Addressing Demographic Disparities in Clinical Trials

Disparities in health care are not just limited to the delivery of care. They also exist in the way that clinical trials are structured. Women, people of color, and the elderly often are not adequately represented. This article offers three strategies for addressing this deficiency: (1) know what “representative” means for your disease and set intentional recruitment goals; (2) use patient registry datasets as the basis for trials; and (3) go beyond traditional academic medical centers (AMCs) to access patients.

Most clinical trials are not representative of the general population or of those with a particular disease. Clinical trials primarily enroll white, male patients, with consistent underrepresentation of women, the elderly, and people of color — especially Black and Hispanic patients. While people of color make up about 39% of the U.S. population, these groups represent from 2% to 16% of patients in trials.

This underrepresentation emerged as an issue during the trial of Moderna’s Covid-19 vaccine. When it was reported that Black Americans represented only 7% of the trial versus 13% of the U.S. population, Moderna slowed the trial to quickly recruit more people of color as participants.

Moderna was the exception. Companies rarely face public scrutiny, don’t need a representative trial to secure the approval of the U.S. Food and Drug Administration (FDA), and believe that recruiting representative patients adds time and expense. Conducting diverse trials has not been mandated and has not been a priority.

Yet overnight the world changed. The pandemic has shined a spotlight on disparities in health care. Addressing them has rapidly become a priority for the health ecosystem, including the research community.

How to Create Representative Patient Data

With disparities now on the agenda of cure-seeking organizations, the challenge is how to create representative patient data and how to access and engage patients who are people of color.

Based on interviews with dozens of leaders throughout the health care ecosystem, including disease foundation CEOs who participate in leadership forums hosted by the Harvard Business School Kraft Precision Medicine Accelerator , we offer the following framework.

1. Know what “representative” means for your disease and set intentional recruitment goals.

Every disease affects different populations — with differences in age, gender, race, and ethnicity. What is representative for sickle cell disease, which is more common in people of African descent and Hispanics, is very different than what is representative for breast or prostate cancer.

By knowing what is representative for a disease, researchers can set intentional goals for the trial population. For example, since the Multiple Myeloma Research Foundation (MMRF) has learned that about 20% of patients with this disease are Black, it strives to ensure that all of its studies include a representative population.

2. Use patient registry datasets as the basis for trials.

Trials tend to be finite in scope and transactional. Patients are recruited from scratch, are induced to participate, the trial is conducted, and then the trial ends.

Disease registries — which are often managed by independent disease foundations — are fundamentally different. Registries are longitudinal in nature, can aggregate broad and deep real-world data, including a wealth of personal and demographic data (such as race and ethnicity), as well as clinical data, genomic information, and biological samples. Importantly, when done right, through multiple interactions with patients over time and by providing participants with value by sharing data and insights, registries engage patients, establish relationships, and build trust.

By tapping robust registries as the source of patients for trials, researchers can define precise participation criteria and can identify patient segments and sub-segments that match these criteria. In this way, researchers can avoid many of the problems that plague clinical trial recruitment and can access representative patient populations.

3. Go beyond traditional academic medical centers (AMCs) to access patients.

The ability to recruit diverse patient populations has been limited because the source of patients in trials has been largely confined to AMCs. More diverse patient populations can be engaged by expanding where and how patients are recruited for registries and trials.

We recommend that organizations planning to conduct trials “go where the patients are.” This means thinking broadly and creatively about all possible channels for recruiting patients — both within and outside of health care — and formulating a channel strategy most appropriate for your disease and patient population.

In addition to going to AMCs, we recommend considering the following:

Health systems. They have access to large numbers of patients and typically have good data about patients’ disease state, clinical history, gender, ethnicity, and more. For example, Ochsner Health , the largest health system in Louisiana, cares for about one million individuals, encompassing a diverse population. Ochsner is building community health centers in disadvantaged neighborhoods and has established a clinical research arm. With a strong presence in “cancer alley” — the stretch along the Mississippi River between Baton Rouge and New Orleans with dozens of petrochemical plants and high rates of cancer — Ochsner represents a potential partner for researchers interested in enrolling this system’s diverse patients.

Another example is a partnership between the Prostate Cancer Foundation (PCF) and the Veteran’s Administration (VA), America’s largest integrated health care system. The VA has large numbers of patients with prostate cancer, 40% of whom are African American, and has one of the country’s best electronic health records (EHR) systems . Used by all of the VA’s 1,700 hospitals and clinics, it collects and stores consistent longitudinal data on more than 23 million patients who have received care from the VA.

Community networks. These are grassroot networks that use creative approaches to engage and enroll patients. For example, organizations that work with thousands of churches, which are vital trust brokers in underserved communities of color, can help recruit diverse patients for registries and trials.

Direct to consumer (DTC). This involves going to patients directly, using modern consumer marketing approaches and social media to reach and engage patients. The Follicular Lymphoma Foundation — founded by a Facebook executive — has harnessed DTC marketing and social media to build a global online Follicular Lymphoma community . This is the antithesis of working solely through AMCs.

The idea is not to be limited to AMCs or to pick just one channel but to think strategically about which channels and combinations can help your organization build a large, representative registry for your disease. Understand the capabilities of potential partners, the populations they can reach, their level of trust in the community, and their ability to execute.

Why Addressing Disparities Matters

In this era of precision medicine, the key to accelerating cures is data. But without data that is representative of all patients with a disease, researchers won’t be able to determine which segments and sub-segments of patients are helped most by a treatment. Representative patient data and the insights from this data are essential in ensuring that no patients are left behind in efforts to cure a disease.

Kathy Giusti is a faculty cochair of the Harvard Business School Kraft Precision Medicine Accelerator and is the founder and chief mission officer of the Multiple Myeloma Research Foundation.

Richard G. Hamermesh is a faculty cochair of the Harvard Business School Kraft Precision Medicine Accelerator and is a professor of management practice (retired) at Harvard Business School.

Mark Krasnow is founder and CEO of BullsEye Resources, a communications firm that helps leaders and experts convey their most important ideas.

Copyright 2021 Harvard Business School Publishing Corporation. Distributed by The New York Times Syndicate.