Spontaneous abortions occurring after the 13 weeks of gestation will be considered as these sometimes represent problems for the general surgeon. By strict definition, abortion includes all pregnancies expelled up to the 28th week of gestation, but modern obstetric and gynaecological practice views threatened abortions at 24 weeks' gestation and later as labour, with the fetus having a chance of survival with intensive neonatal care. Therefore management is, to a certain extent, dictated by the circumstances at presentation and, in particular, the gestation period.

In most cases, pregnancy will have been confirmed or suspected by the patient. The primary clinical presenting feature is of vaginal bleeding, often with some associated lower abdominal pain, usually described as similar to menstrual cramps. The degree of discomfort is often a pointer to the stage that the process has reached, although this can be misleading. Similarly, the volume of blood lost through the vagina, as reported by the patient and observed by the physician, is often indicative of an inevitable abortion but, on occasion, may be misleading. Unless the pregnancy had already ceased growing and the fetus has died (a missed abortion) the usual systemic pregnancy symptoms and breast signs are likely to be positive at the time of presentation.

Differential diagnoses for the vaginal bleeding to be considered from the history include an intrauterine molar pregnancy (trophoblastic tumour); an extrauterine pregnancy, which is likely to have implanted into the abdominal cavity once the second trimester has been reached; and sources of bleeding from the lower genital tract such as cervical polyps, cervicitis, and cervical carcinoma.

The clinical examination must include lower abdominal palpation, to exclude signs suggestive of intraperitoneal bleeding (as would be found in an abdominal pregnancy or from a traumatized uterus resulting from a criminal assault upon pregnancy). The uterine size must be assessed, to compare with that expected from the calculation of duration of pregnancy using the first day of the last menstrual loss. Bimanual pelvic examination is necessary to confirm pregnancy, determine uterine size as accurately as possible, and to assess the degree of cervical dilatation; if the cervix is closed, the abortion is threatened and if open, the abortion is inevitable. Lesions of the lower genital tract involving the vagina and cervix will be palpable or visible with the aid of a vaginal speculum.

In all cases of abortion, blood should be checked for haemoglobin, blood group, and Rhesus type. Patients who are Rh(D) negative should receive intramuscular immunoglobulin prophylaxis within 60 h of the onset of bleeding; 250 IU should be given up to 20 weeks of gestation and 500 IU thereafter. A Kleihauer count on a sample of maternal blood taken at this time gives an approximate guide to the volume of any fetomaternal haemorrhage that may have occurred, and further doses of immunoglobulin can be given.

In those cases where abortion is threatened, an ultrasound examination is helpful, particularly if the uterine size is smaller or larger than expected from the menstrual dates. Hopefully, this will demonstrate an apparently normal viable fetus of appropriate size for gestation. It may, however, demonstrate fetal death, with absent fetal cardiac pulsations and movements (missed abortion), or a molar pregnancy, with the classical ‘snowstorm’ sonographic picture. If real-time ultrasound is not available, the small Doppler instruments can be used to demonstrate the presence of the fetal circulation. Circulation should be detected from 14 weeks' gestation and, occasionally, as early as 12 weeks, although difficulty may be encountered in obese subjects. A urine or serum immunological pregnancy test can be performed but is not particularly helpful, especially since the placenta may continue to secrete human chorionic gonadotropin (hCG) after fetal death has occurred. Very high levels of hCG however, are suggestive of trophoblastic disease and, once the diagnosis has been confirmed by ultrasound, the appropriate management should be instituted.

Bed rest with mild sedation forms the initial approach to management. The volume of vaginal blood loss should be noted. In most instances, heavy vaginal bleeding is associated with progression to an inevitable abortion. In a few cases however, marked bleeding may occur and persist over days or weeks; this is usually associated with placental implantation over the cervix. In this instance, conservative measures are maintained and include transfusion when necessary. If blood loss persists and threatens the patient's health or life, termination of the pregnancy may be advised, either as abortion or by inducing labour or delivering by Caesarean section (if gestation has lasted more than 26 weeks and if intensive neonatal care facilities are available.

In the majority of cases, bleeding settles within a matter of days and mobilization can occur, with normal daily activities resuming shortly thereafter. It is common practice to caution against coitus until 2 weeks have elapsed, although there is no good scientific evidence to support this advice. The use of oral tocolysis, such as salbutamol (8 mg three times a day), has been advocated in the later gestational period, to be maintained till fetal viability has been reached (36 weeks' gestation); again, there is no evidence that the prognosis is improved as a consequence. Cases have, however, been described of the abortion of one twin and retention of the other, to be delivered more than 7 weeks later.

Once the cervix is dilated, abortion is inevitable. Some gynaecologists have suggested inserting a cervical circumsuture. The value of this is doubtful and abortion frequently follows, although there are occasional reports of success. There remains the risk of cervical laceration if cervical effacement and dilatation continue and the suture is not urgently removed. Almost invariably, uterine contractions will begin and will result in expulsion of the fetus and placenta. If contractions are slow to establish and the process becomes prolonged, or if the fetal membranes have ruptured and contractions do not start, the cervix being partially dilated, uterine activity can be augmented with intravenous oxytocics, such as oxytocin 100 mU/min, to speed up the process. Local administration of prostaglandin pessaries, such as prostaglandin E&sub2; (25 mg on one occasion) or the prostaglandin E&sub1; analogue, gemeprost (1 mg at 3-hourly intervals), may be used as an alternative approach.

When the fetus is expelled, ergometrine (0.5 mg, intramuscularly) should be given to try to ensure a firmly contracted uterus and help speed placental separation, reducing blood loss. In the absence of excessive bleeding, it is appropriate to wait a few hours for the placenta to be expelled. If it is not forthcoming, or if it has been delivered incomplete, surgical evacuation will be necessary. Following evacuation, whether spontaneous or with the aid of surgery, the uterus should be well contracted, there should be minimal bleeding and vital signs should be stable, and the patient can be discharged. Bromocriptine (2.5 mg daily for 14 days) will prevent lactation in patients who were more than 20 weeks pregnant at the time of abortion.

Follow-up should be offered to discuss possible aetiological factors that have been detected or suspected from the pathological examination of the fetus and placenta, and to advise upon future pregnancies.

This should be suspected if the uterine size is at least 4 weeks smaller than anticipated from the menstrual dates, pregnancy symptoms having disappeared, and fetal movements having ceased if previously present. Immunological pregnancy tests may not be negative, but the diagnosis can be confirmed by ultrasound. If the uterine size is not greater than that equivalent to a 15 week gestation, evacuation should be managed by cervical dilatation and uterine aspiration (using a paracervical block, regional block, or general anaesthetic), and at larger gestational sizes evacuation should be achieved by induction of a miscarriage, using oxytocics.

Following vulval and vaginal preparation with aqueous chlorhexidine or similar, and draping, bimanual examination is performed to confirm the size and position of the uterus. Intravenous ergometrine, 0.5 mg, is given to contract the uterus and reduce associated blood loss. The cervix is then stabilized with a sponge holding forceps and the cervix dilated with graduated dilators to a maximum of 10 mm. Uterine aspiration is then performed with a curette of appropriate size (to a maximum of 10 mm). The procedure is completed when no further products of conception are seen escaping along the suction tubing and the cavity feels empty with the characteristic ‘grating’ feeling experienced with the curette against the uterine wall. Subsequent care is as described for spontaneous inevitable abortion. Histological examination of the tissue removed is generally only able to confirm apparently normal placental tissue and little else.

The most appropriate and effective method of evacuating a missed abortion when the uterus is larger than a 15 week pregnancy involves the administration of prostaglandins intramuscularly or into the vagina. A number of different protocols have been described, and the most efficient are: 15(s)-15-methylprostaglandin F&sub2;&subagr;, IM, 6 hourly; prostaglandin E&sub2;, 25 mg vaginally plus IV oxytocin after 18 h; prostaglandin E&sub1;, 1 mg vaginally 3 hourly plus IV oxytocin after 18 h.

Treatment is maintained until expulsion of the pregnancy has occurred. Adequate analgesia, such as diamorphine (10 mg at 4-hourly intervals), should be provided as necessary: the miscarriage is managed as described for an inevitable abortion. If it is incomplete, a surgical evacuation to remove any retained products will be necessary.

If expulsion of the pregnancy is incomplete, either following spontaneous abortion or after prostaglandin induction in cases of ruptured membranes or missed abortion, surgical evacuation under anaesthesia of any retained pieces of placenta will be necessary. A paracervical block is usually insufficient, especially to remove a placenta retained in the fundus. Epidural, spinal, or general anaesthesia are preferred and should follow appropriate asepsis and draping; the bladder should be empty or catheterized if necessary. Removal of placental tissue is generally best accomplished by digital exploration of the uterine cavity, which reduces the chance of traumatic damage to the uterine wall: sharp instruments, such as a uterine sound, should be avoided for this reason. Following digital removal of all placental tissue, gentle blunt curettage covering all the walls and the fundus of the uterus should be performed to confirm that the cavity is empty and to remove some of the decidua. Ergometrine (0.5 mg intravenously) can be given at this stage to ensure a well-contracted uterus, firm against the curette, to reduce further unnecessary bleeding. On completion, having ensured that the uterus is well contracted and that bleeding is minimal, the volume of blood lost is estimated. If there is doubt about whether the placenta is normal, it should be sent for histological section and, if sepsis is suspected, a sample should be sent for bacteriological culture.

Following recovery from the anaesthetic, and once observations are stable, the uterus is well contracted, and bleeding is minimal, the patient may be discharged from medical care.

Trophoblastic disease, including hydatidiform mole and choriocarcinoma, may be suspected with a history of vaginal bleeding and excessive symptoms of pregnancy. The diagnosis is made by pelvic ultrasound and, once confirmed, a chest radiograph should be performed, followed by evacuation of the uterus. An intravenous line should be in place and a ready supply of cross-matched blood available, since heavy bleeding may occur at the time of abortion. The evacuation is managed as described for a missed abortion; in all cases, if prostaglandins have been used at the later gestational sizes, a postabortion uterine curettage should be performed to ensure that all pieces of placental tissue have been removed. All products of conception recovered should be examined histologically to confirm the preoperative diagnosis and to determine the precise degree of differentiation of the tumour.

Follow-up should be arranged to institute repeated assays of hCG in maternal blood, in order to monitor any persisting or recurrent disease. All cases should be registered with the appropriate tumour centre for advice on further management.

Table 1 422 lists some of the factors that have been associated with spontaneous abortion in the second trimester. In most instances, investigation of possible causes will only be considered once abortion has occurred. Specific postabortion investigations to be considered should include the following.

1.Enquiry about possible maternal infections.

2.Detection and quantitation of maternal antibodies are worth considering, but require the testing of two blood samples at 3-week intervals to detect any change in titre, while some of the viral infections, such as parvovirus, are best diagnosed by culturing samples of placental tissue.

3.A glucose tolerance test is worth performing if there are other stigmata of diabetes.

4.A hysterogram will demonstrate any developmental anomaly of the müllerian duct system, such as a septate or subseptate uterus, or more severe abnormalities, such as a bicornuate uterus (Fig. 1) 1468 or a uterus didelphys: some of these anomalies are amenable to plastic operations. Intramural and submucosal fibroids distorting the endometrial cavity and possibly resulting in abortion will also show up on a hysterogram and could be removed by abdominal or vaginal myomectomy. Cervical incompetence is extremely difficult to diagnose, although it may be suspected if the radio-opaque dye has disappeared from cavity of the uterus on a delayed film.

5.Fetal conditions predisposing to abortion, if not evident at the time, should be detected at postmortem examination, and karyotyping using fetal blood or skin will detect any chromosomal anomaly. When appropriate, parental karyotyping should be performed to check for the presence of chromosomal translocations.

Specific reasons for the abortion will be detected in few cases and, for the majority of couples, reassurance can be given for future pregnancies.

Spontaneous abortions are often quoted to occur in 10 per cent of recognized pregnancies and once 13 weeks' gestation has been reached the chance of spontaneous abortion occurring is 0.7 to 1.0 per cent. This rate appears to decline as pregnancy advances towards 28 weeks' gestation. Although spontaneous losses are more common in older women, the increase for second trimester losses does not seem to be significant. Overall, the risk of the next pregnancy being expelled spontaneously as a late abortion or preterm labour is approximately 15 per cent, and increases to 30 per cent after two consecutive losses.

Charles D, Hurry DJ. Cervical incompetence. In: Fuchs F, Stubblefield PG, eds. Preterm birth. Causes, prevention and management. New York: Macmillan, 1984: 98–111.

Elder MG, Hendricks CH, eds. Preterm labour. London: Butterworth International Medical Reviews, 1981.

Mackenzie IZ, Davies AJ, Embrey MP. Fetal death in utero managed with vaginal prostaglandin E&sub2; gel. Br Med J 1979; i: 1764.

Main DM, Main EK. Management of preterm labor and delivery. In: Gabbe SG, Niebyl JR, Simpson JL, eds. Obstetrics—normal and abnormal pregnancy. New York: Churchill Livingstone, 1986: 689–738.

Stray-Pederson B, Stray-Pederson S. Etiologic factors and subsequent reproductive performance in 195 couples with a history of habitual abortion. Am J Obstet Gynecol 1984; 148: 140–6.

Tabor A, Philip J, Marson M, Bang J, Obel FB, Norgard-Pederson B. Randomised controlled trial of genetic amniocentesis in 4606 low-risk women. Lancet 1986; i: 1287–93.

Wallenburg HCS, Keirse MJNC, Fiere HMP, Blacquiere JF. Intramuscular administration of 15 (s)-15-methyl prostaglandin F&sub2;&subagr; for induction of labour in patients with fetal death. Br J Obstet Gynaecol 1980; 87: 203–9.

Woolfson J, Fay T, Bates A. Twins with 54 days between deliveries. Case report. Br J Obstet Gynaecol 1983; 90: 685–7.