Ethical, legal, and social implications of digital health: A needs assessment from the Society of Behavioral Medicine to inform capacity building for behavioral scientists

Abstract

The ethical, legal, and social implications (ELSIs) of digital health are important when researchers and practitioners are using technology to collect, process, or store personal health data. Evidence underscores a strong need for digital health ELSI training, yet little is known about the specific ELSI topic areas that researchers and practitioners would most benefit from learning. To identify ELSI educational needs, a needs assessment survey was administered to the members of the Society of Behavioral Medicine (SBM). We sought to identify areas of ELSI proficiency and training need, and also evaluate interest and expertise in ELSI topics by career level and prior ELSI training history. The 14-item survey distributed to SBM members utilized the Digital Health Checklist tool (see recode.health/tools) and included items drawn from the four-domain framework: data management, access and usability, privacy and risk to benefit assessment. Respondents (N = 66) were majority faculty (74.2%) from psychology or public health. Only 39.4% reported receiving “formal” ELSI training. ELSI topics of greatest interest included practices that supported participant engagement, and dissemination and implementation of digital tools beyond the research setting. Respondents were least experienced in managing “bystander” data, having discussions about ELSIs, and reviewing terms of service agreements and privacy policies with participants and patients. There is opportunity for formalized ELSI training across career levels. Findings serve as an evidence base for continuous and ongoing evaluation of ELSI training needs to support scientists in conducting ethical and impactful digital health research.

In this article, we uncover important knowledge and experience gaps among practitioners and researchers with regard to the ethical, legal, and social implications of technologies that collect personal health data.

Implications.

Practice: The needs assessment revealed important areas for further ELSI training, such as implementation and dissemination of digital health research and practice, and that a staged approach to ELSI training may be indicated.

Policy: The entire research ecosystem, including organizational, institutional, and community levels, needs to be responsive to unique ELSI-related challenges associated with digital health research and practice; developing better training infrastructures are paramount to these efforts.

Research: Future research includes ongoing assessments to understand ELSI training needs and best practices for evidence-based learning across partners, including those who: (i) develop the technologies (developers), (ii) procure and deploy the technologies (researchers, clinicians), and (iii) are asked to use the technologies (research participants, patients, people).

Introduction

Behavioral scientists are increasingly using digital tools and strategies to enhance participant recruitment, conduct observational and interventional studies, gather and transmit personal data, and return research results to participants [1–5]. Often called “digital health” research, these tools are used across health conditions (e.g. mental health, cardiovascular disease, cancer) to support research on health promotion, disease prevention, and treatment. Some examples are: body worn sensors used to study daily activity with a goal of identifying personalized interventions to improve health [6]; remote sensors placed on pill boxes/bottles to monitor medication adherence [7, 8]; mobile apps used to support real time data collection and feedback [9, 10]; and social network platforms used to deploy behavioral interventions [11].

These new digital approaches have led to an increased interest in understanding the related ethical, legal, and social implications (ELSIs) specific to this work [11–14]. ELSI research seeks “to identify, understand, and make recommendations regarding the challenges posed by genetic information, technologies, and practices for individuals, families, and communities” [15]. The study of ELSIs emerged as a result of the Human Genome Project initiated by the National Center for Human Genome Research at the National Institutes of Health (NIH) and the Department of Energy [16]. As that project was initiated there was a call for proposals to explore the ELSIs of conducting genomic and genetic research. Since then, the NIH has continued to fund ELSI research and has expanded its focus on the ethics of biomedical and behavioral research more generally. While broadly defined and field-dependent, ELSIs typically include considerations that revolve around the protection and preservation of individuals and communities, such as unintended societal consequences of novel technologies, policy and regulation, accessibility, privacy concerns, data ownership, or responsible applications of new technologies [17, 18].

There is a growing need for ELSI research and expertise in the field of digital health. In 2017, there was an estimated 470% increase in the number of digital health research studies funded by the NIH over a 10-year period [19] and it has been established that more diverse, and potentially vulnerable, communities could be reached with these tools [20]. However, studies have also documented the lack of guidance available through the affiliated institutional review boards (IRBs) for researchers who were planning to conduct research using digital health tools (e.g. social network platforms [20]). With little access to guidance and faced with an absence of “best practices,” researchers have begun to explore and share experiences and recommendations within the greater digital health community [21, 22].

One example of such initiatives is the Research Center for Optimal Digital Ethics in Health (ReCODE Health), which aims to improve awareness and application of ethical principles and practices among all involved in digital health research, including participants. ReCODE Health hosts the Connected and Open Research Ethics (CORE) platform [22], a resource library and discussion forum that offers educational resources to the digital health community. In 2018, the founders developed a Digital Health Checklist (DHC) tool to promote consideration of ELSIs among researchers selecting technologies for use in their research [13]. The team iteratively developed the DHC tool by examining key ethical frameworks and engaging field experts. This work led to defining ELSI-related domains (e.g. privacy, data management) and developing a checklist to ensure technologies used in research are aligned with features of these domains; the DHC tool can be found at https://recode.health/tools/. These efforts are aligned with recent calls to center research on issues related to ethics and justice (rather than scientific or technical goals) to reduce public mistrust in science and maximize public benefit. Research that prioritizes ELSI from the outset via tools like the DHC, as well as individuals with the education and training to deeply consider ELSIs, are critical to this paradigm shift [23].

Given the lack of regulatory guidance in a rapidly emerging digital landscape and increasing levels of public mistrust in science and technology, understanding educational needs with respect to ELSIs of digital health is a vital step forward [23] and is an important compliment to initiatives such as ReCODE Health and the DHC. To begin identifying ELSI educational needs, researchers affiliated with the Society of Behavioral Medicine’s (SBM) former Digital Health Council carried out a needs assessment among SBM members that was informed by the DHC tool. The assessment goals were to: (i) identify areas of ELSI proficiency and need; and, (ii) examine interest and expertise in ELSI topics by career level and endorsed prior ELSI training history. Results will support: identifying more specific ELSI topics to inform training resources; identifying the levels at which training may be most beneficial (e.g. early vs. late career); and understanding where specific expertise needs to be fostered so more individuals are prepared to train others.

Methods

A 14-item Qualtrics survey was disseminated to approximately 2400 SBM members from 11/15/2021 to 12/31/2021 via weekly society e-mail newsletters. Responses were completely anonymized. The survey was informed by the ReCODE Health DHC tool because it is one of the only rigorously developed set of guidelines for considering the ELSIs of the broad field of digital health and has been applied in several research studies [12, 24]. The survey included items that mapped onto the four domains of the DHC tool: data management, access and usability, privacy, and risk/benefit assessment [13]. Respondents were asked to rate their interest level (scale ranging from 1 to 3) in various ELSI topics relating to the four domains, as well as to rate their perceived expertise (scale ranging from 1 to 3) on each of these topics (all topics, with corresponding DHC tool domains, listed in Table 1). They were also asked about their training history in ELSI of digital health research (formal vs. informal). Respondents were asked to provide basic information about their career level and area of study. Because these existing data were collected in a manner that survey respondents could not be directly identified, this work was given a “review not required” determination by the University of California, San Diego IRB. Given the small sample size, and the goals of this needs assessment, no a priori hypotheses were specified and thus no inferential analyses were conducted. Descriptive information was used to accomplish our proposed aims. Descriptive statistics included means and standard deviations of continuous variables (ratings of interest and expertise levels) and frequencies and percentages for categorical variables. Graphs were used to visually depict levels of interest and expertise by career level and self-reported ELSI training history.

Table 1.

Interest and expertise ratings across ELSI-related topics

DHC tool domain(s)	ELSI-related topic	Interest rating (M, SD)	Expertise rating (M, SD)
Access and usability	Evaluating practices that support short- and long-term engagement strategies (i.e. Keeping individuals interested in participating in the study)	2.78, 0.81	1.71, 0.63
Access and usability	Dissemination of evidence-based digital health methods to settings outside academia (e.g. community centers, schools, hospitals)	2.76, 0.53	1.44, 0.59
Access and usability	Implementation of evidence-based digital health methods in practice and research outside academia (e.g. community centers, schools, hospitals)	2.72, 0.54	1.50, 0.66
Risk/benefit assessment	Having discussions about the ELSIs of digital health	2.63, 0.69	1.29, 0.49
Access and usability Privacy Data management	Recruiting participants via digital methods (e.g. via a mobile app or Facebook ads)	2.53, 0.70	1.83, 0.75
Data management Privacy	Working on practices for data management (e.g. collection, secure storage, and data sharing practices)	2.48, 0.66	2.03, 0.70
Access and usability Data management Privacy	Considerations for returning research results to participants	2.42, 0.70	1.58, 0.63
Risk/benefit assessment Privacy	Conveying how a technology will be used in an informed consent document	2.32, 0.75	1.95, 0.66
Risk/benefit assessment Privacy Data management	Deploying informed consent using a web platform or mobile device	2.27, 0.81	2.03, 0.76
Privacy Risk/benefit assessment	Reviewing terms of service and privacy policies when using or developing a commercial product or service	2.27, 0.73	1.38, 0.52
Risk/benefit assessment	Evaluating risks of harm and mitigating risks	2.25, 0.71	1.86, 0.58
Privacy Data management	Managing collection of “bystander” data from people who are not enrolled participants in your study	1.9, 0.73	1.14, 0.34

Note: Interest and expertise ratings ranged from 1 to 3.

Results

Respondent characteristics

Respondents (n = 66) were 74.2% faculty (n = 49), 12.1% graduate students (n = 8), 9.1% postdoctoral fellows (n = 6), and 4.6% staff (n = 3). Of the faculty and postdoctoral fellows, 47.3% identified as “early career” (n = 26), 36.4% were “mid career” (n = 20), and 16.3% were “late career” (n = 9). Most respondents classified their fields of study as psychology (50.0%; n = 33), followed by public health (13.4%; n = 9), behavioral medicine (4.5%; n = 3), and epidemiology (4.5%; n = 3). A minority of individuals endorsed careers in biomedical informatics (n = 1), dentistry (n = 1), kinesiology (n = 1), nursing (n = 1), nutrition (n = 1), and psychiatry (n = 1). Most respondents (89.3%, n = 59) indicated that they had used technology to facilitate their health research in the prior 12 months, and 100% of respondents stated that they plan to use technology in their research in the next 12 months. Individuals endorsed using, or planning to use, the following technologies in their health research: Mobile applications or “apps” (83.3%; n = 55), social media (62.1%; n = 41), wearable devices and sensors (62.1%; n = 41), websites (62.1%; n = 41), electronic health records (56.1%; n = 37), machine learning (31.8%; n = 21), natural language processing (16.7%; n = 11), and augmented/virtual reality (16.7%; n = 11). Respondents reported using these technologies to support their health research in the following ways: data collection (96.9%; n = 64), recruitment (77.3%; n = 51), intervention (74.2%; n = 49), enrollment (51.5%; n = 34), data analysis (43.8%; n = 29), implementation (33.3%; n = 22), dissemination (27.3%; n = 18), predictive analysis (19.7%; n = 13), and augmented/virtual reality (15.2%; n = 10).

ELSI expertise and topics of interest

According to Table 1, ELSI topics of greatest interest included practices that support short- and long-term participant engagement, and dissemination and implementation of digital tools outside of academia. Respondents indicated that they were least interested in managing “bystander” data, evaluating and mitigating risks, mobile/web informed consent, and reviewing terms of service/privacy policies in commercial products. Respondents rated themselves most experienced in mobile/web informed consent, secure data management practices, and describing how technology will be used in informed consent. Respondents rated themselves as least experienced in managing “bystander” data, having discussions about ELSIs, and reviewing terms of service and privacy policies when using commercial products. Only 39.4% (n = 26) endorsed having received “formal” training in ELSI (as opposed to “informal” or self-taught).

ELSI expertise and topics of interest by career level

Fig. 1a depicts interest in ELSI-related topics by career position. Among postdoctoral fellows and staff, there appeared to be more interest in reviewing terms of service and privacy policies, and having ELSI discussions. Postdoctoral fellows also rated greater interest in learning about the process of returning results and managing “bystander” data. Data management practices were of high interest to postdoctoral fellows and graduate students, compared with faculty and staff. Explaining risks and potential harm was of greater interest to faculty and postdoctoral fellows.

Figure 1.

Interest ratings (panel a) and expertise ratings (panel b) for specific ELSI topics by career level.

Fig. 1b depicts expertise in ELSI-related topics by career position. Faculty and staff rated themselves with higher expertise in explaining technology during the informed consent process, remote consenting, and evaluating risks of harm. Staff and students rated themselves with more expertise in recruitment than faculty and postdoctoral fellows. Faculty and students endorsed higher expertise in data management practices than staff and postdoctoral fellows. Staff rated themselves with more expertise in returning research results to participants and all of the postdoctoral fellows rated themselves with the lowest expertise (beginner level) in implementation, dissemination, and having ELSI discussions.

ELSI expertise and topics of interest by ELSI training history

Fig. 2a and b depicts interest and expertise in ELSI-related topics by prior ELSI training history (formal vs. informal/self-taught), respectively. Those endorsing a history of primarily informal ELSI training generally reported greater interest in receiving training in most ELSI-related topics, with the exception of evaluating risks of harm. Conversely, those endorsing a history of primarily formal ELSI training generally rated higher expertise in all ELSI-related topics. Descriptively, the largest expertise gaps between formally and informally trained individuals appeared to be related to engagement, evaluating risk of harm, implementation, and dissemination.

Figure 2.

Interest ratings (panel a) and expertise ratings (panel b) for specific ELSI topics by ELSI training history.

Discussion

The results of this needs assessment identified areas of self-reported ELSI proficiency and need for further training and educational opportunities among behavioral scientists, surveyed from one professional society (SBM). Our assessment is unique because it included graduate students, postdoctoral fellows, faculty and staff of the organization, reflecting ELSI education needs across career levels. Data indicated that interest and expertise in ELSI topics may vary across career levels. While the patterns in this sample were fairly intuitive given the typical roles and responsibilities of these positions (e.g. data management practices were of greater interest to students and postdoctoral trainees, who are often in charge of the data management), these data could show that that future ELSI training opportunities may be more impactful if the topic areas were tailored to career level and years of experience in digital health research. Unsurprisingly, those self-reporting an informal ELSI training history endorsed the greatest interest in most ELSI topics. While attention to the ELSIs of conducting digital health research has increased in recent years, findings from this study highlight the remaining need for additional and formal training in specific ELSI topics. Fewer than half of respondents reported that they engaged in formal ELSI training, reflecting a potential opportunity to develop ELSI trainings for behavioral scientists.

Experience managing bystander data was absent among respondents, but this is not necessarily unexpected. In the USA, the protection of human subjects is centered on the human subject [25]. The definition of “protection of human subjects” does not focus on bystanders and therefore may lead researchers to focus primarily on their research participants when planning and conducting studies. Ethical questions about whether this narrow definition is limiting the protections of people in research have been raised [26]. There are many nuanced ELSI-related risks to consider when conducting research (and the bystander issue is just one example of these nuanced risks). Unless the IRB specifically asks questions about bystanders, it might not be considered by researchers. Regardless of whether there are IRB requirements for the protection of bystanders, the opportunity for researchers to fully consider the ELSIs of conducting this work has broader scientific and societal implications. Focusing on these implications can illuminate facilitators and barriers that downstream can affect research participation, translation and sustainability of findings into real-world settings, and creation of trustworthy science [23].

Another important finding was that a limited number of individuals rated themselves as having intermediate or advanced expertise in discussing ELSI topics. ELSI research emerged from genetic research and the need to have multi stakeholder engagement in discussions, thereby elevating awareness of the complexities of research and the related ELSIs involved [16]. Bell et al. suggest that researchers might find themselves isolated from others within their own discipline or lack support structures for ELSI due to the nature of ELSI research being interdisciplinary and international [27]. Indeed, issues of ethics and justice are typically considered and taught as tangential to the research process for most research scientists, rather than a key central component, and there is a dearth of resources and/or program structures to encourage a paradigm shift [23]. Thus, perhaps lack of opportunities for discussion and training might be a reason for lower levels of comfort discussing ELSI. This would be consistent with other studies that have shown lack of training opportunities and lack of a network to be potential barriers for career development [28].

In sum, these findings and ongoing conversations highlight the need to further cultivate awareness of and capacity to consider the ELSIs of digital health research. Researchers could use this improved capacity to partner with ELSI experts (e.g. research ethicists, bioethicists, historians) to examine and shape ethical practices as technologies evolve [23]. Our needs assessment contributed to a sparse literature on ELSI training in the area of digital health, utilized an extant published checklist for digital health research, and surveyed across career levels. These strengths must be weighed against the limitations of conducting this assessment in one professional society. The results are from a relatively small sample and limited to behavioral scientists, although SBM is interdisciplinary. Some questions were intentionally broad, and so respondents’ answers likely depended on their own interpretation of the construct being assessed (e.g. respondents may have different definitions of what qualifies as “formal” training in ELSI). There may also be other aspects of ELSI training needs that our survey did not address.

Implications for practice, research, and policy

We are in a pivotal period where scientific ethics is of utmost importance [23]. It is critical that we use our growing knowledge in ELSI training to influence the research and practice “ecosystem” at multiple levels (e.g. individual training, organization infrastructures, scientific community, public policy) [29]. At the individual level, our needs assessment revealed that the greatest interest for training is in implementation and dissemination of digital health research outside of academia. Additionally, a focus on mobile apps is paramount for ELSI training given more than 80% of respondents use or will use mobile apps in their research. At the institutional level, IRB members may lack sufficient expertise to evaluate digital health research risks and risk management practices. Institutional policies also may have gaps that increase potential risks specific to data management and privacy protections. At the organizational and community levels, ELSI training infrastructure is needed. Within SBM, the new Scientific Education Council plans to prioritize creation of ELSI training opportunities within the organization as one of its strategic goals. The Council will continue to assess training needs and engage relevant partners in the development of educational resources. At the policy level, funding agencies can both enhance and hinder advancement in our understanding of ELSIs associated with digital health research. For example, the NIH has issued supplemental awards to study ethical issues in biomedical and behavioral research. Journals can also support more explicit ethical considerations and practices, similar to the way most journals have instituted requirements for trial registration, reporting guidelines, and data transparency [30–32]. At all levels, alignment between and across all partners and entities is necessary.

Future endeavors may consider a staged approach to ELSI training. Focusing initial ELSI training in areas of highest interest within digital health may result in greater uptake and participation, which can serve as an opportunity to address other ELSI educational needs and raise more interest in ELSI training. Future research in ELSI capacity building includes ongoing assessments to understand training needs among those who: (i) develop the technologies, (ii) procure and deploy the technologies (researchers, clinicians), and (iii) those who use the technologies (participants, patients). Concurrently identifying, applying, and evaluating best practices for evidence-based learning for these groups is needed.

Conclusion

While ELSI topics have been the focus of genetic and genomic research for over three decades [16], there is an opportunity now to increase awareness of ELSIs in social, behavioral, and environmental contributors to health. Formal education in ELSI is limited among behavioral scientist training programs. With the increasing use of technologies to gather data and deploy interventions combined with the related ELSIs, it is time to prioritize education. There is an opportunity to increase ELSI training at all career stages, and programs may benefit from tailoring ELSI training topics to career level.

Conflict of interest statement

None declared.

Funding

This time spent conducting this work was supported by the National Heart Lung and Blood Institute [R01 HL153543], the National Institute of Diabetes and Digestive and Kidney Diseases [R01DK132210], and the National Institute of General Medical Sciences [P20 GM139743] of the National Institutes of Health.

Ethical Approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the University of California, San Diego Institutional Review Board and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed Consent

This research involved the study of existing data that was collected (i) in a manner that survey respondents could not be identified directly or through an identifier and (ii) that was not originally conducted for research purposes. The University of California, San Diego Institutional Review Board therefore determined that a review was not required for this research (Protocol #807819).

Welfare of Animals

This article does not contain studies with animals performed by any of the authors.

Transparency Statements

Study registration: This study was not formally registered because it originated as a needs assessment to inform programming within an academic organization. Analytic plan preregistration: The analysis plan was not formally preregistered because the study originated as a needs assessment to inform programming within an academic organization. Analytic code availability: Analytic code used to conduct the analyses presented in this study are not available in a public archive. They may be available by emailing the corresponding author. Materials availability: Materials used to conduct this study are not available in a public archive. They may be available by emailing the corresponding author.

Data Availability

Deidentified data from this study are not available in a public archive. Deidentified data from this study will be made available (as allowable according to institutional IRB standards) by emailing the corresponding author.