AI in reducing maternal harm is implemented ethically, with embedded bias management

Researchers will use AI in reducing maternal harm to analyze maternity investigation reports in the hope of reducing disproportionate levels of harm among black mothers and babies.

Experts from Loughborough University will work with the Healthcare Safety Investigation Branch (HSIB) to identify patterns in more than 600 of its recent investigations into adverse outcomes during pregnancy and birth.

The independent patient safety body has conducted more than 2,000 approx. maternity investigations since April 2018, with around approx. 10% concerning black, Asian, and ethnic minority families.

These examined cases such as the death of the mother during pregnancy or childbirth, miscarriages, stillbirths, and neonatal deaths.

Black mothers are four times more likely to die from pregnancy complications than white women, according to a report published by a survey earlier this month.

It found that pregnancy is safe, with 8.8 women per 100,000 dying during pregnancy or up to six weeks after childbirth or the end of pregnancy between 2017-2019.

But it said outcomes are not equal, with the maternal mortality rates of women from Asian and mixed ethnicity backgrounds around twice as high compared with white women, and an increased risk for women in more deprived areas.

The research team will develop an AI in reducing maternal harm and a machine learning system capable of identifying factors, based on a set of codes, that contribute to harm during pregnancy and birth experienced by black families.

These include biological factors, such as obesity or birth history; social and economic factors such as language barriers and unemployment; and the quality of care and communication with the mother.

It will look at how these elements interact with and influence each other, and help researchers design ways to improve the care of black mothers and babies.

But the researchers believe the findings could be used to reduce harm among patients of all ethnicities.

The researchers said involving black families is “essential”, with two groups representing patients and the general public who will be consulted throughout “to minimize any potential bias”.

The project is led by Dr. Patrick Waterson, from the university’s human factors and complex systems group in the School of Design and Creative Arts, and Dr. Georgina Cosma, an expert in Artificial Intelligence and Data Science from the Department of Computer Science.

He said that ultimately, he believes the outcomes from their research have the potential to transform the NHS’s ability and help AI in reducing maternal harm amongst mothers from black ethnic groups.

And in the longer term, the research could improve patient safety for all mothers, regardless of ethnicity. 

Kevin Stewart, the HSIB’s medical director, is delighted to be working with Loughborough University to better understand the factors that lead to poor outcomes and experiences for some women from black ethnic groups.

They believe their data, gathered from so many of our maternity investigations, will help develop the learning required to automate analysis and identification of themes.

The project will run for two years from January and is funded by NHSX and the Health Foundation.

Jabeer Butt, chief executive of the Race Equality Foundation, said AI in reducing maternal harm has the potential to improve outcomes if “used wisely”. A woman’s ethnicity should have no impact on her chances of having a safe and healthy pregnancy, yet the reality for black pregnant women is that it does.

While they’re welcome to use AI in reducing maternal harm to better address the problem through examining key factors, like the care experiences of black mothers, safeguards will be crucial to ensure that the use of AI doesn’t end up exacerbating the inequalities it aims to solve. 

Dr Edward Morris, president of the Royal College of Obstetricians and Gynaecologists, also welcomed the research and said it is exciting to see how AI can progress work in the field.

An AI app developed with the pregnancy charity Tommy’s, which uses routine data from antenatal appointments to assess a woman’s chance of developing complications during pregnancy, is being piloted in some areas, he said.

Brhmie Balaram, head of AI research and ethics at NHSX, expressed they’re excited to support innovative projects that demonstrate the power of applying AI to address some of our most pressing challenges; in this case, they’re keen to prove that AI can potentially be used to close gaps in minority ethnic health outcomes.