Dataset of survey with Research Ethics Committees (RECs) and ethics experts on their approach to ethical reviews of emerging technologies: AI in health and healthcare, extended reality, biobank, genome editing

To understand how Research Ethics Committees (RECs) and ethics experts approach ethical reviews of emerging technology (AI in health and healthcare, extended reality, biobanking, and gene editing) research, an online survey was conducted from 2nd February 2024 to 2nd April 2024. The survey has been developed to match the main challenges identified by the SCORE analysis of irecs Task 2.2 Deliverable.

Survey design and topics

The survey was designed to collect data on experiences, needs and gaps in the ethics review process and preferred methods of training. It contained several key sections, each for specific information from participants. Participants provided demographic information and professional background, including their country of work, years of experience in ethics assessment, main disciplinary background, and their roles in the ethics assessment process. This information helped contextualize the responses based on participants' backgrounds and expertise.

The survey was designed to allow experts to select one of the initial multiple-choice questions relevant to the technology they work with. Those who did not select a specific technology were not asked questions related to it. In total, we received responses from 242 participants, out of which 149 completed the entire survey by selecting all the technologies.

In the biobanking section, participants were asked when an ethics review is required in their country. They self-assessed their competence in addressing ethical issues such as consent, data protection, data privacy, data sharing, return of individual health-related findings, involvement of children, and commercialization. Participants also shared their opinions on when ethical review is most necessary for biobank-related research.

In the AI in Healthcare section, participants evaluated their competence in addressing ethical issues, including consent, data protection, data privacy, data sharing, bias, justice and fairness, transparency/explainability, and responsibility for the consequences of AI applications. They also considered the criteria that determine when AI-related research falls within the scope of RECs.

Regarding extended reality (XR), some questions were similar to AI's: participants considered the criteria determining when XR-related research falls within REC scope. Also, they assessed their competencies in addressing ethical issues such as consent, data protection, data privacy, and data sharing. Some questions were specific to XR - like violence or abuse in XR environments, mental health issues, AI usage in XR research, involvement of children, and cybercrime. Experts self-assessed their understanding of XR technologies' inner workings and differences, including virtual reality, augmented reality, and XR using AI. 

In the genome editing (GE) section, participants reflected on their competencies in addressing ethical issues related to somatic, germline, and non-human genome editing. For somatic genome editing, they considered issues such as consent, safety, accessibility, enhancement, and misuse. Similar ethical issues were addressed in germline genome editing, acknowledging the additional considerations due to heritable changes. For non-human genome editing, participants focused on safety, accessibility, enhancement, misuse, and protection of species.

The survey also explored cross-cutting ethical issues, where participants reflected on broader ethics governance and assessment practices across all technologies. They provided insights into the types of ethics governance applied—such as self-assessment, REC reviews before starting projects, and ongoing monitoring—and shared opinions on the sufficiency of current governance structures, offering suggestions for improvement. Key factors hindering effective ethics assessment were identified, including a lack of ethics expertise, insufficient scientific or technical understanding, inadequate guidelines, limited training opportunities, and scarce funding. Additionally, participants considered the feasibility of addressing broader societal impacts, like risks of discrimination and stereotyping, within research involving each technology. They reflected on the potential for implementing continuous ethics oversight, or "ethics by design," throughout the research lifecycle, beginning from the design stage. 

The final section of the survey assessed training needs, evaluating participants' desires for additional training in each technology area. Participants selected the training content they believed would be most beneficial, and preferred modes of training and learning styles for online training were specified.

The survey target groups were the European Network of Research Ethics Committees (EUREC), the European Network for Research Ethics and Integrity (ENERI) e-community, the European University Association (EUA) Research & Innovation Strategy Group (Rectors or Vice-Rectors for Research) and the European Association of Research Managers and Administrators (EARMA). Thus, the survey was disseminated among members of ENERI, EUREC, EUA and EARMA In addition, participants were recruited by other project partners who distributed the survey invitation with a link. The same recruitment strategy was implemented in the EU and non-EU countries.

The questionnaire was hosted online and managed by RAIT GROUP, an independent market research company, ensuring confidentiality and professional handling of the data. Quantitative data were analyzed using descriptive statistics to identify trends and patterns among participants. We conducted additional analyses by categorizing participants into EU and non-EU groups and further segmented the data for specific regions, including the EU, China, Africa, and others. Where possible, statistical significance tests were performed to determine if observed differences between groups were meaningful.

Cross-tabulations were used to identify relationships between variables, and qualitative data from open-ended responses were analyzed to identify recurring themes, concerns, and suggestions.

This comparative analysis across different technologies and regions helped identify unique challenges and commonalities. A regional breakdown provided a better understanding of how ethical assessment practices and needs might vary in different parts of the world.

Challenges and limitations

A significant issue was the overrepresentation of participants from the European Union, which may have skewed results toward EU perspectives and limited the global applicability of our findings. The low number of participants from regions like Africa meant that specific regional insights were underrepresented, affecting the diversity of perspectives. 

Response bias, including self-selection and social desirability bias, is a common issue with any survey and might have influenced the results. Participants with a particular interest in emerging technologies may have been more likely to respond, and some may have provided answers they thought were expected rather than reflecting their true opinions.

Data limitations included incomplete responses, resulting in varying sample sizes for different analyses. Self-assessments of competencies may not accurately reflect actual expertise, presenting potential inaccuracies.

Despite these limitations, the survey provided valuable insights into the current state of ethical reviews for emerging technology research and highlighted areas where further support and resources are needed.