Health Disparities: Why Inclusion and Health Equity Matter in Clinical Research

Health Disparities

Health disparities are differences in the occurrence of a disease, injury, exposure, and the opportunity for obtaining medical treatment, which have a negative impact upon a population¹ (Table 1). Persons who experience health disparities are treated differently than those who do not experience health disparities. Often, persons with health disparities are not included in clinical research or face more barriers to participation. In clinical care, even with health insurance, a person who experiences health disparities may not receive comparable medical treatment. Health disparities can cause negative outcomes or have a negative impact upon persons who are disadvantaged. Thus, persons who are already marginalized by society may be affected more by inadequate health care and may be less likely to be included in research. Persons can experience health disparities due to their geographic location with greater impact in rural settings and in communities where there are fewer resources.² The United States is unique in that it is a high-income country with large health disparities.³

Vulnerable Populations

The Code of Federal Regulations defines vulnerable populations as categories of research subjects who are vulnerable to coercion or undue influence such as children, prisoners, individuals with impaired decision-making capacity, and economically or educationally disadvantaged persons (21CFR 56.111). This revised definition was effective on Feb 14, 2019.  Regulatory standards for conducting research with children, adolescents, and vulnerable subjects (45 CFR 46 Subpart D) are different than the general regulations. A legally authorized representative can provide informed consent for these populations.

Historical, Social, and Cultural Inequality

There are historical, social, and cultural reasons for health disparities. Historically, individuals have been viewed based upon factors such as race or cultural background, education, and socioeconomic status. This was the case with the Tuskegee Study, a longitudinal study of African American males in the south who had syphilis that was conducted by the U .S. Public Health Service from 1932 to 1972.4 The study participants were not informed that they had a diagnosis of syphilis; when treatment with antibiotics became available, it was not provided. At the time of the Tuskegee Study, the racial segregation laws in the south and the general acceptance of health disparities in the society made the study possible. In response to the Tuskegee Study, the Belmont Report was published, which established the ethical principles of autonomy, justice, and beneficence along with guidelines to protect persons involved in human subjects research.⁵ Although there are regulations to prevent unethical clinical research, health disparities still continue to occur in studies in the United States.

The terms “health disparities” and “vulnerable populations” are not interchangeable terms. A person who belongs to a certain racial group or who has a low socioeconomic background is not necessarily part of a vulnerable population. When looking at the cultural context of that person, perhaps there are other gaps that can be bridged instead of simply trying to identify that person as part of a vulnerable population. Some persons who are part of a vulnerable population experience health disparities; however, not all vulnerable populations experience health disparities.

The Institute of Medicine (IOM) defined 13 predominant categories of persons who experience health disparities:

• Women
• African Americans
• Appalacian poor
• Asian Americans
• Elders
• Immigrants/refugees
• Latinos/Hispanics
• Persons with disabilities
• LGBTQA Community
• Native Americans
• Overweight Persons
• Prisoners
• Religious minorities

The IOM categories are not hierarchical and are not rigid or exclusive; males could also experience health disparities, depending on their circumstances. Women are considered a category of persons who experience health disparities; this may be due in part to economic circumstances.

Work is underway by the National Institute of Minority Health and Health Disparities (NIMHD) to strengthen research, evaluation, and reporting about health disparities, and to understand the etiology of health disparities. The dialogue about health disparities and funding for activities to reduce health disparities had been ongoing. Yet, challenges to reducing health disparities persist. The NIMHD takes a systematic approach to improving research methods and developing interventions to reduce health disparities and change the research culture from within. This includes increasing investigator-initiated research that will be inclusive of persons of minority and culturally diverse backgrounds who had not been included in research historically. The National Institute of Minority Health and Health Disparities website is a useful resource for researchers.⁶ Individuals and groups within society who represent a broader diversity in the U.S. culture will be engaged so that all stakeholders are represented.

Research will no longer be conducted as it was in the past, where Caucasian males were the only subjects and the outcome of that research was presumed to apply to all persons. The Framingham Heart Study showed researchers that women who have cardiovascular disease manifest cardiovascular disease differently than men.⁷ It was not until 1992 that the Food and Drug Law Institute and the FDA expressed the importance of including women as research participants.⁸ Inclusive research, which is inclusive of the entire population, will capture information that identifies subgroups who have different health needs or health conditions than the overall public. Strengthening research is a dynamic process.

Disparities In Research

When researchers are enrolling subjects, publishing papers, and disseminating information, there may be bias in the conduct of their studies. Disparities and bias do exist in research. African Americans comprise about 12% of the population yet they participate in less than 5% of clinical trials.⁹ Disparities are even greater among Hispanics, who represent 16% of the population yet participate in less than 1% of clinical trials.⁹ The reasons for these disparities are unknown, but it could be attributed to the extra cost or effort required to recruit inclusive populations. Perhaps with a focus on rapid enrollment, researchers may not recognize that they are not being inclusive when enrolling subjects. Researchers can do something to start bridging the gaps. Otherwise, some of the research data that are generated are not generalizable to the general population.

Expression of Disparities in Research

Health disparities can be expressed in different ways, including implicit bias (Table 2). Stereotyping means automatically categorizing individuals based upon previous experience or stereotypes of certain groups of persons. Examples of stereotypes are assuming that certain racial groups are good at handling information-technology problems, that women with blonde hair are not intelligent, that some racial groups are better at sports than others, or that every elderly person does not know how to manage technology.

Implicit bias is subtle. It is not overt, intentional behavior like explicit racism. Implicit biases can emerge over time without researchers being aware of negative feelings towards a group. We are all influenced by our family of origin and our place in society. There is a tendency to stereotype persons with similar characteristics. Researchers ought to convey mutual respect and believe that others have the capacity to be capable and competent. It is beneficial for a researcher to be aware of his/her preconceived notions or implicit bias about a person or a group. If an implicit bias cannot be overcome, it would be better to step away from the situation.

When researchers see a new patient, there may be a tendency to begin to make assumptions about the person with expectations about their behavior. They may categorize persons as different based upon racial, ethnic, or social characteristics. In these circumstances, researchers do not see or treat the person as an individual but rather as a member of a categorized group based on preconceived notions of race, gender, age, or other characteristics.

More discrimination occurs based upon weight than based upon any ethnic, religious, or other factor. There are implicit and explicit biases and stereotyping against persons who are overweight and obese.¹⁰ As a result, overweight or obese persons will have adverse health outcomes and are more likely to be at risk for chronic medical conditions. Stereotyping of persons who are overweight and obese means that health care providers tend to expect persons to do something about their weight first and then interventions are provided later. Poor health outcomes become a self-fulfilling prophecy.

If clinicians and researchers challenge the idea that discrimination of any type is unacceptable, they can change their approach to clinical care and research and influence the values and behaviors of others. Somewhere along the line, researchers develop behaviors and assumptions that carry them through life in a certain way. This results in decision-making processes that do not treat everyone equally.

Health Outcomes and Health Disparities

There are many disparities in health outcomes (Table 3). Infants born to African American mothers are 1.5-3 times more likely to die than other infants.¹¹ African American males have higher rates of cancer than other persons, with nearly 600 cases per 100,000 persons.¹² American men, especially teenagers and men over the age of 70, are four times more likely to commit suicide than women.¹³ Even though women have health disparities, it has been suggested that biological factors contribute to longevity and resilience among women.¹⁴

Caucasian women have higher rates of hip fractures. However, African American women have poorer outcomes after a hip fracture, including more mobility limitations and mortality. In a study that looked at clinical outcomes in Medicare-eligible women with hip fractures with osteoporosis, thirty-three percent of Caucasian women received comprehensive osteoporosis screening versus 5% of African American women.^{15, 16} Health care providers were less likely to recommend a diagnostic DEXA (Dual-energy X-ray absorptiometry) scan for African American women when compared to Caucasian women. This study demonstrated the clinical effects of health disparity in African American women with osteoporosis. Because of this study, in order to address the disparity in service delivery, a preventive clinical care model was designed to overcome health disparities.  Preventive care screenings are recommended; providers are reminded to conduct routine for all women regardless of their race.

Asian American males have the highest rate of inpatient acute myocardial infarction mortality. Mortality after an inpatient acute myocardial infarction is higher for Asian American males than for African Americans, Hispanics, and Caucasians. ¹⁷ Even though this finding has been documented in the literature, cardiologists and primary care physicians may not recognize the additional risk factors in treatment of Asian American males.

The underlying factors that affect decisions that lead to poorer health outcomes must be addressed. If the implicit bias is that health disparities occur among persons of certain racial categories or socioeconomic backgrounds, then poor outcomes in these patient populations become a self-fulfilling prophecy. Persons cannot experience good health outcomes if they do not have access to treatment.

While health disparities are more likely to occur among certain social, racial, and cultural groups, it is possible for any person to find themselves in circumstances where health disparities occur. For example, a person who is employed and has health insurance and a good social support system who is not physically or mentally impaired has many strengths and assets to utilize. However, if that same person loses his/her job, private insurance, and due to reduced resources then moves to a different community and loses her/her social support system, then he/she has become marginalized and is more likely to experience health disparities.

Epidemiology Factors

Epidemiologists might categorize persons demographically based on certain characteristics like race, gender, socioeconomic status, or health status; however, epidemiology is mathematical, not inter-personal. Historically, public health, anthropology, and population genetics have identified groups based upon race (Table 4). As a result, the concept of “predisposition” to a condition, disorder, or disease was developed. Inequality in health care availability can play a role in characterizing a group’s predisposition to a condition, disorder, or disease. Researchers must begin to recognize the relationship between the risk factors, exposures, environment, and subsequent disability and disease rates and then analyze the prevalence of certain conditions based upon inclusive research data.

The National Human Genome Research Institute defines genomic medicine as “an emerging medical discipline that involves using genomic information about an individual as part of their clinical care (e.g., for diagnostic or therapeutic decision-making) and the health outcomes and policy implications of that clinical use.”¹⁸ Genomic medicine focuses on identifying subgroups who may need different therapeutic interventions. Genomic medicine can provide objective data for use in informing decisions about health care.

In research, external validity relates to how generalizable the study results are to the relevant population. This is different from internal validity, which shows how accurately the study was conducted. For example, if we are conducting a research study in the US that intends to capture a representative sample from the population, then the enrollment would be approximately similar to the US census racial demographics. Accuracy might look at the racial composition of the individuals in society and their representation in clinical trials, such as enrolling 15.9% of Hispanics, 13.4% of African Americans, and 5.8% of Asians and other persons of different backgrounds.¹⁹ Research that includes persons from urban, rural, and suburban settings would strengthen the representative sample and reduce bias. Accurate representation of the study population in research recruiting would provide results that are generalizable to the actual population.

Under the clinical uncertainty cycle, since less is known, less is understood.²⁰  Therefore, if little is known about a patient population; past research obtained from an exclusive study sample cannot be adapted with the same effect. For example, results of studies in cardiovascular disease conducted in males cannot be extrapolated to women. The clinical uncertainty cycle works against inclusion of persons of different racial and ethnic groups in research. If cardiovascular disease studies do not enroll any African Americans, researchers will not understand how cardiovascular disease specifically affects African Americans.

Strategies to Reduce Disparities in Research

Disparities in research must be reduced. When sponsors and contract research organizations assess clinical research sites for feasibility, they should consider asking not only whether the site can conduct the research, but whether the site has the capacity to conduct inclusive research. This means ensuring that the investigators and staff at the chosen sites do not have a strong implicit bias that reduces diversity in the enrolled participants. Researchers can be trained to increase their cultural awareness to learn how to engage with persons from different backgrounds more effectively. Site staff may need to conduct community outreach with local leaders to engage with diverse study populations or to develop new or different advertising and recruiting approaches.

Strategies to reduce disparities in research include recognizing that differences exist between cultures (Table 5). Researchers have the capacity to work to bridge these differences. Good peer support within the research team is necessary in order to reduce health disparities in research. Bridging gaps interpersonally can be an uncomfortable process. Learning about persons’ different backgrounds and cultures and how to interact effectively with diverse cultural experiences takes time. Persons who point out their colleagues’ biases will be less effective than those who strive to eliminate health disparities to improve the quality of the research.

Experience, beliefs, values, and language affect researchers’ perceptions and worldviews as well as the way in which we interface with other persons and cultures. Researchers ought to grow to understand and recognize the influence of their own and others’ customs, beliefs, and behaviors. As long as health disparities exist, there will be persons who know that they are not treated equally. Researchers are perceived as persons with more social advantages with more opportunities than many of the persons we serve. Persons who have social and cultural advantages can take a leadership role to be proactive about social justice in research. Researchers are encouraged to use their training and education to do what they can every day to be inclusive. Each person is different; their experience as a research participant is unique to them and is influenced by how researchers interact within the team and in their communication with research participants. It is important for researchers to be aware of health disparities. Cultural competency training can be valuable in improving the attitudes of members of the research team to prevent bias. This can help researchers to develop skills to recognize and promote equality in clinical research.

---

Table 1: Description of Health Disparities

• Difference in the occurrence of disease, injuries, or exposure to violence
• Differential opportunity in obtaining medical treatment
• Negative outcomes or impact upon persons who are socially disadvantaged
• Persons who experience health disparities are treated differently

Table 2: Expression of Health Disparities

• Implicit bias
• Stereotyping
• Colonial mentality
• Institutionalized racism
• Gender bias
• Sexual orientation
• Health literacy and interpretation quality
• Cross-cultural communication

Table 3: Disparities in Health Outcomes

• Health disparities and weight:
  • Persons who receive poorer care are more inclined to be:
    • Overweight or obese
    • Exposed to risk factors for chronic health conditions
  • Overweight and obese persons are stereotyped
  • Overweight and obese persons have poorer health outcomes ¹⁰
• Health disparities, race and gender
  • Infants born to African American mothers:
    • 1.5-3 times more likely to die
  • African American males:
    • High rates of cancer, nearly 600 cases per 100,000 persons ¹¹
• Suicide risk: American men are 4 times more likely to commit suicide than women¹³

• Osteoporosis outcomes:
  • White women have higher rates of hip fractures, but African American women have poorer outcomes
  • 33% of white women had osteoporosis screening versus 5% of African American women 
  • White women were screening and diagnosed early twice as often as African American women^15,16
• Inpatient acute myocardial infarction outcomes:
  • Asian Americans have the highest rate of mortality—7.2%—even though they comprise only 3% of the population
  • Mortality rates is higher than:
    • African Americans (5.4%)
    • Hispanics (5.5%)
    • Whites (6.3%)¹⁷

Table 4: Epidemiology and Health Disparities

• Historically, public health, anthropology, and population genetics identified groups based on race
• Epidemiology led to “predisposition” to a condition, disorder, or disease
• Epidemiology categories and analysis in Europe and the United States usually compare ethnic minority populations with whites
• Genetic, social factors, and differences among groups have an effect on health outcomes and treatment response
• Descriptive epidemiology can model the relationship between risk factors, exposures, and subsequent disability and disease rates

Table 5: Strategies to Reduce Disparities in Research

• When assessing site feasibility, strive to ensure enrollment of an inclusive research cohort
• Recognize that differences exist between cultures
• Provide patient-centered, culturally-competent approaches
• Realize that researchers’ experiences, beliefs, values, and language affect perceptions and world view
• Recognize and value different customs, behaviors, communication styles, traditions, and family systems
• Adapt research to be inclusive
• Recognize that each person is unique with a new voice
• Develop inclusive research teams
• Use a social psychological approach to counteract negative stereotypes
• Use cultural competency training to improve staff attitudes and prevent bias

References

1. Braveman, P. A., Kumanyika, S., Fielding, J., Laveist, T., Borrell, L. N., Manderscheid, R., & Troutman, A. (2011). Health disparities and health equity: the issue is justice. American journal of public health, 101 Suppl 1(Suppl 1), S149-55. doi: 10.2105/AJPH.2010.300062.
2. Maina I, Belton T, Ginzberg S, Singh A, Johnson T. A decade of studying implicit racial/ethnic bias in healthcare providers using the implicit association test. Social Science and Medicine. 2017; doi:10.1016/j.socscimed.2017.05.009.
3. Hero, J. O., Zaslavsky, A. M., & Blendon, R. J. (2017). The United States Leads Other Nations In Differences By Income In Perceptions Of Health And Health Care. Health Affairs, 36(6), 1032-1040. doi:10.1377/hlthaff.2017.0006.
4. Mandal, J., Acharya, S., & Parija, S. C. (2011). Ethics in human research. Tropical parasitology, 1(1), 2-3. doi: 10.4103/2229-5070.72105.
5. Las Vegas: University of Nevada; 2010. History of research ethics. Retrieved February 27, 2019 from: https://www.unlv.edu/research/ORI-HSR/history-ethics
6. National Institue on Minority Health and Health disparities. (n.d). Retrieved February 27, 2018, from https://www.nimhd.nih.gov/
7. Wong, N.D and Levy, D. (2013). Legacy of the Framingham Heart Study: rationale, design, initial findings and implications. Elsevier. doi: 10.1016/j.gheart.2012.12.001.
8. Liu, K. A., & Mager, N. A. (2016). Women’s involvement in clinical trials: historical perspective and future implications. Pharmacy practice, 14(1), 708. doi: 10.18549/PharmPract.2016.01.708. 
9. Coakley, M., Fadiran, E. O., Parrish, L. J., Griffith, R. A., Weiss, E., & Carter, C. (2012). Dialogues on diversifying clinical trials: successful strategies for engaging women and minorities in clinical trials. Journal of women’s health (2002), 21(7), 713-6. doi: 10.1089/jwh.2012.3733.
10. Flint, S. W., Čadek, M., Codreanu, S. C., Ivić, V., Zomer, C., & Gomoiu, A. (2016). Obesity Discrimination in the Recruitment Process: “You’re Not Hired!” Frontiers in psychology, 7, 647. doi:10.3389/fpsyg.2016.00647
11. Frieden, T. R. (2011). Health disparities and inequalities report–United States, 2011(Vol. 60, pp. 1-2). Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention.
12. Cancer Statistics. (2018, April 27). Retrieved March 4, 2019, from https://www.cancer.gov/about-cancer/understanding/statistics
13. Suicide. (2018, October 22). Retrieved March 4, 2019, from http://www.mentalhealthamerica.net/suicide#11
14. Curtis JR, Carbone L, Cheng H, Hayes B, Laster A, Matthews R, Saag KG, Sepanski R, Tanner SB, Delzell E. Longitudinal trends in use of bone mass measurement among older Americans, 1999– 2005. J Bone Miner Res. 2008; 23:1061–1067.
15. Curtis JR, Carbone L, Cheng H, Hayes B, Laster A, Matthews R, Saag KG, Sepanski R, Tanner SB, Delzell E. Longitudinal trends in use of bone mass measurement among older Americans, 1999– 2005. J Bone Miner Res. 2008; 23:1061–1067.
16. Cauley J. A. (2011). Defining ethnic and racial differences in osteoporosis and fragility fractures. Clinical orthopaedics and related research, 469(7), 1891-9. doi: 10.1007/s11999-011-1863-5.
17. Kim EJ, Kressin NR, Paasche-Orlow MK, et al. Racial/ethnic disparities among Asian Americans in inpatient acute myocardial infarction mortality in the United States. BMC Health Services Research. 2018; 18:370. doi: 10.1186/s12913-018-3180-0.
18. U.S. Census Bureau Quick Facts: UNITED STATES. (2017, August 1). Retrieved February 27, 2019, from https://www.census.gov/quickfacts/fact/table/US/PST045217
19. What is Genomic Medicine? (2018, November 6). Retrieved March 5, 2019, from https://www.genome.gov/27552451/what-is-genomic-medicine/
20. Cristancho, S. M., Apramian, T., Vanstone, M., Lingard, L., Ott, M., & Novick, R. J. (2013). Understanding clinical uncertainty: what is going on when experienced surgeons are not sure what to do? Academic medicine: journal of the Association of American Medical Colleges, 88(10), 1516-21.  doi: 10.1097/ACM.0b013e3182a3116f.