Preventing Maternal Mortality and Morbidity With AI

“Giving birth should be your greatest achievement, not your greatest fear.” —Jane Weideman.

The United States has the highest maternal morbidity and mortality rate among all high-income countries. This crisis is especially pronounced in racial minority and ethnic populations. In fact, Native American and Black women are much more likely to die from pregnancy and childbirth-related complications in the United States than their white counterparts, while also more likely to suffer severe maternal morbidity due to postpartum hemorrhage, hypertensive disorders, and sepsis. The impact of the COVID-19 pandemic on maternal mortality and morbidity is not yet known. Yet, this global crisis is expected to have worsened the situation, especially with exacerbated racial and economic inequities. However, data from the Centers for Disease Control and Prevention (CDC) suggest that nearly 60% of maternal deaths are preventable through better public health programs, training, and access to care. But the world of tech can also play a pivotal role here. There is growing interest in using artificial intelligence (AI) in healthcare, including childbirth, to find enhanced health outcomes for babies and their mothers.

In a recent article published in the Harvard Business Review, Cesar Padilla, Gillian Abir, Mark Zakowski, and Brendan Carvalho share a three-pronged strategy about how healthcare agencies, academic medical centers, technology companies, and state and federal institutions can work together to solve the growing problem of maternal mortality and morbidity in the United States. Their approach includes using artificial intelligence and electronic medical records to predict which pregnant women are at high risk of experiencing complications during childbirth. The authors also emphasize the use of technology to monitor these women and improve their access to both routine and acute care facilities. The final aspect of this strategy is to refer pregnant women to hospitals for specialized care in accordance with the stipulations of the American College of Obstetricians and Gynecologists.

Artificial intelligence is often spoken about in the healthcare industry, where algorithms are developed based on training datasets – typically electronic health records – to predict patient risks and manage health outcomes. Additionally, some corporations including Philips, PeriGen, and Early Sense are building obstetric decision support systems, with the attempt to use artificial intelligence in maternal and even prenatal care. There are several methods of applying artificial intelligence to obstetrics. For example, Hewlett Packard used artificial intelligence to create a monitoring system, where physicians can ask patients questions based on their recorded medical history. This resource was particularly developed for women with high-risk pregnancies.

More modern forms of artificial intelligence use advanced techniques, such as pattern recognition, to predict preeclampsia, hypertension, and severe maternal morbidity in expecting mothers. For example, the Early Warning System (one of PeriGen’s products), uses artificial intelligence to monitor vital signs and other health diagnostic indicators to identify possible medical threats to the mother or child. In addition, Early Sense developed a tool that monitors a potential mother’s ovulation cycle and helps to predict fertility dates. However, racial and gender biases in the utilization of these technologies must be effectively addressed, in order to enhance the health outcomes of all mothers and their children through technological means. To combat algorithmic and real-life bias, Timnit Gebru, a researcher at Microsoft, proposes that machine learning models should include enough training data for Black and Native American women with maternal morbidity. This is an important step in preventing racial inequality in medicine because according to Gebru, the foundation of U.S. healthcare is based on the belief that African American and Native American women are “inherently uninsurable due to their low life expectancy.” However, uninsured expecting mothers are nearly three times as likely to experience maternal death compared to women with insurance.

In summary, technology can play a critical role in helping to reduce both existing disparities and the digital divide in obstetrics. By partnering with state and federal government agencies, large healthcare programs, university medical centers, and leading physician-researchers, artificial intelligence resources can be funded and developed to serve diverse communities, particularly marginalized ones, in a way that is free from bias. Furthermore, these players can create an infrastructure that would significantly reduce the U.S. maternal mortality and morbidity rate for all women, regardless of their socio-economic, racial, or ethnic backgrounds.

Sources:

https://www.expresshealthcare.in/blogs/guest-blogs-healthcare/preventing-maternal-and-infant-mortality-with-ai/425262/
https://hbr.org/2021/08/how-ai-could-help-doctors-reduce-maternal-mortality
https://journals.sagepub.com/doi/full/10.1177/17455065211046132
https://www.vice.com/en/article/ywkk75/bias-in-maternal-ai-could-hurt-expecting-black-mothers