Ethical Considerations for Conducting Health Disparities Research in Community Health Centers: A Social-Ecological Perspective

Abstract

Community health centers (CHCs) provide optimal research settings. They serve a high-risk, medically underserved population in the greatest need of intervention. Low socioeconomic status renders this population particularly vulnerable to research misconduct.

Traditional principles of research ethics are often applied to participants only. The social-ecological model offers a comprehensive framework for applying these principles across multiple levels (participants, providers, organizations, communities, and policy).

Our experience with the Trial Using Motivational Interviewing, Positive Affect and Self-Affirmation in African-Americans with Hypertension, a randomized trial conducted in CHCs, led us to propose a new platform for discussing research ethics; examine the social, community, and political factors surrounding research conducted in CHCs; and recommend how future research should be conducted in such settings.

Community health centers (CHCs) serve approximately 20 million people in more than 7500 communities nationwide and function as an important safety net for medically underserved and socioeconomically disadvantaged patients. More than half of those served are racial/ethnic minorities, almost two thirds live at or below the federal poverty level, and approximately 38% are uninsured or rely on Medicaid.^1–3 CHCs provide care for large numbers of patients with multiple comorbid conditions and complex social and economic circumstances. Their focus on serving a high-risk medical population with complex medical and social conditions makes CHCs an ideal setting to develop and conduct public health or health disparities research that focuses on improving the health of underserved racial/ethnic minorities. A survey conducted by the Association of Academic Health Centers reported that 72% of academic health centers conduct research in CHCs.⁴ For academic researchers, CHCs offer the potential to reach populations at greatest risk for poor health outcomes; these populations therefore may stand to benefit the most from health behavior interventions. However, the often-precarious health, social, and economic status of some of the patients cared for in this setting renders them vulnerable to being taken advantage or exploited in research studies.^5,6

Interest in CHCs as settings for health disparities research is likely to increase. Under provisions of the Affordable Care Act, CHCs are expected to undergo tremendous expansion and growth. The law will provide additional resources to deliver care to approximately 20 million more people.¹ As the number of CHCs continues to increase, demand for research in these settings may also increase, testing their ability to provide quality care, narrow health disparities, and contain cost.⁷ Researchers and policymakers may wish to examine best-practice models of providing service and interventions targeted at improving the health of underserved minority populations and may seek to conduct their studies in CHCs. This calls for a renewed and more nuanced examination of ethical concerns that may arise in the conduct of research in CHCs.

When conducting research in large safety net systems of care, researchers must consider the complex interactions among patients as research participants, the providers caring for them, the community settings in which care is rendered, and the organizational policies and practices that govern them. Care must be exercised to avoid placing undue burden on the health care system that operates within CHCs. This requires discussion of research ethics to focus not only on the impact on individual participants but also on the potential impact on the entire system that operates a CHC.^7,8

All public health research requires ethical considerations at multiple levels. However, CHC patients who participate in research deserve special consideration because of the well-documented potential for exploitation, abuse, and other harm among underserved racial/ethnic minority populations.⁵ The social-ecological model provides a framework for examining and understanding potential ethical challenges to conducting research in settings that provide care to medically underserved racial/ethnic minority populations. This framework, often used for health behavior change and prevention research, represents a confluence of multiple integrated factors:

1. intrapersonal factors such as the knowledge, attitudes, and behaviors of an individual;

2. interpersonal factors such as the provider–patient interaction and social network involvement;

3. organizational factors that dictate rules and regulations for operation;

4. community factors such as relationships among organizations, institutions, and informal networks; and

5. public or local policies that set standards for how different organizations and constituents operate or interact.⁹

The comprehensiveness of this model makes it applicable to understanding ethical challenges to conducting health disparities research at multiple levels.^10,11

Our experience conducting a randomized clinical trial served as a case study to explore the ethical considerations in conducting health disparities research in a CHC. The Trial Using Motivational Interviewing and Positive Affect and Self-Affirmation in African-Americans with Hypertension (TRIUMPH) is a randomized controlled trial among adult African Americans whose blood pressure is above the target range despite their use of medication.¹² TRIUMPH uses the behavioral cognitive approaches of positive-affect induction, motivational interviewing, and self-affirmation to motivate participants to adhere to their hypertension medications and to adopt healthier lifestyles.^13–16 The protocol used in developing TRIUMPH is described elsewhere.^13,17–19 TRIUMPH recruited participants from a large tertiary care urban health center and its affiliate CHCs. Here we focus on our experience of recruiting participants from the primary CHC in TRIUMPH.

ETHICAL CONCERNS IN DEVELOPING TRIUMPH

We reviewed detailed field notes and held discussions with key stakeholders—patients, community representatives, providers, administrators, and staff. The box on this page depicts the 5 levels of factors for ethical consideration that we identified and organized according to the social-ecological model:

1. intrapersonal or research participant factors, such as the social and economic context in which patients or potential research participants live;

2. interpersonal factors, such as ensuring that the study design does not undermine the provider–patient relationship or adversely affect physicians’ work patterns;

3. organizational factors, such as the availability of facility space and time for research and the impact the study may have on these logistic factors;

4. community factors, such as making sure that eligibility criteria do not inadvertently contribute to perceptions of disparate treatment in research and do not promote community mistrust of research; and

5. policy factors, such as respect for the autonomy of existing institutional review board (IRB) processes at the site.

Ethical Considerations for Conducting Research in Community Health Centers

Intrapersonal factors

What is the socioeconomic context in which participants live?

What are the major social determinants of health that are at play in their lives?

What are ways these determinants may be addressed in the context of the study, without significantly changing the research study protocol?

Interpersonal factors

Who are the key stakeholders who should be involved in this process (practitioners, nurses, administrators)?

How can the physicians and staff in this setting be engaged in the study?

What impact will the study as it is currently designed have on the provider–patient relationship or patients’ perspectives of the care that they receive?

How will the study protocol affect providers’ pattern of work?

Organizational factors

What are the logistical factors that need to be considered in implementing this study? Is there sufficient space to conduct study operations?

What impact will study implementation have on practice site work flow?

How can disruptions of work flow be minimized?

Community factors

Who are the dominant racial/ethnic groups in the community?

How will the study be viewed by different racial/ethnic groups who receive care in the community health center?

How will the eligibility criteria affect the community’s view of research?

Will some community members be excluded from the study by the eligibility criteria? How will this be perceived by the community?

Policy factors

What is the policy of the community health center toward patient consent forms? Does it have an institutional review board?

What are center's policies regarding ownership of data files?

Intrapersonal Factors

The primary research setting for TRIUMPH is a CHC located in Central Harlem, a community in New York City that is 67% African American and 20% Latino and where approximately 25% of residents report not having a doctor, nearly 24% are without a primary source of care, 32% do not have a high school diploma, and approximately 36% live below the poverty line.²⁰ Early on during study implementation, CHC providers and administrative staff voiced concern that the socioeconomic circumstances of the participants were not being considered in the process of research. The overwhelming sentiment was that study participants should be compensated for their time because they had limited socioeconomic resources. These CHC personnel questioned the practice of taking up participants’ time without any immediate compensation, especially because future benefits from practice change or dissemination of research knowledge to the community could take years to arrive. The consensus was that despite the value of participating in a research study and contributing to science, participants should also be compensated, especially, in the words of one provider, “given these economic times.” In setting a reimbursement amount, we also had to avoid offering an amount that could be viewed as exerting an undue influence on the decision to participate in the study. As a result of these considerations, we made 2 amendments to the protocol. The first incorporated reimbursement in the form of a $20 gift card at enrollment and another at closeout, as well as a $4.50 MetroCard (covering the cost of a round trip subway ride in New York City). This change was approved by the IRBs.

The second amendment regarded the type of item that was given as part of the positive-affect induction protocol. A component of TRIUMPH is to provide participants in the experimental arm with small gift items as positive induction. We intended to offer clocks, blankets, and umbrellas, gift items provided in our previous study.¹⁷ However, our community partners suggested that in light of the socioeconomic status of the participants, an additional $20 gift card, gift cards to the local neighborhood retail stores, or a $20 MetroCard would be more welcome. As one of the providers said, “These are real patients with real problems, so they need real money.” We therefore decided to provide $20 gift cards as the form of positive-affect induction in this study. This change took into consideration the social and economic context in which participants lived without changing the intent of the study, which is to provide a small gift item to motivate health behavior change.

In the process of conducting the study, we also learned that some participants had lost their jobs or their apartments. Although it was beyond the scope and resources of the study to directly address these problems, they propelled the research team to work with the administrative staff on-site to compile a list of social service resources that could assist participants in addressing these concerns. We made this list of resources available to all study participants, regardless of their randomization group. Addressing the social determinants of health was not a primary objective of the study. However, this experience heightened our awareness of the impact that the social or economic context in which participants live may have on the recruitment and retention process and led us to consider this factor in this study as well as in our subsequent studies.

Interpersonal Factors

Providers in CHCs care for a patient population with multiple chronic comorbid conditions.²¹ Both providers and patients in this setting raised questions: providers asked whether the study would alter their relationships with their patients and patients asked whether their participation in the study would change the type of care they received or their ability to continue to see their providers. This issue led us to reevaluate our approach to recruitment. It was important for us to preserve and promote the provider–patient relationship rather than undermine it. At the outset, we had to be clear that the goal was to build on what the providers were doing and not to replace their work. We encouraged participants to share the study material with their providers, and we assured them that the study would not change the care that they were receiving.

Another factor we considered was making sure that the study did not infringe on the limited time that patients had with their providers. Time is a factor that must be accounted for in all studies; however, it may be of greater concern in settings where providers see a large number of patients who are high risk and who have a multitude of comorbid chronic medical conditions. In the process of recruitment, we took care to avoid interfering with providers’ schedules by not interrupting or delaying appointments or work flow. In an effort to overcome this issue, we modified the protocol and conducted all research interviews after the medical visit. This helped the study to proceed with minimal disruption to providers’ work flow and minimal impact on provider–patient encounters.

Organizational Factors

The primary TRIUMPH site is a CHC where high volumes of patients are seen. This raised a valid concern that the presence of research assistants would take up space and occupy the limited number of examination rooms available to see patients. A staff concern was that the presence of a research team could disrupt practice flow and ultimately delay the care of patients. To accommodate the need for space to interview participants, we conducted study recruitment on weekends and during evening hours rather than during the busiest hours of the practice. This also helped to accommodate participants whose work schedules prevented them from coming to the practice during the traditional working hours on weekdays.

Clinicians and staff were also concerned about participants’ privacy. They did not want prospective participants to be approached in the waiting room, seeing this as too intrusive and possibly disruptive to the practice. The staff recommended that the research assistants set up a recruitment table near the entrance of the site, which would not interfere with the process of care. This table would be visible upon entering and leaving the practice, and prospective participants who were interested could voluntarily approach the table; if they expressed an interest and met eligibility criteria, they would be taken to a private area to obtain IRB consent and collect baseline data.

Community Factors

It is important that researchers conducting studies in settings such as CHCs that serve multiracial/ethnic populations consider the impact that the study could have on community trust and how inclusion or exclusion criteria will be viewed by the community at large. TRIUMPH is being conducted in a practice setting in which 53% of patients are non-Hispanic African American, 38% are Hispanic, 4.5% are non-Hispanic White, and 2% are Asian. TRIUMPH was originated to recruit African Americans with hypertension because of consistent data showing that non-Hispanic African Americans have a greater risk of developing hypertension and are more likely than Hispanics and non-Hispanic Whites to experience adverse outcomes.^22–26

The ineligibility of Hispanics for the study raised concerns among staff and Hispanic patients, who deemed it unfair that they were ineligible to participate because of their ethnicity. They raised their concerns to the providers and staff in the practice setting as well as to members of our Community Advisory Board, who recommended that we provide an explanation of the rationale for the study eligibility criteria and offer those participants ineligible for TRIUMPH information on other studies. We used this as an opportunity to work with the Community Advisory Board to develop studies in which race/ethnicity was not an exclusion criterion. The experience with TRIUMPH emphasized that in a community that is already vulnerable to disparate treatment, ethics requires that study designers ensure that eligibility criteria do not inadvertently contribute to these disparities or promote further mistrust. Furthermore, the exclusion of Hispanics may send an unintentional message that race and ethnicity are biological constructs, disregarding their important social underpinnings.

Policy Factors

In the process of study implementation, concerns about the role of the IRBs arose. Both the academic institution and the urban health center with which the CHCs are affiliated have federal-wide assurances of compliance (required approval from the US Department of Health and Human Services issued to institutions that conduct research with human participants, ensuring protection of participants) and IRBs that govern research. Clinicians and staff in this CHC felt strongly that the primary institution named in the final IRB consent form should be the community site rather than the academic institution. This was also consistent with their IRB policy.

The CHC staff also felt that asking participants to sign 2 IRB consent forms was burdensome. We amended the protocol and approved 1 consent form, after review by both the academic research site and the CHC. This form took into consideration the policies and procedures of both institutions. Streamlining the IRB process was necessary to recognize and show respect for the CHC’s autonomy as a research entity, and it helped to reduce respondent burden.²⁷

APPLYING THE SOCIAL-ECOLOGICAL FRAMEWORK

Our experience with implementing TRIUMPH in a CHC provided invaluable lessons on community engagement and led us to consider the ethical implications of conducting research in this setting through a social-ecological lens. The box on the previous page organizes factors for consideration along 5 domains and provides questions that can be used as a guide by researchers who are considering conducting research in these settings. For optimal transparency between researchers and CHC staff and providers, these questions should be discussed at the beginning of a research project in collaboration with key CHC stakeholders. For example, in TRIUMPH, we engaged the participation and solicited the advice of the clinical director of the research site, the nurses, the patient care technicians, and the practice administrator.

Intrapersonal factors concern participants and the socioeconomic contexts in which they live. In TRIUMPH, an important question raised was whether researchers conducting studies in underserved communities have an obligation to at least consider the social and economic context in which their participants live. Researchers should also understand the impact that the social determinants of health may have on research participation and retention. Asking the question, What are ways in which these determinants may be addressed in the context of the study? may help researchers and community partners to develop creative approaches that do not undermine or substantively change the research study protocol. Although the issue of compensation for study participants may be contentious, it should be considered and discussed with community representatives. When participants are provided with incentives, steps must be taken to avoid undue inducement or coercion.^28,29 Care must also be taken to ensure that participants’ autonomy is preserved and that the form or amount of compensation does not affect their freedom to decline participation.

Interpersonal factors include the impact that the study may have on the provider–patient relationship. It was important to convey that the providers made the decisions about patient treatment and that the study would not override providers’ clinical judgment. As in community-based participatory research, asking Who are the key stakeholders that should be involved in this process? enables researchers to engage a broad range of stakeholders, who may hold different perspectives on research. Engaging stakeholders may also help to improve recruitment and retention strategies. Researchers should identify key stakeholders who are responsible for maintaining quality care and consider optimal ways of engaging them. Similarly, at the organizational level, researchers should consider whether conducting research in the setting would disrupt the practice environment, a potential harm that would extend beyond harm to individual participants.

Ethical considerations also involve the community. Researchers conducting studies in these practice settings must understand who are the dominant racial/ethnic groups being cared for or that compose local communities. Asking the question, How will the study’s eligibility criteria affect the communities’ view of research? is important, especially if these criteria render 1 group ineligible. An unintended consequence of neglecting this factor may be a widening of the gap between researchers and communities. Care must be taken to make sure that individuals who are not the focus of the research do not feel that they are being unfairly excluded. Conducting pilot surveys to assess the culture of research or the relative acceptance of racial/ethnic-specific studies may help to anticipate the response to such studies. This step also helps to further engage communities in research. Finally, considerations at the policy level provide an opportunity for researchers to address the potential impact of their research on the local regulations and policies that govern the practice. Before researchers submit a research protocol to the IRB of their home institution, they should determine whether the research site has its own IRB or IRB process and, if so, endeavor to develop a protocol acceptable to both IRBs.

Our experience with TRIUMPH reinforced the critical role of involving providers, community members, and staff in the research project. In research among populations that are vulnerable because of either sociopolitical or health conditions, the goals and interests of researchers do not always align with the interests of the prospective participants. Therefore, it may be advantageous to engage stakeholders or representatives from the community who are well positioned to safeguard participants’ interests and may promote greater understanding of their social context.^30,31 In TRIUMPH, we worked closely with members of the local community board, community advocates, a member of a local law enforcement team, and several CHC patients with hypertension who lived in the community. The involvement of multiple stakeholders helped to cultivate a trusting relationship between academic investigators and CHC providers and staff. The social-ecological model’s framing of ethical considerations may be especially valuable to public health researchers working in CHCs.

This framework can also be applied when conducting studies in other research settings. As an example, when the setting for a study is a faith community, stakeholders to consider engaging might include pastors, ministers, lay health educators or parish nurses, and members of the congregation. The impact the study may have on the congregation as a whole should be considered, as well as whether the study might disrupt religious services. The goal of the framework is to serve as a tool that prompts researchers to think about the broader implications of their studies beyond the level of the individual and to consider the social context in which research participants live.