Introduction

Because clinical trials evaluate medical product safety and efficacy in support of regulatory and clinical decisions, participants should reflect target populations across clinically relevant characteristics. Diverse enrollment can also promote trust and fair distribution of benefit [1]. However, certain populations—racial, ethnic, and linguistic minorities; adolescents and older adults; pregnant people; individuals with disabilities; and individuals of low socioeconomic status—are often underrepresented in trials [2]. Because overly strict eligibility criteria not justified by scientific need or participant safety can contribute to this problem, we examined eligibility criteria that may impact equitable inclusion in trials supporting US Food and Drug Administration (FDA) drug approvals.

Methods

We examined pivotal trials supporting all new molecular entities and therapeutic biologics approved by FDA’s Center for Drug Evaluation and Research (CDER) in 2022 [3]. We selected this year given increased attention to trial diversity, including a 2022 FDA draft guidance document and federal legislation requiring diversity action plans [4]; accordingly, 2022 approvals based on earlier pivotal trials can provide a baseline against which future approvals may be assessed. We identified pivotal trials by reviewing summary and/or medical review documents in approval packages posted to the Drugs@FDA website. We matched these trials on Clinicaltrials.gov using their National Clinical Trials (NCT) identification numbers and extracted language describing eligibility criteria.

We then qualitatively coded these criteria, using a codebook developed through inductive and deductive methods, beginning with broad, pre-defined categories (e.g., exclusion based on age), then refining and expanding codes based on the data (e.g., more specific age categories) [5]. The codebook consisted of categories relevant to equitable inclusion across race, ethnicity, socioeconomic status, pregnancy, disability, and age. Explanations of the potential relevance of these categories to equity, as well as potential scientific and regulatory justifications, are included in Table 1. These categories were selected because they were either directly related to individual demographics (e.g., age, pregnancy) or potentially associated with disproportionate impact on certain groups based on epidemiologic patterns (e.g., HIV disproportionately impacts Black and Hispanic populations) [6]. Several categories were divided into “general” and “specific” subtypes to acknowledge that more specific eligibility criteria may have stronger justifications. Coding was performed by a single author (KS), with audits by other authors (JM, HFL) to ensure consistency and validity; challenging cases were discussed and resolved by all authors.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Coded eligibility criteria potentially related to equity examined in pivotal trials.

https://doi.org/10.1371/journal.pone.0324807.t001

We calculated the frequency of equity-related eligibility criteria overall and across several drug categories (orphan; oncology; first-in-class; expedited). We included expedited programs (breakthrough, fast track, and accelerated approval) as FDA has deemed these drugs promising and likely to address unmet treatment needs, increasing the importance of equitable opportunities for trial participation and appropriately generalizable study populations. Due to the small number of trials with relevant exclusion criteria in these categories, we did not assess for statistically significant differences between them.

Results

In 2022, CDER approved 37 novel drugs based on a total of 56 pivotal trials. As described in Table 2, the most common equity-related exclusion criteria were of adolescents (45, 80.4%), pregnant/lactating individuals (30, 53.6%), and individuals with limited literacy (20, 35.7%); this pattern held across drug categories. About 1 trial in 5 excluded individuals based on weight/obesity (11, 19.6%), had general exclusions for hepatitis, i.e., without specification beyond current infection (10, 17.9%), or excluded adults aged 80 and up (11, 19.6%). General HIV exclusions were present in 7 (12.5%) trials. With few exceptions, exclusions based on literacy, HIV, and hepatitis were proportionately more common for trials supporting specific drug categories than they were in pivotal trials overall. Pregnant/lactating individuals were most often excluded from trials supporting drugs with fast-track designation (10, 71.4%) and those granted accelerated approval (4, 66.7%).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Frequency of equity-related eligibility criteria examined in pivotal trials, overall and by drug/approval type.

https://doi.org/10.1371/journal.pone.0324807.t002

Criteria allowing exclusion based on an individual’s general social condition were infrequent (4, 7.1%), as were those excluding adults aged 65 and up (4, 7.1%), young adults (3, 5.4%), or individuals with a general mental health condition (2, 3.6%). Although individuals with intellectual disability were infrequently excluded (2, 3.6%), only 1 trial explicitly permitted a Legally Authorized Representative to consent. No trials excluded individuals based on general substance use disorder or smoking; only 1 excluded for cannabis use. While no trial explicitly required English language, this likely overlaps with common literacy exclusions.

Discussion

Our analysis found that eligibility criteria in pivotal trials supporting 2022 drug approvals were broadly inclusive of individuals with HIV, diabetes, hypertension, mental health conditions, and substance use disorder, as well as those who smoke, use cannabis or are intellectually disabled. Most trials avoided vague “catch-all” exclusions based on general social condition that could lead to bias based on factors like race, ethnicity, or socioeconomic status. For oncology studies and accelerated approvals (which often overlap), several trials had no equity-related exclusion criteria beyond literacy, pregnancy/lactation, and age.

Exclusion of adolescents was common but likely reflects distinct safety considerations and regulations applicable to pediatric research, as well as distinct disease epidemiology; in addition, children may be more often included in post-approval studies that are encouraged or required under FDA policy. Exclusion based on pregnancy/lactation status was also common, likely for similar reasons, despite expert recommendations encouraging routine inclusion of pregnant/lactating people in later stage studies, such as pivotal trials supporting drug approval [7,8]. Frequent exclusion based on weight/obesity documented in our sample could have clinical justification, for example related to dosing or various risk factors, but more specific rationales for exclusion would be most appropriate. While current diversity discussions have largely focused on race/ethnicity, sex, and age, there has been less attention to exclusion based on body size despite high rates of overweight and obesity in the American population. Similarly, the substantial exclusion of adults aged 80 and up found in our sample sometimes may be clinically justified but warrants attention given the aging population and need for data to support use in Medicare beneficiaries. Finally, the frequent exclusion of individuals with limited literacy reflected in our findings could be addressed by utilizing recommended literacy supports such as ensuring that all research materials, including consent forms, are written in plain language, utilizing visual aids like diagrams and videos, facilitating reading/writing assistance, and offering individualized support from research team members. Similar strategies can be employed for individuals with cognitive impairment, as well as permitting reliance on legally authorized representatives (LARs) [9].

Conclusion

FDA guidance recommends broadening eligibility criteria when scientifically appropriate to improve diverse participation, recognizing that some criteria have become “commonly accepted over time or used as a template across trials… excluding certain populations... without strong clinical or scientific justification” [10]. While overt exclusion based on race/ethnicity and socioeconomic status is not reflected in eligibility criteria, it is important to also consider associated factors that could lead to the same outcome, such as exclusion based on comorbidities that reflect health disparities. It is also essential to address trial diversity across all domains that influence the generalizability of study findings and relate to fair distribution of benefit. To minimize unnecessary equity-related exclusions, eligibility should focus on specific issues expected to influence safety or efficacy (e.g., liver function, drug interaction, uncontrolled comorbid disease), rather than blanket exclusion based on broad diagnoses or characteristics. Accommodations also may be possible. Investigators, sponsors, institutional review boards, and regulators should scrutinize eligibility criteria with equity in mind.