Cultural Issues in the Retention of Minority Populations in Clinical Trials

Neil Powe, M.D., M.P.H., M.B.A., Professor, Johns Hopkins Medical Institutions; Director, Welch Center for Prevention, Epidemiology, and Clinical Research, Baltimore, Maryland

Dr. Powe's presentation addressed cultural issues in the recruitment and retention of minorities in clinical trials. In a two-pronged approach to the subject, Dr. Powe (1) examined the demographic information related to NIH extramural clinical trial recruitment, disease burden, and mortality rates in minority populations and (2) laid out a conceptual framework regarding issues that hinder minority participation.

Demographic Information

Data from the NIH Blue Report, published in December 2002, reveal information about minority participation in NIH extramural clinical trials by racial and ethnic group. Data from fiscal year 2000 can be used to compare the percentage of the population in different minority groups with the percentage of the minority groups in NIH studies. The data show that 12.1 percent of the U.S. population is African American, and 11.3 percent of subjects in NIH clinical studies are African American. The data also reveal that 12.5 percent of the U.S. population is Hispanic, but only 7.9 percent of subjects in NIH studies are Hispanic. However, NIAID did significantly better in the recruitment of both groups: 28 percent of patients in NIAID studies are African American and 11.2 percent are Hispanic.

The question involves the appropriate target number. The census reflects the overall proportion of patients in the population, but the burden of disease varies by race and ethnicity. The burden of disease can be described by many measures, including incidence, prevalence, total mortality, years of life lost, hospital days, and disability adjusted life years. Regarding HIV burden, 50 percent of AIDS cases are African American, 19 percent of cases occur in Hispanics, 64 percent of female HIV cases are African American, and 54 percent of new HIV infections occur among African Americans. Those statistics reveal that NIAID is falling below the appropriate target number for enrollment of participants in clinical trials. An examination of prevalence of disease by race and ethnicity also reveals a disproportionate burden of asthma among African Americans, of diabetes among Hispanics and African Americans, and of alcohol consumption among whites and Hispanics. Regarding death rates by race and ethnicity, heart disease, lung cancer, and breast cancer kill a disproportionate number of African Americans. Dr. Powe called for an examination of conditions and use of a variety of measures of burden of disease to arrive at appropriate targets for minority enrollment in clinical trials. The statistics suggest that NIH's efforts in minority enrollment may not be as successful if examined from the point of view of appropriate targets.

Conceptual Framework and Barriers to Minority Participation

Dr. Powe offered a conceptual framework to achieve the full participation of minorities in clinical trials. The framework includes the participant's reasons for joining a study, the investigator's role, the sponsor's role, and the institutional role, as depicted in the figure below.

Dr. Powe described the three steps that participants must take to enroll in clinical studies and gave examples of various factors that might influence each of the steps:

1. The patient must be aware of the study. However, studies have shown that race, ethnicity, and socioeconomic status can affect a person's awareness of studies. A household telephone survey in Baltimore¹ to determine the health information sources used by women (print health media, regular news media, broadcast media, computer resources, health organizations, and organized health events) found that whites are much more likely than African Americans to use print news media (64 percent versus 44 percent), computer resources (45 percent versus 21 percent), and health organizations (21 percent versus 7 percent). Dr. Powe pointed out that use of information sources affects utilization of health care.

2. The patient must be aware of the terms of the study and be willing to enter it. Acceptance is based on perceived harm and perceived benefit, which often depend on health status. Patients with poor health status will try many interventions and will enroll in a trial if it means the possibility of a cure. Other patients may not be so inclined. Past experiences with the health care system also can affect acceptance or willingness to enter a study. In addition, trust in the investigator or sponsor, altruism, and religiosity can affect acceptance. In a recent study, Dr. Giselle Corbie-Smith² at the University of North Carolina designed a national cross-sectional telephone survey, randomly sampled 527 African Americans and 382 whites from the general population, and administered the 45-minute survey. On a seven-item scale of distrust developed by Dr. Corbie-Smith, the survey found the following results:

• 15 percent of African Americans versus 8 percent of whites do not believe they could freely ask their physicians questions.
• 37 percent of African Americans versus 20 percent of whites believe their physician would ask them to participate in medical research even if he or she thought it would harm them.
• 47 percent of African Americans versus 23 percent of whites do not trust their physician to fully explain research participation.
• 45 percent of African Americans versus 34 percent of whites believe that physicians sometimes expose patients to unnecessary risk.
• 79 percent of African Americans versus 51 percent of whites believe that it is very likely or somewhat likely that people might be used as guinea pigs.
• 62 percent of African Americans versus 8 percent of whites believe physicians very often or fairly often prescribe medication as a way of experimenting.
• 24 percent of African Americans versus 8 percent of whites believe that physicians have given them treatment as part of an experiment without their permission.

In general, the study found that African Americans have higher distrust scores than whites (mean score of 3.1 among African Americans and 1.8 among whites). Distrust was associated with lower education, unemployment, male sex, and geographic region. After adjustment for sociodemographic factors that differ between the races, African Americans still had almost 5 times the odds versus whites of having a distrust score of 5 or greater. Dr. Powe commented on this national study's profound results.

Another cross-sectional study of willingness to participate in a hypothetical heart disease prevention trial³ used a random/consecutive sample of outpatients treated at 13 Maryland internal medicine and cardiology clinics during 2002. The patients were adults who presented for a medical visit, who spoke English or used an interpreter, and who were able to comprehend the nature of a 15-minute, self-administered questionnaire. The questionnaire included a description of the research study and its objectives, risks and benefits to the patient, the voluntary nature of the study, requirements to be met by the patients during the length of the study, alternative treatment options, the right to withdraw, the study sponsor, the presence of potential investigator conflicts, and the rewards for joining. Respondents were asked about their willingness to join the study. When Dr. Corbie-Smith's medical research distrust index was used, it was found that those respondents with a higher level of distrust were less likely to join the trial than those who had none or a low level of distrust; therefore, distrust is associated with willingness to join. More African Americans were unlikely or very unlikely to join the hypothetical trial than whites. African Americans were much more likely to have higher distrust scores than whites. Thus, addressing the issue of distrust might address why African Americans are unlikely to join trials.

Another study addressed African Americans and participation in AIDS research⁴. This cross-sectional study of 301 African Americans in Durham, North Carolina, found through in-person interviews that 50 percent of the African Americans would not participate in AIDS-related clinical trials. Distrust was the strongest inverse predictor of willingness to participate, along with a variety of other factors that were also associated with willingness to participate. Dr. Powe pointed out that these findings indicate that distrust is a very important issue that must be addressed.

3. The patient must be retained in the study over the life of the study. A framework to achieve full participation in clinical trials addresses participant barriers, investigator barriers, sponsor barriers, and institutional barriers:

• Participant barriers to full participation in clinical trials include knowledge and education (limited opportunities to learn about research, concerns about changing medical regimens, fatalistic attitudes about chances for recovery), sociocultural factors (beliefs and moral values, religion, language, mistrust, negative experience with hospital or injury due to treatment, family and friends' influence, beliefs about alternative medicine), and economic factors (caring for children or other ill relatives that might prevent participation in trials, working multiple jobs, transportation, lack of medical resources in the community).

• Investigator barriers to full participation in clinical trials include issues of study design and implementation (ineffective guidance to study staff, recruitment based on convenience, study run-in periods and selection processes, ineffective informed consent processes), communication (limited knowledge about methods to increase awareness, ineffective study staff communication, stereotypes and attitudes of investigators, lack of feedback), and resources (lack of community liaison efforts, limited knowledge of appropriate retention methods).

• Institutional and sponsor barriers to full participation in clinical trials include the lack of investigator incentives, rewards, and penalties for not enrolling in a clinical trial; lack of funds to address difficult recruitment and participant incentives; and institutional stereotypes and attitudes that must be dismissed to achieve full participation.

Dr. Powe described some of the promising practices or interventions to address the barriers. In terms of awareness, promising practices might be a physician taking time to explain and complete the study paperwork, adding additional staff or monetary incentives, providing free or reduced care, doctors encouraging other doctors to involve their patient populations in clinical trials, and effective communication or advertisement from community leaders, local newspapers, and churches. In terms of illiteracy and language barriers, interventions might include using translators and alternatives to written communication, providing taxi vouchers or vans to address transportation issues, using community members as research staff, using peer groups and community-based participatory research methods, using outreach efforts to emphasize the value of participation, providing child and elder care, and creating simplified informed consent forms. In terms of retention, promising practices include allowing more flexibility in scheduling examinations and followup, making the clinical environment more friendly, involving minority research staff, involving physicians in the study, creating newsletters that track the progress of the study, and acknowledging patients through cards and gifts.

Dr. Powe stated that some issues in enrollment can be addressed through community-based participatory research strategies, such as involving members of the target population in the planning efforts of the trial, taking the message to the target population, using a community spokesperson to empower patients to join research, and giving back to the community through jobs, gifts, screenings, and clinical tests.

Dr. Powe called attention to the existence of fear and mistrust. He pointed out that the recruitment strategies used in the Tuskegee study (1932-1972) included a community-based approach with local black churches, the Macon County Medical Society, local plantation owners, local public schools, the Tuskegee Institute, black physicians, and boards of health. The Tuskegee study offered free physicals, food, transportation, and burial stipends to gain permission for autopsies from family members. Similar strategies are being advocated for HIV education and AIDS risk reduction programs today, but the value of these community-based strategies is not diminished by their association with the Tuskegee study. The issues of mistrust must be addressed.

Ethical considerations in recruitment involve patients' understanding of the consent form, doctor referrals, use of incentives, and minority-only studies, which heighten fear and mistrust in some minority populations because of suspicion of a study targeting only one minority population. However, minority-only studies result in clear evidence of efficacy for specific minority populations and justice in monetary distribution of research funds.

Dr. Powe ended his presentation by stating that recruiting minorities remains a challenge even though NIH policies on tracking recruitment have led to more attention and progress in this arena. A variety of barriers exist to awareness, recruitment, and retention of minorities in clinical trials. Distrust is a significant barrier among the African American population, and language is a significant barrier among the Hispanic population. However, promising strategies are emerging to overcome barriers in a way that will maintain ethical standards.

Discussion

Dr. Powe's presentation raised the following issues and questions:

• An epidemiologist raised the question of the distinction between mistrust and stigma. Mistrust comes from actual events that occurred in past research trials and experiments involving African Americans. Participants were correct to mistrust a health care system that operated without honesty or integrity. However, the institutional review board (IRB) process and community advisory boards have now succeeded in removing the stigma associated with clinical trials.

• Another participant commented on the provider barrier and the disrespectful treatment of community providers by institutions that perform clinical trials. Dr. Powe responded that one way to address this problem is with focus groups of community physicians who plan the objectives of a study and give input about recruitment and enrollment. NIH's idea of reengineering the clinical enterprise, referred to by Dr. Fauci, involves establishing a national core of community physicians who would be involved in research and empowered to be part of studies.

• Another participant asked Dr. Powe to elaborate on the institutional barriers, in particular, regarding stereotypes and attitudes. Dr. Powe explained that there are very few studies that track provider attitudes and stereotypical views. One study of providers' attitudes toward minority patients of low socioeconomic status showed that providers assumed that minorities would not comply with research protocols. Their "fatalistic attitude" has a negative impact on their inclination to enroll minorities in research studies.

• Another participant mentioned barriers in terms of funding, in particular, regarding outreach and the use of plain language in informed consent documents. Dr. Powe remarked that these issues can be dealt with through more involvement with institutional leadership roles (e.g., on IRB committees) to call attention to the difficulties and ways to address them. The direct involvement and experience of individuals in positions of leadership can influence the issues that are addressed.

^1.Nicholson WK, Grason HA, Powe NR. The relationship of race to women's use of health information resources. Am J Obstet Gynecol 2003;188:580-5.

^2.Corbie-Smith G, Thomas SB, St George DM. Distrust, race, and research. Arch Intern Med 2002;162:2458-63.

^3. Braunstein JB, Schulman SP, Powe NR. Effect of Ethnicity on Distrust Toward Medical Researchers and Willingness to Join Clinical Trials. J Am Coll Cardiol 2003; 41:534A.

^4. Sengupta S et al. Factors affecting African-American participation in AIDS research. J Acquir Immune Defic Syndr 2000;24:275-84.