co-workers (1978) observed that 73 percent of women with threatened abortion given hormonal treatment did abort, but on the average

20 days later. In women who received no hormonal support, 67 percent aborted at a mean of 5 days.

RECURRENT ABORTION. This has been defined by various criteria of number and sequence, but probably the most generally accepted definition refers to three or

more consecutive spontaneous abortions. Repeated spontaneous abortions are likely to be chance phenomena in the majority of cases. Accepting an independent

risk of miscarriage occurrence to be 15 percent, a second loss could be calculated to occur at a rate of 2.3 percent and a third loss in 0.34 percent of women. In a

study of women doctors, the occurrences of one, two, and three miscarriages were reported to be 10.4, 2.3, and 0.34 percent, respectively ( Alberman, 1988).

Approximately 1 to 2 percent of women of reproductive age will experience three or more spontaneous, consecutive abortions, and as many as 5 percent will have two

or more recurrent abortions (Blumenfeld and Brenner, 1999).

Epidemiologically, there is some uniformity in the proportion of recurrent pregnancy loss with chromosomal anomalies. Despite this, however, there are significant

discrepancies when compared with the relative prevalence of other categories. There are several reasons to explain this: First, differences in definition make direct

comparisons between various studies questionable. For example, some authors included women with only two losses into their analysis. Second, methods used to

assign women to various diagnostic categories differ. Complicating this issue is that many diagnoses under consideration are debatable, both as to the criteria used

and to their actual contribution to pregnancy loss. Third, the intensity of evaluation applied prior to categorizing a woman as "unexplained" varied among these

studies. In general, the majority of studies have observed that women with three or more miscarriages more commonly are determined to have a chromosomal

anomaly, endocrinological disorder, or an altered immune system ( Table 33-4).

PROGNOSIS. With the exception of antiphospholipid antibodies and an incompetent cervix, the apparent "cure rate" after as many as three consecutive spontaneous

abortions will range between 70 and 85 percent regardless of treatment. That is, the loss rate will be higher, but not a great deal higher, than that anticipated for

pregnancies in general. In fact, Warburton and Fraser (1964) reported that the likelihood of recurrent abortion was 25 to 30 percent regardless of the number of

previous abortions. Poland and associates (1977) noted that if a woman had previously delivered a live-born infant, the risk for each recurrent abortion was

approximately 30 percent. If, however, the woman had no live-born infants and had experienced at least one spontaneous fetal loss, the risk of abortion was 46

percent. Women with three or more spontaneous abortions are at increased risk in a subsequent pregnancy for preterm delivery, placenta previa, breech

presentation, and fetal malformation (Thom and colleagues, 1992).

INDUCED ABORTION

Induced abortion is the medical or surgical termination of pregnancy before the time of fetal viability. In 1996, a total of 1,221,585 legal abortions were reported to the

Centers for Disease Control and Prevention (1999) . Approximately 20 percent of these women were 19 years of age or less and the majority were less than 25 years

of age, white, and unmarried (Centers for Disease Control and Prevention, 2000 ). About 88 percent of abortions were performed before 13 weeks, 55 percent before 8

weeks, and 16 percent at 6 weeks or less.

LEGAL ASPECTS. Until the United States Supreme Court decision of 1973, only therapeutic abortions could be performed legally in most states. The most common

legal definition of therapeutic abortion until then was termination of pregnancy before the period of fetal viability for the purpose of saving the life of the mother. A few

states extended their laws to read "to prevent serious or permanent bodily injury to the mother" or "to preserve the life or health of the woman." Some states allowed

abortion if pregnancy was likely to result in the birth of an infant with grave malformations.

The stringent abortion laws in effect until 1973 were of fairly recent origin. Abortion before quickening—the first definite perception of fetal movement, which most

often occurs between 16 and 20 weeks' gestation—was either lawful or widely tolerated in both the United States and Great Britain until 1803. In that year, as part of a

general restructuring of British criminal law, a statute was enacted that made abortion before quickening illegal. The Roman Catholic Church's traditional

condemnation of abortion did not receive the ultimate sanction of universal law (excommunication) until 1869 ( Pilpel and Norwich, 1969).

It was not until 1821 that Connecticut enacted the nation's first abortion law. Subsequently, throughout the United States, abortion became illegal except to save the

life of the mother. Because therapeutic abortion to save the life of the woman is rarely necessary or definable, it follows that the great majority of such operations

previously performed in this country went beyond the letter of the law. Borgmann and Jones (2000) have extensively reviewed legal issues in providing abortions.

INDICATIONS. Some indications for therapeutic abortion are discussed with the diseases that commonly lead to the operation. Well-documented indications are

persistent heart disease after previous cardiac decompensation and advanced hypertensive vascular disease. Another is invasive carcinoma of the cervix. The

American College of Obstetricians and Gynecologists (1987) established guidelines for therapeutic abortion:

• When continuation of pregnancy may threaten the life of the woman or seriously impair her health. In determining whether or not there is such a risk to health,

account may be taken of her total environment, actual or reasonably foreseeable.

• When pregnancy has resulted from rape or incest. In this case, the same medical criteria should be employed in evaluation of the woman.

• When continuation of pregnancy is likely to result in the birth of a child with severe physical deformities or mental retardation. Issues such as maternal HIV-13

infection are less clear-cut but problematic ( Araneta and colleagues, 1992).

ELECTIVE (VOLUNTARY) ABORTION

Elective or voluntary abortion is the interruption of pregnancy before viability at the request of the woman but not for reasons of impaired maternal health or fetal

disease. Most abortions done today fall into this category; in fact, there is approximately one elective abortion for every three live births in the United States.

LEGALITY. It was only about 25 years ago when elective abortion was again legalized in the United States. Several Supreme Court decisions are noteworthy in the

history of abortion.

ROE VERSUS WADE. The legality of elective abortion was established by the United States Supreme Court in its 1973 decision in Roe v Wade. It defined the extent

to which states might regulate abortion:

(a) For the stage prior to approximately the end of the first trimester, the abortion decision and its effectuation must be left to the medical judgment of the attending

physician.

(b) For the stage subsequent to approximately the end of the first trimester, the State, in promoting its interest in the health of the mother, may, if it chooses, regulate

the abortion procedures in ways that are reasonably related to maternal health.

(c) For the stage subsequent to viability, the State, in promoting its interest in the potential of human life, may, if it chooses, regulate, and even proscribe, abortion,

except where necessary, in appropriate medical judgment, for the preservation of the life or health of the mother.

WEBSTER VERSUS REPRODUCTIVE HEALTH SERVICES. Since the Roe v Wade decision, many different pieces of legislation, both state and national, have been

introduced, and some enacted, to regulate or even dismantle its three provisions. All such attempts were unsuccessful until the United States Supreme Court ruled in

the 1989 Webster v Reproductive Health Services that states could place restrictions interfering with provision of abortion services on such items as waiting periods,

specific informed consent requirements, parental/spousal notification, and hospital requirements. Based upon this decision, numerous challenges have arisen to limit

a woman's choice and access to abortion services.

PLANNED PARENTHOOD OF SOUTHEASTERN PENNSYLVANIA VERSUS ROBERT P. CASEY. In 1992, the United States Supreme Court considered whether a

state could require, prior to an abortion, a woman's informed consent, a 24-hour waiting period, spousal consent, parental consent in the case of a minor, and a

physician's description of the fetus and the abortion technique to be employed. In a 5 to 4 decision, the court upheld a state's right to require all these except spousal

consent. The court also reconsidered and reaffirmed the Roe v Wade decision by a 5 to 4 vote.

STENBERG VERSUS CARHART. The United States Supreme Court in 2000 considered whether the state of Nebraska, as represented by Attorney General Don

Stenberg, could by state law prevent the performance of so-called "partial birth abortions," and specifically those performed by Dr. Leroy Carhart. The court, in a 5 to

4 decision, struck down the Nebraska law. The court ruled that the Nebraska law was too broad, and that such a prohibition could jeopardize abortions that involve

more common surgical procedures. The court also noted that the Nebraska law "would place some women at unnecessary risk" because it did not include a health

exception.

HILL VERSUS COLORADO. In this 6 to 4 decision rendered also in 2000, the United States Supreme Court upheld the 1993 Colorado statute that is used to regulate

speech-related conduct within 100 feet of the entrance to any health-care facility. The specific portion of the statute that was challenged by Leila J. Hill and others

dealt with the provision that made it unlawful to "knowingly approach" within 8 feet of another person without that person's consent "for the purpose of passing a

leaflet or handbill to, displaying a sign to, or engaging in oral protest, education, or counseling with such other person. . . ."

The Colorado law had been enacted to prevent the disruption of normal functions at medical facilities by protests conducted for any reason. The Supreme Court's

decision to uphold the Colorado law does not prevent protests against abortion facilities, but it certainly reduces the likelihood of confrontations between patients and

demonstrators. That is, the court noted that the Colorado law insured "the right to be left alone" and did not violate the First Amendment because the law is "content

neutral."

COUNSELING BEFORE ELECTIVE ABORTION. There are only three choices available to the woman considering an abortion. These are continued pregnancy with

its risks and responsibilities; continued pregnancy with its risks and arranged adoption; or the choice of abortion with its risks. In any event, knowledgeable and

compassionate counselors are invaluable. For instance, women who delay abortion decisions to later gestational ages have been shown to exhibit greater

disturbances in the basic sense of self as noted in indications of gender/sexual conflict and lower achievement or striving orientation ( Cancelmo and associates,

1992).

ABORTION TECHNIQUES

Abortion can be performed either medically or surgically. In a randomized comparison of the efficacy and acceptability of these techniques, Creinin (2000) reported

that medical abortion seemed to be slightly less costly than surgical techniques. Paul and colleagues (1999) have provided detailed summaries for each method

shown in Table 33-5.

Prior to an elective abortion, if bacterial vaginosis is found, the woman should be treated with metronidazole to reduce the postoperative infection rate ( Larsson and

colleagues, 1992). Vaginal application of 2 percent clindamycin cream for 3 days reduces postabortion pelvic infection fourfold compared with placebo ( Larsson and

co-workers, 2000). Treatment of D-negative women after abortion with anti-D immunoglobulin is recommended, because about 5 percent of D-negative women

become sensitized after an abortion ( Chap. 39, p. 1066).

Goldstein and associates (1994) emphasized the utility of both ultrasound and immediate tissue examination in 674 women undergoing elective first-trimester

abortion. Each woman was scanned with an empty bladder using the transabdominal technique. If no gestational sac was seen, endovaginal sonography was

performed. All specimens underwent modified gross examination at 33 times magnification and tissue was sent for histopathological analysis. Transabdominal

sonogram demonstrated uterine gestational tissue in 612 women (91 percent) and 595 of these were less than 12 weeks. There were 17 women (2.5 percent) who

were 13 weeks or greater. In 62 women (9 percent) with no sac seen transabdominally, endovaginal ultrasound showed 34 uterine and two unruptured ectopic

pregnancies. Of the remaining women, chorionic villi (17) and decidua without villi (4) were seen on tissue examination. Of the latter four, two had rising serum hCG4

levels and were diagnosed with ectopic pregnancy, and two with falling levels were managed expectantly.

SURGICAL TECHNIQUES FOR ABORTION

The pregnancy may be removed surgically through an appropriately dilated cervix or transabdominally by either hysterotomy or hysterectomy.

DILATATION AND CURETTAGE. Surgical abortion is performed by first dilating the cervix and then evacuating the pregnancy by mechanically scraping out the

contents (sharp curettage), by vacuum aspiration (suction curettage), or both. The technique for early manual vacuum has been recently reviewed by MacIsaac and

Jones (2000). The likelihood of complications—including uterine perforation, cervical laceration, hemorrhage, incomplete removal of the fetus and placenta, and

infection—increases after the first trimester. For this reason, curettage or vacuum aspiration should be performed before 14 weeks.

After 16 weeks, dilatation and evacuation (D & E) is performed. This consists of wide cervical dilatation followed by mechanical destruction and evacuation of fetal

parts. With complete removal of the fetus, a large-bore vacuum curette is used to remove the placenta and remaining tissue. A dilatation and extraction (D & X) is

similar to a D & E5 except that with a D & X6, part of the fetus is first extracted through the dilated cervix to facilitate the procedure.

In the absence of maternal systemic disease, pregnancies are usually terminated by curettage or by evacuation or extraction without hospitalization. When abortion is

not performed in a hospital setting, it is imperative that capabilities for effective cardiopulmonary resuscitation be available, and that immediate access to

hospitalization be possible.

HYGROSCOPIC DILATORS. Trauma from mechanical dilatation can be minimized by using a device to slowly dilate the cervix. These devices draw water from

cervical tissues and also are used for preinduction cervical ripening ( Chap. 20, p. 473). Laminaria tents are commonly used to help dilate the cervix ( Fig. 33-8). They

are made from the stems of Laminaria digitata or Laminaria japonica, a brown seaweed. The stems are cut, peeled, shaped, dried, sterilized, and packaged according

to size (small, 3 to 5 mm diameter; medium, 6 to 8 mm; and large, 8 to 10 mm). The strongly hygroscopic laminaria are thought to act by drawing water from

proteoglycan complexes, causing them to dissociate and thereby allowing the cervix to soften and dilate.

Synthetic hygroscopic dilators have also been used. Lamicel is a polyvinyl alcohol polymer sponge impregnated with anhydrous magnesium sulfate ( Nicolaides and

co-workers, 1983). Stornes and Rasmussen (1991) reported that although both lamicel tents and gemeprost pessaries were effective for cervical dilatation

preparatory to first-trimester abortions, further dilatation was significantly easier after gemeprost.

Dilapan is made of hydrogel polymer, and while it was used for some time, it is not available now in the United States. It has been claimed that it dilates the cervix

more rapidly than dilators made of traditional seaweed ( Blumenthal, 1988; Chvapil and co-workers, 1982). Patsner (1996) found same-day Dilapan insertion

successful in preparing the cervix for dilatation and evacuation for second-trimester abortion. Hern (1994) compared Dilapan with laminaria in 1001 women as

overnight dilators. Although both were equally effective dilators, women receiving the Dilapan were at least twice as likely to experience problems in cervical dilatation

or problems resulting from poor dilatation or disintegration of the device than women who used the seaweed preparation.

An interesting dilemma is presented by the woman who has an osmotic dilator placed overnight preparatory to elective abortion, but who then changes her mind.

Among seven first-trimester and 14 second-trimester pregnancies in which this occurred, the dilator was removed and 14 had term deliveries, two had preterm

deliveries, and one had a spontaneous abortion 2 weeks later ( Schneider and associates, 1991 ). None suffered infectious morbidity, including three un-treated

women with cervical cultures positive for chlamydia.

TECHNIQUE FOR INSERTION. The cleansed cervix is grasped anteriorly with a tenaculum. The cervical canal is carefully sounded, without rupturing the

membranes, to identify its length. A laminaria of appropriate size is then inserted so that the tip rests just at the level of the internal os using a uterine packing forceps

or a radium capsule forceps (Fig. 33-8). Later, usually after 4 to 6 hours, the laminaria will have swollen and thereby dilated the cervix sufficiently to allow easier

mechanical dilatation and curettage. The laminaria may cause cramping.

PROSTAGLANDINS. Rather than using a hygroscopic dilator to effect cervical softening, prostaglandin pessaries (suppositories) have been inserted into the vagina

against the cervix 3 hours or so before attempting dilatation. Chen and Edler (1983) reported good results from applying 1 mg of prostaglandin E 1 methyl ester.

Several newer prostaglandin products have been used to induce labor or to efface the cervix prior to labor induction. Many of these same products have also been

used to prepare the cervix prior to mechanical dilatation for induction of abortion. These prostaglandins are also discussed in Chapter 20 (p. 471).

TECHNIQUE FOR DILATATION AND CURETTAGE. The anterior cervical lip is grasped with a toothed tenaculum. A local anesthetic such as 5 mL of 1 or 2 percent

lidocaine is injected bilaterally into the cervix. Alternatively, a paracervical block may be used.

The uterus is sounded carefully to identify the status of the internal os and to confirm uterine size and position. The cervix is further dilated with Hegar or Pratt dilators

until a vacuum aspirator suction curet of appropriate diameter can be inserted. As shown in Figure 33-9, the fourth and fifth fingers of the hand introducing the dilator

should rest on the perineum and buttocks as the dilator is pushed through the internal os. This provides a further safeguard against uterine perforation.

Suction curettage then is used to aspirate pregnancy products. The vacuum aspirator is moved over the surface systematically in order eventually to cover all the

uterine cavity. Once this is done and no more tissue is aspirated, then gentle sharp curettage is done if it is thought that any placenta or fetal fragments remain. A

sharp curet is more efficacious, and its dangers need not be greater than those of the dull instrument. Uterine perforations rarely occur on the downstroke of the curet,

but they may occur when any instrument is introduced into the uterus. As shown in Figure 33-10, manipulations should be carried out with the thumb and forefingers

only.

In cases advanced past 16 weeks, the fetus is extracted, usually in parts, using Sopher or similar forceps and other destructive instruments. These late abortions are

unpleasant for medical and nursing personnel and more dangerous for the woman undergoing the procedure. The risks of uterine perforation and laceration are

increased due to the larger fetus and the thinner uterine walls.

It is reemphasized that morbidity, immediate and remote, will be kept to a minimum if:

1. The cervix is adequately dilated without trauma before attempting to remove the fetus and gestational tissues.

2. Removal of the pregnancy is accomplished without perforating the uterus.

3. All pregnancy tissue is removed.

UTERINE PERFORATION. Accidental uterine perforation may occur during sounding of the uterus, dilatation, or curettage. The incidence of uterine perforation

associated with elective abortion varies. Two important determinants of this complication are the skill of the physician and the position of the uterus, with a much

greater likelihood of perforation if the uterus is retroverted. Accidental uterine perforation is recognized easily, as the instrument passes without hindrance further than

it should have. Observation may be sufficient therapy if the uterine perforation is small, as when produced by a uterine sound or narrow dilator.

Considerable intra-abdominal damage can be caused by instruments passed through a uterine defect into the peritoneal cavity. This is especially true for suction and

sharp curets. In this circumstance, laparotomy to examine the abdominal contents, especially the bowel, is the safest course of action. Unrecognized bowel injury

causes severe peritonitis and sepsis ( Kambiss and associates, 2000). We have also cared for a woman transferred to us after much of her right ureter had been

removed at the time of attempted abortion using suction curettage! Similar cases have been observed by others ( Keegan and Forkowitz, 1982).

Some women may develop cervical incompetence or uterine synechiae following dilatation and curettage. The possibility of these complications should be explained

to those contemplating abortion. In general, their risk is very slight. Unfortunately, more advanced abortion performed by curettage may induce sudden, severe

consumptive coagulopathy, which can prove fatal.

MENSTRUAL ASPIRATION. Aspiration of the endometrial cavity using a flexible 5- or 6-mm Karman cannula and syringe within 1 to 3 weeks after a missed

menstrual period has been referred to as menstrual extraction, menstrual induction, instant period, traumatic abortion, and mini-abortion. Problems include the woman

not being pregnant, the implanted zygote being missed by the curet, failure to recognize an ectopic pregnancy, and rarely, uterine perforation.

A positive pregnancy test will eliminate a needless procedure on a nonpregnant woman whose period has been delayed for other reasons. MacIsaac and Jones

(2000) recommend the following technique for identifying placenta in the aspirate. First, the syringe contents are placed in a clear plastic container and examined with

back lighting. Tap water is used to wash the tissue held in a strainer. The tissue is immersed in clear water. Placenta is macroscopically soft, fluffy, feathery, and

villous. A magnifying lens or culposcope provides visualization. If there is doubt as to whether the tissue is placenta or decidua, microscopic examination of a small

piece under a cover glass with high-light contrast will allow differentiation. Placental villi are obvious.

LAPAROTOMY. In a few circumstances, abdominal hysterotomy or hysterectomy for abortion is preferable to either curettage or medical induction. If significant

uterine disease is present, hysterectomy may provide ideal treatment. If sterilization is to be performed, either hysterotomy with tubal ligation or hysterectomy on

occasion may indicated. At times, hysterotomy or hysterectomy becomes necessary because of failure of a medical induction during the second trimester.

MEDICAL INDUCTION OF ABORTION

Throughout history, many naturally occurring substances have been tried as abortifacients by women desperate not to be pregnant. Most often, serious systemic

illness or even death has been the result rather than abortion. Even today, there are only a few effective, safe abortifacient drugs.

OXYTOCIN. Successful induction of second-trimester abortion is possible with high doses of oxytocin administered in small volumes of intravenous fluids. One

regimen we have found effective is to add 10 1-mL ampules of oxytocin (10 IU/mL) to 1000 mL of lactated Ringer solution. This solution contains 100 mU oxytocin per

mL. An intravenous infusion is started at 0.5 mL/min (50 mU/min). The rate of infusion is increased at 15- to 30-minute intervals up to a maximum rate of 2 mL/min

(200 mU/min). If effective contractions are not established at this infusion rate, the concentration of oxytocin is increased in the infused solution. It is safest to discard

all but 500 mL of the remaining solution, which contains a concentration of 100 mU oxytocin per mL. To this 500 mL is added an additional five ampules of oxytocin.

The resulting solution now contains 200 mU/mL, and the rate of infusion is reduced to 1 mL/min (200 mU/min). A resumption of a progressive rate increase is

commenced up to a rate of 2 mL/min (400 mU/min) and left at this rate for an additional 4 to 5 hours, or until the fetus is expelled.

Similar regimens have been shown to be highly effective by Winkler (1991) and Owen (1992) and their associates. In a retrospective comparison of prostaglandin E 2

(PGE2) vaginal suppositories and high-dose oxytocin, Winkler and colleagues (1991) reported successes of 93 percent and 91 percent, respectively. The mean

duration of labor was 13.1 hours with PGE 2 and 8.2 hours with oxytocin. The mean dose of PGE2 was 65 mg and of oxytocin was 200 units. Side effects were limited

to the PGE2 group, with nausea (46 percent), vomiting (37 percent), fever (64 percent), and diarrhea (20 percent).

In a subsequent randomized trial, Owen and co-workers (1992) concluded that concentrated oxytocin is a satisfactory alternative to prostaglandin E 2 for midtrimester

abortion. The same group also compared concentrated oxytocin plus low-dose prostaglandin with prostaglandin E 2 vaginal suppositories for second trimester

terminations (Owen and Hauth, 1996). The women in the prostaglandin-only group received a 20-mg PGE 2 vaginal suppository every 4 hours, and those in the

combined group received a 10-mg PGE2 suppository every 6 hours. The success rate was 81 versus 89 percent, but the side effects were significantly higher for the

vaginal PGE2-only group. In a more recent trial, the Alabama group compared vaginal misoprostol (200 ug vaginally every 12 hours) with a regimen of concentrated

oxytocin plus 10-mg PGE2 suppositories every 6 hours (Owen and Hauth, 1999). They concluded that misoprostol vaginal tablets in this dose were not satisfactory for

second-trimester pregnancy termination.

With concentrated oxytocin, careful attention must be directed to the frequency and intensity of uterine contractions, because each increase in infusion rate markedly

increases the amount of oxytocin infused. If the initial induction is unsuccessful, serial inductions on a daily basis for 2 to 3 days are almost always successful. The

chance of a successful induction with high-dose oxytocin is enhanced greatly by the use of hygroscopic dilators such as laminaria tents inserted the night before.

INTRA-AMNIONIC HYPEROSMOTIC SOLUTIONS. In order to effect abortion during the second trimester, 20 to 25 percent saline or 30 to 40 percent urea have been

injected into the amnionic sac to stimulate uterine contractions and cervical dilatation. These techniques are used infrequently in the United States, and according to

the American College of Obstetricians and Gynecologists (1987) , they have been replaced by dilatation and evacuation. Benefits of the latter cited included speed,

lower cost, and less pain and emotional trauma.

In a study from Thailand, among 125 pregnancies undergoing midtrimester termination using hypertonic saline, the mean induction-to-delivery time was 31.7 hours

(Herabutya and O-Prasertsawat, 1994). Retained placenta developed in 63 percent and pyrexia in 39 percent. In a study from India, Allahbadia (1992) reported

success rates of 96 percent in pregnancies ranging from 14 to 20 weeks when 200 mL of 20 percent saline was instilled. This compared favorably with a success rate

of 90 percent with intramuscular PGF2a, and 100 percent with extra-amnionic instillation of 5 percent povidone-iodine mixed with normal saline.

Hypertonic saline may result in serious complications, including death ( Jasnosz and colleagues, 1993). Other complications include:

1. Hyperosmolar crisis following entry of hypertonic saline into the maternal circulation.

2. Cardiac failure.

3. Septic shock.

4. Peritonitis.

5. Hemorrhage.

6. Disseminated intravascular coagulation.

7. Water intoxication.

HYPEROSMOTIC UREA. Urea, 30 to 40 percent, dissolved in 5 percent dextrose solution, has been injected into the amnionic sac, followed by intravenous oxytocin

at about 400 mU/min. Urea plus oxytocin is as efficacious an abortifacient as hypertonic saline, but is less likely to be toxic. Urea plus prostaglandin F 2a injected into

the amnionic sac is similarly effective.

PROSTAGLANDINS. Because of shortcomings of other medical methods of inducing abortion, prostaglandins and their analogues are used extensively to terminate

pregnancies, especially in the second trimester. Compounds commonly used are prostaglandin E 2, prostaglandin F2a, and certain analogues, especially

15-methylprostaglandin F2a methyl ester, PGE1-methyl ester (gemeprost), and misoprostol. Mechanisms of action of the prostaglandins are considered in detail in

Chapter 20 (p. 471).

TECHNIQUE. Prostaglandins can act effectively on the cervix and uterus when:

1. Placed in the vagina as a suppository or pessary immediately adjacent to the cervix.

2. Administered as a gel through a catheter into the cervical canal and lowermost uterus extraovularly.

3. Injected intramuscularly.

4. Injected into the amnionic sac by amniocentesis.

5. Taken orally.

Various treatment regimens are outlined in Table 33-6.

Christin-Maitre and colleagues (2000) provided a recent update on the medical termination of pregnancy. They reviewed numerous studies regarding the efficacy and

side effects of prostaglandins and methotrexate utilized alone or in various combinations. They also reviewed the efficacy and side effects of mifepristone utilized with

a prostaglandin. Kahn and associates (2000) provided a meta-analysis regarding medical abortion with mifepristone and misoprostol, mifepristone with other

prostaglandins, and methotrexate with misoprostol. They concluded that these regimens had high levels of success for early gestations. Parenteral approaches

reduce appreciably, but do not eliminate, the unpleasant systemic effects, especially gastrointestinal, that accompany oral administration. Repeated doses of

prostaglandin are often required and hygroscopic cervical dilators are often used concurrently.

Effectiveness of the various treatment regimens has ranged from 86 to 95 percent. Induction-to-delivery intervals ranged from 4 hours to more than 48 hours. In one

study of 932 second-trimester terminations using gemeprost, the median induction-to-abortion interval was 18 hours in nulliparas and 15 hours in parous women

(Thong and associates, 1992).

Prostaglandin vaginal suppositories applied to the cervix are also used in a lower dose during the first and early second trimesters to ripen or soften and dilate the

cervix before curettage or as an adjunct for mifepristone termination ( Healy and Evans, 1994). The safety of late induced abortion after a previous cesarean delivery

using prostaglandins or mifepristone was reported by Boulot and associates (1993). At a mean gestational age of almost 24 weeks, vaginal evacuation was achieved

in 20 of 23 women. At hysterotomy in the three treatment failures, one uterine rupture was found and successfully repaired. Chapman and colleagues (1996) reported

a uterine rupture rate of 3.8 percent in 79 women undergoing termination by induction at a mean age of 21 weeks.

MIFEPRISTONE (RU 486). This oral antiprogesterone has been used to effect abortions in early gestation, either alone or in combination with oral prostaglandins

(Baird and colleagues, 1992; el-Refaey and associates, 1995; Newhall and Winikoff, 2000; World Health Organization Task Force, 1994 ). The effectiveness of the

drug as an abortifacient is based upon its high receptor affinity for progesterone-binding sites ( Healy and colleagues, 1983). A single 600-mg dose of RU 486

administered prior to 6 weeks results in an 85 percent abortion rate. In first trimester nondeveloping pregnancies, a single dose of 600-mg of mifepristone induced

expulsion in 82 percent of women (Lelaidier and co-workers, 1993).

Ulmann and associates (1992) reported their results from over 16,000 women in whom RU 486 was given followed by a prostaglandin analogue for medical

termination. Overall success rate was 95 percent, with no difference regarding the nature or dose of prostaglandin used. Median duration of bleeding was 8 days and

it was 12 days or less in 90 percent of women. Bleeding necessitated vacuum aspiration or curettage in 0.8 percent of cases. Transfusion was required in 1 per 1000

women.

As discussed in Chapter 58 (p. 1548), RU 486 is also highly effective for emergency postcoital contraception if given within 72 hours ( Glasier and associates, 1993 ).

After 72 hours, the agent is progressively less effective. The addition of various oral, vaginal, or injected prostaglandins to this regimen results in abortion rates over

95 percent.

Side effects of RU 486 include nausea, vomiting, and gastrointestinal cramping. The major associated risk is hemorrhage due to partial expulsion of the pregnancy

and due to intra-abdominal hemorrhage from an early unsuspected ectopic pregnancy. The duration of vaginal bleeding is approximately 2 weeks after RU 486 alone

and approximately 1 to 2 weeks after RU 486 is given with a prostaglandin.

EPOSTANE. This 3ß-hydroxysteroid dehydrogenase inhibitor block the synthesis of endogenous progesterone. If administered within 4 weeks of the last menstrual

period, the drug will induce an abortion in approximately 85 percent of women ( Crooij and associates, 1988). Clinical responses are likely related to circulating

endogenous progesterone levels. Nausea is a frequent side effect, and hemorrhage is a risk if the abortion is incomplete. Other antiprogestins such as ZK 98,734 are

under study, and appear to be promising for induction of early abortion ( Swahn and colleagues, 1994).

CONSEQUENCES OF ELECTIVE ABORTION

MATERNAL MORTALITY. Legally induced abortion is a relatively safe surgical procedure, especially when performed during the first 2 months of pregnancy. The

risk of death from abortion performed during the first 2 months is about 0.6 per 100,000 procedures ( Berg and co-workers, 1996; Centers for Disease Control, 1986;

Grimes, 1994). The relative risk of dying as the consequence of abortion is approximately doubled for each 2 weeks of delay after 8 weeks' gestation. Atrash and

colleagues (1988) reported that the proportion of abortion-related deaths caused by general anesthesia had increased from 8 percent in 1975 to 29 percent in 1985.

This likely reflects an absolute decrease in deaths from nonanesthetic complications. LeBolt and co-workers (1982) estimated that during the 1970s, overall risk of

death from legal abortion was 15 percent of the risk from childbirth. Indeed, in 1987 six maternal deaths were reported among 1.3 million legal abortions ( Grimes,

1994). In his recent review, de Swiet (2000) emphasized that prior to the United Kingdom Abortion Act in 1968, 40 percent of maternal deaths were due to criminal

abortion. In 1994 through 1996, only 4 percent of deaths were related to abortion, and none were illegal.

IMPACT ON FUTURE PREGNANCIES. Hogue (1986), in a scholarly review of the impact of elective abortion upon subsequent pregnancy outcome, summarized data

from more than 200 publications. She emphasized that consideration must be given for the method of inducing abortion, and that women chosen as controls should be

nulliparous because parous women had a reduced risk of complications in subsequent pregnancies.

Fertility is not altered by an elective abortion. A possible exception is the small risk from pelvic infection. Vacuum aspiration results in no increased incidence of

midtrimester spontaneous abortions, preterm deliveries, or low-birthweight infants in subsequent pregnancies. Dilatation and curettage, however, in primigravidas

results in an increased risk for subsequent ectopic pregnancy, midtrimester abortion, and low-birthweight infants.

Subsequent ectopic pregnancies are not increased if the first termination is done by vacuum aspiration. Possible exceptions are in women with preexisting chlamydial

infection or those who develop postabortion infection. Multiple elective abortions do not increase the incidence of preterm delivery and low-birthweight infants

(Mandelson and associates, 1992). Placenta previa was reported to be increased following elective abortion ( Barrett and associates, 1981). Hogue (1986) discounted

this study because of failure to control for maternal age.

Induced midtrimester abortions apparently carry little risk to subsequent pregnancies if injection techniques are used. There are not enough procedure-specific data

available to permit valid conclusions regarding the risks to future pregnancies following any midtrimester abortion.

SEPTIC ABORTION

Serious complications of abortion have most often been associated with criminal abortion. Severe hemorrhage, sepsis, bacterial shock, and acute renal failure have

all developed in association with legal abortion but at a very much lower frequency. Metritis is the usual outcome, but parametritis, peritonitis, endocarditis, and

septicemia may all occur (Vartian and Septimus, 1991). In 300 septic abortions at Parkland Hospital, a positive blood culture was found in a fourth. Almost two thirds

were anaerobic bacteria and coliforms were also common. Other organisms reported as causative of septic abortion include Haemophilus influenzae, Campylobacter

jejuni, and group A streptococcus (Denton and Clarke, 1992; Dotters and Katz, 1991; Pinhas-Hamiel and associates, 1991). Treatment of infection includes prompt

evacuation of the products of conception along with broad-spectrum antimicrobials given intravenously. If sepsis and shock supervenes, then supportive care is

essential as discussed in Chapter 43 (p. 1167). Septic abortion has also been associated with disseminated intravascular coagulopathy ( Chap. 25, p. 657).