South Africa's Choice on Termination of Pregnancy Act,¹ which became law on February 1, 1997, established a woman's right to an abortion on request through the 12th week of gestation. It allows abortion from 13 through 20 weeks' gestation if a continuing pregnancy would pose a risk to the woman's mental or physical health, if it would end in the birth of an infant with a severe mental or physical abnormality, if the pregnancy resulted from rape or incest or if carrying it to term would significantly affect the woman's social or economic situation. For longer gestations, abortion is permitted only if continuing the pregnancy would endanger the woman's life or result in a severe fetal malformation or a risk of fetal injury.

When the parliament passed the law in October 1996, it intended to ensure accessible and available abortion services for all women, particularly those who are poor or who were disadvantaged during apartheid, who are the most likely to suffer complications or die from an unsafe abortion. In part, the government was responding to national-level research conducted in 1993 indicating that approximately 425 women died each year in public hospitals while being treated for complications resulting from clandestine, unsafe abortions. This research also showed that, of the nearly 45,000 women admitted to public hospitals with incomplete abortions each year, at least one-third had medical complications related to unsafe abortion.²

The law stipulates that registered midwives—key staff in health care facilities, particularly at the primary level—as well as medical practitioners can perform abortions up to 12 weeks' gestation. Thus, training and certification of registered midwives throughout South Africa were identified as critical steps toward making high-quality abortion services accessible to all women.³

In this article, we summarize the main activities of the Midwifery Abortion Care Training Programme, which is the first effort in South Africa to train midwives to provide abortion services. We also outline the major findings of a process evaluation of the quality of midwives' practices and offer recommendations for continuing and strengthening the training and supervision of midwives in abortion care throughout South Africa.

BACKGROUND

As is true with any new and progressive law, implementing the Choice on Termination of Pregnancy Act posed challenges to national and provincial authorities,⁴ such as a large demand for the new services throughout the country, opposition among some health care professionals to abortion services and to those who provide them, and a lack of health care personnel trained to provide abortions.

As expected, after the law was passed, increased numbers of legal abortions were requested and performed, indirectly indicating that the incidence of unsafe abortion was declining. However, women's access to safe services remained restricted and unequal, as not all hospitals and clinics in the country were prepared to meet this demand. In the first three months after the law was passed, 60% of all legal abortions were performed in Gauteng Province, where South Africa's capital, Johannesburg, is located. After a year, only one-third of the hospitals and clinics designated by the Department of Health to provide abortion services were actually doing so. Of the 31,312 legal abortions performed in 1997, almost all were carried out in hospitals located in urban areas, as services were not available at that time in the community-based clinics that provide most primary care in South Africa.⁵

To counteract resistance among health care workers, the Planned Parenthood Association of South Africa, the Reproductive Health Research Unit (RHRU) of the University of the Witwatersrand and the Reproductive Rights Alliance conducted workshops with more than 4,000 health care providers throughout the country during the year after the law was passed. The workshops addressed providers' feelings about abortion, educated them about the provisions of the law and encouraged them to approach abortion in a nonjudgmental way and to treat women seeking an abortion with dignity and respect. In Cape Town, 70% of workshop participants said they would be able to interact with patients having an abortion "quite a bit" or "a lot" better than before attending the workshop.⁶

In 1997, authorities in all of the provinces reported difficulty in implementing the new Act because of the lack of health care providers trained to provide abortion care. To ensure that providers would be trained in all provinces and at all types of facilities, in 1998 the government established the National Abortion Care Programme, which encompassed the Midwifery Abortion Care Training Programme. The RHRU was responsible for coordinating the national program, which was carried out through a partnership among the Maternal, Child and Women's Health Directorate of the Department of Health, the RHRU and provincial health departments and academic institutions. Ipas, an international nongovernmental organization with extensive experience in training and research on abortion care, collaborated in the design of the training content and process as well as in the evaluation of midwives' skills.

The main purpose of the program was to develop the capacity of public clinics and health centers to provide safe, high-quality and accessible abortion services, treatment of abortion complications, contraceptive services and counseling, and other reproductive health services abortion clients need. In doing so, the program aimed to bring services closer to the communities where women live, particularly poor women and women living in rural areas.

The Midwifery Abortion Care Training Programme consisted of three major activities: developing an abortion care curriculum and training manual; training registered midwives in comprehensive abortion care services; and conducting a process evaluation of the quality of midwives' service provision after they had been trained.

CURRICULUM AND MANUAL DEVELOPMENT

RHRU was charged with developing a core curriculum for training registered midwives in abortion care services, and in July 1998, it submitted the curriculum to the South African Nursing Council.

After the curriculum was approved, RHRU developed a training manual, which was reviewed by, among others, representatives of the nursing council, midwifery training institutions, medical schools, national and provincial reproductive health departments, practicing midwives, academic institutions involved in reproductive health programs and reproductive health and reproductive rights organizations. The manual was finalized in October 1998.⁷

The curriculum emphasizes that abortion care services should not be provided in isolation, but rather as an integral component of comprehensive reproductive health care. It can be used to train registered midwives to:

• provide comprehensive abortion services to women with a normal pregnancy of no more than 12 weeks' gestation and to treat incomplete abortion among women with an equivalent uterine size;

• stabilize and refer women with abortion complications and a uterine size of more than 12 weeks' gestation, and refer women who are seeking an abortion who are more than 12 weeks pregnant to trained medical practitioners in hospitals;

• provide postabortion contraceptive services and follow-up care to women receiving abortion services; and

• provide other reproductive health services (including identification and treatment of sexually transmitted diseases and reproductive tract infections), as needed, to women receiving abortion services.

ABORTION CARE TRAINING

As required by the South African Nursing Council, midwives are considered for certification in abortion care after completing 160 hours of training—80 hours of theoretical training and 80 hours of clinical training under the supervision of experienced practicing physicians in accredited hospitals. The clinical training must be completed within three months of the theoretical training.

Training of registered midwives took place from November 1998 through May 1999. During the first course, RHRU and Ipas trained 22 midwives to provide abortion services and to act as trainers for other midwives in their provinces. The two-week course emphasized both clinical and psychosocial skills and included a didactic introduction to clinical issues and abortion techniques, classroom instruction in counseling and interpersonal communication skills, and training in clinical techniques using pelvic models. RHRU conducted similar national-level workshops in March and May 1999. (Participants in these workshops were not trained to be trainers.) After the theoretical training, all midwives received manual vacuum aspiration kits to use in their clinical training and to provide services once they were certified. Provincial authorities were responsible for ensuring that clinics provided midwives with space and equipment for their services.

By the end of 1999, at the time of our evaluation, 81 (88%) of the 92 midwives who had participated in the two-week theoretical training had completed their clinical training and were certified to provide abortion care services (Table 1). Sixty-nine (85%) of the certified midwives were providing services in 39 public-sector health care facilities and in three Marie Stopes International clinics in South Africa's nine provinces. Gauteng province had the most trained, practicing midwives (16), mainly because a very dedicated obstetrician in Gauteng had taken the initiative to train some midwives using the approved manual outside of the national workshops. The other densely populated provinces, KwaZulu Natal and Western Cape, had the fewest (five each).

Twelve trained midwives were not practicing for various reasons: One had died, one had left South Africa, one had left the public sector for the private sector, one had been transferred to night duties, one was on sick leave, others had not yet been able to establish services and a few had experienced administrative barriers that prevented them from providing services.

EVALUATION

As part of the process evaluation of the Midwifery Abortion Care Training Programme, we conducted a comprehensive evaluation of the quality of trained midwives' abortion care service provision. The results have been used to inform training and supervision activities throughout the country. Because the National Abortion Care Programme and the Midwifery Abortion Care Training Programme are still relatively new, an evaluation to determine the impact of the services midwives are providing has not yet been conducted.

Methods

We conducted our evaluation from October 1999 through January 2000. We visited 27 of the public facilities where certified midwives were providing abortion care services—eight primary care facilities (clinics or health centers), 13 secondary care facilities (district hospitals for general care and first-level referrals) and six tertiary care facilities (academic and teaching hospitals for general care and referrals) (Table 2). Forty-two midwives were providing services in these facilities.

The evaluation team consisted of a provincial trainer (an obstetrician-gynecologist), RHRU's clinical director and midwife trainers, an Ipas researcher and the reproductive health coordinators for each of the provinces we visited. In addition to examining the quality of certified midwives' abortion care services and their clinical competence, we documented the settings in which midwives were providing services.

Data collection involved observing abortion procedures and counseling sessions, reviewing records and a random sample of patient charts, and interviewing patients and providers. We obtained verbal informed consent from patients prior to observing the procedure and again prior to observing any postprocedure contraceptive counseling. The interviews consisted mostly of closed-ended questions administered by members of the evaluation team.*

The day-long visits to each facility were prearranged to coincide with the normal operating hours of each facility and included a feedback period during which the evaluation team discussed preliminary findings with facility staff and administrators. Team members offered immediate feedback and guidance to health care providers regarding abortion services. They also delivered supplies to each facility, including extra manual vacuum aspiration kits or replacement parts when needed, and information, education and communication materials developed by RHRU and the Reproductive Rights Alliance. (The program arranged with the provincial authorities to provide a continuous supply of materials and equipment through the normal hospital supply systems after the program ended in 2000.)

RESULTS

Our review of records in the 27 facilities indicated that 3,927 women had obtained a first-trimester abortion during the period June-August 1999 (Table 3). The number of abortions varied from province to province, with Gauteng serving the most women (1,491). Overall, the majority of abortions were performed in primary and secondary facilities.^† Unfortunately, because facility registers in secondary and tertiary hospitals did not record the type of provider who performed each abortion, we cannot report the number of procedures midwives provided, as an abortion could have been performed by either a midwife or a doctor. Results from the evaluation visits, however, indicate that midwives were very active in providing first-trimester abortion services, and doctors in facilities with midwives trained in abortion care tended to let them manage most of the first-trimester procedures. Midwives provided all services in primary care facilities.

Assessment of Midwives' Services

• Technical competence. The midwives' level of clinical skill was assessed by practicing physicians experienced in performing abortions using a variety of techniques, including manual and electric vacuum aspiration as well as dilation and curettage. (Physicians did not evaluate midwives they had trained themselves.) Physicians observed 96 abortions performed by 40 midwives in 25 sites. (No procedures were observed at two sites because no clients came in for abortions on the day of the evaluation.)

Physicians assessed midwives' skills in listening to their clients, informing clients about the procedure, performing infection-prevention activities, using manual vacuum aspiration, examining and managing aspirated tissue, monitoring for postprocedural complications, responding to clients' questions, and informing clients about the warning signs of possible health problems and about the risk of pregnancy if they did not use contraceptives. Midwives were rated as having good clinical skills if they successfully performed all these tasks.

Midwives used manual vacuum aspiration in 85 of the abortion procedures and electric vacuum aspiration in 11. They did not use dilation and curettage for first-trimester abortions in any of the facilities. The physicians concluded that the midwives showed good clinical skills in 75% of the procedures. The only area identified as needing significant improvement was administration of antibiotics—mainly identifying the need to administer them.

• Waiting time for abortion services. We interviewed 90 abortion clients after their abortion procedure (not necessarily the same clients whose abortions were observed). Women reported that they had waited from less than 24 hours (13%) to more than 21 days (4%) from the time they sought an abortion to the time they had the procedure. Most women had waited 1-7 days (52%) or 8-14 days (22%). Those who had waited for more than seven days gave the following reasons: They had first come to the facility too early in their pregnancy; the waiting list at the facility was full; they were referred for a medical consultation; midwives were allowed to perform abortions only when a physician was present, but no physician was available when the client arrived; and personal reasons. When women were scheduled for the abortion procedure, they waited an average of five hours (the range was 1-7 hours) from the time they arrived at the facility to the time they received the service.^‡

• Referral for complications. Clients who experience complications that cannot be addressed at the facility where their abortion is performed should be referred to a facility where the clients can receive the services they need. We asked 27 midwives (one in each facility) if they had made arrangements for such referrals, and if so, where they referred these clients. During the evaluation period, no complications occurred during abortion procedures or as an immediate result of the procedures, and no women died in any of the facilities. However, all midwives had established appropriate referral mechanisms for clients who developed complications. Fourteen midwives (52%) had arranged to refer women with complications to the gynecology outpatient department, 10 midwives (37%) had arranged to refer women to the gynecology ward itself and one (4%) had arranged to refer women to the general outpatient department within the same facility. In the two facilities where such areas did not exist, midwives had arranged to refer women to a secondary- or tertiary-level hospital. Transportation for referrals was available in five of the eight primary-level facilities.

• Information provided to clients. Of the 90 abortion clients we interviewed, 91% had received information about the procedure itself. Most women also reported having received information about possible postoperative bleeding (92%), the immediate risk of pregnancy (86%) and the need to avoid sexual intercourse until bleeding stopped (81%) (Figure 1). Fewer women reported receiving information about signs of possible postprocedure health problems, including vaginal discharge (64%), pain (58%) and fever (46%).

• Postabortion contraceptive services. To help abortion clients avoid future unwanted pregnancies, the training program emphasized integrating abortion services with postabortion contraceptive counseling and methods. The 40 midwives we observed in the 25 facilities consistently provided women with contraceptive counseling immediately after the abortion. In 23 of the facilities, registered midwives provided counseling and methods. In 19, the midwives who offered contraceptive counseling and methods were also directly involved in performing abortions. Midwives provided counseling to women individually in nine facilities, in groups in eight and both individually and in groups in eight. Information, education and communication materials to aid postabortion counseling were available in almost all sites, most commonly in the form of posters and leaflets. These materials focused on contraceptive methods and the prevention of sexually transmitted infections (STIs) and HIV transmission.

Of the 25 facilities where abortion procedures were observed, two provided postabortion contraceptive counseling, but not methods. These facilities had formal arrangements with on-site family planning clinics where abortion clients could receive contraceptive methods. Thus, no abortion clients were referred off-site for postabortion contraceptive services, indicating a positive first step toward linking abortion and other reproductive health services.

Our interviews with midwives indicated that the injectable was the only method that was available at all 27 sites (Figure 2). The male condom was available at 25 sites (93%) and oral contraceptives at 23 (85%). Midwives at most of the 27 sites said that a dedicated product for emergency contraception was not available at their sites.

Eighty (89%) of the 90 clients we interviewed accepted a contraceptive method before leaving the facility where they had received their abortion (not shown). Most women (75%) received an injectable contraceptive. Only 1% received condoms: Neither female nor male condoms were routinely offered to women as part of the contraceptive method mix. Of the 80 women who accepted a method, 70 (88%) said that they had received their first choice. The 10 clients who had not received their method of choice had been given an injectable. Almost all of the women who had accepted a contraceptive method (91%) were told of a facility near their home where they could obtain a new supply of their method.

Assessment of Facilities

• Record keeping. All of the sites needed to strengthen their record keeping. All 27 listed the date of the abortion procedure and the client's age on patient registers, and 24 (89%) included the patient identification number. However, they did not record other critical information, such as the client's name or the type of provider performing the abortion. Furthermore, only eight (30%) facilities recorded information about procedural complications, nine (33%) recorded names of antibiotics and pain medications administered, 10 (37%) noted whether misoprostol had been used as a cervical ripening agent and 11 (41%) recorded which contraceptive method the client accepted.

• Procedural protocols. Twenty-two (81%) facilities had abortion care protocols available for staff to review. However, these protocols varied from site to site. At 20 facilities (74%) the protocol contained information about infection-prevention procedures and the administration of misoprostol, and at 18 (67%) the protocols contained information on the use of analgesics and antibiotics. Twenty of the 27 midwives (74%) we interviewed were not aware of the mechanism for reviewing maternal deaths in their facilities, even though a national mechanism existed.

• Supervision of abortion services. Midwives emphasized the importance of ongoing supervision so that their questions and concerns could be addressed in a timely manner. In 21 facilities, midwives reported the presence of a coordinator or supervisor of abortion care services. In 19 of these facilities (90%), supervisors were on-site. In the majority of these facilities, supervisors were either professional nurses (55%) or physicians (25%). In 12 facilities (57%), on-site supervisors visited staff performing abortion services weekly, although, at the time of our evaluation, four on-site supervisors (21%) had never visited the midwives providing abortion services. In two facilities, supervisors were based off-site; one had visited the facility once, and the other had not visited during the six-month period prior to the evaluation.

DISCUSSION AND CONCLUSIONS

The initial aim for our evaluation was to interview and observe the management of at least five clients at every site. However, this was not possible because only one client was present for abortion services during some visits, and there were no abortion clients during others. Thus, a main limitation of our evaluation was our inability to observe and interview a standard number of clients. Despite this, the evaluation team was able to collect a wide array of data from almost all of the sites where midwives were providing abortion care services in South Africa's nine provinces. Our findings provide important insights and lessons learned for the continuation of the program.

Among the most significant findings is that midwives can provide high-quality abortion services in the absence of physicians. The program breaks new ground for midwives in South Africa and around the world by providing an innovative model for expanding midwives' scope of practice to include abortion care services. Nationwide training in abortion and treatment of incomplete abortion will continue to expand service delivery at the primary care level. Special emphasis and resources need to be dedicated to this training so that abortion services can be made accessible to women throughout the country. In addition, the abortion care curriculum and training need to be integrated into midwifery training, and comprehensive documentation of midwives' practice is needed to demonstrate the program's impact on women's health and lives.

In light of the high incidence of HIV infection in South Africa, the finding that so many women did not receive male or female condoms highlights a missed opportunity. All of the abortion clients who participated in the evaluation had had unprotected intercourse and were thus at risk of STIs and HIV infection in addition to unwanted pregnancy. Postabortion counseling should have been used to counsel women about preventing STIs and HIV, and not just the prevention of further unwanted pregnancies. However, because both providers and clients traditionally have considered the condom a method of STI prevention, it often is not routinely offered in the context of contraceptive services.

Finally, the program needs to incorporate an array of community education activities to inform people about the provisions of the new abortion law and about preventing unwanted pregnancy, STIs and HIV infection. In addition, these activities should include information about early pregnancy detection to help women seek abortion services as early as possible in primary-level facilities.

Midwives and other health care providers participating in abortion care services need support and monitoring systems, preferably on-site, that can respond immediately to their questions and concerns. The supervisory system also needs strengthening, as a significant percentage of the supervisors had never visited the midwives after they started providing services. Most of the midwives expressed their appreciation for the support they received from their supervisors and the evaluation team, which they said helped them avoid feeling overwhelmed with the volume and intensity of the abortion services. One midwife said it is important to know "that there are people caring for us," which reflected the sentiments of many midwives as they began to provide abortion services, often without support from colleagues in their facilities. The evaluation visits were educational for everyone involved, and innovative ideas were implemented and shared among the facilities. Future evaluators should use these efforts as an opportunity to strengthen services throughout the country.