abortions, it is necessary to understand the factors behind the persistence of unsafe abortion and the opportunities and barriers to preventing unsafe abortion. A consultation on “Priorities and Needs in the Area of Unsafe Abortion” was convened in Geneva, Switzerland, in August 2000, to bring together internationally prominent researchers from a variety of disciplines to assess the global status of unsafe abortion and to identify a research and advocacy agenda to reduce unsafe abortion and its burden on women, their families, and public health systems. A number of emerging and long-standing issues related to research on unsafe abortion formed the background discussion to the chapters in this volume. A few of these topics are discussed in this chapter. Medical abortion Medical termination of pregnancy can be achieved through the use of mifepristone in combination with a suitable prostaglandin, such as misoprostol, or a prostaglandin analogue. Medical abortion is increasingly available worldwide and is likely to have a substantial impact on the service provision of both safe and unsafe abortions. In countries where abortion is legal, misoprostol may improve abortion services where skilled providers are not accessible and sterile surgical equipment is not available. Misoprostol is inexpensive, simple to administer, easy to store and is therefore particularly appealing to providers in developing countries and to women seeking covert abortions. When provided under supervised medical care, medical abortion has a success rate of approximately 95 per cent (5). However, questions remain about the introduction of medical abortion in settings lacking appropriate medical supervision, B-Chap 1.indd 4 07/02/2006 13:31:34 5 UNSAFE ABORTION : AN OVERVIEW OF PRIORITIES AND NEEDS particularly where women depend on social networks for dosage information and where access to follow-up medical care is limited. To what extent does the unsupervised self-administration of medical abortion provoke complications and how serious are these complications, compared to those from unsafe surgical abortions? The consensus thus far is that medical abortion is most likely safer than other means of self-induced abortion using physical objects or caustic substances and is likely to reduce morbidity and mortality related to unsafe abortion (6, 7). However, women who take incorrect dosages may risk potentially serious complications such as prolonged bleeding if they do not receive prompt medical attention. A case study of the covert use of medical abortion is Brazil, where misoprostol has been available through women’s networks and complicit pharmacists for over a decade. A study of medical abortion in Brazil, Jamaica and the USA (7) found that in the absence of standard or labelled regimens for abortion, providers relied upon networks of colleagues and the medical literature to develop their own regimens which were substantially higher or lower than those recommended by professional organizations. Despite fluctuating regimens and their corresponding efficacy, when misoprostol was removed from the Brazilian market in 1998, the number of septic surgically-induced abortions rose in several major hospitals and clinics. India is an emerging market for medical abortion and it is likely that both unsupervised and supervised medical abortion will rapidly gain ground. The cost of a self-administered medical abortion is approximately the same as a surgical abortion performed by a private sector physician. However, there is no need for a visit to the physician, confidentiality is assured, and the intrusion of surgery is avoided, all of which may lead women who can afford it to favour medical abortion. Abortion, postabortion care and contraception The relationship between contraception and abortion may, at first glance, appear paradoxical. For every country where the incidence of abortion declined as contraceptive use increased, another experienced simultaneous increases in levels of abortion and contraceptive use. Marston & Cleland (8) reviewed the available country data and demonstrated that where fertility remains constant, abortion declines as contraception rises. However, where fertility is in the process of declining, an increase in contraceptive use alone may be insufficient to meet the demand for fertility regulation, and an increase in abortion incidence may result until fertility levels have stabilized. Singh & Sedgh (9) examined the role of contraceptive use on abortion incidence over several decades in Brazil, Colombia, and Mexico. The B-Chap 1.indd 5 07/02/2006 13:31:34 6 INA K.WARRINER results of their analyses, which found substantial variation in the pace and timing of decreases in abortion, broadly support the hypothesis that where fertility rates are changing, abortion rates may increase in conjunction with contraceptive use but ultimately stabilize and then decline. Improved understanding of the relationship between abortion and contraception would provide policy-makers with a clearer picture of the factors influencing fertility change, including the relative impact of abortion, whether safe or unsafe, and the need for family planning policies and programmes. More studies are needed to rigorously assess the long-term effects of postabortion contraception on subsequent unplanned pregnancies. In one of few such studies, Johnson et al.