Allan Templeton, M.D., and David A. Grimes, M.D.
N Engl J Med 2011; 365:2198-2204December 8, 2011DOI: 10.1056/NEJMcp1103639
“A 22-year-old student presents to her primary care physician with an unintended pregnancy at 9 weeks of gestation and requests an abortion. She is aware of both medical (drug-induced) and surgical methods of terminating a pregnancy and wants to know which approach would be recommended. She also asks whether either method will affect her future reproductive health. What would you advise?”
THE CLINICAL PROBLEM
Induced abortion is one of the most common medical interventions. In the United States, approximately 1.2 million abortions were performed in 2008.1 In the United Kingdom, more than 200,000 abortions are recorded annually.2,3 About one of three women will have had an induced abortion by the time she reaches menopause.
Approximately 90% of abortions are performed in the first trimester because the pregnancy is unintended or unwanted. A small but important proportion (1 to 2%) of abortions are performed later because of a fetal abnormality (e.g., anencephaly, trisomy, or myelomeningocele)4 or serious illness (e.g., cancer or pulmonary hypertension) in the woman.
Until recently, the main method of abortion was surgical, but since 1992 in the United Kingdom and since 2000 in the United States, medical abortion has become increasingly available. Medical abortion involves the combined use of the progesterone antagonist RU-486 (now known as mifepristone), which initiates the abortion, and a prostaglandin, which causes uterine contractions and empties the uterus.5 Of all abortions, medical abortions account for approximately 10% in the United States, 40% in England, and 70% in Scotland
Easy access to safe, legal abortion services is important to the general health of women and their families.6 Women will seek abortion whether it is legal or not, and the morbidity and mortality associated with illegal abortion remain high. This review article does not address the debate over abortion or the adverse health effects associated with restricting access to abortion services, but it assumes a woman’s right to have her pregnancy terminated if the relevant legal requirements are met.
STRATEGIES AND EVIDENCE
Medical abortion has evolved considerably since mifepristone was first licensed for use in Europe in the early 1990s. Whereas initial results with mifepristone alone were disappointing, effectiveness improved dramatically when a prostaglandin or prostaglandin analogue was administered 1 or 2 days after mifepristone.5 Today, the most commonly used prostaglandin is the prostaglandin E analogue misoprostol; its advantages include stability at room temperature (which facilitates short-term storage), possible administration by several routes (vaginal, buccal, sublingual, and oral), and low cost. Misoprostol (and other prostaglandin analogues) can be used alone as abortifacients but are less effective than mifepristone plus misoprostol.7 Similarly, in areas where mifepristone is still unavailable, such as Canada, methotrexate followed by misoprostol can be used, although this regimen is not as effective as the mifepristone–misoprostol regimen.
Initially, misoprostol at a dose of 400 μg was administered orally, and this continues to be the preferred route of administration in France, although in the United Kingdom and United States, vaginal administration of misoprostol at a dose of 400 to 800 μg was shown to be more effective, particularly at gestational ages of more than 7 weeks.8 If needed, the administration of a second dose of misoprostol, either vaginally or orally according to the amount of vaginal bleeding, further increased the likelihood of complete abortion.9 Alternative routes of misoprostol administration, including sublingual10 and buccal,11 have been shown to be as effective as vaginal administration, although side effects of prostaglandin, which are mainly gastrointestinal, are more frequent.12However, many women prefer the convenience of these approaches, which do not involve vaginal administration of tablets.
The Food and Drug Administration–approved dose of mifepristone is 600 mg, but in a randomized trial, the effectiveness of a 200-mg dose was similar to that of a 600-mg dose for medical abortion at all gestational ages.13 An initial vaginal dose of 800 μg of misoprostol is widely used; although a lower dose may be sufficient in many women and is associated with fewer side effects, there is no effective way to predict which women will require the higher dose.
Some studies have indicated that the interval between administration of these two medications can be reduced to 24 hours or less, but most clinics wait 24 to 48 hours after mifepristone administration to administer misoprostol. Once medications have been administered, abortion can be completed at home. The safety and efficacy of medical abortion completed at home are similar to the safety and efficacy of medical abortion in the clinic,14 although ready access to emergency facilities is still required. This is now the usual approach to medical abortion in the United States and is increasingly common in Europe.15
Since the 1970s, suction curettage, also known as vacuum aspiration, has been the standard method of abortion in the United States, where more than 80% of abortions in the first trimester are performed surgically (Figure 1). Vacuum aspiration is a safe procedure that is associated with low rates (<1%) of serious complications (e.g., uterine perforation) and of retained products of conception. Preparation with misoprostol is an effective means of softening and dilating the cervix.16Whether this approach reduces morbidity is not yet established, although the use of osmotic dilators (Laminaria digitata or L. japonica) has been associated with significantly lower rates of uterine perforation17 and cervical injury, as compared with no cervical preparation. Given the rarity of these complications, cervical preparation is not routinely performed before vacuum aspiration, although a recently updated guideline recommends such preparation in all women who are undergoing abortion.18
Local anesthesia (usually with lidocaine) is customary for vacuum aspiration in the United States, since large case series suggest that hemorrhage, cervical injury, and uterine perforation occur more frequently when general anesthesia is used.19 Either electric or manual vacuum aspiration can be used, the latter with the use of a syringe with a valve and plunger that lock when the vacuum is created. The two techniques appear to have similar effectiveness and acceptability,20 although manual vacuum aspiration is preferentially used at earlier gestational ages, since the bulkier uterine contents after 9 weeks limit its efficacy. The uterine contents are aspirated with either a rigid or a flexible plastic cannula; the two types of cannulae have similar efficacy. Although a sharp metal curette has been used to check the completeness of the operation, there are no data to provide support for this practice. The aspirated tissue should be immediately inspected to confirm the presence of the trophoblast and to minimize the chance of an undetected ectopic pregnancy.
Deep sedation (with the use of a tranquilizer, narcotic, or both) or general anesthesia is used less often for first-trimester procedures than for second-trimester procedures. Oral analgesics, such as nonsteroidal antiinflammatory drugs, are commonly administered before the procedure. A paracervical block is widely used. Data are lacking to provide support for this practice, although less pain is reported with a paracervical block that is combined with intravenous sedation than with intravenous sedation alone.21 A supportive environment, including frequent explanation and reassurance by the medical and nursing staff, is known to decrease the perception of pain and the need for analgesia.22
Comparison of Medical and Surgical Abortion
Most abortion providers offer both medical and surgical options for abortion at up to 9 weeks of gestation. Few randomized trials have compared the two approaches.23 Trials in the United States24 and the United Kingdom25 (involving abortions at up to 13 weeks of gestation) showed that women found medical abortion less acceptable than surgical abortion. An earlier trial in Scotland showed equal rates of acceptability for medical and surgical abortion at up to 50 days of gestation, with a lower rate of acceptability for medical abortion at 50 to 63 days of gestation.26 A subsequent randomized trial comparing the two procedures late in the first trimester showed that women undergoing medical abortion were significantly, albeit modestly, less likely to say they would opt for the same method again (70% vs. 79%).27
These results may be explained by the observations that in the trials comparing medical with surgical abortion, medical abortion resulted in more pain, more prolonged bleeding (up to 2 weeks after administration of misoprostol), and a slightly higher failure rate. At approximately 9 weeks of gestation, 2 to 5% of women undergoing medical abortion will require a repeat procedure to complete emptying of the uterus, as compared with 1% of women undergoing surgical evacuation. A registry-based study from Finland involving more than 4000 women likewise indicated that bleeding and incomplete abortion were more common with medical than with surgical abortion, although it also showed higher rates of rare complications requiring major surgery after surgical approaches;rates of infection and other serious complications were similar.28 Hemorrhage requiring blood transfusion is rare (occurring in 0.1% of women) after both medical and surgical abortion.18
Although women value having a choice in the method of abortion, the factors that determine an individual woman’s decision are not always clear. Some women prefer surgical methods that are simple, quick, and associated with a low risk of complications or failure. Others may favor medical methods because they do not involve surgical instrumentation and may appear to be more natural (i.e., more like a miscarriage).
Assessment before Abortion
Once the woman’s choice to proceed with an abortion has been clearly established and written informed consent has been obtained, there is no need for further delay, which may only increase the risk of complications. Counseling should be offered only if the woman requests it or there is a perceived need for it.18 The blood-group rhesus type should be determined and Rh immune globulin should be administered in Rh-negative women. Cytologic screening of the cervix and screening for sexually transmitted diseases should be offered as appropriate. Ultrasonographic examination of the uterus is common, but it is not required routinely before a first-trimester abortion is performed. 29
Prevention of Infection
Antibiotic prophylaxis at the time of abortion significantly reduces the likelihood of infection after vacuum aspiration.30 A randomized, controlled trial showed that prophylaxis was more effective and less expensive than a screen-and-treat approach for chlamydia, gonorrhea, and bacterial vaginosis.31 Doxycycline is widely used, and the best evidence provides support for only a single dose for 24 hours of coverage, although some clinicians prefer presumptive treatment of chlamydia with doxycycline (usually at a dose of 200 mg daily for 7 days); a single 1-g dose of azithromycin can be used instead, but it is more expensive. In the United Kingdom, metronidazole is administered in addition to doxycycline or azithromycin at the time of either medical or surgical abortion, but there are no data to provide support for this routine practice.
Data from randomized, controlled trials of antibiotic prophylaxis with medical abortion are lacking. However, a large before-and-after study of clinics providing medical abortion showed a marked decline (93%) in the rate of serious infections after implementation of routine antibiotic prophylaxis and a change in the route of misoprostol administration from vaginal to buccal.32 It is increasingly common for antibiotic prophylaxis to be used at the time of abortion, whether medical or surgical.
Subsequent Health and Reproductive Risks
Few long-term sequelae are evident after abortion, and the morbidity and mortality are lower with induced abortion (either medical or surgical) than with pregnancy carried to term.33
Induced abortion is not associated with an increased subsequent risk of ectopic pregnancy, placenta previa, infertility, or miscarriage.18 An association between induced abortion and a subsequent risk of preterm birth, which increases with the number of abortions, has been reported18; however, data from prospective cohort studies have not confirmed this finding. There are no data to suggest that medical abortion differs from surgical abortion with respect to these risks.
Two further putative risks warrant mention because they are a source of concern to people who are opposed to the availability of abortion. One is that abortion might be associated with an increased risk of breast cancer. In 2003, the National Cancer Institute reviewed the published data and reported that induced abortion is not associated with an increase in breast cancer risk. The Collaborative Group on Hormonal Factors in Breast Cancer34 later published a reanalysis of 53 studies from 16 countries involving 83,000 women with breast cancer and concluded that pregnancies that end in spontaneous or induced abortion do not increase the risk of breast cancer. This group also observed that the retrospective recording of abortion tended to yield misleading results that supported an association between abortion and breast cancer. Subsequently, two large cohort studies, one in Europe35 and the other in the United States,36 showed no increase in the risk of breast cancer among women who had undergone one or more abortions, regardless of the age of the women, gestational age at the time of abortion, or the number of abortions.
A second putative risk is that induced abortion may have adverse effects on women’s mental health. Most women will find it difficult to make the decision to have an abortion, and many will find the experience stressful and unpleasant. However, the most common emotional response after abortion is a profound sense of relief, although some women have lingering feelings of sadness and regret.18The main predictor of mental health problems after abortion is mental health difficulties preceding the pregnancy. Systematic reviews of observational studies37-39 and a more recent large, population-based, cohort study using linked Danish registries40 concluded that there was no increased risk of mental disorders after an induced abortion in the first trimester. Few studies have compared medical abortion and surgical abortion with respect to subsequent mental health problems, although one trial showed that women’s psychological health at 2 years did not differ according to the type of abortion.41
The prevention of a subsequent unintended pregnancy and of the need for repeated abortion is important. In this respect, the insertion of an intrauterine device (IUD) immediately after abortion, which is safe and acceptable to women,42 has been shown in observational studies from three continents43-45 to be associated with significantly fewer repeated abortions than the use of other methods of contraception. Furthermore, a recent randomized trial comparing immediate with delayed IUD insertion after uterine evacuation showed that immediate insertion resulted in higher rates of IUD use at 6 months.46 Although the evidence at this stage relates to surgical abortion, the same approach should be considered after medical abortion.18
This is a very young patient with an unwanted pregnancy. As physicians, our duty is to advise and not talk about whether abortion is right or wrong. When a patient has made a decision about her health, we must advise but never force her to do what we consider to be correct.
When we have a patient with self doubt we must act and give our medical opinion from the medical point of view without mixing too personal.
In this specific case it is important for the patient to be familiarized with all the
methods available to carry out the abortion. We must take into account the patient’s medical history, age, past illnesses and a good scan to recommend the best treatment available to her.
It is very important that the patient understands the risks of abortion. Therefore it is very important to sign the consent where she says she wants to perform this procedure. It is also important that the patient knows the prevention of a subsequent unintended pregnancy and of the need for repeated abortion is important. In this respect, the insertion of an intrauterine device (IUD) immediately after abortion is safe and acceptable to women. This has been shown in observational studies from three continents associated with repeated abortions. Significantly fewer than the use of other methods of contraception. Furthermore, a recent randomized trial comparing immediate with delayed IUD insertion after uterine evacuation immediate insertion Showed That Resulted in higher rates of IUD use at 6 months. Although the evidence at this stage relates to surgical abortion, the same approach should be considered after medical abortion.
Few randomized trials have compared short- and long-term outcomes of medical and surgical abortion; the very low rate of serious complications with either type of early abortion means that very large samples would be needed to detect clinically important differences. Data are lacking from large multinational studies comparing the cost-effectiveness and acceptability in different clinical settings of medical and surgical approaches according to standard protocols (mifepristone–misoprostol and vacuum aspiration). The most cost-effective antibiotic regimens for both medical and surgical abortion remain unknown. Whether routine rather than selective cervical preparation before surgical abortion significantly reduces morbidity has not yet been established, although a randomized trial to assess this question is under way (World Health Organization). More study is needed to resolve the question of whether an increase in the risk of preterm birth is associated with a history of either medical or surgical abortion.
Guidelines for first-trimester abortion have been published by the WHO, the National Abortion Federation, the Society of Family Planning, and the Royal College of Obstetricians and Gynaecologists. The recommendations in this article are generally concordant with these guidelines.
The patient described in the vignette, with a pregnancy at 9 weeks of gestation, should be offered the choice of a medical or surgical abortion. Both are safe with respect to short- and long-term sequelae. Medical abortion is associated with more pain and bleeding and a higher risk of incomplete abortion, whereas the risk of rare complications requiring major surgery is higher after surgical approaches. Antibiotic prophylaxis has well-established benefits in suction curettage and may also be useful in medical abortion, although this is less certain. The patient can be reassured that the best evidence indicates no long-term psychological harm, impairment of future fertility, or increased risk of breast cancer associated with abortion. The insertion of an IUD at the time of the abortion should be recommended to prevent another unintended pregnancy.
Article found in “The New England Journal of Medicine”
CAROLINA PÉREZ BENITO