A pilot recruitment strategy to enhance ethical and equitable access to Covid-19 pediatric vaccine trials

Abstract

Background/Aims:

The SARS-CoV-2 pandemic disproportionately impacted communities with lower access to health care in the United States, particularly before vaccines were widely available. These same communities are often underrepresented in clinical trials. Efforts to ensure equitable enrollment of participants in trials related to treatment and prevention of Covid-19 can raise concerns about exploitation if communities with lower access to health care are targeted for recruitment.

Methods:

To enhance equity while avoiding exploitation, our site developed and implemented a three-part recruitment strategy for pediatric Covid-19 vaccine studies. First, we publicized a registry for potentially interested participants. Next, we applied public health community and social vulnerability indices to categorize the residence of families who had signed up for the registry into three levels to reflect the relative impact of the pandemic on their community: high, medium, and low. Finally, we preferentially offered study participation to interested families living in areas categorized by these indices as having high impact of the Covid-19 pandemic on their community.

Results:

This approach allowed us to meet goals for study recruitment based on public health metrics related to disease burden, which contributed to a racially diverse study population that mirrored the surrounding community demographics. While this three-part recruitment strategy improved representation of minoritized groups from areas heavily impacted by the Covid-19 pandemic, important limitations were identified that would benefit from further study.

Conclusion:

Future use of this approach to enhance equitable access to research while avoiding exploitation should test different methods to build trust and communicate with underserved communities more effectively.

Background/aims

Minoritized groups had some of the highest disease burden from Covid-19 in the United States,¹ where the pandemic disproportionately affected communities with lower access to health care. Vaccines for SARS-CoV-2 were developed in record time and provide impressive protection against severe Covid-19 disease.^2,3 As early vaccine trials in adults struggled to enroll diverse participants, one trial paused enrollment to increase focus on recruitment of minoritized groups.⁴ Ultimately, about 80% of initial adult messenger RNA (mRNA) vaccine trial participants were White, despite 71% of the US population identifying as White, alone or in combination with another group.⁵

The principle of justice in biomedical research requires using “fair procedures and outcomes in the selection of research subjects.”⁶ Selecting participants fairly requires both avoiding the exploitation of groups who have been historically or would be unfairly burdened by research participation, while assuring opportunities to enroll in research that may ultimately lead to potential benefits for all.⁷ One way to enhance equitable participation in research is to offer preferential participation in studies that may provide benefit to communities with the largest burden of disease.

Applying these principles to Covid-19 vaccine trials, our tertiary care pediatric medical center piloted a recruitment approach to preferentially include groups bearing a greater burden from the pandemic. By participating in Covid-19 vaccine trials, communities with high impact of the pandemic could benefit from the chance to receive vaccines with evidence of safety and promising efficacy to provide protection from disease. In anticipation of these studies, our research ethics consultation service helped develop a three-part strategy to enhance equitable recruitment. We describe our experience with this approach and demonstrate the resultant participant enrollment from minoritized communities and underrepresented populations.

Methods

We devised and evaluated a three-part recruitment strategy to enroll participants from communities impacted more severely by the pandemic. This involved (1) creation and publicization of a registry of families interested in Covid-19 vaccine trial participation for their children, (2) application of public health community and social vulnerability indices to location of the registrant’s residence, and (3) prioritization of recruitment of residents from communities experiencing high impact from Covid-19.

Registry creation

Prior to launching SARS-CoV-2 vaccine trials for children at our center, we developed a registry in REDCap⁸ in English and Spanish. Caregivers (parents and guardians) interested in having their children participate in Covid-19 vaccine trials entered limited information about themselves and their children (Table 1) after virtually signing a consent form in their preferred language (Supplementary Appendix 1). Demographic data were used to assess the effect of this strategy on recruitment. The registry and materials publicizing it were approved by our institution’s IRB prior to their use. We informed parents of the registry using methods including social media posts and notifications on the hospital’s publicly facing website (Supplementary Appendix 2). Internally, emails were sent to employees. We presented information to practice leaders from our Clinically Integrated Network and provided fliers with QR codes for posting in offices. In addition, local English and Spanish language television and radio news outlets included information about the registry as part of local coverage of vaccine development.

Table 1.

Registry questionnaire.

SARS-CoV-2 Pediatric Vaccine Trial Registry

Parent/legally authorized representative information

First Name:

Last Name:

Email Address:

Telephone number:

Household zip code:

How would you prefer to be contacted to assess if your child is eligible for CoV-2 vaccine studies?

Email

Telephone

Would you be willing to come to our location for the study (<addresses listed>)?

Yes, I would drive my own vehicle (note the study would pay for parking)

Yes, but I would need transportation (would be arranged and paid for by the study)

No, I would only participate if there was an option close to where I live

Preferred transportation:

Rideshare (Uber, Lyft)

Van service

Public transportation (CTA)

Are you interested in being contacted about other pediatric clinical trials?

Yes

No

Can we contact you to provide updated information annually?

Yes

No

Number of children potentially interested in participating:

Child race/ethnicity? (select all that apply)

Asian

Black Latinx

Black non-Latinx

Caucasian Latinx

Caucasian non-Latinx

Native American

Other

Prefer not to answer

Child date of birth?

(Month-Day-Year format)

Has your child previously had Covid-19?

Yes

No

Maybe, but not tested

If maybe,

Suspected based on symptoms

Exposed to person with known infection

If yes,

Suspected based on symptoms and known contact with infected person

Positive test

Approximate date of infection? (Month-Year format)

Has your child ever needed medicine other than Tylenol (acetaminophen) or Motrin (ibuprofen) because of a reaction to a vaccine?

Please check the box if your child has ever been diagnosed with any of these medical conditions. These do not necessarily exclude your child from participation.

Cancer

Immune deficiency including HIV infection

Bleeding problem (including hemophilia or idiopathic thrombocytopenic purpura)

Inflammatory bowel disease (including Crohn’s disease, ulcerative colitis)

Liver disease (including infection with Hepatitis B or C)

Autoimmune disease such as lupus (SLE or systemic lupus erythematosus), juvenile rheumatoid arthritis (JIA), psoriasis, type 1 diabetes

None

SARS-CoV-2: severe acute respiratory syndrome coronavirus 2.

Incorporation of public health community and social vulnerability indices into trial recruitment

We employed two public health community and social vulnerability indices to stratify the Covid-19 impact into three categories: low, medium, and high. For registrants living in the city of Chicago, we applied the Community Covid-19 Vulnerability Index (CCVI), created by the Chicago Department of Public Health (CDPH)⁹ to identify “communities that have been disproportionately impacted by Covid-19 and are uniquely vulnerable to barriers to Covid-19 vaccine uptake.” The CCVI combines sociodemographic, epidemiological, and occupational factors with the cumulative burden of Covid-19 disease for different communities within Chicago to derive a composite weighted score.¹⁰ CDPH used the scores to create three levels of vulnerability: low (mean: 21.0, range: 4–31.3), medium (mean: 41.1, range: 31.7–47.4), and high (mean: 53.2, range: 47.9–64).

A similar index, created by the US Centers for Disease Control and Prevention (CDC), was applied to addresses located outside of Chicago. The Social Vulnerability Index (SVI), which is not specific to the Covid-19 pandemic, incorporates census data to calculate a numerical value of overall vulnerability by county that ranges from 0 (lowest vulnerability) to 1 (highest vulnerability);¹¹ CDC describes the SVI as identifying communities with less ability to prevent human suffering and financial loss in a disaster based on factors including poverty, lack of access to transportation, and crowded housing.¹² We divided the 86 counties of residence associated with ZIP codes of registrants into terciles and then calculated mean (SD) overall vulnerability scores for each group as low (mean: 0.13, SD: 0.09), medium (mean: 0.39, SD: 0.07), and high (mean: 0.65, SD: 0.10).

Registrant address information was used to assign each child as living in a community with low, medium, or high vulnerability score corresponding to the CCVI for registrant addresses within Chicago and SVI for resident addresses outside of Chicago.

Prioritization of registrants

Since there were many more registrants than available slots in the trial, we randomly sorted the entire list prior to applying the CCVI/SVI vulnerability category to the offer of participation. This mitigated any advantage to those who signed up first, who were likely to be more well-connected and better-resourced.¹³ We also filtered our registry to select participants residing in the Chicagoland area (within about a 2-h drive of the study site), reflecting both our hospital mission to serve local children and to maximize participant retention, an important metric for achieving study scientific endpoints. Finally, we preferentially offered participation to our HIV clinic population (due to their higher risk from Covid-19 and study team access to this population) and to those with another child already participating in the trial, primarily for the purposes of family convenience and improved study retention. This last group comprised a small minority of our participants but were selected independent of their participation in the registry, though they were included in the data analyzed in the “Results.”

Invitations to registrants

After identification of priority populations and categorization of community-level vulnerability, we offered participation in a ratio of 1:2:3 for low, medium, and high areas. This ratio was chosen based on the relative distribution of mean CCVI scores between low, medium, and high scoring areas (i.e. 21.0:41.1:53.2 ≈ a distribution of 1:2:3). Applying this method to the first 50 families invited, 9 (18%) were chosen from registrants living in a low vulnerability area, 16 (32%) from a medium vulnerability area, and 25 (50%) from a high vulnerability area. We found that those in high vulnerability areas were less likely to enroll (Figure 1(a)); subsequent offers for participation prioritized high vulnerability areas. Importantly, race and ethnicity were not explicitly factored into offers of trial participation, but these data were collected from the registry.

Figure 1.

(a) Percent of 6- to 12-year-old participants with addresses in low, medium, or high vulnerability areas (based on CCVI or SVI data) who were invited (orange) and enrolled (gray) to the trial, compared with the target percentage (blue bars). Target percentages reflected the distribution of vulnerability indices that was targeted based on the rationale provided in the “Methods” section (1:2:3 for low, medium, and high areas, respectively, which corresponds to 16.6%:33.3%:50%). (b) Percent of 6- to 12-year-old participants in different demographic categories (based on parental report to the registry) who were invited (orange) and enrolled (gray) to the trial, compared with Chicago census data (blue bars). Those in the “Other” category included mixed race, non-Latinx children: Black/Caucasian (11% of those invited and 7% of those enrolled), Asian/Black (0.7% of those invited and 2% of those enrolled), and Asian/Caucasian (4% of those invited and 5% of those enrolled).

Results

The SARS-CoV-2 Pediatric Vaccine Trial Registry opened in February 2021 in anticipation of our site being selected for one or more Covid-19 pediatric vaccine trials. By September 2021, more than 10,000 parents and caregivers had registered interest in having one or more children participate in a trial.

Children between 6 and 12 years of age were the first group to enroll to Covid-19 vaccine trials at our site and are the focus of this study. This trial involved a placebo-controlled, double-blind, randomized phase 2/3 study of an mRNA vaccine that was authorized for use in adults (over age 18). From July through August 2021, we invited 151 six- to 12-year-old children to participate from 107 families, leading to enrollment of 56 children from 36 families. Figure 1(a) shows the distribution of participants from low, medium, and high vulnerability areas, compared with the recruitment strategy target level. More children from low and high vulnerability areas were both invited to participate and enrolled to the trial, relative to the target distribution. This is partly related to the adjustment to prioritize high vulnerability area invitations, made after the initial 50 invitations were sent. A greater percentage of those from low and medium vulnerability areas enrolled relative to the percent invited, while a smaller percentage of those from high vulnerability areas enrolled relative to the percent invited (Figure 1).

Figure 1(b) shows the racial and ethnic distribution of those invited to participate and those enrolled to the trial, compared to Chicago census data demographics. Although we did not apply racial/ethnic targets to recruitment, the use of geographic vulnerability levels led to a diverse representation of participants within the trial, acknowledging that there are significant overlaps between area of residence and race/ethnicity. For most groups, the percentage of participants enrolled was comparable to or higher than the percent invited; however, a greater percentage of Black, non-Latinx participants were in the invited group compared to the percent which ultimately enrolled.

Discussion

Our three-part recruitment strategy used public health community and social vulnerability indices to enhance the equitable recruitment of participants in a clinical trial of Covid-19 vaccines in children while avoiding concerns about or a perception of exploitation. By relying on caregivers to first indicate their interest in the study by joining the registry, the recruitment approach simply prioritized people who already valued participation, minimizing the risk of the recruitment efforts being perceived as taking unfair advantage of vulnerable groups. Overall, this approach resulted in successful enrollment of participants from high vulnerability areas. The racial and ethnic distribution at our site included higher enrollment of underrepresented populations than that reported in published studies of Covid-19 vaccines in these age groups (Table 2), though we do not have any information on strategies used at other study sites.^14,15

Table 2.

Comparison of enrollment using recruitment strategy accounting for geographic vulnerability with published studies of Covid-19 vaccines in the 5- to 12-year-old age group.

Demographic group	Percent of participants
	Current study	Walter et al.14	Creech et al.15
Asian	0.2	6.0	9.9
Black, non-Latinx	30.4	6.5a	10.0a
Caucasian, non-Latinx	21.4	78.9b	51.9
Latinx	33.9	21.1	18.5
Other (including multiracial)	14.3	8.6	14.4

^aStudy did not separately report ethnicity for Black non-Latinx versus Black Latinx.

^bStudy did not separately report ethnicity for White non-Latinx versus White Latinx.

Enrollment of diverse groups into clinical trials remains an ongoing challenge,^16,17 with potential reasons including lower access to clinical trials and lower rates of participation even when access is improved.¹⁸ The use of location of residence to generate a population-level metric, rather than a specific focus on racial/ethnic groups for participation, avoids directly conflating race/ethnicity with the structural and social factors that result in concentrated populations with increased vulnerability to many morbidities, including the chronic conditions that complicate Covid-19 disease. It must be acknowledged that many cities have a long history of systematic geographic racial segregation, and the vulnerability indices applied incorporate racial and ethnic minority status into their metrics. However, the strategy employed may help alleviate ethical concerns about exploitation along with legal concerns that could arise if underrepresented populations are specifically recruited solely based on race and ethnicity.¹⁹ This latter concern has been the focus of recent controversy regarding consideration of race and ethnicity in other settings involving prioritizing access to a limited resource (college admission).²⁰

We were able to draw from a large registry which pre-selected potential participants based on their self-identified interest in pediatric Covid-19 vaccine trials prior to any vulnerability-based study invitations, and then to prioritize invitations based on community vulnerability level. Creating registries of families interested in clinical trial participation and offering enrollment preferentially based on criteria including community-level vulnerability to negative health effects or overall disease burden may be helpful for other types of clinical trials. The effect of the pandemic on underrepresented populations had been well-documented by the time our registry opened, potentially influencing individuals to consider participation in a trial more strongly than they otherwise might. Indeed, a caregiver for one enrolled participant specifically noted that they were motivated to enroll to be an example to their friends and family that Covid-19 vaccination was very important for the Latinx community they lived in to take seriously. Trials of other therapeutic or preventive interventions with less urgency or publicity might not result in comparable interest. However, it should be noted that this approach did come with additional effort; the registry needed to be created and publicized, and applying the vulnerability indices to the registry and sorting offers of participation on that basis also required time and effort. Without the lead time we had in knowing these trials would become available, it could be more difficult to generalize this approach to other types of trials.

This approach has limitations. First, although CCVI values were generated for Chicago communities, we aligned these data with ZIP codes, which may span communities at different risk levels. CDC data for SVI were generated by county and applied at the ZIP code level. In addition, CCVI and SVI are calculated differently. Second, there will be inherent variation in the degree to which ZIP code–level vulnerability represents the vulnerability of an individual who lives within it. We cannot determine the extent to which this misclassification bias impacts our results because we did not collect data on each participant’s social vulnerability. Third, although the observed racial and ethnic diversity in enrollment is reassuring that our strategy was somewhat successful, we noted that parents working in health care settings were overrepresented relative to the general population.

Conclusion

Applying public health metrics for community vulnerability to recruitment to a pediatric Covid-19 vaccine trial allowed us to meet pre-specified goals for study recruitment and enrollment from high-risk communities. This indirectly led to a racially and ethnically diverse study population that mirrored the diversity of the surrounding community. Whether this approach can be applied to clinical trials for different therapeutic and prophylactic interventions and whether it can be refined to result in a more representative population are topics for further study.

Supplementary Material

Supplemental material for this article is available online.