Women and Health Research: Ethical and Legal Issues of Including Women in Clinical Studies: Volume 2: Workshop and Commissioned Papers (1994)

Vanessa Northington Gamble and Bonnie Ellen Blustein

Federal guidelines now call for the inclusion of women in clinical studies unless a compelling reason is given for their exclusion. Initiatives to recruit women must not forget the diversity of women. Issues of race and ethnicity cannot be neglected. The purpose of this report is to examine the implications of racial differences on research involving women. It will focus primarily on African Americans. The report is divided into four sections. The first section explores definitions of race and the problems associated with using race as a variable in medical research. The second part analyzes how issues of race and racism have influenced past research efforts. It specifically concentrates on the fertility and osteoporosis literature. The third section studies an important obstacle to the inclusion of women of color in medical research—historically based distrust. The final part offers some recommendations about how racial differences in studies should be conceptualized.

Over the last two decades, concepts of race have come under increasingly sophisticated analysis. Historians, sociologists, and philosophers have joined colleagues in the biomedical sciences and anthropology in the attempt to refine or critique these ideas. Interdisciplinary studies have made plain that "race" is often used casually and unreflectively in ways that may seem intuitively obvious but are profoundly flawed. Historian Evelyn Brooks Higginbotham aptly describes this view. She writes, "When we talk about the concept of race, most

people believe that they know it when they see it but arrive at nothing short of confusion when pressed to define it."¹ But in the context of over 250 years of the mixing of gene pools in the United States, race is not always self-evident. As policies are developed and implemented to include more women of color in clinical trials, it is critical that the distinctions and implications of interpretations of race be more clearly scrutinized.

Concepts of race have evolved and changed since natural scientists first began in the seventeenth century to categorize human beings according to physical attributes. Indeed no consensus even exists among physical anthropologists, biologists, and social scientists about the definitions of race. There is, however, a legacy of debate, confusion, and controversy surrounding the meaning of the term. Divergent interpretations have centered primarily on whether race should be considered a biological construct or a sociological one. Biological constructionists hold that races are genetic entities that are fixed, immutable, and genetically determined. Different populations can be distinguished by distinctive traits that are inherited such as skin color, body build, facial features, and cephalic index. Skin color is the most common criterion used to classify race. Race is often viewed as synonymous with skin color. Thus, a phenotypic attribute is seen as an accurate measure of genotypic differences between human beings.

The social construction model holds that race is a social, historical, and political entity without any essential biological coherence. It is not a natural, fixed category; rather it has been created by society to recognize difference and establish social relationships. Race is viewed as a "highly contested representation of relations of power between social categories by which individuals are identified and identify themselves."² Therefore it cannot be understood outside of its historical and social context.

A historical examination of the concept of race clearly challenges the validity of the biological concept of race. Racial classification has been and continues to be an elusive concept. The accepted number of races, for example, has not been constant. It has varied between three and dozens and have included Jewish, Nordic, Amerindian, African, Alpine, and European races. The Swedish botanist Carol von Linneaus, in his classic work Systematic Naturae (1735) grouped humans into four "varieties"—white, red, yellow, and black. He based his taxonomy on skin color and physiognomy and then correlated these traits with temperament and personality type. In 1749, George Buffon, the naturalist who is credited with the use of the term "race" to describe human variation, maintained that there were six races. He used skin color, stature, and physique as the criteria for racial designation. Over 200 years later, in 1950, W.S. Boyd used blood group data to designate six races which he later revised into 13.³

Cross-cultural and historical studies have revealed changing, arbitrary, and inconsistent definitions of race.⁴ These studies provide additional ammunition

against the argument that races are biologically determined categories. Furthermore, they underscore the ways in which social, political, and cultural factors have influenced racial classification. Definitions of race differ according to geography. The same person defined as black in the United States may be considered "colored" in South Africa. Even in the United States, definitions of race have not been static. By the early twentieth century many states—in part as antimiscegenation measures—had adopted numerous criteria to legally assign race. One of the most common gauges was the "traceable amount rule" or "one drop rule," according to which a single drop of ''black blood" made one black. The varying legal definitions of race in one state, Louisiana, illustrate the arbitrariness of racial classifications. The state followed the ''one drop rule" until 1970 when it adopted more than one 1/32 black as the criterion for blackness. This law stood until 1983,when the legislature gave parents the right to designate the race of newborns.

Federal criteria to designate race also reveal vagueness and inconsistency. The situation of biracial children is illustrative of the continued ambiguities. The growing number of such children raises the question as to how they should be classified in federal health statistics. In 1989 the National Center for Health Statistics (NCHS) decided to tabulate new births by the race of the mother. Previously, in the tradition of the one-drop rule, it had used the race of the nonwhite parent as the determinant. It should be noted that scientific evidence did not guide either of the NCHS policies.

In a 1992 JAMA article, epidemiologist Robert A. Hahn analyzed the dilemmas associated with the collection of federal health statistics on racial and ethnic groups in the United States.⁵ These difficulties raise profound question for medical research and clinical medicine. Hahn critiqued the assumptions underlying federal health statistics on racial and ethnic groups and questioned their validity. He stated that the accuracy of these statistics rested on logical assumptions that include the following: (1) The categories of "race"' and "ethnicity" are consistently defined and ascertained by federal data collection agencies. (2) Racial and ethnic categories are understood by the populations questioned. (3) Survey enumeration, participation, and response rates are high and similar for all racial and ethnic populations. (4) Individual responses to questions of racial and ethnic identity are consistent in different surveys and different times.

Hahn analyzed each of these assumptions and found that long-standing conceptual difficulties in the definition of race and ethnicity challenged their validity. For example, he found that the terminology and definition of race differed from agency to agency. Agencies used separate sources and inconsistent procedures to categorize race including policy directives, self-determination, and observer perception. Therefore, the biracial child classified as white by the NCHS might be classified as black by the Census Bureau, which uses self-

identification as the basis for racial designation. In another example, Hahn illustrated that with different indicators, estimates of the 1970 Hispanic population ranged from 5.2 million (persons of Hispanic heritage) to 9.6 million (persons using the Spanish language).

The establishment of public health policy and the setting of medical research agendas are influenced by the morbidity and mortality rates of different racial and ethnic groups. Accurate statistical information is a key component in the development of these programs. As a step toward the improvement of federal health data, Hahn urged researchers to use more precision in their definitions of race and ethnicity.

In recent years there has been a growing trend to reject biological notions of race. For example, 40 years ago the United Nations Educational, Scientific and Cultural Organization (UNESCO) launched a major publicity campaign promoting the idea that "race is less a biological fact than a social myth."⁶ Criticisms of the biological construction of race have increasingly grown since the initiation of this campaign. The voices of opposition have come from numerous disciplines, including biology, anthropology, and medicine. The views of the critics are aptly summarized by scientists Richard Lewontin, Stephen Rose, and Leon Kamin, who in 1984 wrote, "Any use of racial categories must take its justification from some other source than biology."⁷ Physical anthropology, from its origins in the nineteenth century through the middle of the twentieth century, was based on the central assumption of the existence of the "'pure race' as an assemblage of traits manifest in every individual race member, essentially unchanged by time or circumstance."⁸ But even within this profession, the current trend is to reject the proposition that distinct races exist within the human species and consequently to disavow biological concepts of race. Anthropologist Fatimah Jackson in a 1992 article about the use of race and ethnicity as biological constructs attributes this trend to a growing recognition by her colleagues of "the difficulty of making valid taxonomic assessments in long-lived, sexually reproducing, socially complex, and highly mobile species such as Homo sapiens."⁹

The traditional medical and public health view has been to interpret race as a biological and genetic category. This frequently takes the form of attributing observed differences in health status to inherent putatively "racial" characteristics. However, a growing number of researchers have begun to challenge this traditional perspective. They criticize the practice of interpreting racial differentials in health status as primarily the expression of biological factors. They contend that social factors must also be taken into consideration and that the focus on biological considerations obscures the importance of sociological and political ones. As cardiologist Richard Cooper has argued, "in the biologic sense there are no such things as races . . . the concept of race is itself a social category . . . [and] health status of racial groups should be viewed within this context." In particular, he explains, "black people in this society are

imprisoned by institutional racism; this is the attribute of blackness which at bottom determines their health status."¹⁰ Cooper and other authors have criticized the tendency to conflate race with class and to use race as the dominant measure of health disparities.¹¹ This practice is compounded by the fact that the United States does not collect health statistics by class. Dr. Vicente Navarro, a prominent health policy analyst, asserts that health differentials cannot be explained solely by looking at race—class must also be taken into account. "The publication of health statistics in racial terms," he states, "assumes that white unskilled workers have more in common with white lawyers, for example, than with black unskilled workers."¹²

Criticism of the application of the biological model to medicine and public also focuses on the notion that race, that is, skin color, is an accurate measure of the biological differences between human beings. As Richard Cooper has pointed out:

Opponents also criticize the assumption in the biological model that associates the genes for skin color with those that affect health. Drs. Newton G. Osborne and Marvin D. Feit outlined this position in a 1992 JAMA article. They wrote:

Of course biological factors cannot be dismissed when we analyze the health status of various racial and ethnic groups. Various genetic diseases are more prominent in particular racial and ethnic groups. Racial and ethnic differences in responses to drug therapies have also been ascertained. However, a close analysis of these cases demonstrates that biological considerations may not be as clear-cut as they initially appear. For example, medical historian Keith Wailoo has shown how sickle-cell anemia was defined as a "black" disease in the 1920s on the basis of half a dozen cases—all of which involved persons of

obviously mixed African and European ancestry. The disease, however, is not limited exclusively to African Americans. It is also prevalent among people of Mediterranean, Middle Eastern, and East Indian ancestry.¹⁵ In a move that once again underscores the problems of racial classification, a panel assembled by the Agency for Health Care Policy and Research recently urged that all newborns—not just black ones—be screened for sickle cell disease.¹⁶ The chair of the panel, Dr. Jeanne A. Smith, pointed out that racially targeted screening would miss many with sickle cell disease. She noted, "Because of mixing of the gene pool, it's not always possible to be certain of an individual's racial or ethnic background by physical appearance, surname or self report." In other words, America's mixed gene widens the risk for the disease.

Recent pharmacological research has discovered differences between racial and ethnic groups in drug metabolism rates, clinical drug responses, and side effects.¹⁷ Black patients with hypertension, it has been found, respond better to treatment with thiazides than with beta blockers. Preliminary studies have also indicated that smaller doses of haldol are needed to effectively treat Asian patients with schizophrenia. Much of this work has focused on the pharmacogenetics of drug metabolism. However, an emphasis on this area should not blind us to the effects of environmental and social factors, as the controversy surrounding a 1991 study of the effects of zidovudine (AZT) on HIV positive men clearly demonstrates. The study, conducted over a four year period, examined the effects of early therapy with the drug on 338 men—220 whites and 118 African Americans and Latinos. Its findings suggested that no benefit would be gained from early AZT therapy in African Americans and Latinos.¹⁸ Later studies repudiated this conclusion.¹⁹ The complexities in analyzing racial variations are seen in the VA study. Biological factors such as variations in drug metabolism may be involved. Other explanations cannot and should not be discounted. The black and Latino men in the study may have been of a lower socioeconomic status than their cohorts. Consequently they would have had more limited access to health care and their health status may have been poorer even before they became HIV positive. In addition, they may have had more advanced disease than their cohorts.

The use of race as a variable in clinical medicine and medical research should continue to be carefully examined. Race is usually used reflexively and without much thought. For example, one of the first things that American medical students learn is to take medical histories. They are usually taught to identify the patient, first and foremost, by age, race, and sex—"This is a thirty-two-year old black female." This identification holds even if a patient is presenting with a sprained ankle. It is not always clear what the patient's race has to do with diagnosis, prognosis, or therapy. Furthermore, it is assumed that a health care provider will know how to "diagnose" race. This, however, is a social skill learned in a race-conscious society, not as a matter of scientific instruction.

The point is not to ignore racial differences or deny the significance of race as a variable in medical research. Rather it is to urge researchers to make clear what they measuring and in what ways they are using the term "race." As social scientists Doris Wilkinson and Gary King have argued:

Their words serve to remind us that health policy initiatives and research agendas will differ according to what definition of race is employed. These divergent interpretations of race raise profound questions about what is being measured when racial differentials in research are ascertained. Are the effects of physiological or anatomical differences being measured? Are the effects of genetic differences being measured? Are the effects of the lack of adequate health care being measured? Are the effects of racism being measured? Are the effects of socioeconomic status and genetic endowment being measured? These are questions that must be addressed as programs are developed to include more women of color in research programs. The many unanswered questions regarding racial differences in medical research mandates more precision in the use of the term and the development of strategies to include more people of color in clinical studies.

In a recent article, law professor Patricia A. King warned that medical researchers have to be careful when they analyze racial differences between blacks and whites. She writes, "In a racist society that incorporates beliefs about the inherent inferiority of African American in contrast to the superior status of whites, any attention to the question of differences that may exist is likely to be pursued in a manner that burdens rather than benefits African Americans."²¹ King's comments underscore the dangers of difference in medical research. Medicine does not operate in a vacuum. It has reflected and reinforced the beliefs and values of the wider society. Accordingly, it has been influenced by issues of race and racism. History shows numerous examples of the use of medical thought to support the political ideology that black people are inferior. Medical theories, for example, were used to justify the enslavement of Africans.

An analysis of the theories behind the decline in the fertility rates of black women between 1880 and 1940 illustrates this danger of difference. It

demonstrates how biological and medical arguments have been used to reinforce and perpetuate stereotypes about black women. It also shows how biological explanations have overridden alternative hypotheses. The fertility rate for white women also dropped, but not to the same degree. However, the theories explaining a similar phenomenon in the two groups varied. The explanations were viewed through the prism of race.

The traditional view has been to attribute the decline in the fertility rate among black women to biological and medical factors, including venereal disease, puerperal septicemia, rickets, pellagra, and tuberculosis.²² The dominant paradigm to explain the differentials, the so-called "health hypothesis", specifically points to a high rate of venereal disease in black women as the major factor behind the numbers. An alternative explanation, however, was offered to explain the concurrent trend in white women. White women, it was argued, used contraceptive measures such as abstinence, barrier methods, rhythm, and withdrawal. The decline in the African American rate was portrayed as involuntary, while the white rate as voluntary.

Many physicians and social scientists believed that low intelligence and high immorality among black women prevented their use of contraception. In 1932 Raymond Pearl, one of the major proponents of the health hypothesis, alleged that black women's contraceptive use was lower than that of their white counterparts because "the negro generally exercises less prudence and foresight than white people do in all sexual matters."²³ The authors of a 1958 book on childbirth and abortion also emphasized the perceived racial differences in reproductive behavior. They wrote:

The perception of black women as less willing and less able to control their fertility continues to haunt as today, most notably in the debates surrounding the use of Norplant and Depo-Provera.²⁵

The health hypothesis achieved its dominant position, in part, because it fit stereotypical notions about the alleged promiscuity of black women. Roger Lane in his 1986 award-winning book, The Roots of Violence in Black Philadelphia, 1860–1900, suggests that prostitution and venereal disease significantly impaired the fertility of black women at the turn of the century. He writes, "All told, perhaps a quarter of Philadelphia's black women who reached the end of their childbearing years had at some time had exposure to the diseases and habits associated with prostitution. This figure would certainly account almost precisely for the difference between black and white fertility in the city."²⁶ Stereotypical

ideas about black women were even incorporated into medical textbooks. In the early 1970s Novak's Textbook of Gynecology proclaimed, "Currently, it would appear that the newer antibiotics may completely cure salpingitis without the usual residue of "closed tubes" and sterility. Unquestionably, this is a factor in the increased Negro birth rate."²⁷

In recent years, the validity of the health hypothesis has been challenged and substitute explanations have been put forth to explain the decline in the fertility rates of black women. Proponents of these alternate theories argue that biological factors, especially the chronic poor health of black women, may have played a role, but that other socioeconomic considerations must not be discounted. Close historical examination of the lives of African American women reveals images that sharply contrast to the previously accepted view. Jessie Rodrique has discovered that a nationwide, grass roots birth control movement operated in the black community in the years before World War II. An analysis of this movement clearly demonstrates that black women and men used contraceptive methods.²⁸ Additional studies, including those of social scientists Joseph A. McFalls and George S. Masnick, support this contention. McFalls and Masnick conducted extensive interviews with black women and found that they had used a broad range of contraceptive measures throughout the twentieth century. The researchers concluded:

Darlene Clark Hine argues that African American women who migrated to northern cities often practiced abstinence to gain economic security and personal autonomy.³⁰ She also points out how racism and sexism influenced previous studies. She writes, "Only latent acceptance of the myths concerning the alleged unbridled passions and animalistic sexuality of black women prevent serious consideration of the reality and extent of self-determined celibacy." It should also be noted that the health hypothesis ignored issues of class. Middle-class women, regardless of race, had fewer children than poor and working class women. This points to an inconsistency in the hypothesis—black women "with the socioeconomic and educational characteristics most conducive to good health had the lowest fertility and the highest childlessness."³¹ The current consensus is that many factors appear to have influenced the fertility rates of African American women from 1880 to 1840.

The history of the theories surrounding these rates show some of the

dangers involved in explaining racial differentials. The fecundity of both black and white women declined during the period. However, different interpretations were often sought to explain the event in the two racial groups. Racism frequently shaped these interpretations. The health hypothesis, a theory that reflected and reinforced stereotypical notions of African American women, was the predominant paradigm for many years. In addition, once again we see how a tendency toward biological explanations may obscure the significance of socioeconomic factors.

The myth of racial immunity is another dilemma that can result from an overemphasis on race as an independent variable in medical research and clinical medicine. The implications of this myth are clearly demonstrated in studies of race and osteoporosis. The 1984 National Institutes of Health (NIH) Consensus Development Conference Statement on Osteoporosis called the bone disorder a major public health problem.³² Researchers estimate that one-third of all women between the ages of 45 and 75 will develop osteoporosis. In concert with traditional public health practices, physicians and researchers have focused on identifying risk factors and developing intervention strategies. The NIH statement classified white women as the most at risk population. Black women, it noted, developed the condition much less frequently. In the years since the release of the NIH statement, campaigns to prevent the bone disorder have focused primarily on white women.

Age and gender are the primary risk factors for osteoporosis. However, studies have also stressed the importance of race and ethnicity. Osteoporosis is often portrayed as "a unique phenomenon of relatively inactive, postmenopausal, white females, fragile in outer appearance, and given to little consumption of milk or other dairy products."³³

Women of color are frequently pictured as being relatively immune to the condition. One 1987 review went so far as to declare, "It is a well-known fact that blacks do not suffer from osteoporosis."³⁴ The low incidence of osteoporosis in African American women has been attributed to biological factors including increased bone mass and decreased bone resorption.³⁵ These studies have failed, however, to control for other factors known to affect bone density, such as weight, diet, health, alcohol intake, cigarette smoking, exercise level, and reproductive history. Race is also seen in these investigations as a definable genetic category.

Additional studies have concluded that Hispanic women are also less susceptible to the disorder than white women.³⁶ One author has suggested that as Mexican Americans and African Americans are "relatively protected" that they "may benefit less from prophylactic theories."³⁷ The literature on Asians is contradictory with studies presenting them as both less and more susceptible.³⁸ Race and ethnicity, however, do not offer immunity to osteoporosis. Black and Hispanic women do get the disease, albeit at lower rates than white women. Less susceptible is not equivalent to immune.

Targeting susceptible populations for screening and educational programs is a well-established public health practice. However, we must be careful not to allow confused notions of racial susceptibility to influence inappropriately public health policy and the practice of medicine. Factors other than race play roles in the divergent incidences of osteoporosis in black and white women. Race must not be used as the dominant risk factor. We must evaluate each woman on an individual basis and assess other risk factors. An examination of race and osteoporosis suggests the danger of using race to define a disease. Osteoporosis is not a white woman's disease. However, most prevention and screening programs have been directed toward them. The question remains as to what extent the myth of racial immunity has led to unwarranted neglect of the health care needs of women of color.

Although recent studies have criticized the underrepresentation of women in clinical studies, it is important to know that poor and minority women have, at times, been exploited in the name of science and medicine. An understanding of this past history will demonstrate why so many people of color are mistrustful of the medical profession and its institutions. As efforts begin to include more women of color in clinical research, it is imperative that this legacy of distrust be addressed and not be dismissed as paranoia or hypersensitivity. The challenge before us is to attempt to understand and confront the historically based realities behind these sentiments.

The Tuskegee Syphilis Study symbolizes for many African Americans the racism that pervades American institutions, including the medical profession.³⁹ It is often used to demonstrate why African Americans should not cooperate with medical researchers. The United States Public Health Service (USPHS) initiated the study in 1932 in order to document the natural history of syphilis. The subjects of the investigation were 400 poor black sharecroppers from Macon County, Alabama, with latent syphilis and 200 men without the disease who served as controls. The physicians conducting the study deceived the men, telling them they were being treated for "bad blood." For example, the men were informed that lumbar punctures were therapeutic, not diagnostic.

As part of the project, however, the USPHS deliberately denied treatment to the men who had syphilis and went to extreme lengths to ensure that they would not receive any. When the Tuskegee Syphilis study began, the standard therapy for syphilis consisted of painful injections of heavy metal compounds such as arsenic and bismuth which had to be administered for up to two years. Although this therapy was less effective than penicillin would later prove to be, in the 1930s every major textbook on syphilis recommended it for the treatment of the disease at all stages. Published medical reports have estimated that

between 28 and 100 men died as a result of their syphilis. In exchange for their participation the men received free meals, free medical examinations, and burial insurance.

The Tuskegee Syphilis Study continued until 1972. Throughout its 40 year history, accounts of the study appeared in prominent medical journals. Thus, the experiment was widely known in medical circles. As late as 1969, a committee from the Centers for Disease Control examined the study and decided to continue it. Three years later, a USPHS worker, who was not a physician, leaked details about it to the press. Media disclosure and the subsequent public outrage led to the termination of the study and ultimately to the National Research Act of 1974. This act, established to protect subjects in human experimentation, mandates institutional review board (IRB) approval of all federally funded research using human subjects.

The Tuskegee Syphilis Study raises questions not only about human experimentation, but also about racism in medicine. White physicians in the late nineteenth and early twentieth centuries wrote extensively about the health problems of African Americans, especially syphilis. They maintained that intrinsic racial characteristics such as excessive sexual desire, immorality, overindulgence, and anatomical differences—large penises and small brains—caused black people to have high rates of syphilis. The physicians also believed that syphilis was difficult to treat in black patients because they could not be convinced to come in for treatment or, if they did, to follow the treatment regimen.

Historian Allan Brandt has argued that these assumptions regarding black people and venereal disease influenced the physicians who initiated the Tuskegee Syphilis Study. He writes, "The premise that blacks, promiscuous and lustful, would not seek or continue treatment, shaped the study. A test of untreated syphilis seemed 'natural' because the USPHS presumed the men would never be treated; the Tuskegee Study made that a self-fulfilling prophecy."⁴⁰

The Tuskegee Syphilis Study did not occur in a vacuum. It represented the continuing influence of racist thought not only on medical theory, but on physicians' perceptions of a group of people and consequently on the treatment—or lack of treatment—individuals would receive.

After the Study had been exposed, many black people charged that it represented "nothing less than an official, premeditated policy of genocide."⁴¹ Most recently, both genocide and Tuskegee have come up with respect to AIDS. In September 1990, an article entitled "Is It Genocide?" appeared in Essence, a black woman's magazine. The author noted, "As an increasing number of African-Americans continue to sicken and die and as no cure for AIDS has been found some of us are beginning to think the unthinkable: Could AIDS be a virus that was manufactured to erase large numbers of us? Are they trying to kill us with this disease?"⁴² In other words, some members of the black community see

AIDS as part of a deliberate plot to exterminate African Americans. Although there is not any credible scientific evidence to support such claims, these ideas should not be dismissed as being merely those of paranoid extremists. For example, a 1990 survey conducted by the Southern Christian Leadership Conference found that 35 percent of the 1,056 black church members who responded believed that AIDS was a form of genocide.⁴³ The Tuskegee Syphilis Study is presented as evidence of the government's genocidal policies toward African Americans.

A lasting legacy of the study is that it demonstrates to many African Americans why they should not trust the medical profession. Dr. Stephen B. Thomas, co-director of the Minority Health Research Laboratory at the University of Maryland College Park, laments, ''Although everyone may not know the specifics of the Tuskegee experiment, they have enough residual knowledge of it so that they mistrust government-sponsored programs, and this results in a lack of participation in [AIDS] risk-reduction efforts.''⁴⁴ Alpha Thomas, a health educator at Dallas's University Hospital, often confronts the legacy of Tuskegee. She notes that "so many African American people that I work with do not trust hospitals or any of the other community health care service providers because of that Tuskegee Experiment. It is like . . . if they did it then they will do it again."⁴⁵

These apprehensions contribute to the low enrollment rate of African-Americans in clinical trials.⁴⁶ A 1989 study conducted by pharmacologist Craig K. Svensson demonstrated the underrepresentation of African Americans in clinical trials He reviewed 50 clinical trials for new drugs that had been published in Clinical Pharmacology and Therapeutics for the three year period 1984–1986. He discovered that the proportion of black subjects was less than their percentage in the cities in which the research was conducted and less than their percentage in the general population of the United States. More recent studies confirm this underrepresentation of African-Americans in clinical trials for AIDS drugs.⁴⁷

The strengthening of safeguards and the reforms in research standards that followed the public disclosure of the abuses of the Tuskegee Syphilis Study, have been insufficient to change African Americans' historically based fears of medical research. The tenacity of this conviction is understandable if one examines the broader history of race and American medicine. The historical record makes clear that although the Tuskegee Syphilis Study may have been an extreme example, it was not an isolated aberration. It was not the only case in the history of medicine where people of color have been exploited in the name of medicine.

The foundation of modern gynecology, for example, is based on the sacrifices of three slave women. Between 1845 and 1849 Dr. J. Marion Sims, the so-called father of modern gynecology, used the women as subjects of

experiments designed to develop an operative technique to repair vesico-vaginal fistulas. The three slave women on whom Sims operated each underwent up to 30 painful operations. The physician himself described the brutality of some of his experiments. He wrote, "The first patient I operated on was Lucy. . . . That was before the days of anesthetics, and the poor girl, on her knees, bore the operation with great heroism and bravery."⁴⁸ This operation was not successful and Sims later attempted to repair the defect by placing a sponge in the bladder. This experiment, too, ended in failure. He noted, "The whole urethra and the neck of the bladder were in a high state of inflammation, which came from the foreign substance. It had to come away, and there was nothing to do but to pull it away by main force. Lucy's agony was extreme. She was much prostrated, and I thought that she was going to die; but by irrigating the parts of the bladder she recovered with great rapidity. . . ."⁴⁹ Sims finally did perfect his technique and ultimately repaired the fistulas. Only after his experimentation with the slave women proved successful did the physician attempt the procedure on white women volunteers. He found, however, that they could not, or more accurately, would not, withstand the pain and discomfort that the procedure entailed. The black women had no choice but to endure. They were forced to submit because the state considered them property and denied them the legal right to refuse to participate.

Abuse of women of color by the medical profession extended into the twentieth century. Some of the most predominant examples have revolved around contraception. In one study conducted in 1969 at a family planning clinic in Austin, Texas poor women, mostly Chicanas, received what they believed to be contraceptives.⁵⁰ However, 76 patients, without their knowledge, received placebos. Those who became pregnant were not provided with abortion services even after they requested it. In the 1970s numerous reports of sterilization abuses against women of color surfaced.⁵¹ In the 1974 Relf case, a federal district court found that an estimated 100,000 to 150,000 poor women were sterilized annually under federally funded programs. Some of these women had been coerced into consenting to the procedure by threats that failure to do so would result in the termination of welfare payments.

History, however, is not just about the past. It profoundly affects our contemporary lives. A historically based legacy of distrust still influences the relationship of people of color to the medical profession. It is a significant obstacle to the initiatives to include more women of color in clinical research. One of the challenges that the profession faces is to not to dismiss this sentiment, but to confront it and change it. Efforts should be made to find out what women of color need and want from the health care system. Researchers associated with a new research project at the Centers for Disease Control and Prevention have acknowledged that the voices and experiences of African American women are crucial for the project's success. This investigation is designed to analyze the impact of psychosocial factors on the incidence of preterm delivery among black

women. In a radical departure from traditional scientific studies, the investigators have actively solicited advice about the study from the African American lay community. Their goal is to develop a collaborative research strategy that is viewed as a study that is done with black women, not on them.^52,53

It is not inappropriate to take race into account when clinical research programs involving women are developed. However, the use of race as a variable must be used cautiously and judiciously. History has shown that racial differences have been frequently used without precision. Investigators should make clear how race is being defined in their studies. They should also delineate what they are measuring in research designed to analyze racial differences. Are they using race as a proxy for physiological differences, socioeconomic status, inadequate access to health care? Researchers should also make clear what factors prompted them to explore racial differences. Was it based on observations? Or on preconceived notions that differences should exist? Or on past experiments that did not control for other risk factors? Or on stereotypes about racial minorities? History has also demonstrated that research on racial differences has at times had detrimental effects on people of color. We cannot allow racist and sexist notions about groups of people to influence clinical studies. Finally, programs must be implemented to confront the distrust that people of color have toward the medical profession. The development of collaborative research strategies as proposed by the CDC study of preterm deliveries is a much-needed step.