Increasing Black, Indigenous and People of Color participation in clinical trials through community engagement and recruitment goal establishment

Longstanding social and economic inequities elevate health risks and vulnerabilities for Black, Indigenous and People of Color (BIPOC) communities. Engagement of BIPOC communities in infectious disease research is a critical component in efforts to increase vaccine confidence, acceptability, and uptake of future approved products. Recent data highlight the relative absence of BIPOC communities in vaccine clinical trials. Intentional and effective community engagement methods are needed to improve BIPOC inclusion. We describe the methods utilized for the successful enrollment of BIPOC participants in the U.S. Government (USG)-funded COVID-19 Prevention Network (CoVPN)-sponsored vaccine efficacy trials and analyze the demographic and enrollment data across the efficacy trials to inform future efforts to ensure inclusive participation. Across the four USG-funded COVID-19 vaccine clinical trials for which data are available, 47% of participants enrolled at CoVPN sites in the US were BIPOC. White enrollment outpaced enrollment of BIPOC participants throughout the accrual period, requiring the implementation of strategies to increase diverse and inclusive enrollment. Trials opening later benefitted considerably from strengthened community engagement efforts, and greater and more diverse volunteer registry records. Despite robust fiscal resources and a longstanding collaborative and collective effort, enrollment of White persons outpaced that of BIPOC communities. With appropriate resources, commitment and community engagement expertise, the equitable enrollment of BIPOC individuals can be achieved. To ensure this goal, intentional efforts are needed, including an emphasis on diversity of enrollment in clinical trials, establishment of enrollment goals, ongoing robust community engagement, conducting population-specific trials, and research to inform best practices.

both parts is our use of CBPR approaches.
Pg. 4, line 145, "the registry has over 600,000 diverse individuals…" Would be helpful to define what they mean by "diverse" (e.g., race/ethnicity? Sexual orientation? Age?) We removed "diverse" in describing the registry. The data have not been appropriately cleaned to report on demographics or clinical trials representation and will be included in a publication later in the year.
"Part II -Involving Communities"could the authors clarify how this is different than Part I (CBPR), which also is focused on community involvement? For example, Part II describes building relationships with groups through Historically Black Medical Colleges and engaging tribal/indigenous communities, which sounds closely aligned with a CBPR approach. We agree with the reviewer's point and have combined Parts I and II.
"Part III -stakeholder engagement" -it may help readers to replicate the authors' efforts to describe what interpersonal skills/behaviors they used in order to demonstrate "humility and authenticity" to participants. On page 6, lines 228-232, we have added the following to provide guidance: "This involves knowing and understanding community context and needs, actively listening to fears and concerns of community members, being truthful and transparent at all times, ensuring that all information is provided in plain language, and making resources available from trusted sources to help community members make informed decisions." "Part V" mentions in-depth one-on-one interviews and surveys to shape the marketing and communications campaign. If the findings from these interviews and surveys have been published somewhere (even if not in an academic journal), it would be great to reference them. We have not published the data from the audience research, but plan to do so in the future.
The same paragraph mentioned reaching over 500 million gross impressions and 5 million website visitsdo the authors have any data on who (e.g., demographically) was reached?
We added, this sentence on page 7, lines 282-283: "The advertising buy focused on Black and Latino/a adults, aged 45 and older, who live in the US and speak either English or Spanish.
Relatedly, the authors mention developing a blog to reach lay audiences regarding the vaccine. Do they have any data on who the readership (e.g., demographically) of the blog was?
We are not collecting demography of the readership. We are running at about 20,000 readers per blog post. A brief discussion of how data on race and ethnicity were collected in the trials and presented in the paper would be of help. It appears they were collected separately and per OMB and that people were able to select multiple races, but audiences/readers are consistently confused by these data, particularly when race and ethnicity are presented separately as is done here.
We have provided the following sentence in the methods section on page 7, lines 300-303: "Racial and ethnic category data were collected based on the established NIH-required racial and ethnic categories and definitions. 24 Race focuses on physical characteristics, particularly skin color, whereas ethnicity attempts to capture a group's cultural identity.

Discuss the fact that enrollment of BIPOC was substantially higher in CoVPN than non-CoVPN sites. Was the difference stat significant?
No analysis was done to determine statistical significance of the difference in pace in the enrollment of white versus BIPOC individuals between CoVPN and non-COVPN sites. We have limited access to non-CoVPN site data across the trials and as such are unable to conduct these analyses. The CoVPN sites have a longstanding history of conducting clinical trials for HIV, another pandemic where racial disparities are pronounced. The CoVPN site staff are accustomed to making this effort and have built the community relationships over many years that support them in being successful.

Sentence on higher enrollment of BIPOC in CoVPN sites belongs in the results.
We have moved this sentence to the results section. It is now on page 8, lines 351-353.
I provided a few suggestions in the tracked-changes document to increase clarity of how findings are described. Thank you. We have accepted all of the proposed revisions.

May want to point out that Week 1 enrollment was low across all trials and groups.
We have added this sentence in the results on page 7, lines 311-312: "Across all trials and racial/ethnic groups, enrollment at week 1 was low." Addressing data sovereignty and ownership is mentioned in 2-3 places in the paper, would be helpful to specify how it was addressed or at least provide an example from one trial.
We have added the following sentence to page 5, lines 182-188: "The NIH Tribal Health Research Office, the CoVPN Regulatory Affairs unit, and the pharmaceutical sponsors worked with Tribal nations to develop contracts that outlined tribal data, material, biospecimen sharing and ownership agreements. These contracts took time to negotiate and were not fully executed until the Moderna and Janssen trials were almost fully enrolled. The noticeable increase in AI/AN participants in the Novavax trial is likely due to these contracts being in place earlier during study accrual." Well written and argued. I only struggled with this sentence because I think it is an overstatement: "When this is the reality across clinical research, the establishment of recruitment goals and population-specific trials may no longer be necessary, as equitable inclusion will be the norm and not the exception." Given that disparities in healthcare treatment and access would persist even in this case, I suggest the authors instead point out that ongoing commitment to these standards and partnerships will decrease the cost involved in community engagement for any one study.
We have addressed this. Lines 399-402 on page 9 now reads: "Ongoing commitment to these standards and partnerships will help communities view researchers and research institutions as trustworthy, and build and strengthen rapport between communities, researchers and research institutions."

Figures: I am unclear where Alaska Natives and Pacific Islanders are in these figures, if at all.
We collected all race and ethnicity data according to NIH-required guidelines. "American Indian or Alaska Native" are in the same category. Similarly, "Native Hawaiian or Other Pacific Islander" are in the same category. Both categories were renamed in the revised figures.
Why is it that there are data provided for people with unknown ethnicity, but not unknown race, especially given that Latinos often select unknown or do not specify race? Race 'unknown' was grouped together with 'Race not reported'. We have separated them out in the revised figures.
On the left panel of the 1st & 5th figure, orange is for Asians. However, on the right panel orange is for non-Hisp/Whites. I suggest using a different color for one. The colors have been standardized in both figures. Asian representation is now illustrated in blue.