Less than one year after the Food and Drug Administration (FDA) approved Mifeprex (mifepristone), commonly known as the "abortion pill," an article in the Los Angeles Times1 predicted a bleak future for the drug in the United States. The reporter had contacted 53 clinics in California and found that fewer than half offered medical abortion to their patients. Reasons commonly cited for not offering the service included cost, the need for training and the obligation of instituting new procedures for counseling and follow-up care. Several providers also seemed to believe that medical abortion was inherently less reliable or acceptable than traditional surgical methods.
One year later, the headlines of two articles presented starkly contrasting assessments of mifepristone's acceptability: An article in the Washington Post2 was headlined "Abortion Pill Sales Rising, Firm Says," whereas a piece in the New York Times3 read "Abortion Pill Slow to Win Users Among Women and Their Doctors." The Post story drew from press releases by Danco Laboratories, the manufacturer and distributor of mifepristone in the United States. Meanwhile, the Times article relied on conversations with providers already performing abortions, who compared the costs and time involved in doing medical abortions with those of the already established surgical procedures. Not surprisingly, the providers interviewed stated that most doctors would probably decline to offer or promote medical abortion for their patients because it was a time-consuming and expensive service, particularly compared with surgical abortion.
Continuing to rely on the perspectives of abortion providers to predict the fate of medical abortion in the United States is problematic. First, these providers, precisely because of their familiarity with an alternative technique, must overcome biases and, in some cases, preconceived notions about a new option. In fact, a review of the experience of medical abortion in France, Great Britain and Sweden concluded that "it can take a decade or longer for mifepristone to be fully recognized and integrated as a method of abortion," and "provider knowledge and acceptance" are key factors.4 Second, focusing solely on the concerns of providers fails to recognize the influence of the other major actors in the public introduction of any new technology—namely, the consumers. Finally, these stories give short shrift to potential providers' underlying hesitations about incorporating a new option in their practices.
In early 2001, shortly after medical abortion using mifepristone became available in the United States, the Reproductive Health Technologies Project, a national advocacy organization based in Washington, DC, undertook a survey of women's health care providers in California. The project sought to understand what potential providers would require to begin offering this method:5 What would it take to make the provision of this drug a benefit instead of a liability? Between February and May 2001, the project interviewed 20 providers working in community clinics, university-based medical centers, health maintenance organizations (HMOs), feminist clinics, student health centers and independent medical practices. All provided primary care services to women, 18 were physicians (although only one was an obstetrician-gynecologist) and none were abortion providers.
Given the controversial nature of abortion, gaining access to the respondents was surprisingly easy, and their willingness to talk was unexpected. All those approached agreed to participate, spent more than an hour in the interview, and expressed an interest in receiving more information about the regimen and how to incorporate it in their clinical practices. The perspectives of the persons we interviewed—their concerns, expectations and interest in medical abortion—provide insight into what it will take to get providers to begin offering the method.
FEARS AND EXPECTATIONS
Every provider we interviewed had heard about the FDA's approval of mifepristone. Although before the interview, none had ever seriously considered offering medical abortion to their clients, all were interested in discussing the topic, and more than half wanted to explore the possibility of providing the drug. All of the providers had typically been referring their clients to other providers for surgical abortions and assumed that they would do the same for clients requesting medical abortion. Some thought that their existing referral systems for surgical abortions worked well and therefore saw no need to change the system to incorporate medical abortion. Specifically, the student health and HMO providers believed that women would be better served by being referred to providers who specialized in abortion than by receiving services from the health care personnel on-site. Most providers, however, had not ruled out providing the drug. Instead, as one provider put it, "It never occurred to us [in our practice] to do so."
Some providers indicated that they would be more inclined to offer medical abortion services if they knew that others in their field were providing the service. For example, one provider noted that her community clinic could more easily offer medical abortion if another community clinic in the area also provided the service. Similarly, some providers of student health services said they carefully monitored the services offered by other universities.
The major obstacle for the providers we spoke with was not an inability to meet the FDA requirements for prescribing the drug.* Most already had the necessary systems in place to safely offer this method. Rather, the problem was an assumption that medical abortion can be offered only by providers of surgical abortion. A related perception held by some providers was that provision of the pill is complicated, requiring extensive training and sophisticated backup services.
All of the providers interviewed thought that they could dispense the drug only if their facility had 24-hour emergency backup care. Providers commonly asked, "Where will women go if they hemorrhage in the middle of the night?" and stated, "We don't have hospital privileges." Even providers at clinics with established protocols for treating patients who develop complications after normal business hours (such as community clinics and academic medical centers) believed that they could not dispense the drug because their facility provided insufficient round-the-clock access for their clients. Most providers thought that hemorrhage (rapid and potentially life-threatening blood loss) was a common risk associated with medical abortion.
Although medical abortion is a "low-tech" procedure, provision of this service may require several pieces of equipment—notably, those needed to perform ultrasonography (to date a pregnancy) and manual vacuum aspiration (to treat incomplete abortions). All the providers had used ultrasound equipment, and most had the equipment on-site. However, one provider's facility, a community clinic, had the ultrasound equipment but was not certified to use it. Although only three providers had ever performed manual vacuum aspiration for pregnancy termination, most had used similar equipment for endometrial biopsies. Many providers were impressed by the simplicity of manual vacuum aspiration and its use as backup for incomplete medical abortions.
SOURCES OF INFORMATION
Any provider who can date a pregnancy and can provide backup and follow-up care—on-site or through a referral—can in principle provide medical abortion. However, few providers we interviewed knew this. Their information about medical abortion had come largely from popular media sources; several interviewees mentioned the Los Angeles Times. These sources tend to focus on the obstacles, difficulties and challenges of incorporating a new service.
Several providers mentioned that it would be helpful if their professional organizations (for example, the Association for Family Physicians or the Association of Community Clinics) published guidelines on performing medical abortion. Meanwhile, few organizations for U.S. professionals in non-reproductive health specialties have published information or guidelines on mifepristone use, and only one provider belonged to the American College of Obstetricians and Gynecologists, the National Abortion Federation or Physicians for Reproductive Choice and Health. Thus, although some professional organizations are expanding their efforts to make information on medical abortion available to clinicians who do not provide surgical abortion, the providers in this survey had yet to receive any such information.
In addition, many of the providers were interested in receiving training in providing medical abortion. A half-day training course run by an accredited entity—for example, the University of California, Los Angeles, or the California Family Health Council—was the preferred approach. Two providers, both private practitioners, mentioned that it would be easier for them to attend a course on new reproductive technologies than a course specifically on abortion. These two providers believed that attending such a course could be more easily justified and explained than could attending a course with abortion in the title. All providers commented that a training course lasting more than one day would be difficult to attend.
THE DEMAND SIDE OF THE EQUATION
None of the providers interviewed had ever been asked to provide medical abortion. "Frankly, no one has asked for it" typified their responses. Further questioning by the interviewers revealed that providers were relying on clients to request the service and were interpreting the absence of such requests to mean a lack of demand. The following comment demonstrates one of the assumptions providers were making about patients' interest in the service: "Our clientele is largely [Hispanic], so I don't think this would be popular."†
The issue of demand is complicated because it is closely intertwined with that of availability. That many women want to have the option of medical abortion is undeniable—the drug would not be marketed in this country, or elsewhere, were it not for the demand by women that it be made available. Furthermore, anecdotal evidence from clinics at which medical and surgical abortion are equally accessible shows that an increasing number of women choose medical abortion. In addition, clinical trials have shown that when given a choice, a substantial proportion of women (at some clinics, more than 50%) choose medical over surgical abortion.6
In recent years, demand for new pharmaceutical products in the United States has often been generated through advertising. The amount of money spent promoting Viagra and the resulting demand for that drug is a perfect example. Unfortunately for women, Danco Laboratories has a limited advertising budget, very little of which is going into direct-to-consumer outreach. Thus, U.S. physicians are not receiving sufficient information about this drug through their normal channels of pharmaceutical representatives, and their patients do not even know to ask for it. Moreover, women who do request medical abortions probably ask providers who do not know much more than the women themselves know or who have already formed an opinion about the method and subtly discourage women from considering it further.
Patients deserve to receive abortion care from their own provider, and medical abortion makes this feasible. However, neither U.S. women nor their providers seem to have grasped this possibility. Clients assume that their providers "do not do abortions" and therefore do not ask them for this method. As a result, providers are not experiencing a demand for the service. Without a clear demand from clients, the providers we spoke with are unlikely to make the necessary adjustments to offer medical abortion.
In a world of increasing pressure on time and profitability, health care providers and agency administrators must evaluate the costs, benefits and feasibility of incorporating any new service or technology in their services. To tip the scales toward a more favorable cost-benefit ratio for primary care providers, several changes are needed.
The comments of the providers we interviewed are instructive in determining effective strategies for reaching out to health care providers who are currently not offering abortion services. First, they illustrate that many of the concerns providers have about offering medical abortion are based on exaggerated fears about possible complications and risks associated with medical abortion, and on misperceptions of the level of technical expertise required to offer this method. Second, these interviews suggest that although the providers are willing to learn more about medical abortion—and that some may be interested in providing it—they are not yet receiving the information they need.
The interviews also indicate a greater willingness of general health care providers to consider incorporating medical abortion in their practices than stories in the popular press have suggested. In fact, several respondents offered examples of other services or technologies they had successfully incorporated in their clinical practices despite initial skepticism. For example, providers at community health clinics are offering more prenatal services than in the past. Moreover, providers in family practice and those at community clinics are performing loop electrosurgical excisional procedures (LEEP), a laser procedure for treating early cervical cancer that was previously done exclusively by obstetrician-gynecologists.
Finally, the responses reveal that providers base decisions on whether to incorporate new services on several factors: demand for the service, the providers' commitment to offering comprehensive health care to their clients and economic considerations. Whether a similar pattern can evolve with medical abortion will depend largely on two factors: convincing generalists that they can safely provide medical abortion and proving that a demand for these services exists.
On the basis of what we have learned from these providers' perspectives, we recommend the following strategies for persuading health care providers not currently providing abortion services to begin offering medical abortion to their clients:
• Get information to the generalists. The providers' interest expressed in these interviews clearly suggests a latent demand for audience-specific information and outreach strategies. Because most of these providers do not belong to reproductive health-related organizations, it is up to relevant professional journals, networks and channels of communication to supply this information.
• Improve access to training for primary care providers. The interest demonstrated in these interviews strongly suggests that a latent demand for training exists among primary care providers. To reach these providers, training courses will need to be short (no more than one full day) and conducted by a respected medical authority. In addition, courses should discuss medical abortion in conjunction with other new reproductive technologies, thus allowing a cover for persons who are uncomfortable attending courses on abortion techniques alone. Moreover, to build a broad-based cadre of medical abortion providers, training efforts must go beyond simply building their competence in performing the procedure. Training must also address providers' willingness to discuss options for early medical abortion with their patients and their ability to answer questions about the method accurately.
• Identify and disseminate the work of innovators and leaders. Some comments providers made during the interviews—like "we don't want to be the only ones to offer this service"—reflect the political stakes, and the risks, associated with offering abortion care. Advocates need to find providers who are committed to meeting their patients' needs by adopting new technologies, and need to work closely with them as they establish these services and deal with administrative, attitudinal and practical challenges. We also need to document and learn from the experiences of U.S. providers already offering medical abortions.
• Increase demand. It is clear that providers will offer medical abortion directly to their clients, rather than refer women to existing abortion providers, only if their clients begin to request this method. Economic factors may in time kick in. For example, HMOs may find that it is more cost-effective for their clinicians to prescribe mifepristone directly than to refer out; as a result, insurers may create economic incentives for affiliated clinicians to begin offering this service. Moreover, physicians in family practice might come to recognize abortion care as a potential niche. Ultimately, however, consumers' requests for the drug and demands for the service may be the most crucial factors in promoting access to this method. Public advocacy efforts to increase access at the primary care level will be key.
We believe that the potential of mifepristone to increase U.S. women's access to safe abortion services can still be achieved. When the method was first introduced, the expectation was that, given its "low-tech" nature, mifepristone could be provided in new health care settings. Moving abortion care beyond existing clinics to private physicians' offices, community clinics and college health centers would make safe abortions more easily accessible and decrease the disparities in access between urban and rural areas. These preliminary findings and experiences of other members of the medical establishment show that innovation in practice—even in incorporating medical abortion—is possible. But we must act quickly. Only when these new providers begin dispensing the abortion pill will we have increased access to abortion services. Should we fail to move abortion care beyond existing abortion providers, we will have squandered the opportunity this new technology offers us to increase access to safe abortion for women everywhere.