Governance, Ethics, and Trust in Algorithmic Health Systems

Algorithmic-supported decision-making tools are increasingly being implemented in health-care settings. The use of algorithms to help or replace human decision-making in health has the potential to improve the quality, efficiency, and effectiveness of health services and to avoid boredom, fatigue, bias, and other cognitive errors that may lead to inappropriate choices and outcomes. However, instead of increased safety and lower risks, algorithmic decision-making in health care may also lead to discrimination and social inequality. Concerns include issues of trust (in the algorithms themselves and in the entities deploying them), patient safety, and governance mechanisms that ensure both safety for users of algorithms and safe, equitable health outcomes for patients and communities.