The most recent data available show that six in 10 pregnancies are unintended, and that of these, about half end in abortion. In 1988, 1.6 million American women had induced abortions, but even more--1.7 million--had births resulting from unintended pregnancies. ¹ Since 1988, the abortion rate has fallen slightly and the birth rate has increased slightly, suggesting that the proportion of unintended pregnancies ending in abortion may be decreasing. This will only become clear when data on the proportion of births since 1988 resulting from unplanned pregnancies are available from Cycle V of the National Survey of Family Growth, being fielded in 1995.

We have little quantitative information on why unintended pregnancies are carried to term. Some become wanted or at least accepted by the couple or the woman, and some women or couples decide against abortion in spite of a strong preference not to have a child. In other cases, financial, physical, psychological, social or other obstacles to obtaining an abortion are undoubtedly important, as is lack of knowledge about the availability of abortion services. The extreme variation in the states' resident abortion rates, which ranged from 9.0 abortions per 1,000 women aged 15-44 in South Dakota to 45.8 per 1,000 in California in 1988, ² suggests that barriers to services affect the abortion rate.

The lack of an abortion provider within easy traveling distance is an important barrier for many women. In 1992, 41% of women of reproductive age lived in a county without a facility that performed at least 400 abortions during the year. ³ The greater the distance a woman lives from an abortion provider, the less likely she is to be able to use the provider's services.⁴

Other barriers have been created by legislation. The policy that has probably had the most impact is the almost total exclusion of abortion from federal Medicaid coverage. Although some states pay for Medicaid abortions with their own funds, most women live in states that do not do so. The most rigorous study to date of the impact of the federal funding restriction found that 18-23% of women on Medicaid who would prefer to have an abortion instead carry their pregnancies to term when Medicaid abortion funding is unavailable; ⁵ in another study, that figure was 35%.⁶ State legislative barriers that apply to all women, regardless of their income, include waiting periods, counseling requirements that involve more than one visit to the provider and, for minors, parental notification and consent requirements.

Although the impact of many of these barriers is impossible to quantify, this article uses data from The Alan Guttmacher Institute's (AGI) 1993 Abortion Provider Survey to provide information on the percentage of women who travel long distances to obtain abortion services, the availability of abortion providers for women who need services during the second trimester of pregnancy, the need to make more than one trip to the abortion facility and the amount abortion providers charge for services. In addition, it presents a measure of antiabortion harassment of abortion providers, which also affects women seeking an abortion, and providers' views of the problems that cause them the most difficulty in offering their services.

Methodology

AGI's 1993 Abortion Provider Survey, the 11th survey of all known abortion providers in the United States, updates and adds to information obtained in earlier years.⁷ Its methodology is described elsewhere. ⁸ In each survey, we ask the same questions about the number of abortions provided and add questions that vary from year to year about provider characteristics and conditions of service. In the 1993 survey, the questionnaire for nonhospital providers included questions about the proportion of patients who travel 50-100 miles or more than 100 miles for abortion services, the maximum gestation at which abortions are provided, the days of the week services are offered, whether abortions can be performed during a patient's first visit, charges, and problems affecting the facility's ability to offer abortion services. These questions refer to the time the questionnaire was completed (mid 1993). We also asked about the frequency of harasssment by antiabortion activists during 1992 and the number of abortions performed during 1991 and 1992.

Because the additional topics covered in the hospital questionnaire were limited to gestation limits and antiabortion harassment, most of the information on barriers to services pertains only to nonhospital providers. The results nevertheless represent the experience of most women who had an abortion in 1992, because nonhospital providers performed 93% of all abortions in 1992.⁹

Of the 1,525 nonhospital abortion providers in 1992, 1,106 (73%) provided information on gestation limits, 914 (60%) on charges, and 835-925 (55-61%) on other items. Some of the variation in response occurred because respondents did not have the requested information. Because response rates were higher for nonhospital facilities and facilities with large caseloads than for hospitals and small facilities, we weighted the results to reflect the correct national proportions according to facility type and caseload.

The questionnaire asked respondents to estimate the distribution of their abortion patients by distance of the patient's residence from the facility (within 50 miles, 50-100 miles or more than 100 miles). We asked a similar question in our 1989 survey, but the results are not comparable because of differences in the response categories. To reflect the experience of the typical patient, we weighted variables related to distance by the number of patients served as well as by provider type and size.

We divided nonhospital facilities into three groups: abortion clinics, nonspecialized clinics and physicians' offices. If at least half of a provider's patient visits in 1992 were for abortion services, we classified it as an abortion clinic. This category also includes facilities with clinic names that did not provide information about the percentage of patient visits for abortion care but reported providing 1,000 or more abortions in 1992, as well as physicians' offices that performed 1,500 or more abortions in 1992 and did not indicate that fewer than half of patient visits were for abortion services.

We classified facilities reporting that fewer than half of patient visits were for abortion services as either nonspecialized clinics or physicians' offices. The category called "nonspecialized clinics" includes group practices and other providers with clinic names, surgical centers and health maintenance organizations. Physicians' offices were put in this group if they reported providing more than 400 abortions in 1992 and did not indicate that at least half of patient visits were for abortion services. The category called "physicians' offices" is made up of solo or group practitioners who performed fewer than 400 abortions in 1992.

Distance

According to the estimates of abortion providers, 8% of women having abortions in nonhospital facilities in 1992 traveled more than 100 miles for abortion services, and an additional 16% traveled 50-100 miles (Figure 1).

The need to travel long distances was greatest in the East South Central census division, where an estimated 15% of those having abortions in nonhospital facilities lived more than 100 miles from the provider and about 31% lived 50-100 miles away. This census division consists of Alabama, Kentucky, Mississippi and Tennessee, states in which the relatively few abortion providers are concentrated in the largest cities; 61% of the women in this census division lived in a county with no abortion provider. Distances traveled were also above average in the other noncoastal census divisions except the East North Central region.

Where travel distances were shortest--in the Pacific and Mid-Atlantic census divisions--15% and 17%, respectively, of patients traveled 50 miles or more. The Mid-Atlantic states are dominated by large urban areas well-supplied with abortion clinics, as is California, which dominates the Pacific census division. However, even given these circumstances, 15% of women in Mid-Atlantic states and 6% of those in the Pacific region live in counties with no abortion provider.

The larger the facility, the higher the proportion of patients who travel long distances for services. Of patients of providers reporting 1,000 or more abortions in 1992, 25% came at least 50 miles, compared with only 5% of patients of the smallest providers (fewer than 30 abortions in 1992). Women who must go to a distant facility or prefer to do so are probably more likely to go to large providers, which advertise more widely and are on average less expensive than are small providers.

Women may travel long distances not because their area lacks an abortion facility, but because they desire anonymity or because they are minors who wish to avoid state parental involvement or court bypass requirements. ¹⁰ If mandatory waiting periods are adopted by more states, travel to states without such restrictions will probably increase. ¹¹ Other women may choose a distant provider to take advantage of lower fees or to obtain services such as general anesthesia that may not be available from small local providers.

Gestation Limits

Even if a woman has located a nearby abortion provider, services may not be available to her from that provider if her pregnancy has passed the earliest stages. The later in gestation an abortion is performed, the more cervical dilation is needed, the more complex the procedure and the greater the risk of complications; not all providers choose to perform these more difficult procedures.

The maximum gestation at which providers will perform abortions varies widely, depending on their skills, preferences and other factors. For pregnancies of no more than eight weeks' gestation (dated from the start of the last menstrual period), 98% of facilities will provide services; the remaining 2% are mainly physicians who perform only menstrual extractions or the earliest abortions in their offices (Table 1).

Many other physicians set their limits at between eight and 11 weeks; only 54% will serve patients at 12 weeks in their offices. Many hospitals and nonspecialized clinics also provide services only during the first 12 weeks, sometimes considered the dividing point between the first and second trimester of pregnancy. Fewer than half (48%) of facilities offer services at 13 weeks, and the proportion declines rapidly beyond that point to 22% at 20 weeks and 7% at 24 weeks.

Although abortions after 26 weeks of gestation are unrestricted in many states, they are rarely performed. Only three providers are known to perform third-trimester abortions, although others may occasionally perform such procedures. These providers accept patients past 26 weeks only under certain conditions, the most common of which are known fetal abnormalities and severe medical complications. ¹²

Hospitals and abortion clinics are more likely than other facilities to provide abortion services for women past 14 weeks of gestation. Even hospitals that provide only a moderate number of abortions may feel obligated to serve patients who need a later abortion because of serious health risks, abnormal fetuses or other pressing needs.

The larger a facility's abortion caseload, the later its maximum gestation limit is likely to be. For example, 57% of abortion clinics that performed 2,000 or more abortions in 1992 offered services at 17 weeks, compared with only 35% of those that reported fewer than 1,000 abortions.

Between our 1989 and 1993 surveys, the proportion of facilities that offered services into the second trimester increased, continuing a trend from earlier years. The change was greatest at the highest gestations; for example, the proportion accepting patients at 19 weeks rose from 19% to 23%, and the absolute number increased by 11%; at 21 weeks, the proportion increased from 9% to 13%, and the number increased by 33%. ¹³ These changes occurred among all provider types except physicians' offices. At 15 weeks and earlier, the changes were smaller, and the absolute number of providers changed little, because the total number of providers fell (from 2,582 in 1988 to 2,380 in 1992).

In addition to setting a maximum gestation limit, most providers set a minimum gestation before which they will not perform an abortion because studies have found a higher rate of complications in very early procedures and because the small size of the embryo makes it difficult to be sure that the pregnancy has been ended. ¹⁴ The most common requirement is six weeks since the last menstrual period, the criterion used by 43% of nonhospital facilities (not shown). Nineteen percent of facilities require seven weeks and 5% require eight weeks, although 19% set their minimum at five weeks and 7% at four weeks. Other facilities use criteria based on relatively insensitive pregnancy tests (4%), and 2% will evacuate the uterus if the amniotic sac is visible on ultrasound. No minimum was specified by 2%. All facilities presumably require positive pregnancy tests.¹⁵ Facilities with high abortion caseloads tend to set higher minimum gestation limits than do those that perform only a few abortions.

Appointment Availability

How soon a woman is able to have an abortion depends not only on how long it takes her to locate and get to a provider and on the provider's gestation limits, but also on how often the facility is open and how long it takes to get an appointment. Nonhospital facilities were asked, "What is the average interval between the first telephone or walk-in contact with a woman and the procedure?" Most facilities reported short average periods between first contact and the abortion procedure. Thirty-three percent estimated the interval at 1-3 days, 17% at four days, 17% at 5-6 days, 18% at seven days, and 14% at more than seven days (not shown).

Abortion clinics reported much less time between first contact and the procedure than did other facilities. Periods of one week or more were indicated by only 17% of abortion clinics, compared with 43% of nonspecialized clinics and 35% of physicians' offices. Similarly, periods of one week or more were reported by 25% of clinics with 1992 abortion caseloads of 400 or more, compared with 38% of smaller facilities. The period between first contact and the pregnancy termination is longer now than in 1981, when a survey of providers of 400 or more abortions indicated that only 11% typically needed seven or more days, ¹⁶ compared with 25% in 1993.

A certain amount of delay is inevitable if facilities perform abortions only a few days a week, as is the case with many smaller providers. For example, a woman seeking services from a provider that performs abortions only one day a week might have to wait up to six days for an appointment if the facility has appointment slots available on the next scheduled service day, and longer if the appointments are all taken. Of all nonhospital providers in our survey, 11% performed abortions on only one day per week and 15% on two days. Only 38% performed abortions five or more days each week. Physicians with small abortion practices were most likely to perform abortions any day of the week (51%), while nonspecialized clinics were most likely to offer the procedure on only one or two days (41%). Thus, the facilities most likely to perform abortions on only one or two days a week appear to be those that are large enough to have specialized abortion sessions but small enough to need only one or two such sessions a week.

Once an appointment has been made, a woman is usually able to have her pregnancy terminated in one trip to the facility unless state law requires two visits.¹⁷ As Table 2

shows, 60% of nonhospital providers said it is their usual procedure to perform first-trimester abortions on the woman's first visit, and all but 14% said they would provide services on one visit under some circumstances--if the woman would have to travel a long distance for a second visit (24%), if a delay would push her beyond maximum gestation limits or into a higher cost category (20%) or if she had already received counseling or preliminary evaluation from another source (19%).

The availability of abortion services during the first visit is highly dependent on the abortion caseload of the provider. Almost all clinics (91%) that perform 2,000 or more abortions a year routinely provide services in one visit, while only 25% of facilities with the smallest caseloads do so. Because most women go to a facility with a large caseload, 86% of women having a nonhospital abortion are able to obtain services in one visit if they choose to do so.

Charges

Although exact data are not available, abortion providers report that a large majority of abortions are paid for by the patients themselves rather than by insurance. There are several reasons for this. About one-third of women do not have employer-based insurance, and although Medicaid covers some of these patients, in most states it rarely pays for abortions. One-third of private insurance plans do not cover abortion or cover it only for certain medical indications.¹⁸ Women with insurance that covers abortion may not have met the required deductible, or they may not use their insurance because of concerns that confidentiality might be jeopardized by benefits statements sent to their home, the need for someone else's signature (if the patient is covered under another person's insurance policy) or reports sent to the employer. Therefore, cost can be a significant barrier for some women.

As Table 3 shows, the average nonhospital facility charged $341 in 1993 for an abortion at 10 weeks with local anesthesia; the median fee was $298, and the range was $140-$1,700.

Clinics performing at least 400 procedures per year reported the lowest average charges ($287-$296), followed by facilities reporting 30-390 abortions ($391) and those performing even fewer procedures ($463). Because most women go to the larger clinics with lower fees, the average patient paid about $296 for the abortion itself, not including other expenses such as travel, time missed from work and any additional medical services needed by a particular woman.

The average amount paid in 1993 was 18% higher than the average amount paid in 1989. This increase is much smaller than the 35% increase in the consumer price index for medical care over the same period and about the same as the 17% rise in the consumer price index for all items. In the period between the two previous surveys, 1986 to 1989, abortion charges rose more than the cost of living.¹⁹

Prices increase sharply with gestation after about 12 weeks. In 1993, on average, nonhospital providers charged $604 for an abortion at 16 weeks and $1,067 at 20 weeks; the median charges were slightly lower. Fees ranged as high as $2,500 at 16 weeks and $3,015 at 20 weeks.

Charges at abortion clinics, which averaged $288 at 10 weeks, were distinctly lower than charges at nonspecialized clinics ($335) or physicians' offices ($410), but these differences largely disappear when patient caseload is taken into account. At 16 and 20 weeks of gestation as well, the mean and median charges of physicians operating in their offices were higher than those of nonspecialized clinics, which in turn were higher than those of abortion clinics, but the differences may be attributable to variations in the volume of services.

Although no data were collected on 1993 charges for abortion services in hospitals, the average hospital charged $1,757 for a first-trimester outpatient abortion in 1991.²⁰ These charges, which included the fees of the hospital, the surgeon and, when required, an anesthesiologist, were almost six times the charge in the average nonhospital facility. Because nonhospital facility charges have risen more slowly than the consumer price index for medical care, the average charge for a hospital abortion today is probably at least six times that of a nonhospital abortion.

Charges varied somewhat according to the region of the country. Services were least expensive in the South, where the average amount paid for an abortion at 10 weeks in a nonhospital facility was $271, compared with $292 in the West, $306 in the Midwest and $329 in the East.

For women who seek abortions with general anesthesia, costs may be significantly higher. Relatively few nonhospital providers--33% of abortion clinics, 17% of nonspecialized clinics and 17% of physicians' offices--reported offering general anesthesia. Sixty-three percent of facilities that offered general anesthesia charged extra for it. The extra charge varied according to provider type, averaging $114 at abortion clinics, $136 at nonspecialized clinics and $306 in physicians' offices.

Harassment

Another barrier facing many women seeking abortion services is harassment by antiabortion protesters, which also affects the ability of facilities to offer services. As in earlier rounds, our 1993 survey asked providers to indicate the number of times they had experienced various types of harassment during the previous year.

In all, 55% of nonhospital providers reported experiencing at least one of the 10 listed types of harassment during 1992 (not shown). Forty-nine percent reported experiencing harassment during 1988, but the statistics are not exactly comparable because we asked about only six types of harassment in the earlier survey. Harassment is strongly associated with the abortion caseload, with 86% of facilities that performed 400 or more abortions in 1992 reporting some harassment, compared with 29% of providers with a smaller caseload. As in the previous survey, reports of harassment varied by region of the country, with the Midwest reporting the highest level and the West the lowest. In the Midwest, 48% of the larger clinics (nonhospital providers of 400 or more abortions during 1992) experienced four or more types of harassment, compared with 22% of those in the West.

The experience of larger providers is shown in Table 4.

Picketing, the most common type of harassment, was reported by 83% of these clinics. Many were picketed regularly; 62% reported 20 or more such incidents during the year. Although the change in the proportion reporting any picketing from 1988 to 1992 is small, the increase from 53% (not shown) to 62% in the proportion reporting 20 or more incidents of picketing is statistically significant. Similarly, the proportion that experienced picketing with physical contact or blocking of patient access increased from 46% to 50%, and the proportion that reported experiencing this type of harassment five times or more increased from 18% to 23%.

Demonstrations resulting in arrests were reported by 34% of clinics, almost the same proportion that reported blockades (30%). The proportion of clinics reporting demonstrations resulting in arrests was slightly lower in 1992 than in 1988 (38%), suggesting that the use of violent demonstrations has peaked and may be on the decline.

Other antiabortion activities have increased, however. Vandalism affected 42% of clinics in 1992, up from 34% in 1988 and 28% in 1985. Picketing of homes of staff members rose from 17% in 1988 to 28% in 1992, and tracing of patients' license plates increased from 16% in 1985 to 24% in 1992. Two relatively new activities, stalking of staff or patients and attacks with butyric acid,* were fairly frequent, at 30% and 12%, respectively. The survey responses indicate that approximately 80 chemical attacks occurred in 1992, all or almost all of which involved butyric acid. Bomb threats have become markedly less frequent; 48% of clinics experienced bomb threats in 1985 but only 24% did so in 1992.

Information compiled by the National Abortion Federation on harassment experienced by its members and others confirm some of the trends noted in our survey. Their data indicate increases in arson, attempted arson, clinic invasions, vandalism and picketing and decreases in bomb threats and blockades between 1988 and 1992.²¹

We asked hospital providers about only three types of harassment; 15% reported picketing, 2% reported picketing with physical contact or blocking of patients and 1% reported vandalism. These activities were strongly associated with the number of abortions performed in the hospital: Thirty-six percent of hospitals that reported an annual caseload of 1,000 or more abortions experienced harassment, compared with 10% of those that provided fewer than 30 abortions.

Problems Faced by Providers

The questionnaire for nonhospital providers concluded with a question asking for a list, in order of importance, of the major problems, if any, that had affected their ability to provide abortion services over the previous 12 months. Their responses indicate that providers see harassment and other expressions of antiabortion sentiment in the community as their most important problem with respect to abortion. Although 37% specified that they had no major problem, 30% cited conditions directly associated with harassment as their most important problem. These problems included picketing (8%), demonstrations and blockades (5%), vandalism and other direct action (8%) and other types or effects of harassment (8%). An additional 11% mentioned the physician shortage or other staffing problems that may be indirectly related to harassment, and 2% named lack of cooperation of police or other authorities.

Four percent of providers cited more general problems, including lack of support in the community, boycotts of their practice by nonabortion patients, fear of becoming known as a major abortion provider, and medical practice partners who oppose the provision of abortion services. Another type of problem, named by 6%, concerned regulation and legislation at the national and state level, especially the denial of Medicaid payment for abortion. Six percent identified business problems, including reduced demand, competition and rising costs, and 1% cited insurance problems.

Discussion

Abortion is semiurgent care for several reasons: The risk of complications increases with gestation, abortion becomes impossible if it is delayed too long, and most women who have chosen to terminate their pregnancies want to do so as early as possible. Yet, women seeking abortions must cope with a number of barriers that do not obstruct other kinds of semiurgent medical and surgical care. This study presents information on some of the key barriers but is not meant to cover all difficulties that women have in obtaining abortions or that providers have in making services available.

In some ways and for some women, the abortion service system provides efficient and, judging from the low mortality and morbidity rates, ²² high-quality care. Women who live in large urban areas and have the necessary financial resources can usually obtain abortion services in a single visit without a long wait for an appointment. Prices have risen along with the cost of living but have increased less than for other types of health care.

Many women who have unintended pregnancies, however, have a low income and lack health insurance that covers abortion services. For these women, the average charge of $296 for a first-trimester nonhospital abortion may be a major barrier. Charges approximately double for a woman who is delayed to 16 weeks of gestation and may almost double again at 20 weeks. The exclusion of abortion from Medicaid coverage in most states is perhaps the most severe legislative restriction now in effect.

Nonfinancial obstacles affect some women of all income levels, although those with few financial resources may have the most difficulty overcoming them. Distance from a provider is often a problem, especially for women who need a second-trimester procedure. Evidence of an increasing provider shortage may be seen in the increase over the past 10 years in the average time between a woman's first contact with a provider and the day the abortion is performed; 14% of nonhospital providers cite an average delay of more than one week. (These data were collected before a 24-hour waiting period was in effect in any state except Mississippi.)†

Harassment of women and providers, always a problem, has become more widespread. Although bomb threats and demonstrations resulting in arrests have decreased, these appear to be exceptions to a pattern of increasing antiabortion activism. The rise in the proportion of clinics subject to frequent picketing suggests that in 1992, more people than ever were engaged in the most visible antiabortion activity. Vandalism increased by 50% between 1988 and 1992, and a new form has been devised, attack with butyric acid. The fatal shootings of two physicians, a clinic escort and two clinic staff should be viewed within this context of gradually escalatin violence.

Legislation and government regulation have created additional barriers for both women and providers. Laws requiring two clinic visits cause a delay, often of more than one day, for most of the women affected. At present, services are commonly provided in one clinic visit, and almost all women who travel long distances for abortion services or have other reasons to need immediate care can have their pregnancies terminated in a single visit. Because a majority of providers do not perform abortions every day of the week, the need for two visits can easily cause a delay of more than one day.

Most of the difficulties in providing and obtaining access to abortion services would disappear if abortion were integrated with other health care for women, but health care providers are unlikely to adopt this approach as long as opposition to abortion is widespread. Alternatively, availability of medical methods of early abortion, such as mifepristone, also known as RU-486, or methotrexate in combination with a prostaglandin, has the potential to change both the numbers and types of abortion providers and to reduce some of the barriers. For example, early gestational limits might be dropped and an emphasis placed on women's coming to a provider as early in pregnancy as possible. If abortion services become available in more locations, distances and waiting times for services would be reduced.

Medical abortion methods, however, will not solve all the problems of access. The method can be used only very early in pregnancy, surgical backup facilities are needed, charges are likely to be at least as high as for surgical abortion, two or more physician visits will be required and antiabortion harassment will still be possible in most cases. In addition, legal restrictions on abortion will remain, regardless of changes in abortion technology.