Introduction

Questions regarding which patients are most likely to receive benefits from medical screening in terms of their age raise complex issues in clinical care and research alike. In the clinical realm, age questions commonly arise in preventive screening (for diseases such as cancer or cardiac conditions) regarding when it is medically justifiable and beneficial both to initiate and to end screening. In research, age comes into play when deciding whether to exclude children, who cannot provide full legal consent, or older adults, who may have an increased potential for comorbidities. Thus, screening programs in both clinical practice and research often establish age criteria for patients or participants that exclude children and/or older adults. The topic of inclusion versus exclusion of children in such studies is hotly debated,1,2,3,4,5 but there is little literature regarding older age groups.

What literature there is focuses more generally on the exclusion of older adults from clinical research and indicates that upper age limits often exist without justification.6 Clinical trials commonly have upper age limits for participation, even when the trials involve illnesses that are associated with old age, such as Alzheimer disease,7 Parkinson disease,8 heart failure,7,9 diabetes,10 and colorectal cancer.7 Upper age limits also exist for clinical trials involving medications.11 One review of more than 300 clinical intervention research protocols sent to a research ethics committee found that one-third had an upper age limit that was arbitrary and applied even when the condition under study was prevalent among older adults.12 Justifications for excluding older people in research studies include suspected comorbidities,13 communication difficulties,8,14 cognitive impairment,8 and inability to attend follow-up appointments.8 Without further investigation, it is difficult to know if exclusion criteria are based on chronological age or involve social assumptions related to aging.

Research initiatives that promote preventive genomic screening programs in the general population provide an opportunity for examining tenets underlying inclusion and exclusion age criteria. These programs seek to screen individuals for a targeted set of conditions which can be prevented or ameliorated if found early.15,16,17,18,19,20 One review of existing studies of preventive genomic screening found five studies that returned results to participants.21 All had a lower age limit for study eligibility, but only one had an upper limit (which was 80 years old).21 Another study, the Geisinger MyCode Community Health Initiative, through its partnership with Regeneron Pharmaceuticals, sequences DNA samples of all biobank contributors in their research biobank, regardless of age, for preventive genomic screening of 76 genes associated with 27 medically actionable conditions.22

Age is an important criterion for screening, but calls for population-based preventive genomic screening programs15,16,17,19 do not mention whether upper age limits should be a consideration in developing such programs. Questions remain whether such programs will employ upper age limits, as used with other clinical screenings, on the assumption that older individuals would not benefit. In this paper, we analyze data from GeneScreen, a research study of preventive genomic screening aimed at adults to examine how both the researchers who designed the study and the individuals who joined understood and valued age in relation to screening. For the researchers, we focus on the assumptions and assertions involved in deciding what ages to include. For participants, we examine how their age affected their desire to undergo genomic screening and calculation of its risks and benefits. Social understanding of age is an underappreciated factor at play in how genomic screening progams are designed and offered and in decisions to participate. We propose that this factor be recognized in conjunction with evidence of clinical utility based on chronological age as genomic screening programs are developed.

Materials and methods

GeneScreen, which was based at the University of North Carolina at Chapel Hill (UNC), used a screening panel consisting of 17 genes associated with 11 rare conditions for which treatment and/or prevention options were available.23,24,25 There are currently no guidelines regarding age-related clinical utility for screening the genes related to the GeneScreen conditions. Participants were recruited at two sites: a general internal medicine clinic at UNC Hospitals, where clinicians nominated 436 patients as potential participants, and the Kaiser Permanente Northwest research biobank, which selected 650 individuals to be contacted. As far as possible, we attempted to recruit equal numbers at both sites by gender, race/ethnicity, and age group. The participants in both groups received a brochure with general information about the study and a letter directing them to the GeneScreen website, where extensive information about the study and the conditions being screened was provided. Informed consent was obtained through the website. This process resulted in 262 adults enrolling, all of whom were asked whether they would be willing to be contacted to participate in a telephone interview. Seventy-seven percent indicated they would be willing to be interviewed, and these participants did not differ demographically from those who declined.

We report on data collected from two sources. The first was observations and audio-recordings of 11 GeneScreen research-team meetings that occurred from July 2013 through April 2014. In these meetings, researchers (including bioethicists, social scientists, genomic researchers, and clinicians) debated which conditions and genes would be included on the targeted screening panel and what the eligibility criteria should be, especially regarding age and insurance status. All authors are GeneScreen researchers, with one (R.J.C.) serving as an ethnographic observer during these meetings.24

The second source of data was a series of semistructured telephone interviews conducted with individuals who joined GeneScreen. Fifty interviewees were purposively selected, from those who indicated willingness to be interviewed, to represent three age categories, two gender categories, and three race/ethnicity categories (see Table 1).25 Interviews took place from March through June of 2016 and were conducted approximately two to four weeks after participants had joined GeneScreen. All participants were still awaiting their genetic-screening results at the time of the interview. The interview was designed to elicit in-depth information about participants’ motives for joining, how they made the decision to join, and their understanding of the risks and benefits of targeted genomic screening and various facets of the GeneScreen protocol. All interviews were audio-recorded and transcribed verbatim.

Table 1 Demographic characteristics of GeneScreen interviewees (N = 50)
Full size table

To conduct the analysis of researcher meetings and participant interviews, the transcripts were read in their entirety and, using conventional content analysis,26 coded through an iterative, collaborative process. Codes captured segments of text pertaining to perceived benefits and risks of participating in GeneScreen, participants’ decision-making processes, and references to age in relation to these topics. We reviewed the coded segments to derive the themes reported below.

The Institutional Review Board of UNC approved all procedures. The Institutional Review Board of Kaiser Permanente Northwest approved the procedures related to the participants recruited from the Kaiser Permanente Northwest biobank.

Results

Researchers’ discussions of age

GeneScreen researchers deliberated about eligibility criteria, specifically participants’ age as it related to the potential risks and benefits of being screened for the 11 conditions. Initial suggestions were that older individuals might not benefit from the information they could receive from screening. Clinicians in the group were more inclined to think that screening should not be offered to anyone who, based on population statistics, had fewer than 10 years of life ahead of them, which is a general rule of thumb for clinical screenings. Discussions positioned younger and older individuals differently in relation to the potential benefits of genomic screening, as shown in this example:

Researcher A: And I think we’d all agree that 80-year-olds would be stupid to test.

Researcher B: For anything.

Researcher A: No offense to 80-year olds, but—

Researcher B: No. You shouldn’t do colonoscopies in an 80-year-old.

Researcher A: Right.

Researcher B: You shouldn’t do mammograms in an 80-year-old. You shouldn’t do this either.

Researcher C: Speaking as a person in my 20s, I think that all of this is more useful to folks in my period of life.

Researcher B: I agree. I mean if nothing else because of competing morbidities.

Researcher A: Right.

Researcher B: Right. By the time you’re 60 there’s all kinds of things that can kill you.

Because older individuals would not receive the same benefits from screening as younger people, an upper age cutoff for participation was proposed. Several possible cutoff points were discussed, including ages 60, 65, and 80, but with no agreement. One researcher worried that without a cutoff, GeneScreen would run the risk of reinforcing the desire for screening among older adults as well as the misconception among lay individuals that screening “does a lot of good when you’re 80.”

This initial assertion that older individuals should be ineligible to participate was reconsidered when notions of familial benefit entered the discussion. Researchers stated that one benefit of including older adults would be the information screening could provide to their children, grandchildren, and other relatives who could be tested if a pathogenic variant was found. As one researcher said, participation “might not actually save the 80-year-old that we test, but [it] could save his grandchildren.”

In addition to the identification of familial benefit, researchers had to remind themselves that they were conducting research and not implementing clinical practice. Their comments reflected awareness of blurred boundaries between the two. On the one hand, researchers wanted screening to mirror as much as possible what they envisioned genetic screening for healthy adults to look like in a clinical setting. On the other hand, the reminders that GeneScreen was a research study, addressing as-yet-unanswered questions about an upper age cutoff, prompted researchers to recognize the benefits of including and studying older adults. One researcher’s comments brought the research and clinical realms together in a rationale supporting the inclusion of older adults:

I think for a research study, [including older adults] makes a lot of sense because we would like to have some people who are positive and identify what their feelings are and those sorts of things. [And] for a clinic, there’s good arguments that a broader range would still benefit people especially as we learn more about the penetrance of this.

The recognition that the inclusion of older adults in a genomic screening research study could lead to familial benefit and to understanding their perspectives on screening, as well as provide data to inform clinical practice, ultimately motivated the decision against excluding adults based on age.

Participants’ perceptions of benefits

The 50 interviewees included equal numbers of men and women, ranging in age from 23–84 years, with the majority being white, non-Hispanic, and highly educated, with relatively high incomes (see Table 1). Participants of all ages had similar responses when asked about the possible benefits of participating in GeneScreen, noting five primary categories of benefits (see Table 2). Each benefit was endorsed by members of all age groups.

Table 2 Participants’ perceived benefits of participating in GeneScreen
Full size table

Participants across all age groups noted the societal benefits of advancing science and helping other people. They also mentioned familial benefit as a reason to participate. Grandparents discussed wanting the information for their children and grandchildren. Those who did not yet have grandchildren talked about wanting the information for future generations. Younger people’s responses echoed those of older people. For instance, a young mother talked about being motivated to join because of her young children who depended on her. She wanted to learn if she had any of the GeneScreen mutations to know if she had transmitted a genetic risk to her children for any of the conditions. Participants who did not have children talked about the genetic information being potentially helpful to siblings and other relatives.

Participants saw the identification of conditions that could be prevented or treated as a personal benefit of GeneScreen. Interestingly, while participants both young and old cited prevention and treatment as reasons to join, some older individuals did not think they would personally benefit from GeneScreen because of their age. For example, when asked “Do you see any possible benefits to you personally or for your family of joining the study?”, one participant answered, “Not for me personally. You know, I’m 60 years old.” Other older participants, however, saw as a major benefit information they might learn that could prolong their life. For example, one older participant commented, “On a more selfish side, I’m curious about my own structure, and although I’m 85 now, maybe there are some things I can do to prolong my life.” Participants of all ages also stated that just having the GeneScreen test results would be a benefit, particularly because they could gain peace of mind if the results were negative.

Participants’ perceptions of risks

As they had for perceived benefits, participants of all ages expressed awareness of the same risks when asked about any hesitations they had in joining GeneScreen. Of the 50 participants interviewed, 14 (28%) saw no risks. For example, when asked whether he saw any possible risks or harms, one 66-year-old said, “No, I don’t. Absolutely not.” A 35-year-old participant echoed this sentiment, saying, “No, I don’t really see any risks or harm. It’s just having more information about your health.”

Others participants did cite possible risks (see Table 3). The most commonly mentioned potential risk was discrimination. It is possible that this perception was intensified by GeneScreen’s policy that a positive screening result, once it was CLIA-confirmed, would be placed in the participant’s medical record. Despite the protections of the Genetic Information Nondiscrimination Act, individuals across all age groups voiced fears that the revelation of a positive result could cause them to be denied health insurance in the future or to have their current health insurance canceled. Individuals also worried that health insurance companies could consider a positive result as constituting a preexisting condition, possibly raising their insurance rates. Others brought up the fact that they could be denied life insurance or experience employment discrimination if they received a positive result.

Table 3 Participants’ perceived risks of participating in GeneScreen
Full size table

Participants from all age groups also mentioned that joining the study carried a potential risk for worry. Individuals were worried about waiting for sequencing results as well as being at risk for a condition associated with a positive result. This risk of worry was alleviated somewhat by the perceived benefits of participating, particularly knowing that the conditions being screened were treatable and manageable.

While individuals across all age groups mentioned these risks, older participants (aged 60 and over) thought that their risks were minimal compared with those of younger individuals because of their age. For example, people over 65 viewed insurance discrimination as less of a threat because of Medicare coverage. As one stated,

“I did have a bit of concern that if I have a condition, that it could adversely affect health insurance or something like that. Now given that I’m 66 and on Medicare it’s less of a concern to me, but it does—it did cause me some second thoughts.”

Similarly, older individuals saw life-insurance discrimination as less of an issue as most already had it. It is important to note, though, that most of those interviewed were financially secure, with an average household income of over $100,000.

Older adults also cited their age as lowering their risk of employment discrimination. As a 60-year-old participant explained, “I’m already approaching retirement age, and so I don’t think I could be discriminated against for job placement because of it. So no, I don’t think there’s any real serious issues.” Not everyone in their 60s may have the desire or means to retire, but for individuals such as this participant, social status combined with age may be viewed as reducing the potential for employment discrimination based on genetic information.

Finally, older individuals stated that their risk of worrying about their future health was smaller than that of younger adults. An older participant explained:

I know some people don’t want to find out something because then they think they’ll worry about it, or it’ll influence how they live their life, or something like that. But maybe cause of my age—I am 75—I’m like, “Hey. Whatever. I’ll shoot for the moon. Whatever.” You know, it doesn’t matter.

Discussion

We have investigated the ways in which the meanings individuals attach to age matter (or do not matter) when considering participation in a preventive genomic-screening research project. GeneScreen researchers originally debated having an upper age limit as an inclusion criterion for participation, but struggled with how to consider age and eligibility in the context of GeneScreen. Although a research study, GeneScreen had strong clinical implications for those who ultimately tested positive for a pathogenic variant. This blurring of lines between research and clinical practice, especially when it comes to deliberations about what results are returned and to whom, produces various challenges.27,28 It is therefore not surprising that the researchers had difficulties at times trying to sort out appropriate frameworks within which to discuss the implications of returning medically actionable results to older adults who may not clinically benefit from the information. However, once they thought beyond individual benefit to familial benefit and learning about how older individuals perceived risks and benefits of participation, researchers ultimately decided not to implement such a limit. As it turned out, over half of GeneScreen participants were 60 and older, and 7% were 80 and older. This supports the findings of Mählmann and colleagues29 that older adults are interested in participating in genomic research, and, as in the GeneScreen project, both young and old individuals saw familial benefit as a reason to participate.

GeneScreen researchers were initially concerned that not having an upper age limit might reinforce the misconception that screening has clinical benefit for older people. Interview data demonstrate that many older adults understood that participating in GeneScreen might not yield personal benefit, given their age, in terms of prevention and treatment. Other older adults said they saw the benefits of participation as the potential for gaining peace of mind from negative results and having the possibility of prolonging their life if they received positive results. However, they may have overestimated the potential for clinical benefit for themselves. That is, the older the person who screens positive, the less likely he or she is to experience symptoms of the conditions being screened. This likelihood translates into potentially fewer health concerns to worry about when undergoing genomic screening. Thus, if older adults are included in future genomic screening programs, the diminishing personal clinical benefit should be clearly communicated to them.

Not only do the types of potential personal benefits vary by age, but so do the risks. GeneScreen participants saw their older age as a protective factor against some of the potential risks for insurance and employment discrimination, especially if retired and on Medicare. The Genetic Information Nondiscrimination Act prohibits employers and health insurers from discriminating on the basis of genetic information, so most individuals regardless of age should be protected. However, recent Congressional activities threaten some of the protections in health insurance for manifested symptoms and in employment for wellness programs. If passed, these types of legislation may create greater risks for younger individuals who are not yet retired and not eligible for Medicare.

In contrast, insurers selling life, long-term care, and disability policies are allowed to take genetic information into consideration. If older adults have not yet secured these types of insurances, they may effectively be barred, given the prohibitive costs associated with their age alone, but the results of a preventive genomic screening program could potentially affect the insurability of individuals of all ages for these types of insurance.

Limitations

Despite our attempt to interview a diverse group of GeneScreen participants, the majority were female, white, and non-Hispanic. Over half had annual incomes of $100,000 or more, and over 90% had at least a college degree. High socioeconomic status may have influenced their minimization of risk. Additionally, everyone interviewed had joined GeneScreen and therefore presumably had an interest in screening. Taken together, these interviewees may well be typical of “early adopters” of genomic technologies.30 A study of ClinSeq participants suggests that early adopters of genomic technologies also have high levels of optimism.30 While our study did not assess optimism, it is possible that the reason the vast majority of interviewees, regardless of their age, saw great benefit and few risks for themselves in joining GeneScreen was this selection bias. Additional research is needed to investigate how a diverse public may understand age in relation to the risks and benefits of genomic screening.

Implications

Inclusion of older adults in research programs has implications for cost, communication during informed consent and return of results discussions, and risks of overdiagnosis and overtreatment. In research projects with limited resources, sequencing participants primarily for familial benefit may not be feasible or central to study aims. GeneScreen paid for testing for family members of UNC participants who could come to UNC facilities. This fit into the overall research aims of assessing the ethical, legal, and social implications of an adult genomic-screening program, and researchers’ valuing of cascade testing for familial benefit. For other research studies, however, these costs could be prohibitive or deviating from research goals. Cost issues are mirrored in the clinical realm, where questions remain regarding who should pay for sequencing and for what purpose. For example, should an individual’s health insurance pay for sequencing done for the benefit of family members?

Inclusion of older adults in research projects also requires researchers to consider working with clinicians to tailor discussions of clinical benefit to minimize the risks of participants’ misunderstanding clinical benefit and subsequent overtreatment. Enrollment of diverse participants requires consideration of how to tailor informed consent, pretest education materials, and communication of return of results for individual circumstances. As found in GeneScreen, many older adults understood that they were unlikely to receive personal clinical benefit from results, but some overestimated the potential clinical benefits. Without thoughtful and tailored communications, older adults participating in genomic research may misunderstand the clinical benefits of the research and may undergo preventive screenings that increase risks of physical complications and overtreatment.

As preventive genomic screening enters clinical settings, will practice recommendations suggest upper age limits? GeneScreen researchers engaged in robust discussion about upper age limits, suggesting that this could be a point of contention for other preventive genomic-screening programs. Clinical evidence of benefits and risks is traditionally interpreted on the basis of chronological age. Our exploration of perceived benefits and risks introduces the social value of screening for older adults, a topic we argue also deserves attention.

There are currently no age guidelines for preventive genomic screening, but we can extrapolate from other types of screening programs to determine how upper age limits may come into play. Clinical screening guidelines often recommend that older adults should not undergo preventive screenings because the risks may outweigh potential benefits, given their life expectancies and the likelihood of comorbidities. For example, a survival benefit from breast or colon cancer screening is unlikely unless an individual is expected to live at least another 10 years.31 The United States Preventive Services Task Force considers age when making recommendations regarding screenings—for example, recommending mammograms for women from 50 to 74 years of age, cervical cancer screenings for women from 21 to 65 years of age, and colorectal cancer screening for men and women from 50 to 75 years of age. The report for colorectal cancer for those 76 to 85 years of age recommends selective screening for this population, noting, “Screening would be most appropriate among adults who 1) are healthy enough to undergo treatment if colorectal cancer is detected and 2) do not have comorbid conditions that would significantly limit their life expectancy.”32

Age cutoffs may serve population health in other clinical contexts, but it is not clear if this is the case for preventive genomic screening, which is associated with a familial cascade benefit. For example, the Evaluation of Genomic Applications in Practice and Prevention Working Group recently recommended that all individuals diagnosed with colon cancer have their tumors sequenced for variants associated with Lynch syndrome, owing in part to the potential for familial benefit.33 As the concept of clinical utility expands to include familial utility, it is likely that excluding older adults will be seen as antithetical to a screening program’s goals.