Advancing Sexual and Reproductive Health and Rights
 
Perspectives on Sexual and Reproductive Health
Volume 35, Number 1, January/February 2003

The Accessibility of Abortion Services in the United States, 2001

By Stanley K. Henshaw and Lawrence B. Finer

CONTEXT: A woman's ability to obtain an abortion is affected both by the availability of a provider and by access- related factors such as cost, convenience, gestational limits and the provision of early medical abortion services.

METHODS: In 2001-2002, The Alan Guttmacher Institute surveyed all known abortion providers in the United States, collecting information on their delivery of abortion services and on the number of abortions performed.

RESULTS: A minority of abortion providers offer services before five weeks from the last menstrual period (37%) or after 20 weeks (24% or fewer), but the proportions have increased since 1993. Providers estimate that one-quarter of women having abortions in nonhospital facilities travel 50 miles or more for services, and that 7% are initially unsure of their abortion decision. The majority of providers (59%) say that these clients usually receive abortions during a single visit. An average self-paying client was charged $372 for a surgical abortion at 10 weeks in 2001, up from $319 in 1997; only 26% of clients receive services billed directly to public or private insurance. Early medical abortions are becoming increasingly available but are more expensive than surgical abortions. More than half (56%) of providers experienced antiabortion harassment in 2000, but types of harassment other than picketing have declined since 1996.

CONCLUSIONS: Abortion at very early and late gestations and early medical abortion are more available than before, but charges have increased and antiabortion picketing remains at high levels. Thus, many women still face substantial barriers to obtaining an abortion.

Perspectives on Sexual and Reproductive Health, 2003, 35(1):16-24

Unintended pregnancies and induced abortions are common and occur among women of all social and economic groups.1 Yet the availability and accessibility of abortion services have long been a concern for reproductive health professionals, as women seeking an abortion have a fairly narrow time period during which they can obtain the procedure. Measures of availability have generally declined since 1982: The number of abortion providers in the United States has fallen by 37%, and the proportion of women who live in counties with no abortion provider has increased from 28% to 34%.2 In 2000, 86 of the country's 276 metropolitan areas and almost all nonmetropolitan areas had no abortion provider.3

Accessibility is harder to measure than availability, because of the variety of possible barriers, both tangible and intangible. Besides distance from a provider, cost is the most obvious tangible barrier. The provision of specific services, such as second-trimester pregnancy termination, can determine accessibility for individual women. Among the barriers that are less tangible, and therefore more difficult to quantify, are women's lack of accurate information about the legality of abortion and about where and how to obtain abortion care, misinformation about abortion, intimidation by protesters, state-required waiting periods and mandated counseling topics that may not be relevant to a woman's personal situation, and antiabortion attitudes among family or friends.

Although it is difficult to measure the impact of the accessibility of abortion services on abortion incidence, lack of access likely prevents some women from terminating unintended pregnancies. An estimated 46% of unintended pregnancies in 1994 were carried to term, excluding those ending in miscarriage.4 While many women who continued such pregnancies undoubtedly came to want to give birth, others may have been deterred from having an abortion by the difficulty of obtaining one. For example, distance from services may help to explain why the abortion rate among nonmetropolitan women is half that of women who live in metropolitan areas.5

In addition, as with other types of health care, women are likely to be more satisfied with their abortion-related care if they have access to the types of services that best meet their physical and personal needs. Thus, for a woman who decides early in pregnancy that she wants an abortion, quality of care may be enhanced if she can choose between a medical and a surgical procedure, and if she is not required to wait for one or more weeks to meet minimum gestation limits.

This article documents the current status of abortion service accessibility in the United States, on the basis of data collected in a survey of all known U.S. abortion providers conducted in 2001-2002 by The Alan Guttmacher Institute (AGI). (Information on availability from the same survey is presented elsewhere.6)

METHODS

The 2001-2002 AGI Abortion Provider Survey was the 13th such survey of all known abortion providers in the United States. This survey is the only national source of information on the numbers and types of providers and their geographic distribution. In addition, for the United States as a whole and for most states, the AGI survey produces the most complete data available on the numbers of abortions performed.

The survey methodology has been described in detail elsewhere.7 Briefly, we identified potential providers, beginning with facilities found to be offering abortion services in 1996 through the prior AGI Abortion Provider Survey. Possible new providers were added on the basis of information from a variety of sources (e.g., telephone yellow pages, Planned Parenthood affiliates and World Wide Web listings), and past providers that had closed or that no longer offered abortion services were purged from the list. After four mailings and extensive telephone follow-up of nonrespondents, we identified 1,819 facilities where abortions were performed in 2000; this article analyzes information from these providers.

All respondents were asked about the number of abortions provided at their facility in 1999 and 2000, the minimum and maximum gestational ages at which abortion services were offered, the availability of early medical abortion (using either mifepristone or methotrexate) and the conditions under which this service is provided.

We classified providers into four types: hospitals; abortion clinics (defined as nonhospital facilities where half or more of patient visits are for abortion services); other clinics (all other nonhospital facilities with names suggesting a clinic structure, as well as physicians' offices where 400 or more abortions were performed in 2000); and physicians' offices (or facilities with names suggesting that they are physicians' private practices) where fewer than 400 abortions were performed. Facilities were also classified according to the number of abortions provided, rounded to the nearest 10.

As in the earlier surveys, we asked nonhospital providers how far they thought abortion clients lived from the facility, how much they charged for abortions at various gestations, what sources women used to pay for services and whether the facility experienced any antiabortion harassment. To track trends in these aspects of accessibility, we compare the results of the current survey with data from AGI surveys covering 1992-19938 and 1996-1997.9 (Information on the latter survey's methodology has been published elsewhere.10)

We obtained more detailed information from nonhospital providers that offered early medical abortions: the number performed during the first half of 2001; gestational limits; for mifepristone abortions, the dosage used; and practices regarding home administration of misoprostol, counseling and follow-up. Other new questions for nonhospital providers were whether services were provided in a single visit and what proportion of clients they thought were initially unsure about the abortion decision.

While two questions rely on providers' impressions, rather than on records or policies, both involve information that staff are likely to be aware of, because it is relevant to their service provision. The distance of clients' homes from the facility is known from their addresses and from planning for follow-up care; additionally, providers know what area they have targeted, and often advertised in, as their service area. Similarly, staff typically have an impression of the proportion of clients who are unsure of their abortion decision when they arrive, because assessing the firmness of a woman's decision is part of abortion counseling, and the administrators and physicians who completed the questionnaire would be aware of when extra counseling has been required and when women have left without having an abortion.

Most survey questions were worded in the present tense and referred to the time at which the survey was completed. For convenience of presentation, we considered questions asked in the present tense to refer to 2001, since the large majority of responses arrived in that year. The questions about harassment, however, referred specifically to providers' experiences in 2000.

Other than the number of abortions, gestational limits and the provision of early medical abortion, the items reported here were asked only of nonhospital facilities, because the medical records staff who completed many of the hospital questionnaires are unable to answer questions about the functioning of the abortion service. The results nevertheless represent the experience of most women having abortions, since nonhospital providers performed 95% of all abortions in 2000.11

We obtained information on gestational limits from 77% of the 1,216 nonhospital abortion providers surveyed in 2000, information on charges from 72% and data on other items from 52-55%. The large majority of providers with missing data did not return the questionnaire; through follow-up telephone calls to the facility or from other sources, we were able to obtain information on the number of abortions performed, gestational limits and charges, but not on most other questions.

Response rates were higher for nonhospital facilities and for facilities with large caseloads than for hospitals and smaller facilities. For example, information on gestational limits was obtained for only 35% of hospitals, compared with 77% of nonhospital providers, and for 81% of facilities performing 1,000 or more abortions, versus 65% of those performing fewer than 30. We therefore weighted all results to reflect the correct national proportions according to facility type and caseload. To account for item-specific nonresponse, we used different weights for each variable. Most data presented here refer to the characteristics of providers. However, where the focus is on the experience of the typical woman obtaining an abortion, we also weighted the results by the number of abortions reported by each facility in 2000, since numbers of abortions in 2001 were not available.

FINDINGS

Gestational Limits

Providers typically set a minimum and maximum gestation at which they are willing and able to perform an abortion. These limits are expressed as the number of weeks since the woman's last menstrual period (LMP). Thirty-seven percent of facilities that offer abortion services provide either surgical or medical abortions at four weeks or less LMP (Figure 1), often for any pregnancy that can be confirmed by ultrasound or even a pregnancy test. This represents a sharp increase from the level of 7% reported in 1993 (not shown). Eighty-two percent of abortion facilities perform abortions at six weeks LMP (Figure 1). Abortion clinics are more likely than other types of facilities to offer abortions at five and six weeks LMP.

More than 90% of all abortion providers offer services at 8-10 weeks LMP. However, the proportion drops with each additional week of gestation after eight weeks LMP (typically four weeks after the woman's first missed period) and declines steeply after 12 weeks. At 20 weeks, for example, only 33% of all providers offer abortion services, and at 21 weeks, 24% still do so.

Hospitals and abortion clinics are much more likely than other providers to offer services past 12 weeks. At 13-15 weeks LMP, a higher proportion of abortion clinics than of hospitals perform abortions, while at 17-23 weeks LMP, the reverse is true. (Many hospitals, however, provide very few abortions and do so only in extraordinary circumstances, such as when the fetus has an abnormality or the pregnancy poses severe health risks to the woman.) After 24 weeks LMP, the number of providers offering abortion services again drops off sharply. Only 2% of all abortion providers (approximately 11 hospitals and 19 abortion clinics) provide abortions at 26 weeks.

The proportion of all facilities offering abortion services at 20 weeks increased from 22% in 1993 and 25% in 1997 to 33% in 2001, for a 50% increase over the period. However, because the overall number of abortion providers has declined, the number offering such services increased only 16%, from 524 in 1993 and 511 in 1997 to 607 in 2001. A majority of facilities providing abortions at 17-24 weeks LMP are hospitals, although a majority of abortions at each gestational age are probably performed in abortion clinics (not shown).

Distance Traveled

Respondents estimated that 8% of women having abortions in nonhospital facilities travel more than 100 miles to obtain this service, and that an additional 16% travel 50-100 miles.* Travel patterns appear to have changed little over time. In both 1993 and 1997, providers also reported that 24% of clients traveled at least 50 miles, including 8% and 7%, respectively, who traveled more than 100 miles.

The proportion traveling long distances varies by geographic region. In the East South Central and the West North Central states, 43% and 37%, respectively, of women travel at least 50 miles to obtain an abortion, including 14-15% who travel more than 100 miles. In contrast, only 11% of women in the Middle Atlantic states travel 50 miles or more.

In general, clients of large providers are the most likely to travel great distances. Only 6% of clients of providers that perform fewer than 30 abortions a year travel 50 miles or more, compared with 26% of clients of providers that perform 1,000-4,990 abortions and 18% of women seeing providers that perform 5,000 or more abortions annually. Some 26% of clients of abortion clinics (which tend to have larger caseloads than other providers) travel at least 50 miles to obtain an abortion, compared with 16-18% of clients of other types of facilities.

Traveling a long distance to a provider can be difficult for women who need to make two or more trips to the abortion facility. In 2001, four states had legislation requiring most or all clients to receive specified in-person counseling at least 24 hours before the procedure is performed. Such requirements usually necessitate two trips to the abortion provider. In other states, most women can obtain abortion counseling and medical services in a single visit.

Fifty-nine percent of nonhospital providers nationally and 60% of providers in states that do not require in-person advance counseling said that their clients usually obtain abortions in a single visit, while 15% and 14%, respectively, said that this never happens. Single-visit service is highly associated with caseload: All facilities that provided 5,000 or more abortions in 2000 usually perform abortions in a single visit, compared with only 20% of facilities providing fewer than 30 abortions (not shown).

Distance may also be a barrier for women who are uncertain about how to resolve a pregnancy and who first seek counseling about that decision at an abortion facility. They may then need time to consider the decision before having an abortion, if that is their choice. Nonhospital providers estimate that 7% of their clients arrive unsure; this proportion is slightly lower at larger clinics. Since some women presumably decide against an abortion, the proportion of all women having an abortion who were unsure of their decision when they first visited an abortion provider is probably lower than 7%.

Charges

We asked each nonhospital provider to indicate the usual charges that a woman would incur at that location for an abortion (with local anesthesia) at various gestations, including fees for any services always required for an abortion client, even if these are not billed through the provider (e.g., laboratory tests). On average, surveyed facilities charge $468 for a surgical abortion at 10 weeks LMP (Table 1).§ The lowest average charge ($364) is reported by specialized abortion clinics, and the highest average charge ($632) is reported by physicians' offices. The range of charges is wide (from $150 to $4,000), and the median ($370) is lower than the mean, which is influenced by a few facilities that have very high charges.

As a pregnancy advances into the second trimester, the abortion procedure becomes more complex, because it requires more time and more skill on the part of the clinician, and charges increase. At 16 weeks, the mean and median charges ($774 and $650, respectively) are more than half again the amounts at 10 weeks (Table 1). At 20 weeks, the mean and median charges increase to $1,179 and $1,042, respectively. In the second trimester, charges vary relatively little by type of provider, but the range remains wide, with some providers charging 2-5 times the average.

We also examined mean charges for abortions at 10 weeks LMP by the provider's caseload (i.e., the number of abortions performed). Charges are approximately twice as high in facilities that perform fewer than 30 abortions per year as in those that perform 400 or more. Caseload has little relationship to charges among abortion clinics or among other types of clinics that perform 400 or more abortions per year (Table 2).

By weighting the survey responses by caseload, we determined the amount paid by the average self-paying client, rather than the amount charged by the average facility. The mean amount paid by clients ($372) is lower than the mean amount charged by the typical provider, since larger providers (especially abortion clinics) tend to charge lower fees. Clients at abortion clinics and other types of clinics pay about the same as the average ($367 and $376, respectively), but clients obtaining an abortion at a doctor's office pay substantially more ($471).

In current dollars (the amount paid at the time), the average self-paying client's payment for an abortion at 10 weeks LMP has increased steadily over time—from $200 in 1983 to $319 in 1997, and to $372 in 2001 (Figure 2, page 20). When inflation in the cost of living (as measured by the Consumer Price Index for all items) is taken into account, the amount changed little between 1983 and 1997, but increased by 9% ($30) from 1997 to 2001. When compared with the amounts paid for other medical care, the amount paid for abortion services fell from 1983 to 1997, and then increased by 5% ($17) between 1997 and 2001.

How Women Pay

According to estimates by nonhospital providers, a large majority of women (74%) pay for their abortions with their own money or with funds they obtain from their partner, family or others (Table 3). In an unknown proportion of these cases, the women obtain subsequent reimbursement from health insurance. Included among self-paying clients are women for whom the facility reduces the fee or provides the service without charge (12%); some of these may receive partial support from one of the many funds that subsidize abortions for poor women.12 The proportion of clients who are charged reduced fees rises with the facility caseload, from 5% at the smallest providers to 15% at the largest (not shown).

Although the federal government pays for abortions through Medicaid only in cases of rape, incest and life endangerment, 16 states cover abortions under their Medicaid programs, either voluntarily or under court order.** Providers reported that about 13% of abortions are reimbursed by Medicaid; almost all of these occur in Medicaid-funding states (Table 3). An estimated 13% of abortions are covered by private insurance billed directly by the facility.

Other payment sources also vary between Medicaid-funding states and nonfunding states. Specifically, 19% of abortions in Medicaid-funding states are billed to private insurance, compared with only 8% in other states. As of December 2000, four states†† of the 34 that do not fund abortions under Medicaid had legislation prohibiting private insurance from covering abortions except under an optional rider at additional cost, but these could not account for much of the difference between Medicaid-funding and nonfunding states. Evidently, some of the same state characteristics that influence states to cover abortion under Medicaid influence private insurers to cover abortion.

The proportion of providers that bill private insurance for their clients' abortions is higher than average for nonhospital providers performing fewer than 30 abortions per year and for physicians' offices (45% and 27%, respectively—not shown). Direct billing of private insurance is most common in the Northeast (27%) and is least common in the South (5%). The proportion of clients for whom the provider billed private insurance has increased since 1997, when a similar question indicated that only 3% of clients benefited from direct insurance billing (not shown).

Early Medical Abortion

One-third of abortion facilities provided early medical abortions in the first half of 2001, and this proportion was increasing rapidly.13 An estimated 37,000 early medical abortions occurred in this time period; 35,000 of them were provided by nonhospital facilities. One-quarter of early medical abortions in nonhospital facilities were performed using methotrexate, which was available before the Food and Drug Administration (FDA) approved mifepristone in September 2000. However, 82% of medical abortion providers were using mifepristone—although some of these used methotrexate as well. Both drugs are used in conjunction with misoprostol, a prostaglandin that is administered within several days of the mifepristone or methotrexate to cause contractions and expel the products of conception.

The FDA-approved labeling specifies a fairly restrictive protocol for abortion using mifepristone: use within the first seven weeks of gestation; a mifepristone dose of 600 mg; misoprostol administered orally in the physician's office; and a follow-up visit to the provider for an examination to confirm that the pregnancy has been completely terminated. However, experts who have reviewed published studies have concluded that more convenient and less-expensive procedures are equally safe and effective.14 How providers structure their early medical abortion services influences the number of locations where the method is offered, the proportion of women who are eligible, the charges to clients and the convenience of the method from the client's point of view.

Table 4 shows information on the protocols used and experiences of nonhospital abortion providers offering early medical abortion services, including separate data for providers that performed 50 or more medical abortions from January through June 2001 and those that performed fewer.‡‡ Overall, the majority (67%) of medical abortion providers offered the service very early in pregnancy: These providers either set a minimum limit at or before 28 days LMP (45%), required only that the pregnancy be confirmed by a laboratory test (7%) or sonogram (9%), or had no minimum gestation requirement (6%). Seventy-six percent set their maximum gestation limits at or below seven weeks LMP, in line with the mifepristone labeling approved by the FDA; the remaining 24% used the method past seven weeks, on the basis of evidence from clinical studies,15 following mifepristone practice in Britain and elsewhere.16 More experienced providers (i.e., those with larger medical abortion caseloads) were more likely than the less-experienced to offer the method before 29 days and after seven weeks LMP.

The great majority of providers of early medical abortion used a dose of 200 mg of mifepristone (83%), and most permitted the client to take the misoprostol at home rather than requiring her to return to the abortion facility to receive it (84%). Both practices were more common among providers that did 50 or more medical abortions than among less-experienced providers.

Most providers (74%) reported that counseling for medical abortion takes more time than does counseling for surgical abortion. A large minority (43%) said that fewer than 10% of early medical abortion clients called with questions or problems, but one-third reported that 20% or more did so. Calls from one-fifth or more of clients were less common among the more experienced providers.

Unlike a surgical abortion, an early medical abortion is not completed during the woman's appointment with the provider, making follow-up important for ascertaining that the abortion has been completed. Therefore, providers usually attempt to contact clients who miss their follow-up appointments, to make sure that they are no longer pregnant. Respondents said that in early 2001, pregnancy termination was confirmed for almost all of their clients—for 91% in a return visit, and for 3% by telephone or other contact. (Six percent of clients were lost to follow-up.)

The mean charges for a mifepristone abortion and for a methotrexate abortion were $490 and $438, respectively, during the first half of 2001 (not shown). Providers who used 600 mg of mifepristone charged $74 more, on average, than providers who used 200 mg. More than two in five providers (43%) charged between $400 and $499 for mifepristone, and 38% charged $500 or more (Table 4). Providers with relatively large medical abortion caseloads were more likely than those who did few procedures to charge $450-499 and were less likely to charge $500 or more.

Sixty-nine percent of providers who offered both surgical abortions at six weeks and mifepristone abortions charged more for the latter, while 9% charged less. For 15%, the additional charge was $200 or more. More experienced providers were slightly more likely to charge more for mifepristone, but charges varied little by type of provider (not shown).

For 80% of medical abortion providers, the basic charge for a medical abortion included the cost of a subsequent vacuum aspiration, should an incomplete abortion or continuing pregnancy occur. Providers who performed at least 50 early medical abortions were much more likely than less-experienced providers to include surgical completion in their basic charge (95% vs. 71%).

Harassment

Many women seeking abortion face harassment by antiabortion protesters; this also affects a facility's ability to offer services. Each year, 56% of all nonhospital providers experience at least one of five types of harassment—picketing; picketing coupled with physical contact with or blocking of clients; vandalism (such as jamming of locks or other physical damage); picketing of the homes of staff; and bomb threats. Harassment is much more common in facilities with large abortion caseloads than in smaller facilities: The proportion experiencing one or more incidents ranges from 10% among facilities performing fewer than 30 abortions to 70% among those providing 400-990 abortions and to 100% of clinics providing 5,000 or more.

To avoid distortion caused by facilities with small caseloads (which vary in number from area to area and which affect relatively few women), we limited further analyses of harassment to large providers (those that performed 400 or more abortions in 2000). These providers accounted for 94% of all abortions in 2000.17 Overall, 82% of these facilities experience any type of harassment in a given year. The proportion experiencing harassment is greatest in the Midwest (91%) and lowest in the West (78%).

Picketing is by far the most common type of antiabortion activity, reported by 80% of large providers (Table 5). Some 14-28% of large providers experience more extreme forms of harassment. A majority (61%) of facilities experience picketing at least 20 times a year. Other types of harassment usually occur fewer than five times per year at any one facility.

Since 1996, all of these forms of harassment except picketing have become less common. The proportions of large providers reporting picketing with physical contact, vandalism and picketing of staff members' homes have fallen by about half since 1992, when these activities were at their height. The proportion of large providers reporting bomb threats has fallen steadily, from 48% in 1985 to 15% in 2000.

DISCUSSION

For many women, barriers to abortion services are significantly more common than are obstacles to other common types of reproductive health care. For example, only 13% of U.S. counties have an abortion provider,18 while obstetric-gynecologic care is available in half of all counties.19 In 1997, 85% of counties had at least one publicly funded family planning clinic.20 Depending on the circumstances of any given woman needing abortion services, she may have to cope with gestational limits, a long distance from a provider (the effects of which may be exacerbated if she needs to make two trips or is undecided about whether to have an abortion), travel and other expenses that may not be covered by insurance, a lack of choice of method of early medical abortion and antiabortion protesters. These are the potential barriers for which we have information; other factors, including restrictive legislation and attitudes, may also pose important problems for some women.

Gestational limits reduce the number of abortion providers available to specific women. Women frequently encounter such barriers when they seek an abortion during the second trimester. Many providers offer services only up to 12 or 14 weeks, because later abortions require more cervical dilation and greater skill on the clinician's part, the risk of complications is greater than with earlier abortions and the demand is less. When a fetal anomaly is discovered late in the second trimester, the woman may find that facilities where the pregnancy can be terminated are difficult to locate and are far from her home. Nonetheless, the number of facilities where second-trimester abortions are performed has increased in recent years.

A woman who discovers an unintended pregnancy at less than six weeks LMP may find that an abortion provider she contacts will not provide services then, but will ask her to wait until six weeks or later. Studies in the 1970s found a higher rate of continuing pregnancies after very early procedures.21 Recent research has shown, however, that with high-resolution ultrasound and careful examination of the products of conception, early surgical abortions can be performed without an elevated risk of ectopic pregnancy or incomplete abortion,22 and that an increasing number of providers perform such early abortions. In addition, providers of surgical abortion increasingly are offering early medical abortion.

Distance from a provider continues to affect women's ability to obtain access to abortion services, presumably because of the difficulty and expense of arranging travel and unfamiliarity with distant facilities.23 Although a distance of 50 miles may not create any difficulty for some women, for others it may pose a significant problem. Providers estimate that 8% of women who have abortions travel more than 100 miles to do so, a proportion that has not changed in recent years. Women who are able to overcome the barrier of distance may nevertheless suffer consequences. In a survey conducted in 1987-1988, half of a national sample of women who were having an abortion at 16 weeks LMP or later cited difficulty in making arrangements as a cause of delay.24

For many women—especially for those who must travel long distances for services—the option of having the abortion in a single visit is important and lowers travel costs. A large majority of nonhospital abortion providers serve at least some clients in one visit. This practice is not permitted, however, in states that require face-to-face counseling by a physician at least a day before the procedure. The result in one state was a decrease in the number of resident women who obtained abortions.25 Distance is also a factor for women who need to make multiple visits because they are uncertain about their abortion decision. Providers estimate that no more than 7% of clients arrive unsure at their first visit, however, and in such cases additional visits are responsive to the women's needs.

For many women, barriers to abortion services are significantly more common than are obstacles to other common types of reproductive health care.

Another factor affecting access is the fee that clients pay, which averages $372 at 10 weeks. This is a minimum figure, because many clients have additional expenses, such as for other services (intravenous sedation or general anesthesia), transportation, time lost from work and increased costs if the pregnancy is at a later gestation. Nonhospital providers directly bill Medicaid or other insurance for only 26% of their abortion clients, and for only 8% in states where Medicaid does not cover abortion. Whether because of Medicaid funding restrictions, a lack of insurance coverage, women's hesitancy to use insurance coverage for abortion or providers' inability to bill directly, most women pay directly for their abortion care.

While the cost of an abortion may seem moderate to some, many low-income women are likely to find it substantial. Between 18% and 35% of Medicaid-eligible women who would have abortions instead continue their pregnancies if public funding is unavailable.26 The lack of Medicaid coverage may be the public policy that has the greatest impact on the number of women who want an abortion but are not able to obtain one. In addition, a woman's need to secure funds often causes abortions to be delayed; one study found that 22% of Medicaid-eligible women who had a second-trimester abortion would have terminated their pregnancy in the first trimester if Medicaid had covered abortion services.27

After a long period of stability, the average amount paid for a first-trimester abortion, adjusted for inflation, increased by 9% in recent years. These increased fees may reflect increased costs for physicians and other staff (to comply with state and federal regulations or to maintain the safety of the facility, staff and clients) or reduced competition among abortion providers. The increased proportion of women whose health insurance is billed directly for abortion services somewhat offsets the effect of the increased fees.

Early medical abortion is a new development that increases women's options and is preferable to some women. Although it has the potential to make services available in geographic areas that were previously unserved, by mid-2001 there was little evidence of providers' offering only early medical abortion.28 However, the number of surgical abortion providers who also offer early medical abortions has grown rapidly, and as of early 2002, two-thirds of National Abortion Federation members were providing early medical abortion.29

The protocols used to provide early medical abortions have an important effect on the cost of the method and the extent to which it meets women's needs. While practices reported in our survey may have changed since then, the results indicate that many providers, using evidence based on clinical studies, are offering mifepristone abortions with variations in the approved protocol. In particular, a large majority of providers use a smaller dose of mifepristone (200 mg rather than 600 mg) and allow the client to administer the prostaglandin herself at home. In addition, a substantial minority of providers use the method beyond the approved gestational limit of 49 days. If these practices prove to be safe and effective in clinical practice and are seen as preferable to providers and clients, almost all providers of medical abortion are likely to adopt them. The finding that the practices were used most by providers with the largest caseloads supports this expectation.

A majority of providers charge more for medical abortion than for surgical abortion at six weeks. This may reflect the cost of the drug, the greater amount of counseling time required for medical abortion than for surgical abortion, the number of calls from clients with problems or questions, and the greater perceived need for active follow-up of medical abortion clients, to ensure that the abortion was completed without complications. As providers gain experience with early medical abortion, however, these services may become more routine, and the additional expenses for medical abortion could fall.

Picketing remains prevalent at abortion facilities, especially at those with large caseloads. Other types of harassment have declined over time but have not disappeared. This decrease may reflect the impact that federal legislation to protect access to medical facilities has had in deterring illegal antiabortion activity, as well as the fact that by September 2000, 15 states had laws protecting access to clinics that provide reproductive health services.30 Nevertheless, a majority of clinics reported being picketed at least 20 times a year, and many women seeking abortion are exposed to the stress of noisy and sometimes threatening protesters.31 In addition, we did not ask about new types of harassment, such as anthrax threats and photographing clients and staff for publication on the Internet.

In summary, barriers to abortion services remain substantial for many women. Although services are more available at very early and late gestations than in the past and early medical abortion is now available, charges have increased and antiabortion picketing (the type of harassment most likely to affect women directly) remains at high levels. Moreover, because we surveyed abortion providers, we could not address many of the problems that women may face when seeking abortion services, including difficulties in locating services, the prevalence of misinformation about abortion, antiabortion attitudes in some subgroups and in many public forums, and state restrictions such as parental involvement requirements and 24-hour delay laws. Thus, the factors documented in this article present only a partial picture of the barriers women face in seeking abortion services.

References

1. Henshaw SK, Unintended pregnancy in the United States, Family Planning Perspectives, 1998, 30(1):24-29 & 46.

2. Finer LB and Henshaw SK, Abortion incidence and services in the United States in 2000, Perspectives on Sexual and Reproductive Health, 2003, 35(1):6-15; and Henshaw SK, Forrest JD and Blaine E, Abortion services in the United States, 1981 and 1982, Family Planning Perspectives, 1984, 16(3):119-127.

3. Finer LB and Henshaw SK, 2003, op. cit. (see reference 2).

4. Henshaw SK, 1998, op. cit. (see reference 1).

5. Jones RK, Darroch JE and Henshaw SK, Patterns in the socioeconomic characteristics of women obtaining abortions in 2000-2001, Perspectives on Sexual and Reproductive Health, 2002, 34(5):226-235.

6. Finer LB and Henshaw SK, 2003, op. cit. (see reference 2).

7. Ibid.

8. Henshaw SK, Factors hindering access to abortion services, Family Planning Perspectives, 1995, 27(2):54-59 & 87.

9. Henshaw SK and Manzella K, Factors hindering access to abortion services and changes from 1993 to 1997, paper presented at the annual meeting of the American Public Health Association, Washington, DC, Nov. 15-19, 1998; and Henshaw SK, Abortion fees and sources of payment: results from the 1997 AGI Abortion Provider Survey, paper presented at the annual meeting of the National Abortion Federation, Atlanta, Apr. 25-28, 1999.

10. Henshaw SK, Abortion incidence and services in the United States, 1995-1996, Family Planning Perspectives, 1998, 30(6):263-270 & 287.

11. Finer LB and Henshaw SK, 2003, op. cit. (see reference 2).

12. National Network of Abortion Funds, <http://www.nnaf.org>, accessed Oct. 30, 2002.

13. Finer LB and Henshaw SK, 2003, op. cit. (see reference 2).

14. Newhall EP and Winikoff B, Abortion with mifepristone and misoprostol: regimens, efficacy, acceptability and future directions, American Journal of Obstetrics and Gynecology, 2000, 183(2):S44-S53; and National Abortion Federation (NAF), Early Options: A Provider's Guide to Medical Abortion, Washington, DC: NAF, 2001.

15. Ashok PW et al., An effective regimen for early medical abortion: a report of 2,000 consecutive cases, Human Reproduction, 1998, 13(10):2962-2965; Schaff EA et al., Low-dose mifepristone 200 mg and vaginal misoprostol for abortion, Contraception, 1999, 59(1):1-6; and Schaff EA et al., Low-dose mifepristone followed by vaginal misoprostol at 48 hours for abortion up to 63 days, Contraception, 2000, 61(1):41-46.

16. Jones RK and Henshaw SK, Mifepristone for early medical abortion: experiences in France, Great Britain and Sweden, Perspectives on Sexual and Reproductive Health, 2002, 34(3):154-161.

17. Finer LB and Henshaw SK, 2003, op. cit. (see reference 2).

18. Ibid.

19. Special tabulation of February 2000 Area Resource File.

20. Frost JJ et al., Family planning clinic services in the United States: patterns and trends in the late 1990s, Family Planning Perspectives, 2001, 33(3):113-122.

21. Kaunitz AM et al., Abortions that fail, Obstetrics & Gynecology, 1985, 66(4):533-537; and Edwards J, Darney PD and Paul M, Surgical abortion in the first trimester, in: Paul M et al., eds., A Clinician's Guide to Medical and Surgical Abortion, New York: Churchill Livingstone, 1999, pp. 107-121.

22. Edwards J and Carson SA, New technologies permit safe abortion at less than six weeks' gestation and provide timely detection of ectopic gestation, American Journal of Obstetrics and Gynecology, 1997, 176(5):1101-1106.

23. Shelton JD, Brann EA and Schulz KF, Abortion utilization: does travel distance matter? Family Planning Perspectives, 1976, 8(6):260-262.

24. Torres A and Forrest JD, Why do women have abortions? Family Planning Perspectives, 1988, 20(4):169-176.

25. Joyce T, Henshaw SK and Skatrud JD, The impact of Mississippi's mandatory delay law on abortions and births, Journal of the American Medical Association, 1997, 278(8):653-658.

26. Cook PJ et al., The effects of short-term variation in abortion funding on pregnancy outcomes, Journal of Health Economics, 1999, 18(2):241-257; Trussell J et al., The impact of restricting Medicaid financing for abortion, Family Planning Perspectives, 1980, 12(3):120-130; and Chrisman M et al., Effects of restricting federal funds for abortion—Texas, Morbidity and Mortality Weekly Report, 1980, 29(22):253-254.

27. Henshaw SK and Wallisch LS, The Medicaid cutoff and abortion services for the poor, Family Planning Perspectives, 1984, 16(4):170-180.

28. Finer LB and Henshaw SK, 2003, op. cit. (see reference 2).

29. Jones RK and Henshaw SK, 2002, op. cit. (see reference 16).

30. The Alan Guttmacher Institute (AGI), The status of major abortion-related policies in the states: state laws, regulations and court decisions as of September 2000, New York: AGI, 2000.

31. Cozzarelli C and Major B, The effects of anti-abortion demonstrators and pro-choice escorts on women's psychological responses to abortion, Journal of Social and Clinical Psychology, 1994, 13(4):404-427.

Acknowledgments

The authors thank Kathleen Manzella and Jennifer Swedish for coordinating the fieldwork for this study; Jacqueline Darroch and Rachel Jones for reviewing drafts of this article; and Karen D'Angelo, April Fehling, Lori Frohwirth, Sarah Kirshen and Ednesha Saulsbury for providing research assistance. The research on which this article is based was funded in part by the Robert Sterling Clark Foundation and the Educational Foundation of America.

*We weighted these responses by the number of abortions provided in 2000.

The East South Central region consists of Alabama, Kentucky, Mississippi and Tennessee; the West North Central states are Iowa, Kansas, Minnesota, Missouri, Nebraska, North Dakota and South Dakota; and the Middle Atlantic states are New Jersey, New York and Pennsylvania.

Louisiana, Mississippi, Utah and Wisconsin required in-person counseling with the attending or referring physician at least 24 hours before the abortion is performed. Only 2% of all providers, and none of those that performed 5,000 or more abortions in 2000, were located in these states. (Ten other states required a delay after mandatory counseling, but permitted the counseling to be delivered by telephone or another means that did not require a visit.) In some cases, it may be permissible, but often not practical, for a woman to receive the mandatory counseling without visiting the abortion facility. (Source: NARAL Foundation, Who Decides? A State-by-State Review of Abortion and Reproductive Rights, 10th ed., Washington, DC: NARAL Foundation, 2001.)

§We focus on abortions at 10 weeks LMP because almost all providers perform abortions at this gestation. In addition, the charge for a surgical abortion at six weeks (and probably throughout the first trimester) is about the same as at 10 weeks. Charges at 10 weeks LMP are representative of the medical costs for most women having abortions, because 88% of abortions occur within the first trimester (source: U.S. Bureau of the Census, Statistical Abstract of the United States: 2001, 121st ed., Washington, DC: U.S. Bureau of the Census, 2001, Table 93).

**As of September 2000, the following states funded medically necessary abortions through Medicaid: California, Connecticut, Hawaii, Idaho, Illinois, Maryland, Massachusetts, Minnesota, Montana, New Jersey, New Mexico, New York, Oregon, Vermont, Washington and West Virginia. Alaska and Arizona were in violation of court orders to fund medically necessary abortions (source: reference 30).

††Idaho, Kentucky, Missouri and North Dakota.

‡‡We set the division at 50 to separate out providers that are relatively experienced with early medical abortion while maintaining an adequate number of responses in each category. Presumably, providers with relatively large medical abortion caseloads best reflect how services are evolving as providers and women gain experience with early medical abortion.

 

AUTHOR AFFILIATIONS

Stanley K. Henshaw is senior fellow, and Lawrence B. Finer is assistant director of research, both at The Alan Guttmacher Institute, New York.