How Generative AI Will Revolutionize Maternal Health
The recent release of generative AI programs like ChatGPT has already disrupted many industries by providing state-of-the-art large language models. The healthcare industry in the United States needs radical change, particularly in maternal morbidity and mortality rates.
As the fractional Head of Growth for Ema, a mobile app with an empathetic AI companion built using generative AI, I have compiled a list of ways that advancements in AI will revolutionize maternal health.
The Sobering Reality
The United States has one of the highest maternal mortality rates among high-income countries. One in six women experiences mistreatment during pregnancy or postpartum care such as loss of autonomy; being shouted at, scolded, or threatened; and being ignored, refused, or receiving no response to requests for help.
These rates increase among women from underserved communities, specifically women of color and those in poverty.¹
The sobering reality is that 60% of U.S. maternal deaths are preventable.²
The unequal treatment of Black and Native American women contributes significantly to these high mortality rates, as these women are more likely to experience severe maternal morbidity due to postpartum hemorrhage, hypertensive disorders, and sepsis. The outcomes for their infants also reflect these disparities, who have higher mortality rates.³
It's clear that addressing the underlying social and economic factors, as well as the structural and systemic racism and discrimination, will be crucial in reducing these disparities and creating positive health outcomes for pregnant women and new mothers.
AI platforms like ChatGPT can remove many of these barriers to care.
ChatGPT
GPT, or Generative Pre-trained Transformer, is a neural network machine learning model developed by OpenAI that can generate human-like text and memory in any language. This technology will revolutionize maternal health by improving communication, removing barriers to education, and providing more comprehensive care coordination.
Generative AI can assist healthcare providers in tasks such as documentation and chart abstraction, as well as in clinical applications such as natural language processing.⁵ AI can help perinatal health by improving clinicians’ abilities to predict, diagnose, detect, and monitor various health indicators such as maternal mortality and morbidity, low birthweight, and preterm birth.⁶
GPT is so powerful that it even meets the qualifications to be cited as a co-author of a scientific paper published in peer-review journals, according to the ICMJE criteria.⁵
That’s just a taste of what generative AI can do to significantly improve healthcare and make a positive impact on maternal health.
Symptom Tracking and Triage
Remote care coordination involves using technology to facilitate communication between healthcare providers and patients, including sending reminders for necessary appointments or tests and sharing patient information between care providers. Generative AI can improve the patient experience by making it easier to provide information, freeing clinicians to spend more time in meaningful clinical encounters.⁹
AI advances can also assist in symptom tracking and triage. They help in collecting accurate data from patients in a user-friendly way. For example, using AI to assist with triaging non-critical patients in the emergency department can improve the patient experience by making it easier to provide information and enhance patient care.⁸
Additionally, they can help new mothers monitor their postpartum recovery and identify and manage common postpartum complications, such as postpartum depression or fatigue.
The advanced understanding capabilities of AI make it useful in remote care coordination and symptom tracking. They have the potential to improve the patient experience, free up clinicians, and enhance patient care.
Electronic Health Records
These AI-powered systems can help navigate and summarize complex electronic health records (EHRs) and improve patient-provider communication.
Generative AI can analyze large EHR datasets to identify patterns and risk factors for complications like preeclampsia, gestational diabetes, and preterm birth. As a result, these systems can provide valuable insights into the health of pregnant women, prompting interventions and improving maternal health outcomes.²
AI-powered systems can even analyze social determinants of health to identify patients at higher risk of poor maternal outcomes due to poverty, lack of access to transportation, and inadequate housing.⁷
In addition to analyzing EHRs, these systems can improve patient-provider communication, such as virtual consultations to answer pregnancy-related questions or follow-up after delivery or surgery.
They can also predict which patients are most likely to miss appointments or fail to comply with treatment recommendations. This information can help providers identify and support patients who need extra help to stay on track with their care.⁷
By analyzing complex EHRs and improving patient-provider communication, these AI-powered systems can provide valuable insights and support to patients and providers.
Assisted Diagnosis
One of the most exciting areas of AI in healthcare is the development of machine-learning models. Researchers at the Mayo Clinic have developed models that can predict the outcomes of vaginal deliveries based on patterns of change in pregnant patients in labor.⁸ Based on dilation data, these models can predict the probability of poor labor outcomes, such as cesarean delivery, postpartum hemorrhage, and neonatal morbidity or mortality.
Using these models can result in more individualized clinical decisions and be a tool to help remote physicians and midwives transfer rural or remote patients to the appropriate level of care.
AI can also provide personalized and insightful treatment regimens for patients. By incorporating all the knowledge in the universe about a patient's history, social demographic, ethnicity, and more, it could translate this information into a tailored treatment plan with detailed information on side effects and what works best for the individual patient.⁴
Machine learning is already being used in the medical field to predict various perinatal health indicators such as preterm birth, birth weight, preeclampsia, mortality, hypertensive disorders, and postpartum depression.⁶
By incorporating real-time electronic health recording and predictive modeling, artificial intelligence has successfully monitored fetal health and women with gestational diabetes, particularly in resource-limited settings.
The use of AI in medicine has the potential to provide personalized and insightful treatment plans for patients, improving overall health outcomes. The capability of AI to understand context and intent makes it a valuable tool in the field of remote care coordination and symptom tracking.
Data Parsing
Medical research is also poised to experience significant changes because of AI. Junaid Bajwa, Chief Medical Scientist at Microsoft, claims medical knowledge doubles every 73 days. He estimates it could double every three days in the next few years.⁴
Sifting through and making sense of that amount of information seems impossible. But with generative AI, summarizing and translating research papers from around the world can be done in a few short seconds.
In addition to medical research, AI will play a significant role in health apps, particularly in collecting and parsing data. For example, health apps can help monitor heart rate and blood pressure, automatically send updated information to your health records, and alert your provider in case of emergencies.
Serving Underserved Communities
According to the Centers for Disease Control and Prevention (CDC), Black women are three times more likely than their White counterparts to die from a pregnancy-related cause.⁷
In addition to higher rates of pregnancy-related death, Black, American Indian, and Alaska Native women have higher shares of preterm deliveries, low birth weight, and late or no prenatal care compared to White women. This disparity also extends to their infants, who have higher mortality rates.³
These tools can compile and remember specific data missed due to biases and other factors that lead to higher maternal mortality rates in these communities, such as the case of Dr. Shalon Irving.
Dr. Irving was a respected epidemiologist at the Centers for Disease Control and Prevention and a lieutenant commander in the U.S. Public Health Service. She was dedicated to investigating public health disparities and understanding the root causes of health disparities among different communities.
Unfortunately, in January 2017, just three weeks after giving birth, Dr. Irving died of complications related to high blood pressure despite monitoring her health throughout her pregnancy and delivering her child via cesarean section.¹⁰ Dr. Irving's death highlights the need for these systems in maternal healthcare.
Advancements in AI can help address racial and socioeconomic disparities and improve outcomes for pregnant women in underserved communities.
These models can also be used for remote physicians and midwives, providing powerful guidance as they make critical decisions during the labor process.
In low-income areas, maternal mortality is much higher than in high-income regions, and AI-powered models can help transfer patients to the appropriate level of care, reducing the risk of maternal death due to obstetric hemorrhage, infections, hypertensive disorders during pregnancy, preexisting health conditions, complications from delivery, and unsafe abortion.⁸
The integration of AI in obstetrics can improve outcomes for pregnant women in underserved communities, reducing maternal and infant health disparities. By providing powerful guidance to physicians and midwives, AI has the potential to help save lives and improve the health of mothers and their children.
The Future of Maternal Health
OpenAI CEO, Sam Altman, has hinted that the future of generative AI will be multimodal, meaning it can function in multiple modes such as text, images, and sounds. Altman expects this technology to be a platform for new business models and will open up new opportunities. He also spoke of the desire for AI to learn and self-improve beyond the current version paradigm.¹¹
As technology advances, we expect to see increased integration of AI in the medical field, leading to a more efficient and effective healthcare system. AI will revolutionize healthcare in ways we can only imagine. From helping with medical research to predicting outcomes of deliveries, as more companies incorporate this technology into existing systems, it will completely disrupt the maternal health industry.
Sources
[1] https://reproductive-health-journal.biomedcentral.com/articles/10.1186/s12978-019-0729-2
[2] https://hbr.org/2021/08/how-ai-could-help-doctors-reduce-maternal-mortality
[3] https://www.kff.org/racial-equity-and-health-policy/issue-brief/racial-disparities-in-maternal-and-infant-health-current-status-and-efforts-to-address-them/
[4] https://analyticsindiamag.com/finally-microsoft-experiments-with-gpt-3-in-healthcare/
[5] https://assets.researchsquare.com/files/rs-2404314/v1/3984ebcf-c868-41d3-bf27-199df87aa478.pdf?c=1672262984
[6] https://www.pharmacistsconnect.com/ai-in-healthcare-with-openai-gpt3/
[7] https://healthitanalytics.com/features/how-ai-can-help-alleviate-black-maternal-health-disparities
[8] https://www.healthcareitnews.com/news/ai-could-help-deliver-greater-success-birth
[9] https://www.nature.com/articles/s41746-021-00464-x
[10] https://www.americanprogress.org/article/health-care-system-racial-disparities-maternal-mortality/
[11] https://www.searchenginejournal.com/openai-gpt-4/476759/
