Surgical Abortion in the Second Trimester

Abstract

The development of dilatation and evacuation (D&E) as a method of second trimester surgical abortion occurred soon after abortion law reform took place in the 1960s and 1970s in Europe and the United States. Today, D&E is the predominant method of second trimester abortion in many parts of the world. Debate still exists as to whether surgical or medical methods are optimal for second trimester pregnancy termination. A continuing challenge to provision of D&E is the availability of a large enough pool of skilled providers. This article reviews the current surgical methods used in second trimester abortion, as well as their safety, advantages and disadvantages, acceptability and associated complications. Methods used to ensure safe and efficient surgical termination of second trimester pregnancies such as cervical preparation and ultrasound guidance are also reviewed.

Résumé

La technique de dilatation et d'évacuation (D&E) pour l'avortement du deuxième trimestre s'est développée rapidement après la réforme de la législation sur l'avortement dans les années 60 et 70 en Europe et aux États-Unis. Aujourd'hui, dans certaines régions du monde, c'est la méthode la plus utilisée pour les avortements du deuxième trimestre ; néanmoins, l'avortement médicamenteux est également fréquent. Le débat se poursuit pour déterminer lequel de l'avortement chirurgical ou médicamenteux est le plus indiqué en cas de grossesse du deuxième trimestre. Un obstacle récurrent à la pratique de la méthode par dilatation et évacuation est la disponibilité de soignants qualifiés en nombre suffisant. Cet article passe en revue les méthodes chirurgicales d'avortement actuellement utilisées au deuxième trimestre, ainsi que leur sécurité, leurs avantages et inconvénients, leur acceptabilité et les complications associées. Il examine également les techniques utilisées pour garantir une interruption de grossesse sûre et efficace au deuxième trimestre, comme la préparation du col, les agents pour supprimer le fłtus et le contrôle par ultrasons.

Resumen

El desarrollo de la dilatación y evacuación (D&E) como método de aborto quirúrgico en el segundo trimestre evolucionó rápidamente tras la reforma de la ley de aborto en las décadas de los sesenta y setenta, tanto en Europa como en Estados Unidos. Hoy en día, la D&E es el método predominante de aborto en el segundo trimestre en algunas partes del mundo; sin embargo, el aborto con medicamentos también es común. Aún existe debate respecto a cuál es óptimo para la interrupción del embarazo en el segundo trimestre: el aborto quirúrgico o el aborto con medicamentos. Un reto continuo a la prestación de servicios de D&E es la disponibilidad de suficientes prestadores capacitados. En este artículo se estudian los métodos quirúrgicos utilizados actualmente en los procedimientos de aborto en el segundo trimestre, así como su seguridad, ventajas y desventajas, aceptación y complicaciones asociadas. También se analizan los métodos utilizados para garantizar la interrupción quirúrgica, segura y eficaz del embarazo en el segundo trimestre, como la preparación cervical, los agentes para causar la muerte del feto y la orientación en ecografía.

Key Words:

• second trimester abortion methods
• dilatation and evacuation
• fetal demise

The first suction apparatus for uterine evacuation was introduced in the United States in 1966 and quickly adopted for use in surgical abortion. In an early cohort study there, 93% of procedures under 12 weeks gestation were performed using this method.¹ Surgical abortion in the second trimester was, however, relatively unavailable. As gestation advanced beyond 13 weeks, medical induction accounted for 77% of abortions. At 17 weeks gestation and above, the only surgical options were hysterectomy and hysterotomy, and were rarely performed.

Trans-cervical instrumentation of the second trimester uterus was considered controversial in those early days.²³ However, in the same year that these statistics were published in the US, Bierer and Steiner described the successful use of slowly expanding intra-cervical seaweed tents (laminaria) followed by manual dilatation and forceps evacuation for mid-trimester abortion in the British medical literature.⁴ This report, along with others by innovative surgeons in Europe and the United States in the ensuing years, led to the development of what is today known as dilatation and evacuation (D&E).⁵

Safety of D&E

The safety of D&E compared to available methods of medical induction was demonstrated using cohort data collected by the Joint Program for the Study of Abortion under the Population Council and the US Centers for Disease Control (JPSA/CDC) from 1970–78.⁶⁷ Induction of labour by intra-amniotic instillation of saline or urea-prostaglandin F2-alpha was shown to have higher relative risks of serious complications than D&E (RR 2.6, 95% CI 1.9–3.6 and RR 1.9, 95% CI 1.2–3.1, respectively). One randomised trial comparing D&E to intra-amniotic instillation of prostaglandin F2-alpha supported these findings.⁸ Importantly, these studies demonstrated that D&E was a skill that could be rapidly acquired by surgeons familiar with pelvic surgery through focused training and incremental advancement of gestational age thresholds.² Investigators also showed that D&E could be safely performed on an outpatient basis in non-hospital settings.^9–11 With cervical preparation occurring a few hours to 24 hours prior to surgery, the development of D&E made second trimester surgical abortion safer, faster and more cost-effective than the available medical alternative.¹²

Instillation abortion is uncommon in modern abortion care. Prostaglandin analogues such as misoprostol were found to improve the efficacy and speed of induction abortion with a reduction in side effects.¹³ The inclusion of the antiprogesterone mifepristone to medical abortion regimens further shortened the induction time, and decreased both the dose of prostaglandin required and the need for analgesia.^14–16 Use of these medications has largely replaced instillation techniques.

Few studies compare D&E with current methods of second trimester medical abortion. Autry and colleagues¹⁷ retrospectively assessed complications with D&E and a variety of induction regimens in 297 women. The most common medication used for induction abortion was misoprostol. Women treated with misoprostol experienced fewer complications than those induced with other agents, but more than women undergoing D&E (22% vs. 4%, p<0.001). However, women undergoing medical abortion were of a higher gestational age than those undergoing D&E (20.3 ± 2 weeks vs. 18.4 ± 2.2 weeks, p<0.001) which may have biased the outcomes. In addition, most of the complications seen in the medical abortion group were retained placenta, requiring surgical evacuation. This complication is less common with combined mifepristone–misoprostol protocols.

Performance of randomised trials of D&E and current medical abortion regimens has proved difficult. Where surgical termination is offered, women appear less likely to accept the possibility of randomisation to a labor induction. Grimes and colleagues¹⁸ explored the feasibility of such a trial, comparing D&E to mifepristone and misoprostol, in the United States. In 24 months, 18 women were successfully recruited to the study; 62% of eligible women declined participation. Due to the small sample size, the outcomes of the study were imprecise. However, the trial was able to demonstrate that women randomised to medical termination were more likely to experience one or more adverse events than those assigned to D&E (RR 6.0, 95% CI 0.9–40.3, p=0.05). Pain was reported as significantly higher in the mifepristone–misoprostol group (p=0.03); but there were no statistical differences in either acceptability or satisfaction between groups.

Relative frequency of D&E and medical abortion

The relative frequencies of D&E and medical abortion in the second trimester vary. In the United States, D&E predominates, used for 99% of abortions between 13–15 weeks, 95% between 16–20 weeks, and 85% at 21 weeks or later.¹⁹ Similarly, D&E is the method of choice in 75% of abortions greater than 13 weeks gestation in England and Wales.²⁰ D&E has been described as common in the Netherlands, France and parts of Australia.²¹ In contrast, in Nordic countries, Scotland and Viet Nam most abortions in the second trimester are performed medically.²²²³

The reasons for these variations are likely multifactorial. Specialised training and the maintenance of an adequate caseload are required to perform D&E safely.²⁴ Acquiring this skill can be challenging since many training programmes fail to provide adequate instruction in surgical abortion techniques.²⁵²⁶ In the United States, the Accreditation Council for Graduate Medical Education has mandated that residency curricula include induced abortion. This stipulation has increased abortion training nationwide. However, D&E remains less likely to be taught than first trimester vacuum aspiration.²⁵ Individual programmes may also decide whether abortion training is incorporated as a requirement (unless a resident demonstrates a moral objection) or as an elective (residents choose to receive training). More residents in programmes with required training gain exposure to D&E and, importantly, perform more D&Es. Achieving competency in abortion procedures during training is associated with future abortion provision, which emphasises the importance of integrated training.²⁷²⁸

Advantages and disadvantages

Procedural choice may also be affected by the availability of necessary equipment, the ability to book operating room time, facility policies, staff acceptability and patient preference.²⁹ Some providers may find it distressing to perform D&E at advanced gestations.³⁰ Compared to a medical abortion, where a physician may be largely uninvolved during the course of labour and may or may not attend the delivery, a surgeon undertaking a D&E must “deal with the second trimester fetus in an intimate, physical way”.³¹ Although nurses and other support staff may be affected by the conduct of the D&E procedure, they may find it less burdensome as their involvement is less direct. Little is known regarding women's preferences for abortion in the second trimester. Studies comparing medical abortion to vacuum aspiration in the first trimester have demonstrated that many women have a preference between methods and that obtaining their method of choice is a strong predictor of satisfaction.³² However, when randomly allocated, women without a strong preference report vacuum aspiration to be significantly more acceptable, a finding that has been shown to persist up to two years following treatment.³³

Advantages of D&E for both provider and patient are that the procedure can be scheduled on an outpatient basis and operating times are short, as opposed to the unpredictable duration of a medical abortion, which requires hospitalisation. For some women, avoiding surgery or anaesthesia with a medical abortion may outweigh these benefits. For women terminating a pregnancy due to aneuploidy or fetal anomalies, medical abortion has been advocated as it allows for bonding with the fetus after delivery, which may assist in the grieving process. However, one study comparing women self-selecting medical or surgical termination for these indications found no difference in grief resolution.³⁴ Additionally, specimens obtained from D&E abortions are adequate to confirm pathologic and cytologic diagnoses when necessary.³⁵³⁶

An evolving technique

Surgical abortion in the second trimester of pregnancy continues to evolve. Variations in methods of cervical preparation, instrumentation and adjuvant therapies such as feticide are practised to ensure safe and efficient uterine evacuation. Box 1shows the main steps involved in a D&E.

short-legendBox 1

The current nomenclature for second trimester surgical abortion is gestational-age dependent. A D&E refers to any instrumental evacuation of the pregnant uterus through the cervix at ≥13 weeks gestation.⁷ This operational definition was proposed by the US Centers for Disease Control to differentiate second trimester curettage procedures from first trimester procedures (vacuum aspiration or suction curettage). Although vacuum aspiration can be performed to 16 weeks gestation using large-bore cannulae (14–16 mm in diameter),³⁷ by convention this would be referred to as a D&E.

Cervical preparation

A guiding principle in second trimester surgical abortion is to achieve sufficient expansion of the cervical os so that the relatively larger and ossified fetal parts can be removed and the risk of injury minimised. This may be achieved by mechanical dilatation with graduated rigid dilators, slowly expanding cervical tents, or with medications, such as misoprostol or mifepristone.

There are three types of absorptive natural or synthetic cervical tents (osmotic dilators), which after insertion for several hours, either soften the cervix significantly, allowing for easier mechanical dilatation, or exert outward pressure on the endocervical canal, stretching it adequately for instrumentation. Laminaria are the most common osmotic dilator used. Made of compressed stalks of seaweed (Laminaria japonica or Laminaria digitata), they range from 2–10 mm in diameter and 60–85 mm in length. After insertion, laminaria expand to three or four times their dry diameter and may also ripen the cervix by inducing the release of prostaglandins.³⁸ Although some dilation is measurable three hours after insertion, 12–24 hours are required for maximal expansion of the tents. Dilapan-S is a synthetic rod made of polyacrylate-based hydrogel and is available in three sizes (3×55 mm, 4×55 mm and 4×65 mm). Like laminaria, Dilapan expands radially, however this occurs more rapidly, consistently and to a greater degree, meaning that fewer Dilapan can be inserted for shorter periods of time. Lamicel is a stiff polyvinyl alcohol polymer sponge impregnated with magnesium sulfate. The Lamicel sponge rapidly absorbs fluid from the cervical tissue and stimulates collagen breakdown to soften the cervix.³⁹

Prior to D&E at 14–16 weeks gestation, a single Lamicel leads to comparable degrees of difficulty in subsequent dilatation and rates of adequate dilation as multiple laminaria.⁴⁰ A small randomised trial of overnight Dilapan or Lamicel prior to abortion at 13–16 weeks gestation demonstrated that Dilapan produces greater dilation, which is consistent with differences in mechanism of action.⁴¹ Lamicel is approved for use to 16 weeks gestation. It is less commonly used beyond this gestational age because of concern regarding the lack of radial force exerted by the sponge when greater cervical dilation is required.²¹

Laminaria or Dilapan are more commonly used prior to D&E at higher gestations. They are associated with a lower risk of cervical trauma and uterine perforation than mechanical dilatation.⁴²⁴³ Few studies have directly compared laminaria to Dilapan. One series evaluated outcomes in 1,001 women who underwent D&E at 13–25 weeks gestation, and received either laminaria or Dilapan 18–24 hours before surgery.⁴⁴ No differences were found in procedure time, blood loss or need for additional dilatation. However, more laminaria were required to achieve the same amount of dilation achieved with Dilapan. The older formulation of Dilapan used in this study was prone to difficult removals and fragmentation which occurred in 6% of cases. The current, commercially available product (Dilapan-S) elongates with traction in order to avoid fracturing and entrapment.

Preference for Dilapan or laminaria is largely provider-dependent. There is no standard protocol for the use of osmotic dilators. In general, the number used and length of time the dilators are left in situ increases with gestational age. Some providers use a combination of Dilapan and laminaria or perform serial insertions over two or more days, depending on the desired amount of dilation.^*

The placement of osmotic dilators requires insertion of a vaginal speculum, followed by cleansing of the cervix and often, administration of a local anaesthetic. The dilators are then inserted into the cervical canal using gentle pressure. Discomfort can occur during insertion and many women experience cramping as the dilators expand. This process may be viewed by some women and clinicians as unsatisfactory. In addition, osmotic dilators are not available worldwide and are typically left in place overnight, which may present a barrier for women travelling long distances to access care.²³ The prostaglandin analogue misoprostol has been investigated as an alternative. Misoprostol is cheap, widely available, and can be (self-)administered by a variety of routes including oral, vaginal, sublingual and buccal.

The ideal dose, route and upper gestational age limit of misoprostol administration before D&E has not been determined. Like Lamicel, misoprostol softens rather than expands the cervical os. Misoprostol can induce uterine cramping and gastrointestinal side effects, which may affect acceptability. The risk of inadvertent abortion outside of the health care facility is another concern when misoprostol is administered at home.

Buccal misoprostol, in doses ranging from 400–800 mcg, appears to achieve adequate cervical preparation up to 18 weeks gestation.^23,45,46 One randomised trial compared 400 mcg vaginal misoprostol 3–4 hours before surgery to 3–6 medium laminaria placed overnight in women at 13–16 weeks gestation.⁴⁷ The use of misoprostol was associated with less pre-operative cervical dilation, longer operating times and a more frequent need for additional manual dilatation. One woman required an additional dose of misoprostol; however, all procedures were able to be completed and more women preferred a same-day cervical preparation regimen.

The combined use of misoprostol and osmotic dilators has been explored to reduce the number of visits and duration of time necessary for cervical preparation. One placebo-controlled, randomised trial compared overnight laminaria with 400 mcg buccal misoprostol or placebo 60–90 minutes pre-operatively at 13–20 weeks gestation.⁴⁸ Significant improvement in cervical dilation was only observed in the subset of subjects with pregnancies 19 weeks and above, the magnitude of which was two dilator sizes or similar to what would be achieved with a second set of laminaria. One small case series has reported outcomes with 400–800 mcg vaginal misoprostol in combination with Dilapan 4–7 hours prior to D&E at 20–24 weeks gestation.⁴⁹ Of 25 cases audited, all were able to be completed. However, 19 (76%) required further cervical dilatation intra-operatively and three cervical lacerations requiring repair were encountered. Further research is needed to determine the utility of combined regimens, particularly at shorter intervals.

Mifepristone, taken orally 24–48 hours prior to first trimester surgical abortion, has been shown to be as effective as gemeprost⁵⁰ and more effective than vaginal misoprostol⁵¹ at ripening the cervix. Some surgeons exploit the priming effect of mifepristone before D&E, even preferring it to misoprostol (C Paterson, personal communication, 26 June 2008). Carbonell and colleagues have published the only study which examines the utility of mifepristone prior to D&E.⁵² Women at 12–20 weeks gestation were randomly allocated to either 200 mg oral mifepristone 48 hours before 600 mcg vaginal or sublingual misoprostol 1.5–2.5 hours prior to D&E or misoprostol alone. Thirty per cent of women in the mifepristone–misoprostol group and 54% of those in the misoprostol-only groups also received one Dilapan intra-cervical tent. Greater mean pre-operative cervical dilation was achieved with the addition of mifepristone; however, notably, several women in this group aborted spontaneously before surgery, two after receiving mifepristone alone and 15 pre-treated with mifepristone but after administration of misoprostol. There are as yet no studies that assess mifepristone alone for cervical preparation before D&E or which compare mifepristone to other methods of cervical preparation, such as osmotic dilators. In addition, mifepristone currently has a higher cost and more limited availability globally than misoprostol, which may have an impact on its wider use.

Commonest method of D&E and variants^*

The commonest method of D&E involves disarticulation and removal of the fetus through the prepared cervix using strong, elongated extraction forceps. Passive or active drainage of amniotic fluid by electric suction occurs prior to removal of the fetus and placenta and a final suction curettage is performed at the end of the procedure to remove any remaining blood or pregnancy tissue. Castleman and colleagues have described a version of D&E adapted for use in low-resource settings. The protocol uses buccal misoprostol for cervical preparation, forceps and a manual vacuum aspirator (MVA) instead of an electric suction machine.²³ In a demonstration project in Viet Nam, this protocol was shown to be effective between 13–18 weeks gestation. Another variant is known as intact dilatation and extraction (D&X). Wide cervical dilatation is achieved with two or more days of laminaria. This is followed by an assisted breech delivery of the trunk of the fetus, decompression of the calvarium, and delivery of the fetus otherwise intact. The greater degree of cervical dilation (reported as a median of 5 cm, range 2–10) and mode of evacuation mean that fewer instrument passes are made with D&X than D&E, ostensibly reducing the risk of complications and making the procedure faster. In a retrospective comparison of D&X and D&E, procedure time, complications, and blood loss were not statistically different, even though the cases performed intact were of significantly higher gestational ages.⁵³ Three women had complications requiring admission to the surgical intensive care unit, including the only case of uterine perforation in the cohort, all of which were in the D&E group.

Agents for fetal demise prior to D&E

Feticidal agents are used before D&E by some surgeons. The softening of cortical bone that occurs after fetal demise is proposed to reduce the amount of cervical dilation necessary and to make the procedure easier and faster, thus reducing the risk of complications. The true incidence of use is not known and the gestational age at which fetal demise is induced differs among practitioners; however, it is typically reserved for terminations above 18 weeks gestation. The most common medications used are potassium chloride and digoxin. Potassium chloride is injected trans-abdominally under ultrasound guidance into the fetal cardiac ventricle or thorax. Digoxin can be administered by these routes into the amniotic fluid or other fetal tissues.

Intra-cardiac potassium chloride is the most common agent used in the United Kingdom to comply with the recommendation by the Royal College of Obstetricians and Gynaecologists for induced fetal demise before medical abortion at ≥22 weeks gestation.²⁴ Intra-cardiac potassium chloride for this indication in the UK has led to extension of its use before surgical abortion. Digoxin use is more commonly described in the United States.²¹ Administration of either medication is safe.⁵⁴⁵⁵ However, while intra-cardiac potassium chloride has been reported to be uniformly effective, failure to achieve fetal demise has been reported with digoxin, which appears to be related to both the dose and route of administration.⁵⁶

Data supporting the effect of fetal demise on the safety and efficiency of D&E are limited. Large, non-comparative case series utilising intra-fetal digoxin, urea and multiple sets of laminaria for cervical preparation reported a low incidence of complications.⁵⁷⁵⁸ However, one randomised, controlled trial showed no difference between intra-amniotic digoxin and placebo with regard to complication rates or procedure duration when administered prior to D&E at 20–24 weeks gestation.⁵⁹ Symptoms after injection were similar in both groups, with the exception of vomiting, which was significantly more frequent in the digoxin group (16.1% vs. 3.1%, p=.02). Women in this study, did, however, report a strong preference for fetal death prior to the abortion (92% in both groups).

Complications of D&E

Bleeding is the most common complication of second trimester surgical abortion and the risk of haemorrhage increases with gestational age.⁶⁰ Excessive blood loss may occur as a result of injury to the uterus or cervix, an incomplete procedure, or failure of the uterus to contract adequately after evacuation. The risk of disseminated intravascular coagulation (DIC) is also elevated with second as compared to first trimester surgical abortion.⁶¹

Approaches to reducing blood loss include the use of uterotonics, such as oxytocin or ergot derivatives, and local vasoconstrictors. Oxytocin in crystalloid may be initiated at the beginning or end of the operation as an intravenous infusion. Oxytocin or methylergometrine can also be administered as a bolus intramuscular, intravenous or intra-cervical injection. The utility of these agents in reducing blood loss with D&E is unclear, although they are commonly used. One randomised trial of intravenous methylergometrine did not demonstrate a decrease in blood loss.⁶² An alternative is the administration of vasopressin injected para- or intra-cervically before surgery. In a randomised, double-blind, placebo-controlled trial, four units of vasopressin combined with the cervical anaesthestic significantly decreased blood loss during D&E at greater than 14 weeks gestation.⁶³ Compared to placebo, the effect of vasopressin on mean blood loss was greater as gestational age advanced.

Uterine perforation is a potentially serious complication of second trimester surgical abortion, with a reported incidence of 0.2–0.4%.^60,64,65 Although perforations during first trimester abortion may be managed expectantly or with laparoscopy, those occurring with the larger instruments used for D&E typically require exploratory laparotomy and are often associated with bowel injury, haemorrhage, and, in some cases, hysterectomy.⁶⁰⁶⁶ The use of routine intra-operative ultrasound guidance during D&E has been shown to reduce the rate of perforation.⁶⁵

Not all providers use ultrasound during D&E or have this technology available in the procedure room. Completion of the D&E therefore relies solely on a clinician's tactile sensation. Most evacuations can be completed from the lowest part of the uterine cavity and perforations are more likely to occur when instruments are passed high into the uterus. Thus, diligent attention to correct placement of instruments is important, whether or not ultrasound is used. Under-estimation of gestational age is also associated with perforation, so accurate determination of gestational age is essential. This is typically ascertained using sonographic measurements.⁶⁶ Ultrasound is also useful in cases where the surgeon experiences difficulty locating or removing fetal parts. Thus, even if an ultrasound machine is not used regularly in the operating room during the surgery, having one on-site and accessible is helpful.⁶⁷

Use of a “no-touch” technique, which ensures that the tips of instruments never contact non-sterile surfaces before entering the uterus, inspection of the evacuated tissue to verify complete removal, and the provision of routine antibiotic prophylaxis are other measures used to prevent complications. Although some complications such as perforation remain more serious when they occur during D&E, overall complication rates with second trimester surgical abortion have been reported as being no higher than those occurring in the first trimester.⁶⁸

Historically, concern was raised regarding the degree of pre-operative cervical dilation required for D&E and poor pregnancy outcomes such as cervical incompetence, miscarriage, and preterm birth. Recent retrospective case series and one retrospective case–control study did not find an association between D&E and future pregnancy complications.^69–71 Preterm delivery did appear less likely when greater pre-operative cervical dilation was achieved with laminaria, possibly because of less associated cervical trauma than with mechanical dilatation.⁷⁰ There are no data regarding the influence of misoprostol, which commonly requires additional manual dilatation when used for cervical preparation before D&E, and the outcome of subsequent pregnancies.

Surgical abortion in the second trimester can be performed using light or deep intravenous sedation or local anaesthesia injected para- or intra-cervically, often with the addition of oral, non-steroidal anti-inflammatory drugs, narcotics or anxiolytics. Data from the 1970s demonstrated a major complication rate of 0.72 per 100 abortions for general anaesthesia and 0.32 per 100 abortions for local anaesthesia.⁷² Since that time, advances have been made in anaesthetic technique and monitoring. More recent data on the comparative risks of local vs. general anaesthesia with D&E are lacking. However a case series reporting on 170,000 first trimester surgical abortions found comparable rates among anaesthetic techniques.⁷³

Surgical methods of second trimester abortion are common today and the techniques used continue to be refined. Debate still exists as to whether surgical or medical abortion is optimal for second trimester pregnancy termination, and a continuing challenge to provision is the availability of a large pool of skilled providers. However, for those women able to access this option, D&E is a safe, efficient and cost-effective means of pregnancy termination in the second trimester.

Notes

* There are several protocols for Dilapan and laminaria insertion by gestational age, which are described in detail in: Paul M, Lichtenberg ES, Borgatta L, et al, editors. A Clinician's Guide to Medical and Surgical Abortion. New York: Churchill Livingstone, 1999.

* Baird TL, Castleman LD, Hyman AG, et al. Clinician's Guide for Second-Trimester Abortion (2nd ed. Chapel Hill NC: Ipas, 2007) includes service delivery and management considerations and protocols for both D&E and medical abortion procedures, and is applicable in high- or low-resource settings.