Ethnic minorities, women and those from disadvantaged communities are at heightened risk of poorer healthcare due to a bias built into medical tools and devices, according to a recent report.

The Equity in Medical Devices: Independent Review has raised concerns over AI enabled devices and tools used for measuring oxygen levels. Urgent action has been called by the team behind the review.

Junior Health Minister Andrew Stephenson commented, “Making sure the healthcare system works for everyone, regardless of ethnicity, is paramount to our values as a nation. It supports our wider work to create a fairer and simpler NHS.”

Commissioned by the government in 2022 amid the concerns over the accuracy of pulse oximeter readings in Black and minority ethnic people, showcased by the importance of the devices during the Covid pandemic, the report confirmed these concerns and suggested areas for improvement including monitoring changes in reading and developing devices suitable to all patients.

Concerns over AI-based devices were also highlighted by the report, including the potential for the technology to under-diagnose cardiac conditions in women, lead to discrimination based on socioeconomic status and underdiagnose skin cancers in people with darker skin tones.

It also noted issues with polygenic risk scores, which are often used to measure an individual’s risk to disease based on genetics.

However, attempts to correct these biases have shown to be problematic. Race-based corrections highlighted by the report applied to measurements from spirometers that are used to assess lung functions, have themselves been proven to contain bias.

Chief executive of the NHS Race and Health observatory, Prof Habib Naqvi commented “Access to better health should not be determined by your ethnicity nor by the colour of your skin; medical devices therefore need to be fit-for-purpose for all communities.”

Steve Fuller, Global Community Director for Race in STEM, a BioTalent and The IN Group community: “Healthcare is something we all rely on, and the quality of treatment shouldn’t differ due to a person’s background. Organisations should be looking closely at how they mitigate these biases that are impacting millions of people.”

“The issue of bias in training models has been a persistent theme throughout the development of AI and Large Language Models (LLMs), and while the adoption of AI won’t slow down, businesses must have the proper protocols and policies in place to prevent bias from impacting important areas such as healthcare.”

“Part of removing bias comes from the people training the AI models as well as representation and recruitment within the industry as a whole. Organisations should be working to improve their representation, especially in senior roles, empowering diverse thought across the business, which ultimately also carries proven commercial benefits.

“Communities like Race in STEM play a key role in this, connecting STEM professionals with a network of answers, opportunities, and insights for people of diverse multicultural backgrounds to share their experiences, with the goal of advancing equality through STEM.”