Diversity Principles and Suppositions

Basic Principles

The MRCT Center Diversity Workgroup has developed a set of fundamental principles that help to frame considerations of diverse representation in clinical research. While we recognize that a case-based analysis will be required for each clinical research question, we also believe that these principles will help guide those analyses:

Download Principles (PDF document)

The clinical research enterprise—and healthcare – should endeavor to distribute the risks, burdens, and benefits of research fairly and responsibly. The health needs, and responses to interventions, of populations and individuals can only be identified, considered, and managed if those populations and individuals are represented and studied.

Efforts to understand biologic variability and the complex contributions of social determinants of health, disease burden and progression, access to clinical trials, and treatment outcome require careful and diligent study. Approaches for determining and collecting relevant variables for a given disease, condition, diagnostic or therapeutic product, or intervention are necessary.

Diversity and inclusion in clinical research aims to identify subpopulation variability in diagnosis, treatment, and prevention.  Diversity and inclusion also serve to broaden the knowledge base and may identify important group-specific efficacy and safety signals prior to approval of investigational products.  Clinical research in which participants reflect the diversity of the intended population for treatment or intervention is better positioned to develop effective treatments for those most likely to use them. A greater understanding of the barriers that negatively impact diversity and inclusion in research is needed so that data supporting future medical innovation better reflect the intended populations of the intervention.

Efforts to achieve enhanced and representative diversity require consideration of complex scientific, organizational, social, and cultural factors, and intrinsic biases.  Progress requires engagement, commitment, and accountability by all stakeholders, including sponsors, research institutions, investigators, patients and their advocates, regulatory agencies, oversight bodies and others.

Development and adoption of common standards, methodologies, and successful strategies will require global collaboration across stakeholders and scientific disciplines and are necessary to advance medicine and public health.

The size, time, and resource requirements of clinical trials typically preclude their use to detect small but potentially significant differences across all populations of interest.  New research paradigms using real world data, curated data sources, machine learning, bioinformatics, and robust analytics are necessary.

Suppositions

1. Our recommendations and suggestions should not be interpreted as prescriptive. We are not advocating for “quotas” within each clinical trial or across clinical development programs. We do believe, however, that the general intention of recruitment in research is to reflect the population for which the intervention or research question is directed. We also understand that the right answer to the question of which subgroups to study will likely be “it depends.” A case-based analysis, examining trade-offs and opportunities, what is already known and unknown, and other factors will drive the choices made. Further, we believe that careful consideration of appropriate inclusion of diverse populations across the research program is essential; the effort must be advertent and intentional to be successful. We have endeavored to present recommendations that are sufficiently flexible to allow adaptation to the context, not only of the research question, study design, disease, and epidemiology but also of differing cultures, approaches, and technologies.

2. We believe that both the scientific understanding of biological diversity (e.g., “will this product be safe and effective and how generalizable is that determination? What is the heterogeneity of effect?) and health equity (e.g., “how will this study help to advance the ability of each person to attain their full health potential?”) are both important goals, and that their importance may be weighted differently by different stakeholders. However, the objective of human participant research is generalizable knowledge, and thus the two goals are interdependent and complementary. For example, while the mission of health regulatory authorities is to assure the safety and efficacy of medical products and devices for the general population and not, generally, to weigh social justice concerns in their determinations, considerations of generalizability are directly related to the diversity of the population studied. It is a mistake, therefore, to interpret increasing diversity to reflect the population for which a product is intended as solely a health (or social) equity effort; it is equally a scientific concern, directed at more accurate understanding of the health effects of any drug, device, biologic or procedure across a broad population.[1]

3. We start from the perspective that every investigator, sponsor, funder, organization, and participant should support the goals of diversity and inclusion. We therefore offer recommendations that we believe will further that goal.

4. We believe that all stakeholders have responsibility for execution and outcome, although the nature of those responsibilities differ. Further, responsibility is not discharged if certain functions are outsourced to contracted organizations (e.g., academic or for-profit contract research organizations).

5. We hold that the expectations of inclusion of representative participants in clinical research are applicable to all sponsors and funders of research. While the drivers and incentives for academic and industry and other sponsors—and the resources available to support the research—may differ, the value of diversity to research does not.

6. When we speak of diversity in this document, we mean all the dimensions of diversity: demographic factors (e.g., an individual’s sex, gender, race, ethnicity, age, genetic background) and non-demographic factors (pregnancy, metabolism, comorbid conditions, diet, smoking, alcohol use, climate, environment, social determinants of health, local medical practice). We have also tried to draw examples from different dimensions of diversity, although admittedly some are more abstruse than others. When one dimension is particularly relevant, we attempt to make that clear.

7. When we use the term “social determinants of health,” we accept the definition proffered by the World Health Organization (WHO): “The social determinants of health (SDH) are the conditions in which people are born, grow, work, live, and age, and the wider set of forces and systems shaping the conditions of daily life. These forces and systems include economic policies and systems, development agendas, social norms, social policies and political systems.”[2] As such, it includes issues such as economic and educational vulnerabilities, sexual orientation, discrimination and other stressors, and additional attributes that, collectively, affect the health status of an individual and their community. We appreciate that more work is needed to understand how SDH impacts clinical research data, medical and behavioral interventions, medicine and public health.

8. We believe in the primary importance of establishing global standardization of data elements and metadata, and agreement on data collection methodologies. The absence of data and metadata standardization, preferably in a machine-readable format, will hinder progress in data aggregation and analysis.

9. We believe that optimizing inclusion of diverse participants in clinical research requires thought and planning from the earliest conceptualization of the research question and needs an affirmative commitment from all stakeholders. Proactive planning for recruitment and retention of a diverse population will save time, resources, and costs.

10. Engagement of patients and participants, their caregivers and loved ones, patient advocacy groups, and communities throughout the clinical trial lifecycle—from trial design, trial conduct, recruitment, retention, analysis, and return of results—is critically important to optimize the goals of the research, outcomes of relevance to the patient/participant/community, help ensure diverse representation in the trial, and communicate the results. Involvement of individuals, patient organizations, and the community in research implies a continued and long-term commitment by the investigator or sponsor. The optimal goal is a bidirectional partnership in which the purpose of the research serves all those engaged.

11. While clinical research may strive to reflect the population that is likely to receive the intervention, in the end what matters is whether an individual is likely to benefit from that intervention following its approval and whether the anticipated benefit outweighs the risk of harm. Pre-specified subgroup analyses, if designed to have statistical signifance, may be helpful to differentiate those subpopulations that are predictably high (or low) responders. Results derived from clinical research will always require interpretation for the individual.

12. New methods and approaches for appropriate inclusion, represention, data collection, data analysis, and communication are necessary, methods that may involve not only clinical trials but also the integration of observational data, real world evidence, genomic data and other approaches. Investment is necessary until data-driven approaches are found to be effective, and adequate representation, based on evidence, is achieved as a routine expectation of clinical research.

[1] Of note, disproportionate accrual to a study may be required if scientific concerns exist (e.g., it may be necessary to overrepresent—or continue accrual—of a specific subpopulation if early evidence suggests a different safety or efficacy profile).
[2] World Health Organization,https://www.who.int/social_determinants/en/ [Accessed 22 June 2020]