7 minute read

Derrick Aarons

Dr. Derrick Aarons

MB.BS.(UWI); M.Sc. (Bioethics); PhD (McGill) Consultant Bioethicist and Family Physician Member of the Executive Council of RedBioetica UNESCO

Advertisement

MEMBERS’ ARTICLES

Needed For The Caribbean: A Regulatory Framework For A-I Health Research Dr. Derrick Aarons

Introduction:

The convergence of Artificial Intelligence (A-I ), big data methods, and microsystems engineering makes AI-based algorithms for computational neuroscience one of the fastest growing fields of neuro-medical research (IBC, 2021). A-I algorithms in clinical neuroscience research may be used in several ways, e.g., to detect early signs of Alzheimer’s disease and mental illness (IBC, 2021). However, ethical issues such as incidental findings with privacy concerns, transparency and bias, and algorithm discrimination may arise in the process (Social and Human Sciences Commission, 2021). Further, continuous risk monitoring may be needed when A-I medical devices are being researched or being used in research endeavours, and so responsive regulatory mechanisms must be in place (Ho, 2021). Therefore, to protect the participants in such research processes, robust and appropriate regulations for research involving artificial intelligence should be implemented across all Caribbean countries, which would be in keeping with the aims of the research protections proposal that were made in 2015 within the Caribbean Community and Commons Market (CARICOM) (Aarons, 2019).

Regulating Research:

In 2015, the Caribbean Public Health Agency (CARPHA), the regional public health institution in the Caribbean with the responsibility for providing strategic direction in analyzing, defining, and responding to the public health priorities of the 24 member states across the Caribbean, presented the Ministers of Health of the Caribbean Community (The COHSOD) with a ‘green paper’ proposal to regulate the conduct of research with human participants to provide ‘best practices’; ensure consistency and harmonization throughout the Caribbean; and through legislation protect the inhabitants of countries in the region from ‘ethics dumping’ and harmful exploitative research activities (Aarons, 2019). The COHSOD was requested to establish a regional regulatory framework for research involving human participants by approving model legislation for CARICOM countries to regulate research along with regulations for sanctions for non-compliance (See Appendix A – Letter to Ministers of Health). The specific recommendations were: “CARPHA’s Research Ethics Secretariat recommendations to CARICOM Ministers of Health is presented in the draft position paper in Appendix B 1. The establishment of a Regulatory Framework for the conduct of research with human participants based on research ethics principles and standards and including the following requirements:i. all researchers must apply for approval from an appropriately designated research ethics

Needed for the Caribbean……...cont’d

ii. an appropriately constituted research ethics committee with the required legal authority shall review all research proposals that involve human participants; iii. all approved research should be monitored for the well-being of research participants by the research ethics committee. 2. The preparation of a model legislation for CARICOM countries to regulate research with human participants. The legislation should contain the provision of appropriate sanctions for non-compliance. 3. That CARICOM Ministers of Health approve the Proposals for implementation of a regulatory framework in principle, and agree to their adoption as Policy, pending the implementation of the process for enactment of legislation.” (Appendix B ).

CARPHA’s proposal was put to the vote and approved unanimously by the COHSOD, with the approved green paper being sent to the CARICOM headquarters in Guyana for the legislative drafters to draft the legislation (See Appendix B – Draft Position Paper on Human Subjects Research in the Caribbean). To date, nothing further has been heard on the matter.

Artificial Intelligence in health research:

With the current accelerated use of A-I in the collecting and processing of health data for research, including discovering associations in the health data of populations and assessing impacts of algorithms for interventions in health ethical review of such protocols will require research ethics committee (REC/IRB) members to be cognizant of the particular ethical challenges that arise with the use of AI in health research, and will therefore need update training. The legislation and regulations being drafted at the CARICOM Headquarters will also need updating to address these new challenges. In 2016, the Caribbean Public Health Agency developed a network of RECs/IRBs across the Caribbean (called CANREC) and provided research ethics training for the twenty-one member committees of the CANREC (Aarons, 2019). The contents of these training workshops included: i) The ethical principles in research with human participants; ii) Evaluation of value and scientific validity; iii) Fair selection of research participants; iv) Informed Consent, Confidentiality, and Conflicts of Interests; v) Research methodologies; vi) Assessing risks and benefits; vii) Vulnerable populations; viii) Oversight and Monitoring of ongoing research; ix) International Guidelines for Biomedical Research; and x) International Collaborative Research.

Such REC/IRB training has facilitated the conducting of the ethical review of health research protocols across the Caribbean at a high standard (Aarons, 2017). Further, the research ethics review processes in the various countries of the Caribbean have been harmonized to some degree through the work of CARPHA and the CANREC. However, except for Guyana and the Bahamas, national legislation for research with human participants is yet to be enacted in these countries. Consequently, it is opportune to update the model draft legislation for the countries of CARICOM presumably being currently prepared at the CARICOM headquarters to address many of the new ethical issues posed by the use of artificial intelligence in health research.

Needed for the Caribbean……...cont’d

Conclusions:

Most RECs/IRBs across the Caribbean may not have the A-I experience or the expertise needed to conduct an appropriate risk assessment of proposed health research that involves artificial intelligence. RECs/IRBs will therefore need the necessary skills to perform adequate risk/benefit assessments, evaluate the model of consent proposed by researchers, and assess any complex algorithms to be used in the health research. Consequently, research regulations should require REC/IRB members to undergo annual (or biannual, depending on available local resources) update training to keep abreast of new developments in the field of artificial intelligence and their ethical implications. At these update training sessions, REC/IRB members should be required to discuss particular challenges to implementing A-I in Caribbean research, including inadequate access to reliable and affordable internet, unreliable power infrastructure across many of the lower and middleincome communities of the Caribbean, the lack of human resources, too little education and insufficient skills among local populations, inadequate local investments, and the risk of job losses that A-I and automation may produce. These deliberations will provide REC/IRB members with better insights when they seek to provide solutions to the ethical challenges they will meet in reviewing the research protocols submitted to them. The regulations being drafted should also address the specific challenges posed to consent, privacy and security, and the power imbalances between A-I data collectors and research participants to engender public trust in the research endeavours. In addition, the regulations should stipulate the need for public engagement in all such research projects by requiring the involvement from the conceptual stages of not only community representatives but also representatives from vulnerable or marginalized groups where the research may involve them, to inform the design of the research, assist in identifying and mitigating unforeseen risks, assessing possible local and community benefits, and facilitating transparency. With the expansion of the scope of these draft regulations in seeking to protect not only individuals and their interests but also to balance the risks and benefits to society, a more robust, if not comprehensive research ethics framework would have been facilitated for the countries of the Caribbean to address the new realities and challenges posed by the use of artificial intelligence in health-related research.

REFERENCES:

Needed for the Caribbean……...cont’d

1. International Bioethics Committee (IBC). Report of the International Bioethics Committee of UNESCO (IBC) on the Ethical Issues of Neurotechnology. Paris; December 2021: SHS/BIO/IBC-28/2021/3 Rev.

2. Social and Human Sciences Commission (SHS) of UNESCO. Recommendation on the Ethics of Artificial Intelligence. Report of the SHS, 41st UNESCO General Conference, 2021: pp. 14-39. 3. Ho, Wai Loon C. Bridging the Research–Therapy Divide in the Regulatory Governance of Artificial Intelligence as Medical Devices. In Laurie, G et al (Eds). The Cambridge Handbook of Health Research Regulation; Cambridge University Press. 2021; pp. 277-286. DOI: http://dx.doi.org/10.1017/9781108620024.035

4. Aarons D. The historical development of research ethics in the English-speaking Caribbean. CANREC Bull.2019; 1(1): 3-7. 5. CARPHA (The Caribbean Public Health Agency). (2015). Draft Position Paper on Human

Subjects Research in the Caribbean. A submission to the CARICOM Ministers of Health (COHSOD), PAHO Headquarters, Washington DC. 6. Aarons D.E. Research in epidemic and emergency situations: A model for collaboration and expediting ethics review in two Caribbean countries. Developing World Bioethics Journal Jul.2017; 00:1-10. https://doi.org/10.1111/dewb.12157.

APPENDIX A:

Letter to Ministers of Health – to accompany Template for Research Regulation

APPENDIX B:

Draft Position Paper on Human Subjects Research in the Caribbean

This article is from: