Unmasking A.I.’s bias in healthcare: The need for diverse data

Main Article Content

David Bobrowski
Hetshree Joshi

Keywords

Abstract

The Institute for Medicine defines six elements of quality health care, as follows: safe, effective, patient-centered, timely, efficient, and equitable. Health Quality Ontario (HQO) and the Canadian Medical Association envisage health equity as a system in which individuals can achieve their full health potential and receive high-quality care that is fair and appropriate to them and their needs.2,3 Equity in the provision of healthcare services is undermined when preventable systemic conditions, encompassing remediable social determinants of health, constrain life choices. The World Health Organization has identified the social determinants of health as the bio-psychosocial circumstances in which people are born, develop, live, and age, with special reference to marginalized communities.4 HQO has acknowledged that the dimension of equity has not progressed at the same rate as other sociocultural parameters, largely because health equity is complicated by multiple evolving and interdependent causes.2 The Ministry of Health and Long-Term Care report – Patients First: A Proposal to Strengthen Patient-Centered Health Care in Ontario – reaffirms this point. Some Ontarians, particularly Indigenous peoples, Franco-Ontarians, newcomers, and people with mental health and addiction challenges, face barriers in accessing the care they need when they need it.5 Therefore, health equity has been recognized as a goal in addressing the challenges that continue to limit treatment opportunities for a vulnerable segment of the Canadian population.2 Women’s College Hospital’s Health Equity Plan and Health Gap campaign exemplify efforts to close health gaps through policies, programs, and practices of the organization.6 Moving forward, we need to determine how best to leverage information technology such that we might address the quality and equity problems that are endemic at present.1