Equity concerns across pediatric research recruitment: an analysis of research staff interviews

Abstract

Background and Objectives:

Difficulty recruiting individuals from minoritized and underserved populations for clinical research is well documented and has health equity implications. Previously, we reported findings from interviews with research staff about pediatric research recruitment processes. Respondents raised equity concerns related to recruitment and enrollment of participants from minoritized, low resourced, and underserved populations. We therefore decided to perform a secondary coding of the transcripts to examine equity-related issues systematically.

Methods:

We conducted a process of secondary coding and analysis of interviews with research staff involved in recruitment for pediatric clinical research. Through consensus we identified codes relevant to equity and developed a conceptual framework including five stages of research.

Results:

We analyzed 28 interviews and coded equity-related items. We report six implications of our findings. First, inequitable access to clinical care is an upstream barrier to research participation. Second, there is a need to increase research opportunities where underserved and underrepresented populations receive care. Third, increasing research team diversity can build trust with patients and families, but teams must ensure adequate support of all research team members. Fourth, issues related to consent processes raise institutional-level opportunities for improvement. Fifth, there are numerous study procedure-related barriers to participation. Sixth, our analysis illustrates that individuals who speak languages other than English face barriers across multiple stages.

Conclusion:

Research staff members identified equity-related concerns and recommended potential solutions across five stages of the research process, which may guide those endeavoring to improve research recruitment for pediatric patients from minoritized and underserved populations.

Introduction

Difficulty recruiting minoritized, low resourced, and non-English speaking populations for clinical research is a well-documented and longstanding health equity problem.^1–3 Most studies have shown that minoritized and low-socioeconomic status (SES) populations are under-represented within pediatric research. A recent review of 1183 pediatric clinical trials on ClinicalTrials.gov over a 13-year period found that Asian, Black, Native American, and Hispanic participants were under-represented compared to the US population.⁴ Within a multi-center pediatric critical care randomized controlled trial, parents of Black or Hispanic patients were less likely to consent to research compared to parents of white patients.⁵ Racial or ethnic minorities, those with a language of care other than English (LCOE), and those with low income were under-represented in a pediatric biorepository research study.⁶ Several pediatric oncology studies found lower enrollment of Black patients compared to white patients.⁷ Black infants were less likely to participate in Neonatal Research Network studies compared to white or Hispanic infants.⁸ In a survey of parents with infants in the neonatal intensive care unit, those who reported receiving Medicaid, reported lower income, or identified as Black were less likely to enroll their infant in a clinical trial.⁹ A review of human subjects research published within 3 major pediatric journals from 2012–2021 found that less than 10% included non-English speaking participants.¹⁰

There are many potential reasons for the limited inclusion of diverse and representative populations within pediatric research. Prior work has described barriers to research participation among diverse populations of adults. UyBico and colleagues described three levels of barriers to research participation: institutional, researcher, and individual (potential participant).¹¹ Similarly, Schmotzer and colleagues categorized health care barriers, patient barriers, and facilitators to research participation.¹² To our knowledge, such categorization of barriers for diverse populations within pediatric research has not been delineated. Within pediatric populations, commonly identified themes that may lead to differences in inclusion in research include: trust,¹³ relationship-building,¹⁴ burdens of participation,¹⁵ and understanding.¹⁶

Previously, we interviewed pediatric research staff about relationship-building with potential participants and their families.¹⁷ Research team members frequently raised equity concerns related to recruitment approaches and enrollment of participants from underserved and under-represented backgrounds. Hence, we identified a need to describe, at a granular level, the barriers for pediatric research participation in these populations. To better understand considerations for equity throughout recruitment, we performed a secondary analysis of relevant interview data.

Methods

Overview

Previously we published our findings from a study that conducted individual interviews with pediatric research staff involved in recruitment for clinical research to understand and describe how these staff build trusting relationships with research participants and their families.¹⁷ Respondents raised equity concerns related to recruitment and enrollment of participants from minoritized, low resourced, and underserved populations. To better define these issues, our team performed a secondary coding and analysis of existing transcripts using an inductively derived equity framework, which serves as the foundation for the current manuscript. The study was approved by the Institutional Review Board (IRB) at Seattle Children’s Hospital.

Recruitment

Research staff from a single pediatric research institution and affiliated academic medical center were eligible to participate if they were involved in the recruitment, consent, or enrollment process for pediatric research.

Interviews and Initial Analysis

We developed our interview guide based on a framework for researcher-community partnerships¹⁸ and created the initial codebook using inductive and deductive techniques. We made minor revisions to the interview guide after two pilot interviews. One team member (KMP) conducted all interviews between Nov 2020 and Feb 2021. Interviews could last up to 60 minutes. We continued until thematic saturation was reached. Recordings were transcribed, de-identified, and confirmed for accuracy. Dedoose¹⁹ was used for data management and analysis of transcripts. Initial codebook creation, coding by two independent team members, and our process for reaching consensus have been previously described.¹⁷

Secondary Coding and Analysis

During the analysis and interpretation process for the prior manuscript, our team found that many codes explicitly or implicitly highlighted equity and inclusion concerns. For the current study, we therefore decided to perform a secondary coding of all transcripts to examine these concerns systematically. Two coders led the team to identify five temporal stages across the research process at which equity concerns arose (Table 2).

Table 2.

Five identified stages during which inequitable recruitment and enrollment of pediatric research participants may arise

Stages and components

Access to clinical care Opportunities for research Prior relationship with clinical team/institution Geographic location of families Additional research team cost and research staff time to recruit LCOE families Relationship building with research team Perception that it takes additional work and time to build relationships with prospective participants and their families Acknowledgement of historical and contemporary instances of abuse and oppression Study team make-up Staff and participant race or ethnicity discordance Bilingual research staff are an advantage for research team but may present inequitable workload for the individual Informed consent process High medical literacy level of consent documents Limited availability of interpreters Translation of research consent documents is resource-intensive Differential burdens of study procedures Travel to medical sites (e.g., clinic or hospital); time away from work; childcare obligations Incentives and reimbursements Limited availability of validated measures in LCOE

LCOE (Language of care other than English)

We then coded equity-related excerpts that appeared across thematic codes applied in the primary analysis according to stage(s) of research. To do this, we identified 14 codes with equity-related themes, and then reviewed all excerpts to determine equity relevance and stage(s) of operation. This determination was verified by two coders. Excerpts were grouped by their code, which were summarized to facilitate identification of emergent themes.

Results

Participant characteristics

We conducted twenty-eight interviews, representing 72% (28/39) of those who either responded to listserv emails (N=21) or were contacted directly (N=18). Reasons for non-participation were: 5 did not respond; 4 did not meet inclusion criteria; and 2 declined. Participants were largely white, young, female, and highly educated (Table 1).

Table 1.

Participant demographics (N=28)

Characteristic	N (%)
Median years of research experience	5 (range 1–29)
Median age in years	28 (range 22–50)
Gender
Female	21 (75%)
Male	6 (21%)
Non-binary	1 (4%)
Race and/or ethnicity
White, not Hispanic or Latino/a/x	17 (61%)
Asian	6 (21%)
Hispanic or Latino/a/x, no racial category identified	3 (11%)
2+ racial or ethnic categories identified	2 (7%)
Highest degree earned
Bachelors	19 (68%)
Masters	6 (21%)
Graduate/PhD	3 (11%)
Primary research setting
Specialty clinic	10 (36%)
Social/behavioral	7 (25%)
Clinical research core	5 (18%)
Emergency department	3 (11%)
Neonatal or pediatric intensive care	3 (11%)

Items related to equity were identified at five stages, roughly mapping temporally through the research process (see Table 2). These included both concerns and practical solutions identified by respondents. Exemplar quotes are shown in Tables 3–7.

Table 3.

Stage 1: access to clinical team

Contributor	Exemplar quotes
Challenges to get into a clinic in the first place	“There are many stages before a family would be referred to us. It’s about them getting to the clinic, the hospital and seeking the care, all those other factors, I think, and many of those factors, we do not have any control over.” (Specialty clinic)
Specialty care clinics are more likely to be accessed by well-resourced families	“I think people who have more standardized health care and have PCPs that they see regularly and get vaccinations are more likely to notice when a kid’s muscles aren’t responding the same way and maybe refer them to a specialist, and if you don’t have a regular PCP or if you’re transient or something, that’s not gonna get picked up on, and so I think there are always disparities in health conditions in pediatrics, a lot.” (Specialty clinic)
Need for resources to address underlying clinical disparities	“We basically don’t let access barriers get in the way of a patient receiving treatment. We work with insurance companies (insurance processing does at the hospital) so that patients are financially cleared before they come and we make the insurance company sign a contract saying that they will pay for all the services that will come along with the trial, and that if a family can’t afford to fly out, then we have a social services specialist and a social worker on our… team who will get funds to reimburse to pay for flights, lodging.” (Specialty clinic)

Table 7.

Stage 5: study procedures

Contributor	Exemplar quotes
Need to be available during daytime hours for research clinic visits	“You have to have extra time in order to do things. Honestly, most of the time you have to take time off of work. You have to be a stay-at-home mom.” (Clinical research core) “It’s very socioeconomic. Unfortunately, families who have hourly employment, it’s very hard for them to take off to do research stuff without full compensation, or to match or beat what they would normally make in a day for their follow-up. I would just say that the lines are pretty much income-based.” (Intensive care)
Childcare obligations	“Communities of color are more likely to be of a lower socioeconomic status and may not have as much reliable childcare options as wealthier communities do, and at the end of the day, we need parents to have childcare in order to come in, if they have other children, and that’s come up a lot. … It’s safe to say that money definitely was a part of it, and socioeconomic status was definitely a part of why they couldn’t come in. [They couldn’t afford to] take a day off of work, or get a babysitter, or their husband or their partner wasn’t available to take care of the other children because they had to go to work.” (Specialty clinic)
Long travel to research site, extra challenges for rural families	“I think it’s always challenging to recruit rural patients, such a far distance and always a commitment. Even if they’re staying extra for a visit that they already have, if they have to drive four hours, that’s a super-long day, and not every study can provide reimbursements for housing or lodging and that sort of thing.” (Specialty clinic)
Local travel reimbursement modeled on well-resourced	“Parking is reimbursed, but a lot of times travel fees aren’t reimbursed and it’s different, if you have a car and are able to [drive], versus you have to take the bus and with your infant, that’s a big thing.” (Specialty clinic)
Need for stable mobile phone, internet access	“Making sure they have a smartphone that’s able to receive text messages, or at least a phone number that they’re able to talk with, or follow up with a different researcher after visit. If it’s kind of an inclusion criteria to be in the study and continue, we just follow through with that. We don’t really try to break any of those barriers. If they don’t have internet access, it may be hard for us to send a follow-up survey to them.” (Emergency department)
Degree to which team is measuring diversity	“We have an ethics, diversity and inclusion group that meets regularly, once a month, and <Institution> took the lead and they had submitted or are working to submit an IRB of their own site and they were gonna send that out to everybody else so we could then do that here. I think what that would entail is looking at some very simple metrics that we can pull out of Epic using some tool called “SlicerDicer,” which I’m very eager to learn about how that works. I don’t know that much about it, but I think it will be a robust way to capture some information to see where we’re at in terms of outreach and how are we doing in various zip codes and various race and ethnicity areas.” (Specialty clinic)
Modest reimbursement for time as felt particular by lower resourced families	“The folks who are people of color have a lot of barriers and, as a team, I don’t think we were well-equipped to handle the barriers. One of them is transportation, like I said, and we are trying to compensate for their transportation, but another thing is the survey is still about an hour and a half to two hours long. It’s a lot of time to ask of a parent, especially someone who might be lower [socioeconomic status] and requires time to support the family financially. That’s a lot of money, if you translate that into financial means, and we don’t compensate that very much. I think we pay $20 for the whole survey and the baseline visit and another activity that takes a half an hour. In total, you’re looking at three and four hours, and only $20 for that is pretty low.” (Social/behavioral) “The dad like one of my studies, he made $15 an hour as a roofer and every time we had a study visit, he had to take that day off and get his shift changed and make his boss mad and he didn’t get paid, and then the study would be like ‘Oh, can you come back in for labs? We can’t use a nearby lab. You have to come to <Research Hospital>.’ He didn’t have a car that worked, so he had to get a car service, and then he would have to pick his kid up from school. So the kid missed another day of school, so he’s farther behind in school. Had two other boys. They didn’t have a mom that lived with them, so who was watching their kids? And he would have to come up and watch them, or put them in the playroom, and then at a certain point, the study was asking so much, he was like ‘I’m not gonna make that,’ and the studies are like ‘Oh we can’t pay them. We can’t pay them. We don’t want to influence them. We can’t coerce them.’” (Specialty clinic)
Concern about information security	“[We] have to collect social security numbers [to give incentives] … which just I feel like can really rule out people who don’t have a social security number and people with varying immigration statuses.” (Social/behavioral) “Oftentimes, especially under the current administration in the U.S., people are not really comfortable giving their genetic information to NIH databases because they don’t really feel comfortable with the current government having access to that information because of the history of eugenics, and I have not had families say that outright, but it’s kind of like an underlying sort of discomfort that they mention, like, ‘Well with the government right now, I just don’t know.’” (Clinical research core)
Resources available for interpretation	“We’re also starting a new project that’s focusing on non-English-speaking families and seeing how they feel about the type of care and discharge instructions that they receive, so that would be using an interpreter with every family and including every type of language that we can get an interpreter for outside of English.” (Emergency department)
Lack of research tools being available in other languages	“All the standardized questionnaires are normed in English, so I think potentially families who might be more comfortable in a different language other than English to understand a lot of these things might be a little bit not comfortable participating in a project.” (Social/behavioral) “And then there’s the actual measurement tools that would also have to be validated in another language for them to be, I guess, accurate tools to use for manuscripts and things like that.” (Specialty clinic) “We form cohorts, so families are a part of a group, and for that reason, we’re requiring that everybody have a pretty good level of English. It doesn’t mean that their first language has to have been English, but they have to have a good level of speaking and reading, because we provide a lot of materials as well, so it’s important that they speak English.”. (Specialty clinic)
Suggestion: community engagement as potential solution	“We’re trying to make the process as flexible and equitable as possible. We also hold stakeholder groups every three months to have them give us feedback on the process to see if this is too much burden on a parent, etc.” (Social/behavioral) “We are including members of the community in the research process, like developing the surveys and questionnaires. … The aim is to improve overall diversity in clinical trial recruitment in participants.” (Specialty clinic)
Suggestion: increased availability of research visits	“I think it would be amazing if we could accommodate more research visits over the weekend. Not because I want to work over the weekends, but I think it would certainly go a long way in promoting enrollment for families who are working families, can’t time off.” (Clinical research core)

Stage I. Access to clinical team (Table 3)

Some respondents noted that lower SES families, families from minoritized racial or ethnic groups, rural families, and those with less education, are under-represented in both clinical and research settings. Some patients are unable to access medical care at sites recruiting for clinical research, and thus lack opportunities for research invitations. As an exception, some particularly well-funded research studies were able to recruit patients with otherwise poor access to medical care.

Stage II. Opportunities for research (Table 4)

Table 4.

Stage 2: opportunities for research

Contributor	Exemplar quotes
Satellite clinics as being a missed potential for diverse populations	“At satellite clinics … all of those places where we … don’t really have a research presence … the link between what is clinical and what is research needs to be strengthened.” (Clinical research core) “In one situation, my colleague did take a ferry over to <Location> to meet someone at their home to collect a sample, whereas normally we ask people to come in to the clinic, and we’re trying to reach out to satellite clinics to have assessments done there so people can go to places closer to them.” (Social/behavioral)
Lower resourced families as not learning about research opportunities	“Families probably won’t even hear about a research project unless maybe they are enrolled into the participant pool or their school has a flyer posted from somewhere, or someone participated in research, word of mouth. Probably a wider reach would be a really good thing that a research institute could do.” (Social/behavioral)
Lower resourced families as not being able to be present for enrollment discussion	“In terms of this study that I’m currently on, … I think there tends to be more of a bias towards wealthy white families that can afford to come in right now. I think we’ve only seen two families of color since we’ve started visits back up in June, and we’ve had like 40 visits so far.” (Specialty clinic)
Potential bias if not specific inclusion criteria	“One of the things that we do is try to have eligibility criteria so at the end of the day we’re not like ‘Oh, was that a decision that I made because I think it was my opinion, versus just like a fact.’” (Social/behavioral)
Increased time, effort, and money to make non-English speaking families eligible	“I know that it can be quite a process just for being able to enroll Spanish-speaking families because first you need to have your IRB and all the paperwork approved in English, and then you have to go out (and it adds to the budget as well) getting the translation done for all these documents, and it’s not just one. You’ll have a protocol and the invitation letters that might apply and whatnot. You have all that translated, which takes time and effort, and then on top of that, once you do have the paperwork, there’s that added barrier where if you don’t have a certified bilingual-speaking coordinator, then they have to consider that it may take more time to recruit. And it’s not just Spanish-speaking families, but non-English-speaking families. You’ll have to add additional time to the recruitment and enrollment process and any follow-up visits, if applicable, just because you will need that translator present, and that can be difficult.” (Clinical research core)
Increased time for research team member for recruitment discussion in other languages	“We have made a conscious effort to not prioritize English-speakers over non-English-speakers because the [limited English proficiency] families oftentimes take longer to enroll with getting the interpreter on the line, and we’ve identified that as kind of a pitfall we subconsciously have made in the past, so now we take extra effort to really approach those families, even if it means missing other English-speaking families that could’ve been quicker.” (Emergency department)

Beyond the site of care, respondents observed that low resourced families may be less likely to be contacted by study teams for research participation, for example, due to a lack of a reliable phone number or email address. One researcher described a study that recruited exclusively at private schools because their data were shared more readily than from public schools, which may lead to results that are not generalizable. Respondents also noted potential inequities stemming from families’ geographic location, highlighting satellite clinics, which may have more diverse populations but less research presence than the main hospital, as missed opportunities. They also noted challenges connecting with families in rural areas far from a research center.

Respondents noted several factors that created barriers to research participation for families with a LCOE, including increased effort and costs. Additional time needed for enrollment when using an interpreter resulted in the de-prioritization of this patient population. Others noted that recruiting families with cognitive disabilities or sign language interpretation needs added substantial time to recruitment.

Potential solutions relevant to this stage included the creation of clear inclusion criteria. For example, eliminating “soft” exclusion criteria in screening of potential research participants, such as “family unlikely to be available for follow-up,” could decrease bias. One respondent shared that their team actively prioritized non-English speaking families during recruitment in order to build a more diverse research sample, despite increased time and resources required.

Stage III. Relationship building with research team (Table 5)

Table 5.

Stage 3: relationship building

Contributor	Exemplar quotes
Historical mistrust of research—recommend addressing head on	“Certain groups are more wary of research, and rightfully so, [the study PI has been] encouraging us to address those issues head on and acknowledge that those are very valid feelings and that this is just an opportunity that they have full control over, if they would like to participate in that.” (Intensive care)
Historical mistrust of research—addressing head on feels like too much	“I guess I would go even further with the pressure off thing, if they were expressing discomfort and mistrust of research in general, ‘cause I guess I would be willing to have a conversation about it, but that feels like a really big thing to unpack in one little research approach.” (Social/behavioral)
Using family members preferred names and pronouns	“Addressing patients as they like to be addressed, asking about preferences for how we contact them, like which methods and times we use to contact them, but the key is kind of the first thing I was getting at with how we’re communicating with them, giving them a lot of space and empathy.” (Specialty clinic) “In demonstrating respect, another thing is I really try to get people’s names right, and I make it a point of asking them how to spell their names, and then if it’s a difficult name, ‘How do you pronounce it?’ and I really try to practice the pronunciation with them.” (Specialty clinic)
Intersecting identities	“The other identities are not too relevant with my study, but folks who identify as LGBT, especially the T, how we’re trying to make the studies more equitable to them, or someone who identifies as gender queer. Where does that fall into the scope of a coordinator talking to them, or building a relationship with them, also ability.” (Social/behavioral)
Importance of adequate time and patience	“Well sometimes it’s different to recruit when the family is clearly from a different culture, just because they’re coming from a different context and knowledge base. … I think for me, it just comes down to being patient and making sure that I’m not rushing my approach in particular with those families because we’re not coming from the same place.” (Emergency department)
White research staff concern they may be pressuring parents	“I get so stressed about it. I’m just so worried about people feeling coerced into doing research. I mean a big thing that I think about is how I am a white researcher. I have blonde hair, fairly light skin and I just worry about racism, just like white supremacy in <Hospital> and how especially families of color see a white person approaching them for research, that they’re gonna feel more pressured into it and that feels also not totally right, because obviously I don’t want to assume that people can’t say what they want and make their own decisions, but I think I’m hyper aware, sometimes maybe too aware of being a white researcher, particularly. I try really hard to have there not be a power dynamic, but I am often worried about the power dynamic.” (Social/behavioral)
Benefit of diversity in research staff	“I really think that having a diverse group of people enrolling people that look like the population that you’re enrolling really helps.” (Social/behavioral)
Potential participants notice lack of diversity	“[The participant] was right to point them out that all of our interventionists were white, younger than she was, and here she was a single mom trying to make ends meet and she didn’t feel comfortable and included.” (Specialty clinic)
Culture connection (in addition to language)	“Part of the trust starts with when they see you. … Sometimes I do try to think about what I wear. I don’t know if maybe this is more of a personal thing that because of growing up as a Hispanic person, the idea is that you have to dress up to be seen as credible, maybe be educated. … Also acknowledging, at least from what I’ve observed is if I go to the clinic dressed up in a suit, families might be intimidated because families in clinics are living in a different space where they’re trying to make it through their day. Affording clothes that they consider nice is a whole different world sometimes.” (Social/behavioral)
Bilingual research staff as advantage to research team	“The woman that is my coworker, she’s fluent in Spanish and certified. It’s great, and so much better because I don’t have to use the interpreter, who doesn’t know the nuts and bolts of the study. It’s been so great to have her consent the Spanish-speaking patients.” (Emergency department) “Always talking to her through an interpreter, I feel like things get lost and it’s hard to feel like we have a rapport. … I don’t know if there’s trust exactly. … It’s just been a challenge in that way.” (Social/behavioral)
Bilingual research staff member as sometimes taking on burden	“There are individuals that speak two languages, they’re asked to do everything from recruiting individuals, but they are missed out on career development opportunities where they aren’t able to continue developing their skills, learning more about the data analysis and reaching those higher leadership areas of research.” (Social/behavioral)
Suggestion: creation of flier to show composition of team to eligible families	“We actually created this thing called “Meet the Team” flier that we now include in our recruitment envelopes. We have a little blurb about ourselves and a photo, and I think it’s awesome. It’s amazing. It shows families how diverse we are.” (Specialty clinic)

This stage included equity concerns that created barriers to relationship building between research team members and potential participants. Some respondents shared that building relationships with those described as coming from “vulnerable populations” or “from a different culture” takes additional work. They stressed the importance of allowing adequate time and having patience when interacting with people from cultures different from one’s own.

Many respondents mentioned learning relevant family members’ names, correct pronunciations, and correct pronouns as critical, yet simple, ways to support relationship building and demonstrate respect. Respondents acknowledged families’ intersecting identities (e.g., experience of queer Black parents) and identities that might not fit into researcher-created boxes. For example, presenting a survey to a gender queer parent that includes only binary gender options may limit relationship building.

Several respondents acknowledged that family members from minoritized racial groups may have lower levels of trust and comfort during recruitment due to historical abuses of these communities in medical research. They shared that there is no universal approach to addressing these issues. While some suggested that directly calling out the oppressive historical legacy could help build trust, others expressed skepticism about unpacking a multi-generational legacy of abuse during one research interaction—and instead chose not to address the subject. Other recommendations included responding to these concerns with humility, not taking negative responses personally, and listening more (and talking less).

Respondents often discussed study team composition, noting that many research teams do not reflect the diversity of the communities they serve. They shared that parents often notice this lack of diversity. Some white research staff members shared concerns that some minoritized families may feel pressured to enroll due to a perceived or historical racial power dynamic. Some research staff members reported an awareness that their “whiteness” could engender mistrust from populations who have been previously exploited.

Considering families with a LCOE, bilingual research staff were seen as instrumental to help reduce language, trust, and literacy barriers, and optimize research participation. However, if in short supply, these bilingual research staff members could be overburdened with interpretation tasks. If one team member is solely responsible for certain tasks in a language other than English, they could miss out on career development and promotion opportunities. Respondents urged that bilingual research staff be supported by their study teams and have the same career advancement opportunities as their peers.

Stage IV. Consent process (Table 6)

Table 6.

Stage 4: consent process

Contributor	Exemplar quotes
Medical literacy	“[A participant] had some trouble understanding the consent … and I was like ‘I really want to try to help assist you with this,’ and I felt like it was my duty, in an effort to be inclusive, spend more time with her and make sure she’s okay with everything, do a little more I guess you could say hand-holding than a lot of other participants on this study, but in an effort to show that we respect where she is.” (Social/behavioral)
Consent process as a way to promote family autonomy	“Families who are familiar with research, it feels like more of an equal footing, but families who are not familiar with research or maybe are lower English proficiency and speaking to them through an interpreter, those are situations where it’s extra important for me to make sure that I am being very clear in what I’m asking for them, and making sure that they understand that it’s completely separate from their clinical care, because so much goes on in the Emergency Department that it’s pretty easy for families to just be like ‘Oh she’s another person that I have to talk to and tell about my kid,’ when in reality, I am not important.” (Emergency department)
Translation of consent forms into other languages	“Most of our studies are only approved consent forms in either English, or English and Spanish, and so if families aren’t fluent in either of those languages, then they’re just not eligible.” (Intensive care)
English spoken fluency does not guarantee medical literacy	“Even families that are able to speak English and able to engage English might still feel more comfortable reading recruitment materials in other languages where they fully understand what things are” (Social/behavioral)
People with disabilities	“The only one other little group, there are kids that are eligible, but they often have cognitive impairments, or developmental delays, and I do think that’s another group that we often aren’t able to do research with this group, even though some of them want to talk. You have to change how you communicate, but we’ve had families that need sign language interpreters. I do think there’s a disability population that is also being missed.” (Specialty clinic)

Respondents raised equity concerns about the consent process. They described medical literacy as a persistent challenge, due to the advanced reading level of most consent forms. Research staff also noted an increase in concerns from lower resourced populations about the safety of medical research and resulting mistrust. Some advocated that, when done well, the informed consent process should serve to bolster an individual’s autonomy rights, especially for those wary of research.

Respondents noted that their research teams were generally able to translate documents from English to Spanish in accordance with IRB guidelines but had limited ability for other languages. Several noted that document translation is expensive, thus, not all studies may have funds to support it. Others commented that families may use English for medical care but still require translated documents for research. Some noted challenges confirming comprehension when preference for LCOE exists concurrently with limited general or medical literacy. It was also noted that COVID-related staffing shortages made it difficult to obtain in-person interpreters.

Stage V. Study procedures (Table 7)

Respondents noted that research participation differentially burdens families in terms of time and money. Equity concerns were raised related to: childcare obligations; time away from work and school; assumption of ability to pay costs upfront; and assumption of access to a personal vehicle. Travel to the research site was seen as an added challenge for rural families. Respondents also noted that travel costs include opportunity costs (e.g., missed work), creating additional burdens for those with hourly jobs, less job-related flexibility, and lower wages. Several respondents noted the need for reliable mobile phones that can receive calls or text messages for access to study related websites and surveys for data collection. Others shared a belief that privacy and information security was a greater concern for certain marginalized individuals, such as undocumented immigrants, who may have higher risk of legal and social harms resulting from privacy breaches. For example, undocumented families may choose not to enroll if social security numbers are required for incentive payments.

Respondents noted that, to fully include LCOE families, surveys and questionnaires should not only be translated but ideally be rigorously tested for psychometric properties similar to the English versions. Other challenges for families with LCOE included study related videos only available in English, studies that involve focus group components requiring fluency and literacy in English, and follow-up visits requiring added time and cost to have an interpreter present.

Community engagement and stakeholder feedback can help address inequities. For example, one respondent described a project in which children from the target community gave feedback on draft research documents. Other potential solutions to reduce inequities in burdens of participation include increased availability of weekend research visits and research protocols that seamlessly integrate translated study documents into data collection programs.

Discussion

Through our secondary coding and analysis of interviews with pediatric research staff, we identified equity considerations across five stages of clinical research (Table 2). We will discuss six implications of our findings. First, inequitable access to clinical care is an upstream barrier to research participation. Second, there is a critical need to increase research opportunities at locations where underserved and underrepresented populations receive care. Third, diversifying research teams can build trust with research participants and families; yet this must be carefully considered to ensure adequate support of all research team members. Fourth, the consent process raises institutional-level barriers and potential opportunities to address equity concerns. Fifth, research teams must address numerous study procedure-related barriers to participation. Sixth, patients and families who speak languages other than English face barriers to research participation that need to be addressed across multiple stages.

First, respondents discussed equity concerns arising from disparities in access to medical care, a prerequisite to subsequent clinical research participation. Marginalized populations are less likely to access subspecialty care, and most research takes place at institutions that provide this type of care, such as large, academic medical centers and affiliated clinics. Within pediatrics, disparities in referrals have been found among children with autism spectrum disorder,²⁰ growth failure,²¹ and developmental-behavioral disorders.²² Research coordinators we interviewed recognized this upstream problem as a significant contributor to lack of diversity and inclusion in clinical research. Work in genomics has shown that access to care and access to research can be interrelated.²³ Until access to quality medical care becomes more equitable, inequities in research access will continue. Thus, addressing disparities in access to clinical care may rightly be considered a problem for researchers and funders.

Second, respondents discussed inequities in the distribution of research opportunities. Referral to pediatric research by a trusted clinician may increase likelihood of participation.²⁴ Challenges of COVID-19 vaccine trials illustrate gaps between marginalized populations and researchers,²⁵ one of which had to slow recruitment because of failure to recruit adequate numbers of Black, Latino, and Native American participants.²⁶ Mechanisms through which communities have real input into study design and implementation may improve equity and increase access.²⁷ Academic medical centers may improve equity through partnership with community based clinical settings and other locations where diverse populations receive care. For example, the Environmental Influences on Child Health Outcomes Institutional Development Awards States Pediatric Clinical Trials Network is an NIH supported program that aims to increase clinical research participation among infants and children living in rural areas²⁸ and has resulted in major publications including infant²⁹ and pediatric³⁰ participants.

Third, respondents recognized the impact of existing inequities on their ability to build relationships with potential research participants. Relationships with clinical team³¹ and research team members¹⁷ are key factors in enrollment decision-making for parents considering research participation.³² Individuals from minoritized populations report that racial concordance in patient-provider relationships³³ likely benefits patient-researcher relationships.^34,35 Our current analysis highlights the importance of developing a diverse research team and training all research team members to support positive, respectful interactions with diverse patients and families. Several respondents noted challenges and risk of burnout for members of the research team who have a marginalized or minoritized identity. Individuals responsible for recruiting and collecting data from non-English speaking participants must have equivalent professional development opportunities as their peers and resources to avoid burnout.

Fourth, the process of informed consent raises equity concerns. Translating materials into a family’s language of care is burdensome when informed consent document language is complex as it often is. Framing the consent process as a way to empower families—especially those who may otherwise feel disempowered—emerged within several interviews. Notably, these issues may be best addressed at the institutional level. For example, research centers could streamline approaches to consent through a simplified “key information section”³⁶ and facilitate document translation centrally. Noting similar concerns, genomics researchers have developed and tested novel consent processes designed to be better aligned culturally and linguistically to minoritized populations.³⁷ Future work should consider how to improve the consent process in different settings and with different populations.

Fifth, research staff raised concerns regarding inequitable burdens of participation across populations. For example, required follow-up visits and travel for pediatric research participation³⁸ are likely felt unequally due to structural and societal factors that cause socioeconomic disparities across groups, such as access to transportation,³⁹ costs of childcare, and job flexibility. Burden for families is an often underappreciated barrier to participation in pediatric research.^32,40 Previous research has identified differential burdens of participation as contributing to under-representation of minority children in clinical trials for neurodevelopmental disorders.¹⁵ Barriers related to study procedures need to include fuller consideration of lower resourced families, such as people who cannot easily take time off work;⁴¹ individuals who are caregivers of other children or family members; travel time³⁸ and overnight stays in hotels; the need for timely reimbursements; and the opportunity cost of research visits.⁴⁰ While regulatory agencies have traditionally been concerned about the undue influence of incentives,⁴² more recently, many have argued that payments are acceptable recruitment tools⁴³ and even a means towards equity.⁴⁴ Future work must consider these issues within pediatrics, wherein a parent is paid for their child’s participation, and within the context of minoritized populations.

Finally, language barriers were noted across all stages, each requiring different responses. During study implementation (Stage 5), validated measures of tools are needed but require more resources (e.g., funding to validate tools in multiple languages). The need for translation of consent documents (Stage 4) is a mid-level problem (e.g., institutional resources for translation). Support of staff (Stage 3) conducting tasks in a LCOE is an on-the-ground problem. The upstream problems of access to clinical care (Stage 1) and access to research teams (Stage 2) may create additional barriers for those lacking English fluency. Importantly, barriers related to language will vary by institution. Translation and interpretation barriers may be low burden for some studies (e.g., drug studies pulling from EMR data), but almost insurmountable for others (e.g., studies with an educational intervention or group participation). Institutional level pooling of resources and prioritization of commonly spoken local languages may be helpful. In order to include families with a LCOE in pediatric research, the multi-level nature of these barriers must be addressed.

Further work in other clinical research contexts should evaluate whether considerations of our schema may be beneficial outside of pediatrics. For example, working adult children of older adults considering research participation may face similar burdens as parents of infants and children. Future work should examine the impact of and potential solutions to such barriers in these other settings.

Limitations

The generalizability of our analysis may be limited because this was a single-center study with participants who were largely white, young, female, and highly educated. Further, this study did not include other key stakeholders, specifically patients and families facing these barriers. Despite these limitations, we believe the findings provide an important conceptual framework for understanding barriers and potential solutions to equity concerns in pediatric research.

Conclusion

Secondary coding and analysis of interviews with pediatric research staff members identified equity-related concerns and potential solutions across each of five stages of research. Some challenges span multiple stages. Our findings support efforts to improve recruitment of those from minoritized and underserved populations for pediatric research. Future research should build on this conceptual framework by incorporating perspectives from other diverse, key stakeholders, including patients and families, investigators, and regulators in order to develop interventions that meaningfully address equity at multiple stages of the research process.

What’s New?

In this secondary analysis of interviews with pediatric clinical research staff, we identified themes related to equity across five stages of recruitment and strategies which can be used to improve recruitment for minoritized and underserved populations.

Abbreviations:

IRB

Institutional Review Board

LCOE

Language of care other than English

SES

Socioeconomic status