Call for Case studies: Ethics of artificial intelligence in global health research meeting

Deadline for applications: Friday 17 June 2022, 17:00 CET

Background

The Global Forum on Bioethics in Research (GFBR) will hold a two-day meeting in Cape Town, South Africa in November 2022 on the theme: “Ethics of artificial intelligence in global health research”. GFBR seeks to bring researchers, research policy makers and ethicists, among others together to share experiences and promote collaboration around research ethics. The Forum will be built around case study presentations to ensure that discussion of the ethical issues remain grounded in the practical realities of how research is conducted 'on the ground', particularly in low resource settings. There will also be a session on governance issues. Compared to traditional meetings, GFBR is unique in that it is limited in size and built around small group discussions of case studies and governance papers that are submitted by participants. The Forum prioritises the participation of colleagues from low- and middle- income countries (LMICs), encourages networking and mentoring, and creates a venue for open and inclusive discussions. For more information visit: www.gfbr.global.

GFBR is seeking three types of participants for the meeting:

1. Case study presenters will present their research experiences and the ethical issues that have emerged regarding the use of AI in health research in LMIC settings.
2. Governance paper presenters will present on topics such as regulation, policies, guidance, tools and issues associated with ethics and other review and oversight mechanisms.
3. Participants will attend the meeting and actively take part in plenary and small group discussions and networking opportunities.
   

Places are awarded on a competitive basis and successful applicants from LMICs will receive an award to cover the cost of their travel, accommodation and single-entry visa.

All interested applicants should review the information below and submit an application no later than 17 June 2022; case study and governance paper presenters will submit proposals with their applications. All applications will be reviewed by the GFBR Planning Committee and selection will be made on the specific selection criteria (see below). Applicants are not limited to academic researchers; staff from government, non-governmental organizations (NGOs), and private sector organizations are also encouraged to apply if their applications are focused on the theme. If a case study or governance paper is multi-author, and co-authors would like to attend, they must apply separately as participants and state the name of their co-author and title of the proposal in their application.

Selected case study and governance paper presenters will be paired with a member of the GFBR Planning Committee to provide informal mentorship and help them develop their application into a brief paper (2-3 pages) and a PowerPoint presentation. During GFBR, case study and governance paper presenters from around the world will share their presentations and discuss cross-cutting issues, and then participants at the meeting will discuss the challenges and questions raised in both plenary and small group discussion.

This year's meeting is co-organised with local host, the South African Medical Research Council.

Please note: In light of the COVID-19 pandemic, the meeting date and location will be kept under review. Provisional dates being explored are 15&16 November or 29&30 November.

About the theme 

Artificial intelligence (AI) is increasingly being used in global health research but frameworks, policy and best practice for the ethical review and oversight of AI-enabled studies is currently lacking. The Forum will discuss how traditional research ethics regulatory frameworks have responded to the rapid advances in AI technology, and what changes are required, including to the role and responsibility of research ethics committees (RECs). It will explore the ethical challenges such as bias, privacy, data provenance and ownership, along with the need for transparency, accountability and engagement during the design and use of AI in global health research. To date, these discussions have predominantly taken place in high-income countries, and low- and middle-income country (LMIC) perspectives have been underrepresented. The Forum will consider the LMIC context where AI has the potential to address critical skills shortages and improve access to care, but where the ethical challenges are made harder due to existing disparities in infrastructure, knowledge and capacity. The Forum will take a multidisciplinary approach to explore how AI technology is being designed and used in health research, reflecting the range of actors involved in this space and the importance of computer scientists and technologists who apply AI for health to understand research ethics frameworks and considerations.

Please read the background paper for further details on the meeting theme and scope.

Call for case studies

The GFBR organisers are looking for interesting and important real-life cases that are relevant to the theme. The cases could demonstrate the development of good practice; highlight ethical challenges; demonstrate situations in which ethical practice failed; or present unresolved questions for the global community. The organisers encourage cases that address research ethics oversight and cases that address past GFBR topics[1]. However, the actual topics considered at the meeting will be defined by the case studies that are submitted. In this way, GFBR aims to be responsive to applicants and the issues that they consider most important.

The scope includes case studies that:

• Focus on issues around conducting health research in LMICs. (However, we do not want to exclude case studies from high income countries if there could be valuable lessons to learn, and some parallel or relevant ethical considerations. If your case study relates to a high income country please use the commentary section to draw-out the relevance for research in LMICs.)
• Are from any stakeholder perspective, including ethicists, policy-makers, researchers, clinicians, computer scientists, and healthcare workers.
• Are from any organisational perspective e.g. academic, technology companies, government, non-governmental organisation (NGOs) and public-private partnerships.
• Address the ethical issues associated with the lifecycle of developing, validating and using an AI system in the health research context. This could include:
  • Model development:
    • Collecting and processing data on which to train the AI algorithm
    • Designing and developing the algorithm
    • Training the algorithm (e.g. using a ‘test set’ and ‘tuning set’ of data)
  • Model validation:
    • Using an internal or external test set to validate the algorithm
  • Model use in health research. For example:
    • Assessing the impact of an “algorithm for intervention” e.g. a prospective observational trial or an interventional clinical trial to evaluate an AI-based clinical tool in a clinical setting, which could include an assessment of how use of the algorithm may change the outcome for the patient, the behaviour of physicians or the patient/ physician relationship.
    • Using a validated “algorithm for discovery” in research e.g. to generate hypotheses or answer research questions such as discovering associations in population health data that reveal a new disease group or discovering potential drug candidates.
• In addition, GFBR is open to case studies of research on AI to develop systems with health applications e.g. projects that may be framed as data or computer science developed by technology companies and which are not necessarily characterised as ‘health research’ or subject to research governance requirements.

• Focus on the ethical issues that result from the use of the AI system in global health research. Cases studies on health research that happens to use an AI system but the ethical issues relate to an aspect of the study that is not explicitly tied to the use of AI are not in scope. For example, a case about the large scale and variety of health data required to train an AI system, which may increase privacy and security concerns, would be a stronger case than one that discusses data sharing issues that are more common to other types of non-AI research.

In general, case studies should focus on no more than three ethical issues. We indicate below (‘Key themes and questions’) some examples of issues considered important by the GFBR organisers. Please see the background paper for more details on the scope and themes.

Case studies should be 2 pages maximum (excluding references), font Arial size 10.5 (in Microsoft Word or pdf format), clearly articulated in English and contain the following sections:

1. Title of case

2. Your name, institution and country location

3. Brief description of the research project

4. Background - relevant facts about the host country/community and disease studied (if disease specific)

5. Ethical issues with commentary on each issue

6. Conclusions and two recommendations for how to improve the ethics of, and ethical approaches to, using AI in health research. The recommendations can relate to the field broadly or be specific to the case study.

In addition, please provide the following information, in Microsoft Word or pdf document using font Arial size 10.5. Please do not share sensitive personal health information.

• Complete case study presenter application form
• Short CV (2 pages maximum)
  

If you are unsure about the suitability of a possible case study or would like to discuss your proposal, please email gfbr@who.int.

Key themes and questions 

Case study and governance papers could address (but are not limited to) one or more of the following questions and should focus on research in LMICs, though examples from HIC will be considered if they show relevance to LMIC settings. Where the questions refer to a ‘researcher’ this includes academic, commercial and government sector and includes those who develop AI systems and those who validate and use them in a health research context. Example case studies and governance papers can be found in this past GFBR meeting programme.

Some of these questions could be addressed as either a case study or a governance paper. You are welcome to decide which format is best to present your experience and ideas or you can contact gfbr@who.int to discuss which format would be best. In general, a case study is appropriate if the ethical issues are associated with a specific research project and practical research experience. A governance paper would be more appropriate if, for example, you are providing an analysis of a governance issue and a theoretical solution.

Deadline 

All applications should be sent to gfbr@who.int by 17.00 CET on Friday 17 June 2022, in English. Please specify in the subject line whether you are applying to participate, present a case study, present a governance paper or as a journalist. Please ensure you include all the requested information, as incomplete applications can not be considered (see the Checklist below). Applications received after the deadline will not be considered.

Awards: Decision making and eligibility for funding

Successful applicants from LMICs who require full funding will receive an award to cover:

• return travel to the meeting (economy airfare and standard ground transportation costs);
• accommodation (2 or 3 nights maximum, including meals);
• a single entry visa (if required).

  Participants will be expected to meet all other costs.

  The GFBR Planning Committee will select successful candidates (both self-funded and those applying for funded places). The selection committee will consider the following factors when considering the applications:

• Country of origin. We would like to ensure a representative distribution of participants from different regions;
• Background/current area of expertise. Applications will be selected for a diverse representation of many different disciplines, relating to the theme of the meeting;
• Experience of ethical issues related to the use of AI in health research;
• Reasons for attending the meeting. Participants who will be able to actively contribute to the meeting and who expect to achieve impact from the meeting;
• Case study applications and governance papers only: Relevance of the case/paper to the meeting theme and research in LMICs;
• Journalists only: Demonstrated journalistic training and experience and concrete proposals for how the meeting findings will be disseminated, including which media outlets and the format of reporting.

If your case study or governance paper is not selected your name will automatically go forward to be considered in the applications to attend as a participant. Applicants are encouraged to submit a case study or governance paper.

Notification

All applicants will be informed of the Planning Committee’s decision by the 12 August 2022. The decision of the committee will be final.

Privacy

The application form describes the purposes for which WHO will collect and process the personal data you provide in your application and describes your rights as a data subject. WHO’s personal data protection framework includes the WHO Privacy Policy and the UN Personal Data Protection and Privacy Principles, which are available at the following link: https://archives.un.org/sites/archives.un.org/files/_un-principles-on-personal-data-protection-privacy-hlcm-2018.pdf