Returning individual‐specific results of a dementia prevalence study: insights from prospective participants living in Switzerland

Objectives To explore prospective participants' preferences regarding the return of their individual‐specific results from a dementia prevalence study (a probabilistic diagnosis of dementia). Methods/Design We conducted a qualitative study with 22 individuals aged 45 to 86 and resident in the Canton of Ticino (Switzerland). Participants had previously joined the validation phase of an epidemiological study into dementia and its impact. Results We found that individuals welcome the return of their individual‐specific results, provided these meet a number of validity, clinical, and personal utility criteria. They justify researchers' duty to return study findings with the principles of beneficence (eg, providing information that can help participants' medical decision‐making) and justice (eg, acknowledging participants' efforts to help research by sharing their personal information). Furthermore, individuals anticipate societal benefits of the return of individual specific study findings, including improved interpersonal relationships among individuals and decreased dementia‐related stigma. Conclusions Our findings suggest that researchers should address the return of individual‐specific study results early on during study design and involve prospective participants in identifying both the conditions under which results should be offered and the perceived individual and societal benefits returning can have.

in participation rates in epidemiological studies over the past 30 years, which has witnessed an even steeper decline in more recent years. 6,7 The success of population studies depends on the voluntary participation of individuals who donate their time and their personal health information, often accepting a certain degree of risks to their welfare. Therefore, researchers should take into account legitimate expectations of individuals to receive some personal gain from participation, such as research results or information on their health. 8 Evidence suggests that current dementia research ethics policies and norms are not aligned with participants' preferences and may hinder equitable opportunities to take part in epidemiological research. 9,10 Among the reasons to participate in DNA biobank studies, expectations of personal benefit through health information are prominent. [11][12][13] Willingness to participate is higher when participants are offered a chance to know their individual-specific results. Consensus statements indicate that the return of research results should occur when the findings are clinically relevant. [14][15][16][17] Recently, scholars have highlighted the urgent need for guidance in related decision-making contexts, such as when to provide family members with access to health-related data of dementia patients and how to manage the return of individual results from dementia research. 18 Lack of clarity on how to meaningfully interpret positive results in a clinical sense, coupled with fears of causing anxiety or depression to subjects have so far prevented most investigators from disclosing individual-specific research results to dementia research participants. [19][20][21][22] However, little has been done to address such issues in dementia epidemiological research. We conducted a qualitative study to explore preferences regarding and understanding of the return of individual-specific results, in view of an epidemiological study into dementia and its impact, whose primary outcome is prevalence of dementia based on a probabilistic dementia diagnosis.

| Study design
This article presents a follow-up investigation of a validation study conducted between March and October 2019 with 160 dyads in the Canton of Ticino, Switzerland. Each dyad consisted of an old person (aged +65 years) and his/her informant (ie, a carer, spouse, or child).
Therefore, in the original study, the number of informants was equal to the number of older adults (n = 80 in each group). In the present investigation, we conducted semi-structured, in-depth interviews with a sample of the validation study's participants. We included both participants and their informants to generate a more comprehensive understanding of their attitudes, thoughts, and preferences regarding the return of individual-specific results. 23

| Study participants
Inclusion criteria for participants were age (being 65 or older) and place of residence (Ticino). Informants had to be older than 18 years.
In the informed consent for the validation study, participants gave permission to be contacted for a follow-up interview. No individual-specific results were offered in the validation study. Of the 320 participants of the validation study (160 dyads), 35 individuals provided their contact details and were contacted by a member of the research team (RA), who provided a description of the present study, and reference to its ethical safeguards. We offered no financial incentives for participation.
Recruitment lasted between December 2019 and January 2020.

| Data collection
The interviews were conducted in Italian and face-to-face by a member of the RA between December 2019 and January 2020, either at the participant's home or the University, according to participants' preferences. Three couples requested to be interviewed together. In these cases, both participants took turn to answer to all questions posed by the interviewer, resulting in two answers for each question, which were analyzed independently. The interview guide was Participants were asked to imagine that they might be receiving information on whether they are likely to have dementia or not. We followed Holstein and Gubrium's "active interview" model, which conceptualizes the interviewer and interviewee as equal, collaborative partners in the social production of meaning around a given research topic. 24 After permission from participants, we digitally video-recorded all interviews. One member of the RA transcribed them verbatim, while a second member (IF) cross-checked the recorded interviews to guarantee accurate documentation of the discussion. To determine the point of data saturation, we relied on the concept based on thematic  25,26 In this model, which relates to the emergence of new codes or themes, saturation is confined to the level of analysis and data collection is interrupted when no novel insights can be extracted from the data. 26 We reached saturation of the data after the 11th interview. We collected data on participants' sociodemographic characteristics at the end of each interview.

| Data Analysis
Two members of the RA (MF* and IF) independently performed an inductive thematic analysis of the 19 transcripts. We followed the sixstage comprehensive thematic analysis approach developed by Braun and Clarke. 27 The key phases of the coding process include familiarizing with the content of the transcripts, highlighting meaningful quotes regardless of their length, condensing them under a number of labels, organizing the generated labels hierarchically, creating relationships between them, and identifying remarkable quotations to represent thematic similarities, differences, and contradictions. This method allows to unveil themes that may not have been covered by established theory. 28,29 To validate the results, comparisons between the two coders took place multiple times in-between each of the above-mentioned phases, so that themes, labels, and quotations were constantly discussed, and interpretation discrepancies resolved through dialog and reference to the transcripts. We performed the analysis with the qualitative research software NVivo. 30 No translation of the transcripts was needed, as the analysis was conducted in the original language. This article follows the Standards for Reporting Qualitative Research (SRQR). 31 3 | RESULTS

| Participants' characteristics
The final sample included 22 participants (11 women), including three couples that requested to be interviewed together. The mean age was 71 years (SD = 9.3; range = 45-86). Most participants had completed secondary school (n = 11), were retired (n = 17), and resident in urban districts (n = 13; see Table 1 for an overview of participants' sociodemographic characteristics). The focus of this article will be on individual-specific results. Most participants were in favor of the return of their individual-specific results. The majority stated that they expected them, even though they favored different approaches to disclosure. Most (n = 12) reported that they would not be afraid of the disclosure and had a strong desire to know their personal results. Four participants also reported anticipated fear with respect to the disclosure but declared that they would nevertheless prefer to know the results in order to act. Finally, two participants reported they would be afraid and prefer not to know, stating they would not be able to psychologically manage a risk of dementia.

| Preferences regarding the return
Preferences on how results should be returned varied across participants, and according to the type of results. Most participants stated that they would accept their results being communicated via email, postal letter, or phone call in case they were negative (no risk of dementia), but agreed that, in case of positive results, a face-to-face meeting with the investigator would be more appropriate.
With regards to whom they would like their individual-specific results to be shared with, most mentioned their spouse and children.

| A matter of relationship
Almost all participants reported that, for optimal disclosure, it is central that researchers establish a personal relationship with participants early on during recruitment. formed not in the context of the present study], I would probably not do it again, but this is not due to the biopsy itself but rather to the lack of emotional support, and for me emotional support is fundamental.

| DISCUSSION
The goal of this study was to explore participants' preferences regarding the return of their individual-specific results in a prospective study on dementia prevalence. We found that the return of individualspecific results should repose on two main principles, transparency and reciprocity, and be dialogic. Investigators should disclose anything that could inform health-related decision-making, and the benefits that may derive from individual uses of participants' information should be mutual. Next, participants reported that investigators should assess the validity, clinical utility, and actionability of study research results, and researchers recognize that the highest benefit is helping treat or prevent disease. [32][33][34][35][36][37] In the biobanking context, providing participants with their individual-specific study results that are valid and with potential clinical utility, is perceived by both researchers and participants as a means of demonstrating respect and gratitude for their contributions. [38][39][40] Studies have shown that provision of such information may lead to greater trust, accountability and engagement in research, which is, in turn, a strong predictor of research participation. 8 Finally, the obligation to respect participants' ownership of the information they provided-and thus offer "return" rather than "disclosure"-is consistent with the shift towards participant engagement. 41 Traditionally, epidemiologists strive to maximize the potential benefits of research by communicating results to study participants in a timely fashion. 42 This study expands the current interpretation of the principle of beneficence in epidemiological research. 42 Our results indicate that the return of study results can also contribute to reducing stigma, through improved awareness and understanding of dementia at a collective level. This is consistent with the pivotal role of dementia awareness of the WHO public health response to dementia. 1  People might frame data ownership as private property, thus perceiving that data belong first and foremost to themselves. 46  bias. To reduce this, the interviewer adopted techniques such as nominative questions and employed a nonjudgmental approach. Third, each interview was videotaped by a video-maker. To reduce this contextual bias, the video-maker was trained to limit intrusiveness and participants could choose where they would feel more at ease to be interviewed.
Fourth, being dementia a sensitive topic to embark on, this may have had an effect on study participants in terms of difficult emotions and impaired openness. To mitigate this information bias, we established rapport with study participants by telephone before data collection, and fostered reciprocal trust through dialog. Finally, the language (ie, Italian) and nationality (ie, Swiss) of the participants may limit the generalization of our findings, which should be interpreted and applied cautiously to populations of other geographic and linguistic regions.

| CONCLUSIONS
We found that, in the context of a dementia prevalence study, participants expect their rights both to know and not to know their results to be respected, provided the meaning and potential clinical implications of study findings have been previously assessed and clarified.
Considering the implications of the issue of result disclosure for decision to participate and the representativeness of dementia epidemiological study samples, investigators should offer participants an ample set of options on the return of their individual-specific results. Epidemiologists' primary roles are the design and conduct of scientific research and the public health application of scientific knowledge. 42 This includes the reporting of results not only to the scientific community and society but also to research participants. 42 However, it is not clear if this applies to aggregate or individual-specific results. A tradeoff between anonymising data and being able to provide individualspecific results may exist and should be adequately accounted for.
Since formal guidance is lacking, we call for evidence-based guidelines on how to assess the duty to return individual-specific results in dementia epidemiological research. Finally, ethics committees should support the development of plans to return individual research results, and additionally assess whether they were developed in alignment with prospective participants' needs, preferences, and values.