The article “Leveraging RWD With AI To Enable Diverse Recruitment In Clinical Trials” by Lakshmi Sankar and Isabelle Cheung explores the importance and challenges of enhancing diversity within clinical trials, particularly in terms of recruiting participants from underrepresented populations. This issue has become increasingly relevant for pharmaceutical companies, as current statistics demonstrate significant disparities. For instance, in the U.S., only 4%-6% of oncology trial participants are Black, and 3%-6% are Hispanic, despite these groups constituting 15% and 13% of the cancer patient population, respectively. This underrepresentation raises questions about the effectiveness and suitability of drugs across broader segments of the population.
The Importance of Diversity in Clinical Trials
The article underscores the moral and public health obligations for ensuring diversity in clinical trials. Many believe that enhancing diversity in clinical trial recruitment will lead to the development of more efficacious treatments tailored for a broader demographic. This sentiment is backed by findings from a survey of 2,000 respondents across the United States, where three-quarters agreed that diversified clinical trials could increase drug effectiveness and suitability. To address these challenges, the FDA has updated its draft guidance on Diversity Action Plans, aiming to mandate diversity in clinical trials. By requiring clinical trial sponsors to submit these plans, the FDA seeks to ensure adequate participation from relevant and underrepresented groups. This measure is a step towards enabling the analysis of data collected from clinically relevant populations, paving the way for more inclusive medical research.
A central theme of the article is that having a diverse cohort of research participants is crucial for developing studies that represent the target population comprehensively. This inclusivity is essential for understanding the unique healthcare needs and challenges faced by underrepresented communities. Yet, pharmaceutical companies face significant obstacles in achieving diversity in clinical trial research. One viable solution posited is leveraging Real-World Data (RWD) and evidence, combined with artificial intelligence (AI), to more efficiently source data and improve diversity, equity, and inclusion (DEI) in trial designs. The traditional gold standard for evaluating a drug’s safety and efficacy is through randomized clinical trials (RCTs). However, there often exists a significant gap between the populations included in these trials and the broader patient population that will use the medical product post-approval. It’s crucial to consider how study outcomes can be generalized for the intended patient population, especially considering that many clinical trials lack diverse representation.
Leveraging Real-World Data (RWD)
In the U.S., epidemiological data used to inform RCT eligibility criteria often comes from the U.S. Census Bureau’s race and ethnicity data. However, for more inclusive criteria, researchers can utilize RWD from patient registries and electronic health records to identify differences and patterns in diagnosis, treatment, and responses among diverse populations. According to a 2023 publication in Clinical Trials: Journal of the Society for Clinical Trials, data from 495 trials conducted by GSK and ViiV involving over 100,000 participants showed that U.S. Census Bureau data might not accurately represent disease demographics. The study revealed discrepancies, such as higher percentages of Black or African American patients in real-world populations with specific conditions like asthma, COPD, and HIV than those suggested by census-based data. This indicates that integrating external RWD can create a more accurate representation of real-world demographics for diseases.
In addition to informing study design, RWD can pinpoint regions with low enrollment, guide geographic placement of specialty clinics, and help researchers target specific demographic groups to improve trial diversity. Beyond recruitment, RWD and real-world evidence (RWE) can monitor drug or device performance post-approval, track patient outcomes, and monitor behaviors post-launch, especially in underrepresented groups. When utilized correctly, RWD can provide more representative and generalizable evidence for future trials than traditional RCT data alone. Despite the growing awareness of the challenges in accessing robust and reliable data, pharmaceutical companies need to explore the potential for heterogeneous treatments and varied responses to medications across racially and ethnically diverse populations. RWD, especially when paired with AI technologies capable of automating, scaling, and accelerating data analysis, can address these challenges effectively.
Integrating Digital Health Technologies (DHTs)
The integration of digital health technologies (DHTs) such as electronic consent forms, virtual visits, digital tracking, and remote patient monitoring into the clinical trial lifecycle offers a substantial opportunity to harness RWD. When combined with AI and machine learning algorithms, these technologies can significantly enhance the recruitment process. NLP techniques can extract relevant patient eligibility information from unstructured data sources, while AI algorithms can analyze vast datasets to ensure a comprehensive and fair matching process. Machine learning models can predict clinical outcomes, facilitating targeted recruitment of high-risk patients likely to respond to treatments. Moreover, AI can analyze multimodal data, such as genomic information, to select ideal patients for clinical trials, potentially reducing sample sizes while maintaining statistical power.
The Potential of RWD and RWE
The article “Leveraging RWD With AI To Enable Diverse Recruitment In Clinical Trials” by Lakshmi Sankar and Isabelle Cheung addresses the significance and hurdles in achieving greater diversity in clinical trials, especially concerning the inclusion of participants from underrepresented populations. The issue has gained importance among pharmaceutical companies in recent years, given the glaring disparities in current participant demographics. For example, in the United States, only 4%-6% of those enrolled in oncology trials are Black, and a mere 3%-6% are Hispanic. This is disproportionately low compared to these groups’ patient populations, which account for 15% and 13% of cancer patients, respectively. Such underrepresentation casts doubt on the effectiveness and generalizability of drugs for broader populations, highlighting a critical need for more inclusive recruitment strategies to ensure that clinical trials accurately reflect diverse patient populations and improve health outcomes for all.
