Obstetrician-Gynecologists

According to the American Medical Association's Physician Master File, 31,364 physicians in the United States identified themselves as obstetrics-gynecology specialists in 1986 (AMA, 1987). The American College of Obstetricians and Gynecologists, the specialty society for this discipline, counts 27,219 obstetrician-gynecologists as active fellows, 6,587 as junior fellows (still in training), and 2,746 as life and founding life fellows (usually inactive practitioners) (ACOG, 1988).

Like other surgical specialists, most obstetrician-gynecologists work in metropolitan areas.¹ ACOG data show that 17 states, primarily larger states in the southern and western parts of the country, have areas with fewer than 10 obstetrician-gynecologists per 100,000 women age 15 to 44 in the population, 35 states have regions with fewer than 20 per 100,000, and 22 states have areas with no obstetrician-gynecologists at all (affecting close to 400,000 women age 15 to 44) (ACOG, 1988) (Table 2.1).

TABLE 2.1

Active Nonfederal Physicians and Physician-to-Population Ratios for Obstetrician-Gynecologists and Family and General Practitioners, by State, 1985.

Family and General Practitioners

Prior to 1969, there was no separate specialty called ''family practice." General practitioners were physicians who had completed medical school and one year of internship, and specialists were physicians who had completed the longer residency program in their specialty. Concern about increasing specialization and the declining number of primary care physicians led to the establishment of family practice as a recognized program of training and specialization.

By 1980, there were approximately 27,000 family physicians, 18,000 of whom were in office-based practice (AMA, 1987). By 1986, the number in office-based practice had grown by 70 percent, to more than 31,000 (AMA, 1987). During the same period, the number of general practitioners fell, from approximately 52,000 to 30,000, as older physicians retired and younger ones entered family practice or specialty medicine (AMA, 1987). As a result, the number of family and general practitioners in office-based practice in 1986 was nearly the same as in 1970—54,000 physicians—about 30 percent of whom were in nonmetropolitan areas. The AMA estimates that approximately 68,000 physicians designate their specialty as family practice or general practice, but this number includes many who do not engage in office-based practice or who are not in practice at all (AMA, 1987).

Although the number of specialists practicing in rural areas has grown in recent years, general and family practitioners continue to be the principal providers of primary and obstetrical care in these areas. Fifty-three percent of all visits to physicians in nonmetropolitan areas were to family physicians, compared with 10 percent to internists and 7 percent to obstetrician-gynecologists (National Ambulatory Medical Care Survey, 1987). If one considers only visits by adults, the proportion for family physicians rises to 70 percent (National Ambulatory Medical Care Survey, 1987).

Not all family physicians practice obstetrics, however. The American Academy of Family Physicians (AAFP) reported that 71 percent of its practicing members have offered obstetrical care at some time during their careers (AAFP, 1986) and that 35 percent do so currently (AAFP, 1987). Most of those who accepted obstetrical patients had served in residency programs (73 percent) and were board certified in family practice (83 percent). Of these, 33 percent did not perform complicated deliveries or cesarean sections; 43 percent performed complicated deliveries only; and 23 percent performed both (AAFP, 1987). Those performing complicated deliveries or cesarean sections or both were more likely to have completed residencies and to be board certified.

The participation of family and general practitioners in obstetrics varies greatly by region of the country, urban or rural location, and physician age. A survey based on 1977-1978 data found that family and general practitioners in the Northeast saw very few obstetrical patients (only 0.4 percent of diagnoses were prenatal or postpartum care), whereas those in the north central states saw relatively many (4.3 percent of diagnoses) (Rosenblatt et al., 1982). Those in rural areas were twice as likely to see obstetrical patients as those in urban areas (5.2 percent of diagnoses versus 2.3 percent) (Rosenblatt et al., 1982). A study of Michigan family practice residents found that 55 percent of third-year residents planned to practice obstetrics on graduation; plans to practice in rural communities were positively correlated with the decision to include obstetrics, whereas plans to practice in suburban areas were negatively correlated (Smith and Howard, 1987).

Because the data make it clear that family physicians are critical to the provision of obstetrical services in rural areas, the committee was interested in determining precisely how extensive their role in these areas is. Accordingly, it commissioned research on this question as part of its fact-finding. The results were striking and are reported at length in Chapter 3. It should be noted at the outset that, according to estimates prepared for the committee, two-thirds of all obstetrical providers of private obstetrical care in rural areas are family physicians. During the early 1980s—prior to the dramatic increase in professional liability insurance expenses—there were an estimated 16,700 physicians providing obstetrical care in nonmetropolitan areas, two-thirds of whom were family and general practitioners. By contrast, there were only 5,400 obstetrician-gynecologists practicing in nonmetropolitan areas (J. Chapin, director of research, ACOG, personal communication, 1988).

Certified Nurse-Midwives

Certified nurse-midwives are registered nurses with additional training in midwifery. They are certified by the American College of Nurse-Midwives (ACNM), which states that "nurse-midwifery practice is the independent management of essentially normal newborns and women antepartally, intrapartally and postpartally and/or gynecologically. This occurs within a health care system which provides for medical consultation, collaborative management and referral" (ACNM, 1984). The ACNM believes that there are approximately 3,500 CNMs in the United States but that only 2,000 to 2,500 practice. ACNM members number 2,100, and between 200 and 250 nurse-midwives are certified each year. The number of persons entering nurse-midwifery training has remained stable over the past few years; however, schools are experiencing a decline in the number of applicants (ACNM, 1988).

A 1982 ACNM survey found that CNMs, on average, attend 75 births per year (ACNM, 1984). The heaviest concentrations of CNM deliveries were in the Northeast, Southeast, and West (Klerman and Scholle, 1988). According to a 1985 ACNM statement, 75 percent of CNM-attended births occur in hospitals and 15 percent in freestanding birth centers (ACNM, 1988).

The 1982 ACNM survey found that approximately 59 percent of CNMs were employed by organized facilities, including hospitals (36 percent), public health agencies (9 percent), health maintenance organizations (HMOs) (6 percent), the military (6 percent), and university health services (2 percent). The remainder were in private practices operated by either nurse-midwives or physicians (41 percent). Three-quarters of the CNMs were salaried employees; others received income through fee-for-service payments and direct third-party payment (26 percent). The mean 1981 income of CNMs ranged from $18,544 in the Southwest to $25,245 in the Midwest (ACNM, 1984).

The scope of services provided by CNMs varied with their employers. The CNMs working in hospitals and HMOs and those in private practice or maternity services were more likely to work in the largest metropolitan areas. Those working in hospitals, maternity services, and private nurse-midwifery practices were most likely to provide prenatal, labor, and delivery care. Those in practices run by CNMs were also more likely to supervise well-baby care. Those working in public health agencies were less likely to do labor, delivery, and postpartum examinations (ACNM, 1984).

Comprehensive data regarding the characteristics of women whose care is provided by nurse-midwives are lacking. Although the 1982 survey found that CNM patients were older, of lower parity, and better educated than all childbearing women, a 1985 survey of factors affecting the success of nurse-midwifery practice found that more than one-third of CNMs worked in practices in which most of the clients were poor (Rooks and Haas, 1986).

Other Practitioners

The Nurses' Association of the American College of Obstetricians and Gynecologists (NAACOG) is comprised of more than 20,700 nurses, representing approximately 20 percent of those employed in the specialty (NAACOG, 1987). Two-thirds of NAACOG members work in a hospital inpatient setting, and labor and delivery is the clinical area in which they practice most frequently. In 1978 the NAACOG Certification Corporation began certifying nurses for special knowledge in specified areas of obstetrical, gynecological, and neonatal nursing. According to a 1987 General Accounting Office study, less than 2 percent of all medical malpractice claims closed in 1984 were against nurses (GAO, 1987).

Hospitals

Despite the large contribution of hospital clinics to prenatal care, especially for poor women, very little is known about the care provided in this setting. There is no source of national data on hospital clinic utilization that separates obstetrics or gynecology visits from other visits.

Local and State Health Departments

In 1984 the Public Health Foundation reported that 40 state health agencies provided prenatal clinical services to over 361,300 women (Public Health Foundation, 1987). A 1986 Children's Defense Fund survey of officials representing the 51 agencies (50 states plus the District of Columbia) receiving maternal and child health funding under Title V of the Social Security Act found that 48 offered some prenatal care for indigent women, usually through clinics operated by local health departments (Rosenbaum et al., 1988).

Eligibility requirements and distribution of services varied widely from state to state. Eleven states based eligibility on specific conditions, offering services to high-risk, unmarried, teenage, or unemployed women. Thirty-six states used uniform financial eligibility criteria, usually meaning that services were provided without charge to certain groups, such as those with family incomes below the federal poverty level. Services were often available to other women on the basis of a sliding fee scale.

Women who receive prenatal care at clinics subsidized by state Title V agencies are more likely to have incomes below the federal poverty level, to be young, and to be uninsured. An Alan Guttmacher Institute survey of directors of 25 state Title V agencies found that 64 percent of prenatal patients had incomes below the federal poverty level, 34 percent between 100 and 200 percent of that level, and 2 percent at approximately 200 percent of the poverty level. Sixty-four percent were uninsured, 27 percent received Medicaid, and 9 percent were privately insured. Sixty-two percent were between 20 and 34 years old, 34 percent were teenagers, and 4 percent were 35 years and older (AGI, 1987).

Community and Migrant Health Centers

Community Health Centers and Migrant Health Centers are federally funded institutions providing primary health care services, including perinatal services, to medically underserved and disadvantaged populations. Located in areas designated by the U.S. Public Health Service as medically underserved, they provide a broad range of primary and specialized medical and support services to individuals and families who otherwise would not have access to such care. In 1987 there were 567 such centers; 58 percent of them were located in cities.

Standards for the centers are established by the Department of Health and Human Services and require that the centers either provide or arrange for prenatal care and delivery services and that they develop a referral relationship with at least one hospital. In a 1987 survey of health centers 85 percent of the respondents were found to either provide or pay for prenatal care for an estimated 213,000 women (AGI, 1987). Almost two-thirds (64 percent) of these women had incomes below the federal poverty level, and another quarter (24 percent) had incomes between 100 and 200 percent of that level.

Approximately 300 health centers have on their staffs obstetrical specialists, many of them National Health Service Corps (NHSC) physicians repaying medical education scholarships and loans by working in medically underserved areas. As of June 1988, almost 70 percent of the 1,297 NHSC physicians were health center employees, including 419 family physicians, 104 obstetrician-gynecologists, and 50 general practitioners. More than half of NHSC placements are in rural areas. The peak of NHSC placements occurred in 1985-1986, and the last of these scholarships has been awarded. In 1989 approximately 100 placements will be made. The NHSC is attempting to keep physicians in underserved areas after their obligation is met, but the retention rate is currently only between 30 and 40 percent. In addition, the corp is recruiting nonobligated physicians and offering a loan repayment program.

There are grossly insufficient numbers of NHSC obstetrical specialists to meet the needs of the centers, and by 1992 virtually all these specialists are likely to be gone. For this reason, many centers must either provide prenatal care through staff members who are not obstetrician-gynecologists or contract with obstetricians in the community to furnish the care their patients need.

Health centers' budgets are quite restricted. In fiscal year 1987 Congress appropriated $400 million to health centers' programs, yet centers served approximately 5.5 million patients, two-thirds of whom were children and women of childbearing age. Estimates of the number of prenatal patients served by health centers range from 120,000 to more than 200,000 a year. It is estimated that approximately 62,000 babies are delivered in health centers annually. Of those women who received prenatal care at health centers but whose babies were not delivered there, most were referred to an obstetrician early in the course of their pregnancy and were followed jointly by the center and the obstetrician. This arrangement is frequently necessary, because many centers do not have on-site obstetricians. Births to center patients made up 6.5 percent of all poor and near-poor births (approxiamately 1.85 million) in the United States in 1987. A cost-based analysis of perinatal services furnished by health centers in 1986 and conducted by the Public Health Service has estimated that more than $85 million of the program funding for Community and Migrant Health Centers was devoted to such care.

Because the committee was persuaded that Community and Migrant Health Centers are an important source of obstetrical care for low-income women, it commissioned a study on the effects of medical professional liability on the delivery of obstetrical care in Community and Migrant Health Centers. The results are published in the companion volume of this report, and the committee's discussion of them is in Chapter 4.

Evidence of Existing Shortages of Obstetrical Services

In evaluating the effects of medical professional liability on access to obstetrical care, the committee was mindful of the larger problem of constrained access to health services for low-income and minority women generally in the United States and of the fact that the American maternity system is seriously underfinanced (IOM, 1988). Seventeen percent of all women have no health insurance coverage, and others have inadequate coverage (AGI, 1987). Below, the committee summarizes some of the disturbing trends that point to shortages of obstetrical services for certain groups of women. It is important to bear in mind that the effects of professional liability concerns are being experienced in a system that is already falling far short of meeting the public health goals of this nation.

To be sure, the problems associated with the underfinancing of the maternity system in the United States make it difficult to assess the independent effect of professional liability concerns on the delivery of obstetrical care. Moreover, the committee recognizes that, unless these critical finance issues are also addressed, solutions to the problems wrought by professional liability concerns on obstetrics may not be fully realized. Conversely, the committee notes that resolving the professional liability problems will not by itself address all the unmet needs for obstetrical services. However, the committee does believe that professional liability problems are exacerbating the problems faced by the U.S. maternity system.

Insufficient Prenatal Care

The data available to the committee make it clear that, although the United States has made many strides in improving maternal and child health in the last two decades, access to prenatal care is far less than optimal, and the incidence of low birthweight and of certain preventable complications of pregnancy and delivery remains too high (IOM, 1988). In 1985 approximately one-fourth of all infants in the United States were born to women who did not begin prenatal care in the first three months of pregnancy; almost one-third were born to women who did not obtain the amount of care currently recommended by the ACOG (1985). More than 5 percent were born to women who began care only in the third trimester of pregnancy or who had no care at all.

For certain groups, these percentages were higher: for example, only 47 percent of black teenagers began care in the first trimester of pregnancy, and 14 percent obtained no care or care only in the third trimester (National Center for Health Statistics, 1987). Similar patterns were reflected in a study by the Alan Guttmacher Institute (1987), which found that 34 percent of mothers obtained less than adequate prenatal care. Again, certain groups exhibited even higher percentages of insufficient prenatal care. Fifty-one percent of black women and 47 percent of Hispanic women obtained less than adequate care. Women younger than 20 years were more than twice as likely to have received less than adequate care (55.7 percent), as were women aged 35 years and older (26.4 percent). Women who were unmarried, who had relatively little education, or who were poor also were more likely to have obtained insufficient prenatal care (AGI, 1987) (Figure 2.1). These trends are particularly disturbing in light of the broad consensus that prenatal care is an effective intervention that is clearly associated with improved outcomes of pregnancy.

Figure 2.1

Percentage of women who had obtained insufficient prenatal care, by selected characteristics, 1980. Source: Alan Guttmacher Institute. 1987. The Financing of Maternity Care in the United States. New York.

Physician Shortage in Rural Areas

The 1970s saw a major upsurge in physician availability in rural areas. Growth was particularly prevalent among specialists, who increased to almost two-thirds of all physicians in nonmetropolitan office-based practice. The experience of the 1980s indicates that this growth was a short-lived phenomenon fueled principally by the general rise in physician supply. Although the percentage of fourth-year medical students selecting family practice has not varied much in the 1980s (see Table 3.1, Chapter 3), fewer physicians have entered medical practice, and thus growth rates in both metropolitan and nonmetropolitan areas have declined (AMA, 1987). The most recent data suggest that, as a result, nonmetropolitan areas may be losing again in the competition for physicians. Between 1983 and 1986, the absolute number of physicians in nonmetropolitan, office-based practice declined by 2 percent (AMA, 1987). Between 1985 and 1986 alone, nonmetropolitan areas lost more than 4,000 physicians, whereas the number of physicians in metropolitan areas remained stable. In that one year the number of specialists in nonmetropolitan areas declined by 10 percent; family and general practitioners fell by almost 6 percent (Figure 2.2).

Figure 2.2

Physicians in office practice, nonmetropolitan areas, 1970-1986. Source: Lewis-Idema, D. 1988. Professional liability issues affecting family practitioners and delivery of obstetrical services in rural areas. Paper prepared for the Institute of Medicine, (more...)

A 1986 Robert Wood Johnson Foundation survey reported that residents of metropolitan and nonmetropolitan areas experienced approximately equal access to health care but that larger proportions of rural Americans are in poor health (Robert Wood Johnson Foundation, 1987). This national finding, which did not specifically examine obstetrical care, must be considered with other data which suggest that there is considerable variation in access from region to region and specialty to specialty, however. In 1986 the AMA reported that 126 U.S. counties in 25 states had no practicing physician (AMA, 1987); these counties had 466,800 residents (0.2 percent of the U.S. population). All were rural and all had very low population density, an average of 4 persons per square mile; the most densely populated county among these had 60 persons per square mile.

The committee's findings relating to the availability of obstetrical providers in rural areas are discussed in detail in Chapter 3. The implications of the data presented there must be considered against this background of emerging shortages in rural areas.

Projected Increase in Need for Obstetrical Services

Determining the impact of liability issues on maternity care would be easier if there were accepted standards for the number of providers needed to care for pregnant women and projections of the number actually expected to be available. If such standards existed, present and future deviations could be examined and their relationship to the "malpractice crisis" at least surmised.

Several attempts have been made to project manpower needs in medicine, although very few have been specialty specific. The 1980 report of the Graduate Medical Education National Advisory Committee (GMENAC) is the best known effort in this regard (GMENAC, 1981). GMENAC projected that this country would have an excess of 70,000 physicians by 1990, including over 10,000 more obstetrician-gynecologists than would be needed. GMENAC procedures for modeling physician requirements and supply have received harsh criticism from organized medicine, particularly the ACOG, primarily because of their dependence on expert opinions and imperfect data sources. The Bureau of Health Professions (BHP) in the Department of Health and Human Services has used a "demand" model based on health services utilization data and economic and demographic trends to estimate the number of physicians needed (BHP, 1982). These data, which were originally not disaggregated by specialty, do not take into account the chronic shortages of care for low-income women. The committee concluded that neither the GMENAC nor the BHP projections offered a reliable benchmark for evaluating the current situation in obstetrics.

Although it is impossible to project overall need for obstetrical services with any certainty, the committee believes that available data suggest an increase in need in the near future, for three reasons: a rise in the number of births among women who may need additional prenatal visits or prenatal and delivery care from specialists, an increase in the mean number of prenatal visits per pregnant woman, and a continuation of the trend toward more complex perinatal procedures.

The birthrate has risen slowly but steadily since the mid-1970s. Preliminary data indicate that approximately 3.83 million infants were born in 1987, a 2 percent increase over 1986 (Klerman and Scholle, 1988). This increase over the last decade is related to the increase in the number of women of childbearing age and the increasing tendency of women in their twenties and thirties to delay birth. The birthrate to women aged 30 years and older is expected to increase from 25 percent of live births in 1985 to 30 percent in 1995 (AHA, 1987). However, the Census Bureau projects that the birthrate will peak by the late 1980s and then decline through the 1990s as the female population of childbearing age declines (Bureau of the Census, 1984).

The birthrate among women at higher than normal risk of having a complicated pregnancy or delivery—namely older, unmarried, and minority women—is rising and will continue to rise. Seven percent of births were to women age 35 years and older in 1986, compared with 4.6 percent in 1980; 23.4 percent of births were to unmarried women in 1986, compared to 17.8 percent in 1980; 20.9 percent of births were to nonwhite women in 1985, compared with 19.7 percent in 1980 (AGI, 1987) (see Table 2.4). Minority births are expected to constitute 22 percent of all births by the turn of the century (AHA, 1987). This increase among women who are likely to need additional prenatal visits or care from specialists, many of whom are poor and who are already underserved, should increase the need for maternity services.

TABLE 2.4

Births to Minority, Unmarried, and Older Women, 1970 and 1980-1986.

The discrepancy between the accepted standard of prenatal care and the actual receipt of such care in the United States has been well documented, as noted above. If the multiple campaigns and outreach efforts to increase the number of women who receive adequate prenatal care are successful, the need for services should increase.

Finally, the range and number of diagnostic procedures routinely used in obstetrical services continue to increase. These currently include ultrasonography, amniocentesis, chorionic villi sampling, stress testing, electronic fetal monitoring, and cesarean sections. The increased tendency to test prenatally, the rise in "defensive" procedures discussed in Chapter 5, and widespread consumer acceptance of high-technology obstetrics are likely to contribute to an increase in the need for obstetrical services.