Interrogating the Value of Return of Results for Diverse Populations: Perspectives from Precision Medicine Researchers

Abstract

Background

Over the last decade, the return of results (ROR) in precision medicine research (PMR) has become increasingly routine. Calls for individual rights to research results have extended the “duty to report” from clinically useful genetic information to traits and ancestry results. ROR has thus been reframed as inherently beneficial to research participants, without a needed focus on who benefits and how. This paper addresses this gap, particularly in the context of PMR aimed at increasing participant diversity, by providing investigator and researcher perspectives on and questions about the assumed value of ROR in PMR.

Methods

Semi-structured interviews with a purposive sample of investigators and researchers across federally funded PMR studies in three national consortia, as well as observations of study activities, focused on how PM researchers conceptualize diversity and implement inclusive practices across research stages, including navigating ROR.

Results

Interviewees 1) validated the value of ROR as a benefit of PMR, while others 2) questioned the benefit of clinically actionable results to individuals in the absence of sufficient resources for translating findings into health care for diverse and disadvantaged populations; 3) expressed uncertainties in applying the presumed value of ROR as a benefit for non-clinical results; and 4) and debated when the promise of the value of ROR may undermine trust in PMR, and divert efforts to return value beyond ROR.

Conclusions

Conceptualizations of diversity and inclusion among PM researchers and investigators raise unique ethical questions where unexamined assumptions of the value of ROR inform study recruitment efforts to enroll minoritized and under-represented populations. A lack of consideration for resources and infrastructure necessary to translate ROR into actionable information may hinder trustworthy community-research relationships. Thus, we argue for a more intentional interrogation of ROR practices as an offer of benefit and for whom.

Introduction

Over the last decade in precision medicine research (PMR), the return of results (ROR) has become increasingly routine and expected (Delanne et al, 2019). PMR aims to improve prevention, diagnosis, and treatment based on individual differences in genetics, environmental, and lifestyle-related risk factors, making ROR potentially integral to informing an individual’s care. As the field of PMR has advanced and changed, so too have conceptions around the “value of ROR.” From The “duty to report” clinically useful information, expectations of ROR have expanded beyond health-related genetic testing results, to the “new normal” of including non-actionable results and secondary findings (Wolf & Green, 2023; Casalino et al, 2023).

The National Academies of Science, Engineering, and Medicine report, Returning Individual Research Results to Participants, centered participants’ values, engagement, and potential benefits from research ROR, in their recommendations, seeking to reconcile participants’ interest in ROR and researchers’ ability and obligation to meaningfully implement ROR (Botkin et al, 2018). Broader trends also served as a backdrop and sharpened these issues: an increase in public interest in PMR and genetic results (Murphy et al, 2008; Facio et al, 2011; Kaufman et al, 2016), including the availability of direct-to-consumer genetic results (Rubanovich et al, 2021); technological advances that lowered barriers to “big data” collection and analysis combined with high expectations for the impact of such research for the health of individuals as well as communities (Goytia et al, 2018); a shift toward patient-centered care and participant-centered research (Wolf, 2013; Bromley et al, 2015; Nebeker, Leow, and Moore, 2019); and the acceptance of personal utility as rationale for ROR similar to clinical utility (Jamal et al, 2017; Hutchinson, Capron, and Dousseau, 2019).

Yet ROR has generated a great deal of debate and remains a contested area. Researchers have grappled with important ethical issues related to beneficence and respect. Early research on genetics professionals’ support for incidental and non-clinically actionable ROR led to calls for clearer guidelines (Yu et al, 2014). Findings highlight the increasingly blurred distinction between research and clinical care, including concerns over therapeutic misconception (Burke, Evans, and Jarvik, 2014; Jarvik et al, 2014), competing definitions of secondary findings requiring different obligations for ROR (Eckstein, Garrett, and Berkman, 2014), and the use of “pragmatic clinical trials” where research findings inform medical decisions (Stewart et al, 2020) or when participants approach clinical research as an avenue to access care (Gutierrez et al, 2021). Concerns persist over the utility of results, especially given the over-representation of individuals of European ancestry, presenting challenges to the interpretation of variants of uncertain significance (VUS). For example, polygenic risk scores (PRS) created primarily by samples from unrepresentative biobanks have limited portability to other populations, leading some researchers to question whether the promise of benefits can be equitably distributed among members of all genetic ancestral backgrounds (Bentley, Callier, and Rotimi, 2017; Martin et al, 2019; Hutchinson, Capron, and Doussau, 2019).

Research has also focused on not just clinical utility, but on various forms of personal utility (Foster, Mulvihill, and Sharp, 2009), based on participants’ own varying values and beliefs about ROR such as around VUS (Jamal et al, 2017) and a shift toward patient-centered research where participants expect access to their results (Botkin et al, 2018; Wilkins et al, 2019). Extending utility even further, research has demonstrated how value can be co-created between clinicians and participants in pediatric genomic sequencing (Outram, Brown, and Ackerman, 2022), and expanded to familial and social utility, such as by parents leveraging genetic results to obtain benefits and services for a child that might otherwise be out of reach (Childerhose et al, 2021;). Expanding on different stakeholders’ understandings of the value of results must take into account a range of preferences, including the types of results, concerns about feasibility, and the availability of counseling or interpretation (Khodyakov et al, 2019).

At the same time, mandates for increased diversity of participants in federally funded research have highlighted shortcomings in enrolling and retaining historically under-represented groups, including minoritized and marginalized populations. In this context, ROR has been reframed as an inherent benefit of research participation that can help achieve goals of increasing diversity, and in so doing, ultimately address health inequities. Yet this raises further considerations around how the process of return would actually unfold, especially in under-resourced, under-served, and under-represented communities (Henderson et al, 2014; Sabatello, Zhang, and Chen, 2020; Hiratsuka et al, 2020). Multi-sited research programs such as CSER and eMERGE were funded to address these issues around resources for clinical follow-up through development and evaluation of practical strategies for implementing ROR in genomic research (Jarvik et al, 2014). Persistent disparities in access to health care have raised additional questions of whether such results are, or can be, equitably incorporated into clinical care (Landry et al, 2018; Popejoy et al, 2018). Questions of health justice and responsibility continue to be raised where under-served populations face barriers to translating ROR into care (Neuhaus and Crane, 2021; Sabatello, Callier, Garrison, and Cohn, 2018).

We add to this conversation by presenting the perspectives of PMR researchers: How do they understand the wider reframing of ROR as a research benefit, particularly within the context of the push toward greater diversity among research participants? What concerns do researchers raise about the effective translation of ROR into clinically and personally actionable information for participants? What dilemmas do researchers encounter as they navigate ROR practices in their own studies? And what impact does the promise of ROR have on the trustworthiness of the PMR enterprise in the long term?

Materials and Methods

This paper is part of a larger project investigating how the call for more diversity and inclusion has been interpreted and operationalized by PMR funders and researchers and the impact on research practices including enrollment, retention, engagement, data collection and analysis, and ROR. We followed five PMR studies across three consortia funded by the US National Institutes of Health (NIH).

In this paper, we refer to the three consortia, which were explicitly initiated to increase diversity: each study funded under these consortia had target enrollments for historically underrepresented and other underserved populations. The first, Consortium A, is a large-scale, national precision medicine initiative focused on recruiting underrepresented individuals in biomedical research through regional study sites to improve health. This consortium planned ROR for a range of different types of health and genomic results over time, including genetic ancestry and traits, genetic health risks, and pharmacogenomic results. Consortium B generated genomic data for longitudinal cohorts built from diverse populations to address risk factors underlying a range of common chronic diseases. ROR was not initially anticipated for many of the original studies. At Consortium C, the mandate was to study and evaluate practices in translating genomic sequencing ROR into clinical care. Returning genetic results to participants was therefore integral to the study aims.

The study data included document analysis of funding announcements, consent forms, and fliers issued by the NIH (76 documents); approximately 450 hours of participant observations of site and consortium working group calls, and in-person or virtual meetings, where detailed field notes were taken; and 125 in-depth interviews with a purposive sample of PMR funders (n=4), PMR investigators (n=30), research team members (n=64), and research participants/participant advisory board members (n=4), some of whom were interviewed twice. Our findings presented here are based on 94 initial and 23 follow-up interviews, each lasting 60–90 minutes, with PMR investigators and research team members (Figure 1, Demographic characteristics of interviewees). Semi-structured interviews explored the processes of recruitment, engagement, and retention of under-represented participants, in addition to how study teams operationalized diversity mandates in their research, including data collection and analysis, ROR, and translation. Interview guides were adapted to enable us to explore relevant roles and activities. Interviewees were offered a $50 gift card for each interview. Interviews were digitally recorded and professionally transcribed verbatim.

Figure 1.

Demographic characteristics of interviewees.

All data was analyzed according to the principles of constructivist grounded theory (Charmaz, 2014), with documents, observation field notes, and interviews treated as raw data and analyzed as one comprehensive dataset. Information gathered from one source was used to inform other data collection (e.g., using observations to inform interview questions). The research team developed an initial codebook based on study aims, existing literature, and themes and categories that arose consistently throughout initial data collection. The effectiveness of the initial codebook was first tested on a small sample of documents, interviews, and observation field notes uploaded to Dedoose. Codes that were inductively generated from the first test were added, followed by consecutive waves of joint and individual coding and collaborative discussion to create a finalized codebook of 75 codes (some of which have 2–14 sub-codes). For this analysis, we extracted data tagged with the code “Return of Results.” Research staff engaged in two rounds of review and memoing to elicit iterative analysis and interpretation of the data, which were shared with the whole research team to be discussed and refined, leading to this paper’s findings.

In our analysis, we sought to identify the range of positions among our interviewees, specifically investigators and research staff, about the role and utility of ROR in PMR. In what follows, we provide a picture of this range, to demonstrate that the increasing routinization of ROR, under the rationale that results provide benefit to and should not be withheld from research participants, is questioned and contested by PMR investigators and research staff themselves.

Institutional review board approval was obtained from the University of California, San Francisco and Columbia University. Funding was provided through the NIH from NHGRI Award 1R01HG010330–01.

Results

Through our analysis of interview data, we identified a range of perspectives concerning the value of ROR. We found that 1) for some interviewees, ROR took on an overarching value that obscured significant differences in actionability and utility across types of results, consistent with the literature. Other investigators and research staff questioned the value of ROR, by 2) raising ethical concerns about whether clinically actionable ROR constituted a benefit to individuals in the absence of sufficient resources to translate research findings into health care for diverse and disadvantaged populations; 3) expressing uncertainties about the benefit to participants of returning non-clinical results; and 4) debating whether the potential value of ROR should be promised when participants may not be able to utilize them, given the stakes for undermining trust in PMR and diverting efforts to return value beyond ROR.

1). Validating the Value of ROR

Despite early attention to the ethics of returning genetic results in the context of research, the routinization of ROR as an assumed benefit of genomics research, as observed in the literature, is mirrored in our study data. For some of our participants, ROR was seen in general to offer valuable benefits to participants. Researchers saw that benefit as encompassing different types of results that are understood to have varying actionability and utility [see Excerpt 1A, Table 1] and validated the broad value of ROR as providing information that could be useful [1B] and acted upon [1C]. Some research staff shared actual examples where ROR resulted in real, positive impacts for clinical care [1D]. Others felt that the value of ROR was in fact not fully tapped, and that if the general public were more educated about genetic sequencing, the value of ROR would be more fully realized [1E].

Table 1.

Validating the Value of ROR

1) Validating the Value of ROR
Excerpt No.	Theme	Consortium, Interviewee, Type of ROR	Excerpt
1A	Perspective that participants value of any type of ROR	National PMR biobank (Consortium A), Research Staff Genetic health risks vs. genetic ancestry ROR *Same interviewee as Excerpt 3C	So I think... hearing that there’s value, knowing that [ancestry & genetic traits ROR] might be an easier way to start the engagement in education process, the popular interest in genetic ancestry and the view that participants might be waiting for something to come back from the program. That there’s something that we are returning that they may see as valuable. ...And we heard very strongly that, yes, there’s value in knowing back, our individual results. ‘If you tell us the right things and put it in the right way and explain to us what it means and what it does not mean, and what we should do or should not do with this information.’ I believe participants would like to hear back.
1B	All types of ROR as inherently valuable	National PMR biobank (Consortium A), Investigator Genetic health risks ROR	I’m acting on this belief that the principle behind all of this is to share the data with the individuals as much as possible. So, I think there’s been a lot of thought put into how to follow up individuals with data, not just from regular clinical measures, but from genetic and biomarker studies so that they can be usefully interpreted by participants and their caregivers[1]. So, I think the devil’s in the details, but I think the overall belief and principles of the study are to share the data with the participant and I support that.
1C	Value of all types of ROR for clinical care	National PMR biobank (Consortium A), Investigator General ROR *Same interviewee as excerpt 3B	And so now we, we have to change their consent form of course to say, look, there’s a good chance we might find something and we’d like to be able to return this to you in the end as the part of the research that we’re doing. So that’s kind of exciting. That’s what I always talked about when I started doing biomedical research and now we’re finding things, that actually matters. Outpatient care, as a doctor that’s really empowering. So I think that’s really great.[2]
1D	Value of ROR to participants a medically actionable	Disease-specific population genomic data (Consortium B), Research Staff Genetic health risk ROR	I think that especially when you’re giving results to people who aren’t anticipating them, they’ve come out of the blue....you want to make sure you’re giving something back that they’re going to find meaningful and that there is some action or something they can do about it... We’ve seen that in our own biobank where one of my patients that I called, she ended up finding out she had cancer because of her biobank result because it basically prompted her to get mammography earlier than she would have, and she identified that she had active cancer, so we found it earlier. It also informed her decision-making. She opted to have a double mastectomy instead of just having a lumpectomy because she knew about that variant, and that all came just out of her participation in the biobank and us calling her. I guess that’s what I see with value versus the other stuff. Ancestry was mentioned, that might be valuable to some people....I don’t know, I think that everybody determines value their own way, but finding something that is highly actionable seems to me it’s of high value and it’s something that it’s a good place to start when we’re talking about returning this type of information.
1E	Value of ROR related to participant engagement	National PMR biobank (Consortium A), Investigator; General ROR (sequencing and traits)	So, the key question always is what’s the value to participate in [this study] and part of it may be genetic sequencing and traits and Ancestry and the 23andMe kind of aspect of things that people are getting back, in a way.... [B]ut for some people, “Okay. That’s not a big deal either. I don’t want that either. I don’t know what a gene is.” Maybe education around what a gene is what genotyping means and what having a single-nucleotide polymorphism means is an important educational aspect, that if people are educated about that, they actually may then perceive more value from the genetics component of the program, and might make people be more involved.

2). Questioning the Value of ROR: Barriers to Translating Clinical Results for Diverse Populations

However, other researchers raised concerns that in focusing on the need to return any result, little attention is being paid to whether and how results can actually be translated into benefits for individual health. For example, some investigators questioned whether return of clinically actionable results constituted benefit when many participants may misunderstand the meaning of genetic results, especially as related to risk [2A]. One interviewee noted difficulties in meeting the high bar for educating participants about clinically actionable genetic results [2B]. Other investigators raised concerns over how ROR could be utilized by participants when confronting a lack of access to quality health care they viewed as necessary to follow up on clinical results [2C].

Some interviewees expressed doubts that research studies could promise sufficient resources to return results in ethical, responsible ways. Drawing on their direct experiences in their studies, researchers expanded on the shortcomings of PMR broadly, describing the lack of resources for translating research results into clinical care [2D]. Another concern focused on the lack of clinical capacity to interpret results, counsel patients, and follow up with clinical recommendations, which undermined the promise of value of ROR [2E] (Table 2).

Table 2.

Questioning the Value of ROR: Barriers to Translating Clinical Results for Diverse Populations

2) Questioning the Value of ROR: Barriers to Translating Clinical Results for Diverse Populations
Excerpt No.	Theme	Consortium, Interviewee, Type of ROR	Excerpt
2A	Concerns about actionability if participants misunderstand results and risk	Disease-specific population genomic data (Consortium B), Research Staff Genetic health risks ROR	But let’s say that somebody had done that [gotten tested for genetically linked Parkinson’s disease]. And now I think that I’m at increased risk for Parkinson’s and we don’t know anything to do about it. So, there’s an emotional and potentially downside of giving people information that they can’t do anything about. Now with some other diseases like Huntington’s disease... people maybe do need to know because they need to have advance planning, who’s going to take care of them, their families, et cetera. So I think it’s a case by case basis. But in general, my biggest question is what’s the value of getting information about which you can’t do anything about? I do think that we need to think very carefully about when to do that, what the circumstances are. And also truly to have an informed consent process, so that people understand the information that they’re getting. And as a general rule, Americans in general, are not very good at understanding risk. ‘I don’t want to take the vaccine because it might be dangerous.’
2B	Expectation of value of ROR requires educating research participants so that results are interpretable	Disease-specific clinical patients (Consortium C), Investigator, Specific ROR (colon cancer)	[As an internist] I can order any test that’s going to tell my patient they’re going to be dead in six months with a lot less background than we use in genetics to order something that is important, but not as life changing honestly. It’s just an interesting culture genetics has of feeling like everyone has to learn genetics to have a genetic test.... I’m hoping we can train people for a future of finding out what the patient actually needs and studying what the patient actually needs versus what we think they need. [For example] we... gave them their secondary findings at a separate visit because we thought it was too much. They didn’t know what we said at the first visit. We would say, “Any questions about the last visit?” And they had no idea. It was fascinating. So, we’re spending a lot of time and not really communicating with people.
2C	Concern over value of ROR in absence of access to clinical F/U	Disease-specific population genomic data (Consortium B), Research Staff Genetic health risks ROR	I mean, clinically actionable, but the question is, do these folks have access to certain care, or if it was something very serious, do they have the support infrastructure to be able to actually do something about it once they get it? You know, there’s a lot of questions that come up for once you return, you know, returning results.... But the question I would say is actionable for whom?.... And so returning results to whom, and what are the barriers that they might experience in trying to actually address these results? Again, those are the types of things I think you have to take into consideration when you’re dealing with an [under-resourced] population
2D	Lack of resources for handoff to care	National PMR biobank (Consortium A), Research Staff General ROR	That intersection of community and research, that’s been a problem for us for a decade. And before Precision Medicine, the way that we addressed it is that if we were working with people who were found to be food insecure, we would have a system for referring them to people who could help with food insecurity. We couldn’t because of what you said, because we were doing research and that crosses over into kind of a clinical care setting, but you can do a one handoff to a social worker or a food bank or whatever like you can or give people materials that can help bridge that gap. Precision Medicine does not have that because there is no way to do a warm hand off to a Precision Medicine bank. There’s not a mechanism for that. So I don’t like it for that reason. I think we fall short and I think that’s not good.
2E	Value of ROR when procedure, resources are limited for interpreting ROR for research participants	National PMR biobank (Consortium A), Investigator General ROR *Same as Excerpt 1C	What keeps me up at night is, you know, how am I going to deal with the whole volume of, sort of sequence data that’s gonna come in. How are we going to annotate this responsively? How are we gonna not over call anything and not miss things and into sort of this usual thing about reading and X-ray and high volume and it’s the same issues. I think we still haven’t gotten to the point where we can automate this really well on what is that going to mean for the patients in terms of, you know, [our] counseling them....I worry about in terms of [the study], it’s really the scale at which we’re going to wind up doing this is gonna be really the daunting question.

3). Questioning the Value of ROR: The Case of Non-Clinical Results

In recognizing resource challenges, one researcher explained that their current ROR practice had to be limited to non-clinical information due to the costs and challenges of validating research results [3A]. However, for some, the value of the return of non-clinical results, such as the return of genetic ancestry and other trait information, raised further questions about utility for research participants. One researcher voiced concern about the instrumental use of non-clinical ROR related to genetic traits, for example, framed as a long-term benefit of research participation in longitudinal PMR for boosting the lower retention rates among under-represented groups [3B]. Interviewees also expressed concern about promoting the recreational value of these types of ROR, without sufficient attention to cautioning participants about the potential for discovering unexpected or distressing information, such as non-paternity [3C] (Table 3).

Table 3.

Questioning the Value of ROR for Non-Clinical Results

3) Questioning the Value of ROR for Non-Clinical Results
Excerpt No.	Theme	Consortium, Interviewee, Type of ROR	Excerpt
3A	Regulatory constraints for ROR	National PMR biobank (Consortium A), Research Staff Genetic health risks vs. genetic ancestry ROR	So, those are the kinds of constraints I was talking about, saying regulatory constraints that we operate under as well. So that seemed like a right place to start the engagement on return of results [for ancestry and genetic traits], knowing that health-related return results would take more work, more implementation, more time to start doing.
3B	Questioning the utility of non-clinical ROR	National PMR biobank (Consortium A), Research Staff General ROR *Same interviewee as Excerpt 5I	...there’s so many limitations to genetic information, [genetic traits], or information about their ancestry, which I think is like super problematic. But that’s going to be a motivation for people. But then I look at like the number of people who do 23 and Me, and all of the other direct to consumer stuff.... So maybe they have a pulse on the consumer that I don’t have. But... I struggle with the genetics that we can return being the sole return of value for the program. Because there’s nothing else that I’ve seen.
3C	Overlooking unexpected findings in promoting value of non-clinical ROR	National PMR biobank (Consortium A), Research Staff General ROR	I think it’s always a tug and pull where you want to make sure that you are as informative as possible about the potential types of results that you might see, and ultimately, the reason you’re doing it is because you think people will find it entertaining. We’re not trying to harm people. .... there is the idea that people might jump into this and not understand the gravity of the results on the other side. They might not understand that they might find something that they didn’t think they would find, and then be very shocked by that. So, for example, the non-paternity is one of those things where nobody walks into that experience thinking they’re going to discover non-paternity, but it can be very devastating. So, how do we strike that balance of making sure that you understand what the risks are, especially with some of these cases, these risk cases are actually very rare scenarios where most people are going to have just a very fun and an entertaining and somewhat trinket sort of relationship

4). The High Stakes of Leveraging the Value of ROR

These concerns on the part of investigators made some wonder about the ethics of returning results at all, in the context of PMR. They expressed worry about the mismatch between the promised versus actual value of ROR engendering distrust among the very populations needed to advance PMR, as well as detracting from other types of benefit beyond the scope of ROR. The need to temper participants’ expectations in turn placed additional pressures on, and raised the stakes for, PMR-community relationships, with particular emphasis on trust [4A]. One researcher described the importance of building trust given their study population’s experience with historical racism [4B]. Local context, community outreach, input, and engagement were essential to informing the study’s approach to ROR, which was founded on a commitment to protecting participants [4C].

Researchers recognized limits to such an exclusive focus on ROR as a benefit, as well as possible tensions with other values of ROR beyond the individual. They expressed concerns about over-amplifying participant expectations around individual genetic ROR and the negative potential impact of a misalignment between promises and implementing ROR [4D]. More fundamentally, interviewees questioned whether there were alternatives to ROR as a source of value for PMR participation, particularly given the emphasis on recruiting underrepresented populations from marginalized communities. Researchers acknowledged participants’ constraints on time and competing priorities, and that rather than promoting or even inflating the value of ROR, researchers should instead ensure that there are other benefits of PMR that better align with participants’ values [4E, 4F, 4G]. Further, some researchers saw the value of ROR as obscuring possibilities for re-orienting how research results can meaningfully be translated into positive health impacts [4H, 4I] (Table 4).

Table 4.

The High Stakes of Leveraging the Value of ROR

4) The High Stakes of Leveraging the Value of ROR
Excerpt No.	Theme	Consortium, Interviewee, Type of ROR	Excerpt
4A	Value of ROR as requiring trust	Disease-specific population genomics (Consortium B), Research Staff General ROR	No, so, again, it goes back to... what you asked about trust [as an issue related to genome sequencing]. So, in order to do this effectively that there has to be trust from the community by the community that this information how the information will be used.... And the understanding of how you engage community in genetic studies which is really different. The community may view it as it’s very invasive. What are you going to know about me? And what does it mean for my family? So that’s a discussion that we had.... that there is community buy in, community trust. There’s so much information that can be gained from sequencing one’s genome. And what’s the expense? How far can you go? And the communities understanding of that. What can you do with that information?
4B	Community outreach and input essential to building trust and informing ROR	Disease-specific population genomics (Consortium B), Research Staff General genetic ROR	...that [community outreach] group helped to facilitate some things to make people look more trustworthy of being a participant in this study, or either considering being a part of the study.... Because you know with the mistrust, with African Americans having this mistrust with the things that have happened in the past, you had to build that trust up. So you had to have different groups who could be stakeholders in their communities to get people to actually say, yes, I do want to participate in this study.... But at least they know now that it’s okay to participate in these studies because if you don’t, you know, how can you benefit? That it can be beneficial to you.
4C	Community outreach and trust building essential to inform ROR process	Disease-specific population genomics (Consortium B), Research Staff General ROR	I think the commitment that was made to the participants was that we’re not going to do anything that’s going to harm you. We’re going to protect you, if you put your trust in us.... I think that was the most crucial part of that protection. If you give us your blood samples, if you give us your information, we’re going to protect that at all cost, no matter who comes, or you know, ‘I want to do this research and I want to use your DNA.’ No, it’s according to what you’re going to use that for.... Because you don’t know what type of harm or damage that may cause that participant... [and] the more I guess that we get involved [in ROR],, we can have a more clear picture of what is reasonable to return and what is not.
4D	Value of ROR informs high expectations from participants	Disease-specific clinical patients (Consortium C), Research Staff Specific ROR	...[Participants’] expectations are just as high as they were in [an earlier study]. They expect it [ROR] to be useful for everything just about. I mean, everything except for maybe influencing their own reproductive decisions.... And I don’t know that it’s a mismatch or that expectations aren’t being set correctly.... But I think it’s a bigger, broader issue that our whole culture has been built on the idea of DNA as our blueprint... as the thing that guides us and makes us who we are....even if you’re telling them it’s really limited information, I think people just.... They want to save their kids. Their kids have cancer. And they want anything and everything that might help them and I don’t know that their expectations are unrealistic.... I’ve definitely seen people kind of, ‘Oh well, my family has this or oh I used to have this rash when I was a kid. I wonder if that’s related?’ And it’s not at all, but they’re kind of putting those things … They’re putting little pieces together that seem to make sense to them.
4E	Value of ROR as motivation to participate against competing priorities	National PMR biobank (Consortium A), Investigator General ROR, genetic traits ROR	We also need people to remain engaged because we need to continually bring longitudinal data into the system. So how we think about why do people sign up and making sure that that they’re signing up because they believe in the project and they want to contribute. But we recognize that everyone’s busy and they’ve got different priorities. And whether they believe in the project or not, sometimes it just doesn’t get them over the hump of, “I got to care for my kid. I got to go to my job. I got to do 8 million things. I’m sorry, but filling out your surveys online is not the top of my list of what I need to do to put food on the table.” But if there’s some value that we can bring back to you that helps move the priority, so that you’re willing to invest the time because we all have some time that we can choose to do whatever with. So I think that return of value is an element of it. It’s not the only reason people participate. There’s not a lot of value there, to be honest. And there’s different types of value.... There’s a lot of different reasons and there’s the concept of engagement.
4F	Value of ROR in response to limits of altruism as motivation	National PMR biobank (Consortium A), Investigator General ROR *Same interviewee as Excerpt 1E	Altruism only works for so many people, and also for so long. You’re interested when you sign up because you got inspired by something. Some family member just got diagnosed of something or died of something, or a friend or family member, and then time passes, and that grieving and that thing kind of heals, or whatever memory you have goes away or fades a little bit, and then your All of Us email comes, and you’re like, “Well, shit. I’m busy right now. I’m doing 1,000 other things. I’m not going to go do that. You’re not paying me for it. I don’t know what I’m getting from doing it, and I feel less altruistic now than I did six months ago.” Obviously, people don’t necessarily have that whole stream of thought, but I think that is actually what’s happening in reality for a lot of people.
4G	Researcher responsibility for ROR is greater for under-represented groups	National PMR biobank (Consortium A), Research Staff General ROR, genetic traits	[Referring to “benefits in people’s lives”] I’m referring to a host of things actually. I’m referring to return results in not even a timely manner, but I think that we have to... If you were to enroll in the program, I should be able to say, … you’re going to get your results in a year from now. Maybe two years from now. I think you owe people that level of clarity. We still don’t have that tied down for folks. And, I do think that is a barrier at times for the enrollment teams. ... I think, again, this kind of, divide between people... I keep going back to this about people who are underserved, but I think that there is a misconception about what it is to be impoverished or what it is to be... That somehow they have extra time. These are people who, when you’re on the margins, it’s like, time, effort, and energy, you don’t have a lot of that to spend. And, okay, there are all these incredible results that you’re going to get back, but I can’t tell you when exactly.... We claim that we want this population involved. We made a promise, we said it’s the value of the program. I think that’s a great value to have and it’s important that we crack this and figure that out. But yes, I think we have a responsibility, there is going to be a higher bar, and there has to be full recognition of that. And then leaning into that issue to make some change there, I think.
4H	Concerns related to actionability as both potential harm and potential benefit	Disease-specific population genomics (Consortium B), Research Staff General ROR	So I think you have to be very careful, and inform participants of potential benefit, as well as potential harm, before you release genetic information. I personally think it would be much more of use to give people information that they can change their medical care, or change their behaviors, than giving them information, which is unmodifiable.
4I	Value to individuals and communities beyond ROR	National PMR biobank (Consortium A), Research Staff General ROR *Same interviewee as excerpt 4A	I would say that value needs to flow back to communities, right? So it doesn’t need to necessarily flow back to individuals, but it does need to flow back to communities. And community not singularly defined by geography, but also all of our intersectional axis. Like I worry about African-American women with breast cancer, their treatment numbers are so terrible. Like if the program could say to me, we’re targeting health disparities, specifically in these types of communities, I’d be like, oh, health disparities is something I really care about. That would be great. So I would focus less on individual return of value, less transactional, and more meta return of value.

Discussion

PMR relies heavily on amassing biobanks of demographically and genetically diverse populations, often as part of long-term, longitudinal, exploratory and hypothesis-generating studies. Accordingly, recent efforts have focused on the motivations of historically under-represented groups to participate in research and their perceptions of benefit (Sanderson et al, 2013; Lewis et al, 2019; Wilkins et al, 2019). As noted in the existing literature, the value of ROR has become increasingly accepted as an almost self-evident part of this incentive to participate: self-knowledge through individual results has remained an important motivating factor (Facio et al, 2011; Sanderson et al, 2017; Botkin et al, 2018); participants may view research studies that provide genetic testing as contributing to their broader health care (Ackerman et al, 2023); and genetic results may also be valued as knowledge that rightly belongs to participants and their communities (Blanchard & Hiratsuka, 2021).

This increasing expectation to return clinical results has raised ethical questions in terms of responsibility, feasibility, and limitations of what a “clinical handoff” looks like for participants to derive value, where research findings are effectively translated into actionable information for their health. At the same time, research has shown participants’ desire to receive clinically actionable results (Wendler and Pentz, 2007), the public’s interest in receiving secondary findings regardless of clinical actionability (Casolino et al, 2023; Delanne et al, 2019) as well as non-medical traits or VUS (Vears et al, 2021), and the expansion of the notion of ‘utility’ to include personally (and not just clinically) actionable results (Nebecker, Loew, and Moore, 2019; Wilkins et al, 2019). This expanding obligation for ROR from clinical results to information that might be personally actionable, useful, or interesting challenges the meaning of value and criteria that are narrowly focused on likely health interventions. Indeed, we find that untethering value from specific, clinically actionable ROR and instead adopting rhetoric around the self-evident benefits of ROR without reference to specific types of results underpinned our interviewees’ concerns in these areas.

Ultimately, the ethical uncertainties and logistical challenges around ROR raised the question, as our interviewees did, of what “value” really means and why there are such high expectations surrounding ROR. We find that some PMR team members support the increasingly accepted expectation of expansive ROR as a value in and of itself. However, others questioned whether such ROR was a sufficient benefit for participation in research and whether this routinized imbuing of ROR with value might eclipse other means of fulfilling the ethical obligation for research to provide benefit. Where interviewees described logistical and ethical uncertainties around effectively implementing ROR in the absence of necessary resources and infrastructure, as noted in the literature, they also raised concerns over whether ROR was being overly burdened as the locus of value for research participants. Those team members experienced the burden of the expectation to offer ROR, even as the resources required for an effective, ethical, and equitable system for returning results were lacking. Our interviewees pointed out that these obstacles disproportionately affect under-resourced settings and socially disadvantaged populations, for whom the immediate costs of participating in research often outweigh future-oriented promises of value. Further, the overemphasis on the value of individual genetic ROR brought into stark relief the divide between genomic research and clinical care, particularly for populations with limited access to healthcare, or who experience health outcome inequities even with health care access (Jooma et al, 2019).

Our interviewees described how the routinization of the value of ROR created high stakes for their efforts to incorporate and communicate it as a participant benefit. They highlighted the potential positive impact of ROR for recruitment, enrollment, engagement, and retention as a means of building more trusting and equitable community-research partnerships, specifically in the context of increasing participation among historically under-represented, minoritized, and marginalized groups (Kraft et al, 2018; Kwan et al, 2018). However, the emphasis on the value of ROR may lead to language that over-promises the benefits for individual health outcomes to meet these study goals. Moreover, the ubiquity of an assumed value of ROR has raised expectations for returning both clinically actionable and personal, non-actionable information across PMR studies, even when such results do not align with the study design. For example, some studies are funded to build datasets designed to be representative of populations that would therefore only identify rare genetic conditions in a small percentage of participants (Oh et al, 2015; Khodyakov et al, 2019). It is the generalized approach to the abstract value of ROR that in fact can obscure important distinctions across types of ROR.

Efforts to increase diversity in PMR have focused on building trust and strengthening engagement; not attending to the value of specific ROR for specific communities may have a negative impact on researchers’ ability to implement these particular approaches to improve equitable participation. Failing to account for the structural inequities in diverse populations’ ability to access, apprehend, and act on ROR raises fundamental concerns about the approach to, and depth of, PMR’s commitment to diversity and inclusion. Moreover, such over-promise may further erode trust where the value of ROR falls short of participant expectations, undermining the goals of addressing health inequities and justice.

Reframing research participation through the lens of reciprocity that considers the broader social context for understanding researchers’ ethical responsibilities is critical to address questions of value (Lee 2021; Blumling et al, 2021). Through this lens, our interviewees’ concerns that ROR is increasingly serving as a standardized response for the obligation of research to return benefit in turn interrogate what other means might exist for fulfilling this obligation. As opposed to expectations that the primary form of research benefit is focused on individual ROR, what value may be found through a reorientation to other forms of benefits, centered on reciprocity and aligned with community definitions of utility?

Conclusion

In the face of heightened expectations for ROR, PMR team members have confronted additional ethical uncertainties and logistical difficulties in managing their obligations to participants when ROR has been normalized and imbued with inherent value. PMR team members’ conceptualizations of diversity and inclusion as linked to the value of ROR, implicitly or explicitly, gain greater significance in the context of study recruitment efforts aimed at enrolling minoritized and under-represented populations. Leaving unexamined prevailing assumptions of ROR as intrinsically valuable may hinder and even harm the building of trustworthy research relationships. Thus, we argue for a more intentional interrogation of ROR practices as an offer of benefit and for whom.