Abstract
Data are lacking with regard to participants’ perspectives on return of genetic research results to relatives, including after the participant’s death. This paper reports descriptive results from 3,630 survey respondents: 464 participants in a pancreatic cancer biobank, 1,439 family registry participants, and 1,727 healthy individuals. Our findings indicate that most participants would feel obligated to share their results with blood relatives while alive and would want results to be shared with relatives after their death.
Keywords: genomics, return of results, biobank, ethics, family
Introduction
Biobank data and samples frequently endure beyond the life of the individual who provided the sample; this is particularly true for biorepositories that archive data and samples from cancer patients. Those data and samples may be used for research, including after the death of the individual. When the research produces genetic research results (a term used here to include incidental findings and individual research results) that have potential health or reproductive importance for the individual who provided the sample, the results may also have importance for blood relatives. This raises the question of whether the research results should be shared with relatives, at their request or at the initiative of the researchers. The issues are complex even when the research participant is alive, but are particularly challenging after the death of the individual whose data and sample are archived, as the individual may not have been asked their preferences about sharing with family, including after death. Even if the individual’s preferences on sharing have been elicited, investigators and biobank directors may be concerned about withholding genetic research results from relatives that are of potential health significance.
A literature has begun to emerge offering policy and legal perspectives as well as ethical arguments on return of an individual’s results to relatives and the balance of interests between the deceased research participant and the decedent’s family.1 However, few studies report data on the attitudes of cancer biobank participants toward return of genetic research results to relatives, including after death of the participant.2 Empirical data are needed to inform and shape emerging policy and practice. In one recent exploratory study of preferences for disclosure of genetic research results in participants in the OurGenes, OurHealth, OurCommunity biobank, more than half of the individuals surveyed indicated they would want post-mortem return of their own results to a biological relative.3 Other reports, including two qualitative studies addressing public preferences and one population-based study in the Netherlands, have shown that individuals recognize and value learning genetic research information in principle, due to its meaningfulness and potential benefit to family members.4 While these studies provide an important first step in understanding views on return of genetic research results to family, additional research is needed that directly assesses the attitudes and preferences of biobank participants and their family members. To fill this gap, we conducted a survey of perspectives on offering genetic research results to family members. This survey was undertaken as part of a project funded by the National Institutes of Health (NIH) on return of results to relatives, producing empirical research as well as consensus guidance.5 Our study examined the attitudes and preferences of individuals participating in a pancreas cancer biobank, a family member registry on pancreatic, and generally healthy individuals receiving a general medical exam; all three groups were recruited at Mayo Clinic, Rochester, Minnesota. Biobank participants were those with a diagnosis of pancreatic cancer (probands/affected individuals) who participated in Mayo Clinic’s National Cancer Institute (NCI)-funded pancreatic cancer Specialized Program of Research Excellence (SPORE). Family members, including blood relatives and spouses/partners, were participants in a related pancreatic cancer family registry. The question of return of results to relatives is an important one for individuals with pancreatic cancer and their family members, as pancreatic cancer is difficult to detect early and is rapidly fatal, and approximately 10% of pancreatic cancers may be due to susceptibility genes.6 Genetic and genomic research on pancreatic cancer is likely to uncover mutations in genes (e.g., BRCA1 and BRCA2) of participants that may increase risk for other cancers, such as breast and ovarian cancers.7 Genetic and genomic research may additionally discover incidental findings of further potential health importance.8
I. Methods
A. Study population and survey methodology
The sampling frame was comprised of a total of 6,103 individuals: 840 probands in a pancreatic cancer biobank who had a diagnosis of pancreatic adenocarcinoma, 2,471 family members in the pancreatic cancer family registry (reflecting 1,781 blood relatives and 690 spouses/partners), and 2,792 healthy individuals who were patients attending a general medical exam at Mayo Clinic (referred to as “controls” in this study). Controls consented to provide data for pancreatic cancer research and were frequency-matched to pancreatic cancer biobank probands on sex, residence (neighboring five-state area: MN, WI, IA, SD, ND), race/ethnicity, and age (in 5-year increments).9
Survey packets were mailed in three waves between July and September 2013, with two follow-up mailings sent to non-responders at 30-day intervals. Packets included an invitation letter containing a description of the purpose of the study and the voluntary nature of participation, a survey booklet, a postage-paid return envelope, and a toll-free telephone number to allow those not wishing to participate to call, in order to opt out of receiving additional survey mailings. Permission to use responses for research purposes was implied by survey completion and return. The research was approved by the Mayo Clinic Institutional Review Board under expedited review procedures for minimal risk research.
B. Survey development
The content of the survey was developed by our team and guided by qualitative work consisting of 51 in-depth interviews with pancreatic cancer patients and family members sampled from the biobank.10 Interview participants were subsequently excluded from participation in the survey portion of the study. Survey development was also aided by general discussion among members of the NIH-supported Clinical Sequencing Exploratory Research (CSER) Consortium Outcomes and Measures Working Group.11
Initial survey questions underwent pilot testing among 36 individuals at an educational symposium on pancreatic cancer for understandability and to determine the adequacy of response options. Items were revised based on the results obtained during pilot-testing. The revised survey was presented to a 5-member pancreatic cancer patient advocacy group associated with the biobank who considered face validity and clarity of the items and provided feedback on the survey design and appropriateness of the response metrics. Further refinement led to a final survey booklet containing the following nine sections: “Research Participation and Opinions,” “About Your Family,” “Views on Genes and Health,” “Genetic Testing Experience,” “An Example from Genetic Research,” “Practical Considerations in Genetic Research,” “How to Return Genetic Research Results,” “Genetic Research Results and Privacy,” and “About You.” The full content of the survey can be viewed on the website for the University of Minnesota’s Consortium on Law and Values in Health, Environment & the Life Sciences (http://consortium.umn.edu/).
Survey questions were primarily close-ended items with Likert response scales. Questions with “yes” or “no” response options also included “not sure” or “I don’t know” options where applicable; for these items, “yes” responses are reported, with “no,” “not sure” or “I don’t know” collapsed to create a dichotomous variable to reflect absence of endorsement. To assess attitudes across a variety of items, a 5-point Likert rating scale was used with the descriptors “strongly disagree,” “disagree,” “neither agree nor disagree,” “agree,” and “strongly agree.” To reduce reporting complexity, responses were dichotomized; the responses “agree” and “strongly agree” were combined to reflect agreement, while “strongly disagree,” “disagree,” and “neither agree nor disagree” were combined to reflect absence of agreement.
The first four sections of the survey assessed general expectations about participating in genetic research and experience with genetic testing. They also captured the vital status (living or deceased) of selected first-degree blood relatives including biological parents, biological siblings, and biological children.
The fifth section of the survey, “An Example from Genetic Research,” was designed to access perceptions about return of an individual’s genetic research results to relatives for three types of results: 1) a new gene related to pancreatic cancer risk, 2) the BRCA2 mutation which has potential health implications, and 3) a gene indicating carrier-status for the CFTR gene (cystic fibrosis) which thus has potential implications for offspring. The particular gene mutations included in the hypothetical scenarios reflected actual results (incidental findings) that were obtained from testing samples in the pancreatic cancer biobank at Mayo Clinic. Based on the formative interviews, preferences regarding offering these results to relatives were examined using the example of a hypothetical participant named “Pat.” “Pat,” a name chosen purposefully to be gender neutral, was presented alternatively as being: 1) alive at the time of the genetic discovery, or, as the scenario unfolded, (2) deceased at the time the discovery is made, with or without having stated prior wishes about offering genetic results to family. The hypothetical scenarios were presented as “Pat’s Story,” with Pat being described as a 58-year old individual diagnosed with pancreatic cancer in 2009 who enrolled in a cancer biobank and provided a blood sample for research; Pat has a spouse and two biologically-related children, a daughter aged 22 and a son aged 24. In Part 1 of the story, a new gene related to pancreatic cancer risk is discovered and found to be present in Pat’s blood sample. In Part 2 of the story, researchers find a mutation in the BRCA2 gene in Pat’s sample in 2012, and the implications for Pat’s children are described. In Part 3 of the story, Pat’s blood sample tests positive for the cystic fibrosis mutation, although Pat does not have the disease; carrier status and risk of inheritance in Pat’s children and future potential grandchildren are outlined. After each part of the story, a series of eight statements were presented with a 5-point “strongly disagree” to “strongly agree” rating scale. The statements were prefaced with the following: “Please remember that ‘offering’ results means that the person is given the option of saying ‘yes’ or ‘no’ to actually learning the result. The result is not provided unless the person says ‘yes’ to the offer” (emphases in the original). The statements addressed return of results to Pat, privacy and dissemination of information within the family, offering results when Pat’s wishes are unknown, and honoring Pat’s wishes when known, before and after death. After presenting the Pat scenarios, respondents were asked to consider their attitudes toward sharing or keeping private their own genetic research results (assuming they were medically useful).
The section on “Practical Considerations in Genetic Research” addressed two key policy-related issues in genetic research – the right not to know genetic results and the tension that may arise between respecting the source individual’s wishes concerning sharing with family and the potential for family health benefit. This section included “forced-choice” items that instructed respondents to “select only ONE of the following two statements that best reflects their opinion about offering results” (emphasis in the original). The first of these addressed protecting the right not to know (“Not offering results to any research participants, in order to protect those who do not want to know genetic research results, is the right thing to do.” or “Offering genetic research results to all research participants, even at the risk of upsetting those who may not want to be offered results, is the right thing to do.”). The second addressed the potential tension between respecting the individual wishes of the participant on sharing and family benefit (“The most important factor to consider in returning genetic research results is the wishes of the person who provided the sample.” or “The most important factor to consider in returning genetic research results is whether blood relatives will benefit.”). Additional items captured preferences regarding who should make decisions about offering genetic information after a research participant dies, if the participant’s wishes about return to family were not documented before death; whose responsibility it is to offer genetic research results (the research participant or the researcher who makes the discovery); and at what cost to the research.
The following section, “How to Return Genetic Results,” queried the types or characteristics of genetic research results that “should,” “could,” or “definitely should not” be offered to research participants or family members. These items reflected result characteristics such as actionability, disease severity, and result certainty. The last two sections in the survey addressed concerns regarding the privacy of genetic research results within and outside the family and captured respondent demographic information.
C. Statistical analyses
Statistical analyses were performed using SPSS, version 21 (IBM Corp.). Descriptive data are presented as frequencies (n) and valid percentages (%) based on the denominator of those responding to the item (excluding missing data). For ease of presentation, Likert-scale responses are dichotomized as described above, and percent agreement is reported. For selected variables, group differences between pancreatic cancer biobank participants (including affected probands, spouses/partners, and biological relatives) and controls were evaluated to examine patterns in the data. Group differences in categorical variables (such as agreement vs. absence of agreement) were evaluated using z test of proportions, chi-square or Fisher’s Exact tests where appropriate; P values are reported in conjunction with the evaluation of group differences. In light of the multiple comparisons performed, we considered P values ≤0.01 statistically significant and performed a modified Bonferroni adjustment12 to control for inflated error rates due to performing multiple comparisons on the data where appropriate. Preliminary results were presented at the 2014 Annual Meetings of the American Society of Human Genetics and the American Society for Bioethics and Humanities.13
II. Results
A total of 3,630 surveys were returned, reflecting an overall response rate of 59.5%, using the standard definition of response rate involving mail surveys of specifically named persons.14 A lower survey response rate was observed for cancer biobank and family registry participants as compared with controls (57.5% vs. 61.9%, P<0.001), a difference driven primarily by a 55.2% response rate among probands/affected individuals who may have found it difficult to respond due to disease burden. A lower response rate was observed among males as compared with females (56.5% vs. 62.3%, P<0.001). Missing data were minimal, ranging from <1% to 3% across survey items. Missing data were closer to 3% for “forced choice” type items and closer to 1% for Likert-scale responses. No other patterns were observed with regard to missing data.
Survey respondents were predominantly white, non-Hispanic, highly educated individuals (nearly half with 4 years of college or more) with health insurance coverage (Table 1). A greater proportion of cancer biobank and family registry participants (13.1% across probands, blood relatives, and spouse/partners) reported prior experience with genetic counseling as compared with controls (4.6%) (P<0.001). Overall, 85.1% (n=3,044) of respondents had biologically-related children who were still living, suggesting personal relevance of the survey questions; this proportion did not differ between cancer biobank and family registry participants versus controls (84.4% vs. 85.8%, P=0.26). When compared with controls, however, cancer biobank and family registry participants reported higher proportions of living, biologically-related family members: 89.2% of cancer biobank and family registry participants reported one or more biologically-related siblings who were still living (vs. 85.5% controls, P<0.001), 27.5% of cancer biobank and family registry participants reported their biological mother was still living (vs. 23.5% of controls, P=0.007), and 14.5% of cancer biobank and family registry participants reported their biological father was still living (vs. 11.5% of controls, P=0.009). After adjusting the alpha level for multiple comparisons, only the group difference in biological siblings remained statistically significant.
Table 1.
Characteristics of survey respondents (N=3,630).
| Characteristic | All Respondents N=3630 n (%) |
Controls n=1727 n (%) |
Pancreatic Cancer Biobank Participants (Proband/Affected) n=464 n (%) |
Family Registry Participants (n=1439)
|
|
|---|---|---|---|---|---|
| Blood Relative n=1040 n (%) |
Spouse/Partner n=399 n (%) |
||||
|
| |||||
| Sex | |||||
| Male | 1566 (43.1) | 859 (49.7) | 231 (49.8) | 339 (32.6) | 137 (34.3) |
| Female | 2064 (56.9) | 868 (50.3) | 233 (50.2) | 701 (67.4) | 262 (65.7) |
|
| |||||
| Age, years | |||||
| Mean (SD) | 66.2 (12.2) | 69.0 (10.8) | 66.4 (11.3) | 60.3 (13.4) | 69.1 (10.4) |
| Range | 23–99 | 32–96 | 29–94 | 23–99 | 38–94 |
| Median | 67 | 69 | 66 | 61 | 70 |
|
| |||||
| Race | |||||
| White | 3512 (98.2) | 1671 (98.4) | 444 (97.4) | 1012 (98.4) | 385 (98.0) |
| Black/African American | 13 (0.4) | 5 (0.3) | 4 (0.9) | 4 (0.4) | 0 (0) |
| Asian | 18 (0.5) | 8 (0.5) | 3 (0.7) | 4 (0.4) | 3 (0.8) |
| Hawaiian/Pacific Islander | 2 (0.1) | 0 (0) | 0 (0) | 0 (0) | 2 (0.5) |
| American Indian/AK Native | 18 (0.5) | 8 (0.5) | 3 (0.7) | 5 (0.5) | 2 (0.5) |
| Other | 12 (0.3) | 6 (0.3) | 2 (0.4) | 3 (0.3) | 1 (0.3) |
| Missing | 55 | 29 | 8 | 12 | 6 |
|
| |||||
| Hispanic Ethnicity | |||||
| Yes | 30 (0.8) | 12 (0.7) | 6 (1.3) | 8 (0.8) | 4 (1.0) |
| No | 3540 (99.2) | 1686 (99.3) | 446 (98.7) | 1020 (99.2) | 388 (99.0) |
| Missing | 60 | 29 | 12 | 12 | 7 |
|
| |||||
| Marital status | |||||
| Married/Life partner | 2687 (75.5) | 1408 (83.2) | 377 (83.2) | 797 (78.0) | 105 (26.9) |
| Separated/Divorced | 194 (5.5) | 79 (4.7) | 37 (8.2) | 78 (7.6) | 0 (0) |
| Widowed | 538 (15.1) | 143 (8.4) | 27 (6.0) | 82 (8.0) | 286 (73.1) |
| Single/never married | 140 (3.9) | 63 (3.7) | 12 (2.6) | 65 (6.4) | 0 (0) |
| Missing | 71 | 34 | 11 | 18 | 8 |
|
| |||||
| Education | |||||
| High school or less | 688 (19.2) | 338 (19.9) | 111 (24.4) | 157 (15.2) | 82 (20.9) |
| 2 year college/technical school | 1142 (31.9) | 515 (30.3) | 140 (30.8) | 343 (33.3) | 144 (36.6) |
| 4 year college or greater | 1748 (48.9) | 848 (49.9) | 203 (44.7) | 530 (51.5) | 167 (42.5) |
| Missing | 52 | 26 | 10 | 10 | 6 |
|
| |||||
| Employment† | |||||
| Not employed | 229 (6.4) | 87 (5.1) | 62 (13.7) | 67 (6.5) | 13 (3.3) |
| Employed | 1405 (39.4) | 572 (33.7) | 141 (31.2) | 577 (56.2) | 115 (29.4) |
| Retired | 1931 (54.2) | 1037 (61.1) | 249 (55.1) | 382 (37.2) | 263 (67.3) |
| Missing | 65 | 31 | 12 | 14 | 8 |
|
| |||||
| Health insurance coverage | |||||
| No | 43 (1.2) | 12 (0.7) | 3 (0.7) | 22 (2.1) | 6 (1.5) |
| Yes (private, employer, public) | 3530 (98.8) | 1686 (99.3) | 451 (99.3) | 1006 (97.9) | 387 (98.5) |
| Missing | 57 | 29 | 10 | 12 | 6 |
|
| |||||
| Prior experience with genetic counseling | |||||
| Yes | 324 (9.1) | 78 (4.6) | 56 (12.3) | 158 (15.4) | 32 (8.2) |
| No/Unsure | 3239 (90.9) | 1614 (95.4) | 398 (87.7) | 871 (84.6) | 356 (91.8) |
| Missing | 67 | 35 | 10 | 11 | 11 |
Not employed = full time or part time student, unemployed, homemaker, unable to work due to disability; Employed = full time or part time employment, employed but on medical leave; Retired = retired, or retired but working part time.
A. Expectations and attitudes regarding return of genetic research results to participants
Overall, 62.1% (n=2,192) of respondents “expected to learn something about their own genetic results” as a research participant (63.3% of cancer biobank and family registry respondents and 60.8% of controls, P=0.12). Of those who did expect to learn something, 98.8% (99.3% of cancer biobank and family registry respondents and 98.2% of controls, P=0.03) expected to be told “if researchers found something bad (a health risk)” in their blood sample, while 84.6% (88.3% of cancer biobank and family registry respondents and 80.2% of controls, P<0.001) expected to be told “if researchers found something good (I did NOT have a particular health risk)” (emphasis in original) in their blood sample. When asked to assume that the results would be medically useful, 96.2% (n=3,462) of the sample would want to know their genetic research results, and 77.7% (n=2,792) would want researchers to inform their health care provider about their genetic results.
Overall, only 5.4% (n=193) of the sample agreed that they would want genetic research results to be kept private, even after their death (6.5% of controls vs 4.4% cancer biobank and family registry participants, P=0.01). Only 1.8% (n=67) of the sample (1.7% of cancer biobank and family registry respondents and 2.0% of controls, P=0.62) indicated that they “would NEVER want to be offered genetic research results from their sample or a family member’s sample” (emphasis in original). When forced to weigh “offering results to all research participants” (at the risk of upsetting some) against not offering results to any participants” (protecting the participants’ right not know), 86.5% (n=3,038) of survey respondents chose offering results to all, while 13.5% (n=473) chose not offering results to any in order to protect those who do not want to know genetic research results. These proportions did not differ significantly between cancer biobank and family registry participants as compared with controls (P=0.03) based on our threshold of significance.
Over three-fourths (76.7%, n=2,750) of the sample indicated it was the researcher’s responsibility to offer genetic research results to research participants, with the remainder (23.3%, n=835) indicating that the responsibility should rest with the research participant to ask for genetic research results found in the sample. Two additional items probed researchers’ duty to offer results as it relates to cost and their job responsibilities. Overall, 68.5% (n=2,459) agreed that researchers should offer results to participants, no matter how much money it costs. Only 14.2% (n=512) agreed with the statement, “Researchers should NOT be required to offer genetic results because it’s not their job” (emphasis in the original).
B. Responses to the hypothetical “Pat” scenarios
The responses to the statements addressing return of genetic research results were quite similar across the three “Pat” scenarios depicting different genetic findings: a new gene discovery related to pancreatic cancer risk, a BRCA2 mutation, and a CFTR gene mutation, although the responses for the known gene mutations (BRCA2 and CFTR) were more similar to each other than to the responses for the scenario depicting the new gene related to pancreatic cancer risk. The statements and responses for each of the genetic findings are shown in Table 2, with notations denoting patterns in the data between cancer biobank and family registry participants and controls. A large majority of respondents (>90%) favored returning genetic research results to Pat while Pat was still alive (98% in the BRCA2 and CFTR gene mutation scenarios and 95% in the new pancreatic cancer gene scenario). Only about one-third of respondents agreed that Pat should be able to keep the genetic result private from others in the family (32–33% in the BRCA2 and CFTR gene mutation scenarios and 35% in the new pancreatic cancer gene scenario). After Pat’s death, a considerable majority (around 90%) agreed that genetic research results should be offered to Pat’s spouse, and many (77% in the BRCA2 and CFTR mutation scenarios and 70% in the new pancreatic cancer gene scenario) agreed with offering the results directly to Pat’s adult, biological children if Pat’s spouse refused the offer of information. Where Pat was deceased and Pat’s wishes about sharing genetic results were unknown, greater than 80% of respondents agreed that the information should be offered to Pat’s blood relatives. Finally, after Pat’s death in the scenarios, if Pat previously stated wishes not to share genetic information with blood relatives, only about one-third of respondents (31–32% in the BRCA2 and CFTR scenarios and 38% in the new pancreatic cancer gene scenario) agreed that Pat’s wishes should be honored, with a majority of responses reflecting lack of agreement.
Table 2.
Percent agreement across all respondents to offering results across 3 hypothetical scenarios involving the following findings: a new gene related to pancreatic cancer risk, a known cancer risk factor gene mutation (BRCA2), and a mutation revealing carrier status (CFTR) and involving reproductive implications.
| Pat’s Vital Status | Statement (emphases in the original survey) | Percent Agreement when [finding] is: | ||
|---|---|---|---|---|
| Pancreatic cancer gene | BRCA2 gene mutation | CFTR gene mutation | ||
| ALIVE | Researchers should offer Pat the information about the [finding] discovered in Pat’s sample | 94.6 % | 98.2 % | 98.1% |
| Pat should be able to keep information about the [finding] private from others in the family | 34.8 % | 32.7 % | 31.8% | |
| Pat, not the researchers, is responsible for sharing the information about the [finding] with blood relatives (biologically-related family members) | 67.8 % | 59.0 % | 59.3 % | |
| DECEASED | Researchers should ONLY offer Pat’s information about the [finding] to blood relatives if Pat has given EXPLICIT PERMISSION to share genetic results | 72.6 % | 61.7 % | 60.0% |
| If the new discovery is made AFTER PAT’S DEATH, the information about the [finding] should be offered to Pat’s spouse | 88.9 % | 92.0 % | 92.8% | |
| If Pat’s spouse REFUSES the offer of information about the [finding], researchers should offer the results directly to Pat’s children | 69.5 % | 76.7 % | 76.9% | |
| If the new discovery is made AFTER PAT’S DEATH, and Pat’s wishes about sharing genetic information are UNKNOWN, the information about the [finding] should be offered to Pat’s blood relatives | 83.4 % | 87.1 % | 86.8% | |
| If the new discovery is made AFTER PAT’S DEATH, and Pat previously said NOT TO SHARE genetic information, the information about the [finding] should NOT be offered to Pat’s blood relatives | 37.7 % | 32.2 % | 31.2% | |
C. Attitudes regarding the type of genetic research results that should be shared and who should decide
Figure 1 presents respondent’s attitudes (i.e., definitely SHOULD be offered, could be offered, or definitely SHOULD NOT be offered, (emphases in original)) toward offering genetic research results to participants or family members when the result has specified characteristics (e.g., can act on it, is related to a disease that is fatal, etc.). Over 80% of respondents indicated that genetic research results showing an increased risk for a disease that could be prevented or passed to children and a result providing information a person can act on definitely should be offered. These results are internally consistent with the pattern and proportion of responses observed in the Pat scenarios for sharing BRCA2 and CFTR gene mutation results with blood relatives.
Figure 1.
Attitudes regarding offering genetic research results to participants or family members by characteristics of the result.
Respondents were presented with the item, “What if a research participant dies without saying whether his/her genetic information can be offered to family members? Who should make decisions about return of genetic information obtained from the blood sample?” with instructions to mark only one of the following options in response: 1) the research participant’s spouse/partner, 2) blood relatives, 3) personal representative/executor of estate whether or not a blood relative, 4) primary care physician, 5) the researcher, or 6) other. Of the 93.4% (n=3,390) of respondents who adhered to the instructions (checked only one option), three-fourths chose a relative (with 39% choosing the spouse/partner and 36% choosing a blood relative), 8% chose “primary care provider,” 7% chose “personal representative,” 7% chose “researcher,” and 3% selected “other.” Importantly, the remaining 6.6% (n=240) of the sample either checked multiple options (2.4%) or did not check any of the options provided (4.2%), suggesting an unwillingness or inability to identify a single individual to make decisions about their sample after their death.
D. Expectations regarding privacy and sharing genetic research results with relatives
We evaluated individuals’ expectations about balancing proband privacy against sharing with relatives. This series of items was prefaced by the phrase, “assuming the results would be medically useful.” Responses to these items were internally consistent (when phrased either in terms of privacy or in terms of sharing) and favored sharing within families overall, particularly when the respondents were cancer biobank or family registry participants rather than controls (Table 3). After adjusting for multiple comparisons, significant differences between cancer biobank and family registry participants versus controls remained for the items, “I would NOT want my blood relatives to know about my genetic research results” (greater proportion of controls favored privacy) and “I would feel OBLIGATED to share my genetic research results with my blood relatives” (greater proportion of cancer biobank and family registry participants felt an obligation to share with family) (emphases in the original). In addition, 60.5% (n=2,162) of the sample indicated agreement with the statement, “Genetic information belongs to all blood relatives, not just the person who gave the blood sample.” Specifically, agreement was 66.2% among probands, 66.0% among spouse/partners, 59.7% among blood relatives, and 58.3% among controls, reflecting a higher proportion of agreement among cancer biobank and family registry participants as compared with controls (62.6% vs. 58.3%, P=0.009).
Acknowledging that families are comprised of individuals who may not share the same opinions and that offering results is not the same as actually learning them, we queried participant’s beliefs about agreement within their own families. Overall, 67.4% (n=2,422) of respondents agreed that all of their blood relatives would probably have the same opinion about being offered genetic research results. Specifically, agreement was 69.8% among probands, 69.5% among spouse/partners, 69.2% among blood relatives, and 65.2% among controls, reflecting a higher proportion of agreement among cancer biobank and family registry participants as compared with controls (69.4% vs. 65.2%, P=0.007). A somewhat lower proportion, 60.8% (n=2,182), agreed that all of their blood relatives would probably have the same opinion about actually learning genetic research results. Specifically, agreement was 65.2% among probands, 63.0% among spouse/partners, 59.7% among blood relatives, and 59.7% among controls, with no statistically significant difference observed between cancer biobank and family registry participants as compared with controls (61.7% vs. 59.7%, P=0.23).
A forced-choice format was used to ascertain the more important factor to consider in returning genetic research results -- the wishes of the person who provided the sample or whether blood relatives would benefit. Among controls, the two options were fairly evenly divided, with 48% (n=816) choosing the wishes of the individual and 52% (n=884) choosing whether blood relatives would benefit. Among cancer biobank and family registry participants, these options were more split, with 34.6% (n=158) of probands and 34.2% (n=134) of spouse/partners choosing individual wishes and 65.4% (n=299) of probands and 65.8% (n=258) of spouse/partners choosing relatives’ benefit. Among blood relatives, 41.8% (n=427) chose the individual wishes of the person who provided the blood sample as being more important to consider in returning genetic research results over whether blood relatives would benefit (58.2%, n=594). The results of between-group comparisons using the z test of proportions demonstrate that controls significantly differed from probands (z=5.12), spouses/partners (z=4.95), and blood relatives (z=3.13) (all P<0.01).
III. Discussion
Our study findings indicate that a majority of participants expected to learn their own genetic research results, would feel obligated to share their results with blood relatives while alive, and would want genetic research results to be shared with relatives after their death. Consistent with other reports on return of genetic research results15 most respondents desired their own results and expected that researchers would extend an offer to learn them. When given a choice between protecting an individual’s right not to know genetic results versus offering results to all, fewer than one in five respondents favored the former, even at the risk of upsetting some people by offering results.
On return to family, a consistent pattern emerged whereby the controls favored privacy and individual benefit proportionately more than the cancer biobank respondents and family registry participants, who favored family sharing and family benefit. The differences observed between groups warrants further investigation and replication in other research settings. These findings suggest that different populations (healthy patients presenting for medical examination vs. patients and blood relatives affected by a life-threatening cancer with potential heritability of genetic risk variants vs. spouses or partners of cancer patients) may differ in their perspective on the issue of return of results to relatives.
Our results were consistent with other reports suggesting that individuals recognize the potential value of genetic information to relatives16 and highlight that for many research participants, there is a feeling of obligation to share their genetic results with blood relatives. Particularly when genetic research results show an increased risk of diseases that could be prevented or could be passed to children or information a family member could act on, greater than 80% of respondents indicated the result definitely should be offered. This supports widening the scope of research and analysis on return of results beyond return to the participant, to the question of whether and how results should be offered to the family.
Through the use of a hypothetical scenario, we accessed attitudes toward post-mortem return to family that indicated a high level of support for sharing results with family members, even if not authorized by the individual providing the sample or against the sample provider’s wishes. Our sample was split on honoring the previously expressed wishes of the (now deceased) research participant to not share results with family, with about two-thirds agreeing to go against the participant’s expressed wishes to withhold information, particularly in the case of pathogenic and clinically actionable gene mutations, including results with reproductive implications (BRCA2 and CFTR), and one-third expressing lack of agreement. These findings have potential ethical implications for researchers. Our results support the conclusion that the tension between participant control of dissemination of individual results after death and potential family benefit remains unresolved, calls for investigation in multiple populations, and requires an approach that considers differences of view.
As further evidence of support for sharing genetic research results with family members who may benefit, greater than 70% of respondents endorsed offering results directly to the sample provider’s offspring in the case where a spouse refuses the offer of genetic results. This finding suggests that difficulties can arise when an individual designates a single gatekeeper of information within the family, particularly when that individual is not a blood relative, and the designated gatekeeper declines to share the information. Our observation that 6% of our sample was unable or unwilling to designate a single individual as the person responsible for decisions about sharing the participants’ results after their death suggests that some participants may struggle with policies that necessitate such a selection/designation.
A number of survey strengths provide confidence in the data, including a favorable response rate resulting in a large sample size for all three groups of respondents (cancer biobank participants, family registry participants, as well as controls) and minimal missing data. We attribute these strengths to the careful pilot work that informed and shaped our survey design and methodology. Nevertheless, our study has some potential weaknesses including the socio-demographic homogeneity of the overall sample, and the fact that our sample may be biased toward well-educated individuals who were covered by some type of health insurance. These characteristics may limit the generalizability of our results. Lastly, it is possible that individuals who hold ambivalent or overly negative attitudes toward the topic of returning genetic research results may have been underrepresented among survey respondents. Although we are unable to assess this potential bias, we did attempt to minimize its impact by designing survey items and questions that included the ability to express negative, neutral, or unsure attitudes (as well as positive attitudes) regarding return of genetic research results to participants and family members.
The analyses presented in this paper offer primarily a descriptive presentation of our findings, with exploratory comparisons performed between controls and cancer biobank as well as family registry participants. To avoid capitalizing on our large sample size and potentially overstating the significance of our results, we took a conservative analytic approach by setting our initial alpha level at the more stringent P<0.01 versus the more traditional P<0.05, and by applying a modified Bonferroni correction when conducting multiple comparisons (statistical tests) on the data. In addition, we emphasize patterns of responses, de-emphasize statistical significance, and report the results of all comparisons performed to avoid over-interpreting what are sometimes modest observed differences between groups.
Conclusion
Our data demonstrate that a majority of individuals, but certainly not all, who provide blood samples for research prefer to be offered the opportunity to learn genetic research results generated from their sample. For many, this preference extends to offering their genetic research results to family members, particularly blood relatives, after their own death. In a scenario designed to avoid the inherent threat of directly considering one’s own death, our data uncovered a split on whether to honor the proband’s previously stated wishes to withhold genetic information from blood relatives, with about two-thirds favoring sharing with relatives (not honoring the proband’s wishes) and one-third not agreeing to this. Finally, our data show that a number of factors are relevant to attitudes toward family return, including the type of genetic finding, characteristics of the disease associated with the genetic result, as well as the way in which the participant came into genetic research, that is, as a “healthy control,” or as an individual affected by cancer directly (proband), or indirectly (family member of a proband). Taken together, these findings suggest that biobanks and genomic research projects should consider the emerging policy and practice on whether and how a participant’s research results may be shared within the family. In order to fully inform policy and guidelines about return of genetic research results to family, additional empirical studies are needed that assess the attitudes and preferences of more diverse populations across a variety of scenarios. Moreover, longitudinal studies are needed that investigate experiences with different models of return to family, with evaluation of psychological and behavioral outcomes in probands and family members.
Acknowledgments
This project was supported by grants # R01-CA154517 from the National Cancer Institute (NCI) and the National Human Genome Research Institute (NHGRI), # P50-CA102701 and # R01-CA97075 from the NCI, # P20-HG007243 from NHGRI, and # UL1-TR000135 from the National Center for Advancing Translational Sciences (NCATS). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NCI, NHGRI, NCATS, or the National Institutes of Health. We are deeply grateful to the members of RAPPORT for their contributions and guidance throughout this project. We thank Maggie Breslin (graphic design), Catie Erding (masterful logistics), and Krista Sigurdson (survey pilot-testing) for their contributions to this work. We gratefully acknowledge the generosity of our survey participants who gave their time, shared their thoughts, and ultimately made this work possible.
Biographies
Carmen Radecki Breitkopf is Associate Professor of Health Services Research in the Department of Health Sciences Research at Mayo Clinic College of Medicine in Rochester, Minnesota. She earned her Master’s and Doctoral degrees in Psychology from The State University of New York at Albany.
Gloria M. Petersen is Professor of Epidemiology in the Department of Health Sciences Research at Mayo Clinic College of Medicine in Rochester, Minnesota. She earned her Master’s degree in Anthropology from University of Oregon, and Ph.D. in Anthropology from UCLA. She is a Founding Fellow of the American College of Medical Genetics and Genomics.
Susan M. Wolf, J.D., is McKnight Presidential Professor of Law, Medicine & Public Policy; Faegre Baker Daniels Professor of Law; Professor of Medicine; Faculty member, Center for Bioethics; and Chair, Consortium, on Law and Values in Health, Environment & the Life Sciences at the University of Minnesota. She is one of three Principal Investigators on NIH/NCI/NHGRI grant # R01 CA154517 on return of genomic results to family members, including after the death of the proband.
Kari G. Chaffee is an MS Statistician III in the Department of Health Sciences Research at Mayo Clinic College of Medicine in Rochester, Minnesota. She earned her Master’s degree in Statistics from Iowa State University.
Marguerite E. Robinson is the Program Manager for Biomedical Ethics Research at Mayo Clinic in Rochester, Minnesota. She earned her Master’s degree in Religion & Ethics from Yale University Divinity School and Master’s degree in Biotechnology from Columbia University.
Deborah R. Gordon is Assistant Professor of Medical Anthropology in the Department of Anthropology, History, and Social Medicine at the University of California, San Francisco and an Affiliate of the Berkeley Center for Social Medicine at University of California, Berkeley. She earned her B.A. in Anthropology at UC Berkeley, her Master’s degree at the Tel Aviv University and UCSF, and her Ph.D. in Medical Anthropology at UCSF/UC Berkeley. She has conducted qualitative, ethnographic research on a wide range of health-related topics in the United States, Israel, and Italy.
Noralane M. Lindor, M.D., is Professor of Medical Genetics in the Department of Health Sciences Research at Mayo Clinic in Scottsdale, Arizona. She received her Bachelor of Arts degree from the University of Minnesota, her doctorate of medicine from Mayo Medical School in Rochester, Minnesota and did her residencies at Bowman Gray School of Medicine in Winston-Salem, North Carolina, and at Mayo Clinic in Rochester, Minnesota.
Barbara A. Koenig, Ph.D., is Professor of Bioethics and Medical Anthropology based at the Institute for Health & Aging, University of California, San Francisco. Currently, she co-directs a Center of Excellence in ELSI Research that focuses on translational genomics, co-leads an NCI/NHGRI RO1 on return of results in genomic biobanks, and directs the ELSI component of a U19 award focused on newborn screening in an era of whole genome analysis.
Contributor Information
Carmen Radecki Breitkopf, Email: radeckibreitkopf.carmen@mayo.edu, Associate Professor, Department of Health Sciences Research, Division of Health Care Policy & Research, Mayo Clinic College of Medicine, 200 First St. SW, Rochester, MN 55905, Tel.: 507-266-0969.
Gloria M. Petersen, Email: petersen.gloria@mayo.edu, Professor of Epidemiology, Purvis and Roberta Tabor Professor, Department of Health Sciences Research, Mayo Clinic College of Medicine, 200 First St. SW, Rochester, MN 55905, Tel: (507) 538-1563, Fax: (507) 266-2478.
Susan M. Wolf, Email: swolf@umn.edu, McKnight Presidential Professor of Law, Medicine & Public Policy, Faegre Baker Daniels Professor of Law, Professor of Medicine; Faculty member, Center for Bioethics, Chair, Consortium on Law and Values in Health, Environment & the Life Sciences, University of Minnesota, 325 Johnston Hall, 101 Pleasant St. S.E., Minneapolis, MN 55455, Tel.: 612-301-3905,.
Kari G. Chaffee, Biomedical Statistics and Informatics, Department of Health Sciences Research, 200 First St. SW, Rochester, MN 55905, Tel: 507-266-2893, Chaffee.Kari@mayo.edu.
Marguerite E. Robinson, Email: Robinson.marguerite@mayo.edu, Program Manager for Biomedical Ethics Research, Mayo Clinic, 200 First St. SW, Rochester, MN 55905, Tel: 507-538-3402.
Deborah R. Gordon, Email: deborah.gordon@ucsf.edu, Assistant Professor, Department of Anthropology, History, and Social Medicine, University of California, San Francisco, 3333 California St., #485, San Francisco, CA 94118, Tel: (415) 254-4707.
Noralane M. Lindor, Email: nlindor@mayo.edu, Noralane M. Lindor, M.D., Professor of Medical Genetics, Department of Health Sciences Research, Mayo Clinic College of Medicine, Collaborative Research Building, 13208 East Shea Boulevard, Scottsdale, AZ 85259, Tel: 480-301-6817.
Barbara A. Koenig, Email: barbara.koenig@ucsf.edu, Professor, Institute for Health & Aging, University of California, San Francisco, 3333 California St., #340, San Francisco, CA 94118, Tel: (415) 476-3786.
References
- 1.Milner LC, et al. Relationships Matter: Ethical Considerations for Returning Results to Family Members of Deceased Subjects. American Journal of Bioethics. 2013;13(10):66–67;. doi: 10.1080/15265161.2013.828533. [DOI] [PubMed] [Google Scholar]; Rothstein MA. Disclosing Decedents’ Research Results to Relatives Violates the HIPAA Privacy Rule. American Journal of Bioethics. 2012;12(10):16–17. doi: 10.1080/15265161.2012.699588. [DOI] [PubMed] [Google Scholar]; Black L, McClellan KA. Familial Communication of Research Results: A Need to Know? Journal of Law, Medicine & Ethics. 2011;39(4):605–613. doi: 10.1111/j.1748-720X.2011.00627.x. [DOI] [PubMed] [Google Scholar]; Tassé AM. The Return of Results of Deceased Research Participants. Journal of Law, Medicine & Ethics. 2011;39(4):621–630. doi: 10.1111/j.1748-720X.2011.00629.x. [DOI] [PubMed] [Google Scholar]
- 2.See Amendola LM, et al. Patient’s Choices for Return of Exome and Genome Sequencing Results to Relatives in the Event of their Death. Journal of Law, Medicine & Ethics. 2015 doi: 10.1111/jlme.12290.
- 3.Allen NL, et al. Biobank Participants’ Preferences for Disclosure of Genetic Research Results: Perspectives from the OurGenes, OurHealth, OurCommunity Project. Mayo Clinic Proceedings. 2014;89(6):738–746. doi: 10.1016/j.mayocp.2014.03.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Amendola et al., supra note 2; Bollinger JM, et al. Public Preferences Regarding the Return of Individual Genetic Research Results: Findings from a Qualitative Focus Group Study. Genetics in Medicine. 2012;14(4):451–457. doi: 10.1038/gim.2011.66.Meulenkamp TM, et al. Communication of Biobanks’ Research Results: What Do (Potential) Participants Want? American Journal of Medical Genetics. 2010;152 Part A(10):2482–2492. doi: 10.1002/ajmg.a.33617.Murphy J, et al. Public Expectations for Return of Results from Large-Cohort Genetic Research. American Journal of Bioethics. 2008;8(11):36–43. doi: 10.1080/15265160802513093.
- 5.See Wolf SM, et al. Returning a Research Participant’s Genomic Results to Relatives: Analysis and Recommendations. Journal of Law, Medicine & Ethics. 2015 doi: 10.1111/jlme.12288.
- 6.Howell LA, et al. Receptivity and Preferences of Pancreatic Cancer Family Members for Participating in Lifestyle Programs to Reduce Cancer Risk. Hereditary Cancer in Clinical Practice. 2013;11(1):3, 9. doi: 10.1186/1897-4287-11-3. [DOI] [PMC free article] [PubMed] [Google Scholar]; Radecki Breitkopf C, et al. Factors Influencing Receptivity to Future Screening Options for Pancreatic Cancer in Those With and Without Pancreatic Cancer Family History. Hereditary Cancer in Clinical Practice. 2012;108(1):9. doi: 10.1186/1897-4287-10-8. [DOI] [PMC free article] [PubMed] [Google Scholar]; Hruban RH, et al. Update on Familial Pancreatic Cancer. Advances in Surgery. 2010;44 doi: 10.1016/j.yasu.2010.05.011. [DOI] [PMC free article] [PubMed] [Google Scholar]; Petersen GM, et al. Pancreatic Cancer Genetic Epidemiology Consortium. Cancer, Epidemiology, Biomarkers and Prevention. 2006;15(4):704–710. doi: 10.1158/1055-9965.EPI-05-0734. [DOI] [PubMed] [Google Scholar]
- 7.Zhen DB, et al. BRCA1, BRCA2, PALB2, and CDKN2A Mutations in Familial Pancreatic Cancer: A PACGENE Study. Genetics in Medicine. 2014 Nov 20; doi: 10.1038/gim.2014.153. advance online publication. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Grant RC, et al. Prevalence of Germline Mutations in Cancer Predisposition Genes in Patients with Pancreatic Cancer. Gastroenterology. 2015;148(3):556–564;. doi: 10.1053/j.gastro.2014.11.042. [DOI] [PMC free article] [PubMed] [Google Scholar]; McWilliams RR, et al. Prevalence of CDKN2A Mutations in Pancreatic Cancer Patients: Implications for Genetic Counseling. European Journal of Human Genetics. 2011;19(4):472–478. doi: 10.1038/ejhg.2010.198. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.McWilliams RR, et al. Nucleotide Excision Repair Pathway Polymorphisms and Pancreatic Cancer Risk: Evidence for Role of MMS19L. Cancer, Epidemiology, Biomarkers & Prevention. 2009;18(4):1295–1302. doi: 10.1158/1055-9965.EPI-08-1109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Koenig BA, et al. Returning ‘Actionable’ Results to Family Members in a Pancreatic Cancer Biobank” (Session #25). Presented at the 62nd Annual Meeting of the American Society of Human Genetics; November 8, 2012; San Francisco, CA. [Google Scholar]
- 11.Gray SW, et al. Social and Behavioral Research in Genomic Sequencing: Approaches from the Clinical Sequencing Exploratory Research Consortium Outcomes and Measures Working Group. Genetics in Medicine. 2014;16(10):727–735. doi: 10.1038/gim.2014.26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Holland BS, Copenhaver MD. Improved Bonferroni-Type Multiple Testing Procedures. Psychological Bulletin. 1988;104(1):145–149. [Google Scholar]; Holm S. A Simple Sequentially Rejective Multiple Test Procedure. Scandinavian Journal of Statistics. 1979;6(2):65–70. [Google Scholar]
- 13.Koenig BA, et al. Returning Incidental Findings to Family Members of Deceased Research Participants: Perspectives from a Cancer Biobank,” (Program #2360M). Presented at the 64th Annual Meeting of the American Society of Human Genetics; October 20, 2014; San Diego, CA. [Google Scholar]; Koenig BA, et al. Should Return of ‘Incidental Findings’ Include Family Members of Deceased Research Participants? Findings from a Cancer Biobank,” Presented at the 16th Annual Meeting of the American Society for Bioethics and Humanities, October 17, 2014, San Diego, CA [Google Scholar]
- 14.The American Association for Public Opinion Research. Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys. 7. Deerfield, IL: American Association for Public Opinion Research; 2011. [Google Scholar]
- 15.Bollinger JM, et al. Public Preferences Regarding the Return of Individual Genetic Research Results: Findings from a Qualitative Focus Group Study. Genetics in Medicine. 2012;14(4):451–457. doi: 10.1038/gim.2011.66. [DOI] [PMC free article] [PubMed] [Google Scholar]; Meulenkamp TM, et al. Communication of Biobanks’ Research Results: What Do (Potential) Participants Want? American Journal of Medical Genetics Part A. 2010;152A:2482–2492. doi: 10.1002/ajmg.a.33617. [DOI] [PubMed] [Google Scholar]; Murphy JM, et al. Public Expectations for Return of Results from Large-Cohort Genetic Research. American Journal of Bioethics. 2008;8(11):36–43. doi: 10.1080/15265160802513093. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Amendola et al., supra note 2.