Research objective

This project explored research participants’ and their parents or legal guardian’s understanding of the Spanish-language translation of the University of California at Berkeley (UCB) template IC for biomedical research that includes broad consent for clinical-epidemiological and genetic data and sample sharing and the Moore Clause [13]. The Moore Clause, which informs research participants that they will not have the right to claim compensation for commercial products or other intellectual property developed using their biospecimens from human subjects research [14], is related to a lawsuit brought by a California research participant whose biopsy tissue was used without his knowledge for commercial purposes. We included participants from a dengue cohort site in Nicaragua, which has a long-standing collaboration with UCB researchers, and from a dengue cohort in Colombia to compare the understanding of and attitudes toward the template language across the two contexts.

Study population & recruitment

Participants were adults (aged 18 and over) who could provide IC and were the parent or guardian of a child who participated in an ongoing community- or hospital-based dengue virus-related cohort study (N = 42) or, in the case of Colombia, themselves cohort study participants (N = 10). In Nicaragua, we could not enroll cohort participants because they were younger at enrolment into the source cohort study, and the source cohort was launched later than in Colombia. Hence, cohort participants were younger than 18 at the time of this study. In Nicaragua, we excluded parents or guardians of participants younger than six months old or whose child weighed less than eight kilos at the time of enrolment and participants that lived outside the State of Managua from the study because these participants had higher levels of follow-up than the general cohort or could not be reached for the in-person interview, respectively.

In Colombia, we recruited participants from the list of parents and participants in the cohort study "Identification of priority age groups to be vaccinated against dengue in children and adolescents," which began in 2012 and included participants aged 16–25. In Nicaragua, we recruited participants from a current list of parents of participants in the "Prospective hospital-based study of dengue classification, case management, and diagnosis in Nicaragua" study, assembled between 2015–2018. Inclusion criteria for the Colombia [18] and Nicaragua [19] source cohort studies are described elsewhere. To capture the diversity in how cohort participants perceive broad consent language in the IC, we grouped each country-specific randomly ordered list of participants by education level (completed versus did not complete primary school) and gender. Participants were then selected from one of the four groupings at each site (e.g., male, completed primary school) by going sequentially through the previously randomized grouping and selecting one participant from each group for that round. Few men or participants with lower than primary education were included in the lists, so participants were not evenly distributed across these four groups.

Given the ongoing nature of the source cohort studies, cohort staff communicated regularly with participants. A field interviewer from each study approached potential participants during a regularly scheduled study visit or follow-up call and described the project using the recruitment script. Participants that consented to learn more about the cognitive interview study were then contacted by a study team member. The interviews occurred between August 2021 and April 2022 during the COVID-19 pandemic. Contacts took place over the phone in Colombia and in person in Nicaragua in keeping with local directives around maintaining distance and ventilation. Because the assessment of participant understanding of IC was not considered sensitive, we secured verbal IC in Colombia and written IC in Nicaragua. When parents or participants refused participation, we moved to the next parent, legal guardian, or research participant on the list for that group (e.g., completed primary education, female).

Cognitive interview

We used verbal probing and think-aloud methods in the cognitive interviews [16, 20]. Verbal probing was used to clarify participants’ understanding of complex concepts, including genetic data, de-identified data, and data and sample sharing. The think-aloud method was used to understand the respondents’ thought or decision-making process and how or why participants decided to accept or reject broad consent for future use. The classical approach to cognitive interviewing, developed by Tourangeau [21], includes how participants (1) understand the question and answers, (2) retrieve information (recall), (3) weigh different answers (estimation or judgment), and (4) arrive at their decision (response). Because we were using text from the informed consent rather than a questionnaire, participants’ recall and response processes were irrelevant to our analysis and we used a modified version of the Tourangeau model limited to comprehension and judgment.

Before beginning the pilot interviews, interviewers met for several weeks to practice the pre-interview "think aloud" exercise and refine the interview guide. We evaluated how participants understood certain words and phrases and entire interview sections. We used probes like "tell me more about that" to explore different interpretations of the IC. After participants explained what came to mind, we used their language to describe what we saw as the intent behind the related text (e.g., data sharing, genetic data, incidental findings) and asked participants to then describe what they then understood. We iterated this process until the participant could explain in their own words what we understood as the intended meaning of the broad consent-related language. We conducted the cognitive and semi-structured interviews in Spanish. In Nicaragua, interviews were conducted face-to-face and recorded on a digital recorder. In Colombia, interviews were conducted and recorded by smartphone.

Analysis

Interview recordings were transcribed verbatim, and the de-identified transcripts were uploaded to MAXQDA [22] for analysis by the study team who are fluent in Spanish. Interviews were conducted, transcribed, and analyzed on a rolling basis to allow us to identify emerging interpretations and themes. At least two team members independently reviewed all transcripts, and all five team members reviewed and evaluated a subset of transcripts.

We used thematic analysis based on Colaizzi’s model of descriptive phenomenology [23, 24] to explore participants’ attitudes toward data and sample sharing. During the first phase of the analysis, we developed preliminary deductive codes and definitions based on (1) the interview guide; (2) literature identified through a systematic search [25]; and (3) our own experience as researchers in public health research-related data and sample sharing. We developed inductive codes by reviewing interviewer field notes, written transcripts, and analytic memos. In the initial stage, the team read transcripts several times so that we could familiarise ourselves with the data and jointly develop the codebook. Following the joint analysis stage, to ensure reliability, JBC and LML met weekly to resolve differences in the interpretation of cognitive interview findings and the application of codes and identification of emergent inductive codes grounded in the transcript data. During the same period, the entire team held weekly discussions to update and refine the codebook, clarify local expressions with the help of the field interviewers (JBC, HSL), and discuss emerging themes. During the second phase of the analysis, the team met weekly to group the interpretations of the text and codes into meaningful themes and discuss whether themes varied across research site, gender, level of education, and cohort participants or their parents and guardians. We resolved differences of opinion in interpreting the cognitive interview and thematic analysis findings through consensus. Codes are presented in S2 Table.

Ethics statement

The study protocol and forms were approved by the Centro de Atención y Diagnóstico de Enfermedades Infecciosas (INFOVIDA; Bucaramanga, Colombia), Instituto de Ciencias Sostenibles (Managua, Nicaragua), and Universitätsklinikum Heidelberg (Heidelberg, Germany) ERCs. Further ethics considerations are reported in the PLOS Inclusivity in Global Research Questionnaire (S3 Text).

Genetic data

The terms "genetic studies," "genetic data," "DNA," and "RNA" were not well understood. Participants described "genetic studies" as cloning or paternity detection when asked to put these terms in their own words. Thirty-one of the 44 participants across the two sites had trouble explaining what a genetic study referred to.

"Genetic studies? Are they going to look for genes from other people in my son’s blood? That’s more or less what I understand."

(P6-ASU, Colombia, 35–45 years old, female, respondent’s parent)

"Genetic studies are the samples they ask for the parents to know if the child is theirs or not."

(N20-MSMR, Nicaragua, over 45 years old, female, respondent’s parent)

Participants generally understood the reference to commercial products in the IC and that the research results could become the basis for new commercial products but expressed concerns about benefit sharing, including access to medications and vaccines developed through data and sample reuse. The phrase "diagnostic tests or therapeutic agents" was not well understood.

"Well, the diagnostic tests are those of the laboratory, right? I don’t understand the other."

(P6 ASU, Colombia, 35–45 years old, female, respondent’s parent)

Sample storage

The concept of storing biospecimens was generally well understood and accepted by participants. The term "biobank" was associated with "a refrigerator" and a "blood bank" to preserve and safeguard the samples—participants associated biobanking with "preserving the blood" for future research.

"…just like a blood bank, they should keep them in a warehouse, in a totally cold room…so that the samples are not damaged."

(C22-DAR, Colombia, less than 25 years old, male, participant)

In a few cases, participants took storage of blood samples to mean blood donation as for transfusions.

"…it is where they store blood samples or blood types when people need them, and I think that is what donations are for, and they receive them in case of an emergency".

(N6-MPH, Nicaragua, 35–45 years old, female, respondent’s parent)

"Let’s say one bad thing is in the case that, let’s say, she suffers from a chronic disease and you come and, let’s say, the institution came, and they took the sample from my daughter and the person who received that blood, got sick."

(N19 JALC, Nicaragua, 35–45 years old, male, respondent’s parent)

Confidentiality

Most participants understood the confidentiality-related procedures described in the consent form, including removing direct identifiers (e.g., name, address) and assigning a code to data and samples. Participants expressed concern that the steps taken to protect participant confidentiality could mean that clinically relevant results could not be communicated to research participants.

"I am worried because this tells me that all data will be erased. To a certain extent, I do not agree with that idea because it makes sense that there won’t be any negative consequences. But at least I say that if you get hit [you were infected], you should be informed."

(N17-JGG, Nicaragua, over 45 years old, male, respondent’s parent)

In contrast, one participant expressed the concern that sharing detailed results from genetic research could lead to the identification of individual participants.

"Since it’s for future Zika research, the concern was that they would find something and say, ’In such-and-such a person, we found something in their genes!’ Or what do I know, some kind of disease and publish it, and the person’s name would be revealed."

(N12- MMAG, Nicaragua, 35–45 years old, female, respondent’s parent)

Another participant expressed concern that the steps taken to ensure confidentiality might not be sufficient.

"In some cases, I feel good, and in other cases, not because there is always the idea: what will happen to the sample? And if the information is really going to be confidential, what one fears the most as a parent is that something is going to happen to the child because we are giving so much information about him."

(N5-CMAZ, Nicaragua, 35–45 years old, female, respondent’s parent)

Incidental findings

The UCB template allows research teams to choose whether or not to disclose incidental findings from research-related genetic testing but recommends that researchers not disclose these findings to research participants [13]. As such, we chose the option to not disclose incidental findings in the template version used in this study. The phrase "clinically relevant results" was well understood and associated with a diagnosis or a critical health condition.

"Clinically relevant is when the results are bad. When something unexpected comes out, that would be relevant to the child."

(P13-LDB, Colombia, 25–34 years old, female, respondent’s parent)

Research participants generally did not understand that they would not be told about incidental findings from future studies that reused their data or samples. Participants expected that these findings would be shared and disagreed with the language in the template IC once it was clarified.

"I see it as bad that they keep that result if in any case, there is some type of disease that they discover, apart from what we are doing, because the blood sample study is not only for Zika. You are telling me that there will be other types of exams or studies that are going to be done on the samples, then if you tell me that there is not going to be any individual result or that they are going to tell the person I don’t think it’s right, in that sense I put the case when a person is a blood donor."

(N17 JGG, Nicaragua, over 45 years old, male, respondent’s parent)

"I think we have the right to know if something suddenly came out of that study that could affect my son’s health."

(P13-LDB, Colombia, 25–34 years old, female, respondent’s parent)

"…if something relevant comes out, then yes, I would like them to tell me."

(P4-JS, Colombia, over 45 years old, female, respondent’s parent)

"I am aware that you are going to carry out the study. Perhaps that way they will produce a vaccine to cure Zika. But if the child has something relevant, you will not tell me, and that worries me."

(NPILOTO2-ESLO, Nicaragua, over 45 years old, female, respondent’s parent)

"My way of thinking was that if something abnormal happened in her blood type, it would be communicated, but you tell me that it would not be revealed. Practically, [the information] is for you, and that’s it. Is that right?"

(N3-CCBB, Nicaragua, 35–45 years old, female, respondent’s parent)

"That they are going to share my blood, and well, in actuality, I would think that would be wrong because you have to know….It could be that they do something wrong or that there is some illness and you don’t know about it, and afterward, you are sick and, and you don’t even find out. That’s why it would be bad."

(P11-EYP, Colombia, 35–45 years old, female, respondent’s parent)

Broad consent for future use

Participants understood that pseudonymized data and samples could be shared in the future with researchers that study other diseases or viruses and generally agreed with this concept. In one case, broad consent was interpreted as a way for investigators to "clean their hands" of responsibility to the research participants because, when identifying information is removed, investigators can reuse the data or sample without needing to inform the participant of the results or concerning themselves with protecting the participants’ confidentiality.

"Well, I already authorized them to use it for the Zika study, and I also authorized the deletion of the data once, so, well, I authorized it, right? Because it is like a form that is being filled, right? To be able to rotate it, to be able to give it to other researchers, then it is as if it is already there as if they are cleaning their hands, we eliminate it (the data). We can use that sample once they eliminate the information, to avoid a problem there, right? As such, I don’t feel bad, and I’m not going to find out and it’s normal, right? The only thing I’m going to feel is like, ok, right; it’s a study, it was what I authorised. And that’s what I’m going to take with me, and after that, I don’t know what will happen."

(C22-DAR, Colombia, less than 25 years old, male, participant)

Benefit sharing

Trust in the research team

Most participants referenced trust in the research team (UIS in Colombia, Instituto de Ciencias Sostenibles in Nicaragua) as a critical driver of their decision to agree to the future use of their data and samples. This trust was based on their positive experiences with previous research studies and their close relationships with the cohort study staff, which relates to the long-term nature of these cohorts and cohort staff’s sustained efforts to ensure the cohorts benefit the source community. The trust was also associated with the perceived benefit of increased medical attention through their child’s participation in the cohort, including improved access to medical appointments and closer follow-up by the medical team.

"Well, I trust, first in the XX university… and I know that they are not going to deliver (samples, data) to whichever research group or any laboratory because, yes, they must have a project, a study, it is not all [something that happens] overnight. Rather everything is well structured, right? They are professional and very ethical."

(C14-DMM, Colombia, 35–45 years old, female, respondent’s parent).

"I already participated in that dengue study…and it went well for me because [the doctors] stayed there off and on. They were calling me. They were calling people asking how they were doing and all that…So I imagine that this new stage with Zika will be the same."

(C18-NAP, Colombia, 25–34 years old, female, respondent’s parent)

"Until now they have behaved very well, they come and visit me, they look out for the child. So, it must be that the people from XX university are doing something good."

(C20-EM, Colombia, over 45 years old, female, respondent’s parent)

While participants expressed concerns about the possibility of misuse by future groups, this fear was balanced with their belief that the research team would not allow misuse to happen.

"That’s the problem, you don’t know if it is for [a good purpose]…but in any case, if it’s for the benefit of the people…if it is to get some drug, some vaccine, well, yes, I would like other people to benefit as well. Well, in truth, I have faith that you will do things well."

(P6-ASU, Colombia, over 45 years old, female, respondent’s parent)

Perception of the cognitive interview study

Most interviewees said they had gained knowledge about the data, sample sharing, and consent process through the cognitive interview. Participants noted that clarifying the IC in the cognitive interview improved their confidence in the study.

"I had no idea that to carry out a study, they asked for so many permissions, such as taking the samples, storing them, to use them in the future. I had no idea because I thought that they did it without getting your permission. The fact that you are telling me that [the investigator] is going to do it for this or that, to analyse the genome, to analyse the DNA, gives me confidence."

(C17-MSF, Colombia, over 45 years old, female, respondent’s parent)

"I have learned a lot I had never heard before. And I keep thinking about all that information because they should tell you when they will take your blood for those studies. I had never heard all those things."

(C18-NAP, Colombia, 25–34 years old, female, participant’s mother)

"But look, at the moment, sometimes, it happens that you sign and you don’t even realise it, they can even take you to prison, because you don’t know at the moment when they are so rushed, so this becomes an experience we can learn from so that we pay attention to what we are doing."

(N15- LMMU, Nicaragua, 35–45 years old, female, respondent’s parent)

Cross-group differences in the understanding and acceptance of broad consent for future use of data and samples

We did not find meaningful differences in understanding or agreement with key concepts between Colombia and Nicaragua or across participant gender, with the caveat that we could only conduct 11 interviews with men. Participants who had completed high school were better able to understand and describe abstract concepts, like genetic studies, than those who had not completed high school. In Colombia, we interviewed five men and five women, aged 19–21, who were themselves research participants. These participants had all finished high school, and three had post-secondary education. Both parents or legal guardians who had consented to their children’s participation and cohort participants themselves supported research participation. Like their parents or legal guardians, cohort participants expressed concern about sharing data or samples with for-profit companies that they saw as potentially monopolising research data.

"I really would not have to complain because the only thing I’m doing is giving a contribution with a sample, helping them (…) that’s my contribution…But the researchers wouldn’t be able to get to those studies either, to those answers they want to get to, without my help and that of many others, you see?"

(C24-MSB, Colombia, less than 25 years old, male, participant)

"As I told you, I am an administrator, and I am studying administration…many times companies or especially groups, large groups, and in research there also have to be those groups, what they want to do is how to take advantage or, in such a case, be the first to arrive at an answer that no one has yet… and at the moment of having that, if, let’s say, there are several groups that have the same answers, then why do they unite themselves? To make an economic profit, then they raise prices."

(C22-DAR, Colombia, less than 25 years old, male, participant)

Cohort participants also questioned whether their parents understood the information communicated in the informed consent.

"…When I participated in the study… I did not receive all this information. So, you make me think that my mom, for example, has no idea about these things. So, she just said "okay," but because she doesn’t know anything, you understand?"

(C15-CCT, Colombia, less than 25 years old, male, participant)

Participant recommendations

Strengths & limitations

This study has several strengths. This is the first study to use cognitive interviews to assess participants’ understanding of broad consent for future use-related language. We tested participants understanding of broad consent language from the IC template from a large public university that conducts a high volume of biomedical studies in high- and LMICs. While the interpretation of broad consent-related language likely varies across populations and contexts, the finding that participants did not sufficiently understand the language indicates that universities and research teams should consider using qualitative approaches to develop or evaluate the language in IC forms.

This study has some limitations. Ideally, the project interviewers would have been source community members to prevent hierarchical power dynamics during the interviews. Because the project interviewers had long-standing relationships with cohort members, they could develop a good rapport. Additionally, we could not include enough male participants to allow for cross-gender comparisons of responses. In both settings, women are generally responsible for taking care of educational and health-related issues for their children and were most often responsible for consenting to their children’s participation in the source cohort studies. Ideally, the cognitive interviews would have been conducted in person, but we conducted them by phone in Colombia, in keeping with local regulations during the COVID-19 pandemic. We compared field notes between Colombia and Nicaragua and did not find substantial differences in the flow of the cognitive interview session between in-person and phone-based interviews. This finding is in keeping with other work that found that the rapport needed for qualitative research could be built by phone [44].

Recommendations for research and practice

The ethics community has moved away from the paradigm of ERCs in high-resourced settings evaluating the ethics of research conducted in low-resource settings to mandating in-country ethics review for multi-country studies. While this move can help ensure that local values that shape interpretations of shared ethical principles are included in the IC language and structure, local ERCs may be reluctant to change the language of template ICs from ERCs that review the protocols for multi-country studies, which are often managed by researchers based in high-income settings. A 2012 qualitative study with 46 representatives from 34 US IRBs found that IRBs members reported lack of knowledge of local context and how the ethical principles of autonomy, justice, and proportionality are considered in different contexts as major challenges [45]. The same study found that US-based ERCs’ fear of "worse contributions scenarios" related to research management sometimes led them to paternalistic interactions with ERCs in LMICs [45]. While ERCs need to ensure continuity in adherence to widely accepted ethical principles across studies and countries, context is important. ERCs should allow investigators, which need to include representatives from the community participating in the study, sufficient leeway to tailor the consent language and content to reflect community values and preferences.

Participants stressed benefit sharing, providing information on future uses and users, and maximizing the utility of data and samples over privacy concerns. Some legal frameworks related to data and sample reuse, like GDPR, complicate efforts to share data and samples directly or through their interpretation. Researchers or institutions concerned about GDPR compliance may decide to anonymize rather than pseudonymize their data given that sharing anonymous data falls outside of GDPR. When investigators share anonymous data, data where the probability of the reidentification of individual participants approaches zero, researchers cannot inform participants about future uses and users of their data or samples or the incidental findings that arise from those studies. Policy makers should consider whether the current legal framework for data and sample sharing, which is rooted in concerns about the privacy-related harms of data and sample sharing rather than the equitable distribution of benefits from research, aligns with research participant values.