Research News

A chatbot to help patients navigate the complex and evolving field of prenatal genetic testing is being developed by researchers at Cleveland Clinic and will be used in partnership with MetroHealth and Neighborhood Family Practice. The project, which is funded by a $3.15 million, 4-year grant from the National Institutes of Health, will investigate how an artificial intelligence (AI) based chatbot can extend the support genetic counselors and prenatal specialists provide to patients making these decisions. The chatbot has undergone pilot testing with key stakeholders and is now ready for field use. 

The team of co-investigators includes experts in obstetrics, maternal-fetal medicine, genetics, medical decision-making, healthcare communication and AI ethics at Cleveland Clinic, MetroHealth, University of Indiana and Vanderbilt University. 

Prenatal genetic testing provides patients with more information on some congenital genetic conditions, like Down syndrome and spina bifida. Understanding the advantages and disadvantages of testing, and the possible complexities of the results, can cause confusion for patients who are considering it. This can make it difficult for patients to make informed decisions about the results that align with their values and beliefs, says Ruth Farrell, MD, MA, Vice Chair of Research for the OB/GYN Institute.

This project addresses feedback from a previous NIH-funded study on providing access to non-invasive prenatal screening and genetic testing (NIPT) in addition to all testing options. The study, titled the Non-Invasive prenatal genetic testing Education Support Tool (NEST), investigated the impact of a shared decision-making tool that could be used collaboratively between patients and their obstetric healthcare provider. 

"During the NEST study, participants said they wanted 24/7 access to information vetted by their health system," Dr. Farrell says. "They didn't want to just search online for information – they wanted to read what a provider would say in response to these personal and often complex questions. This type of resources would help them make informed decisions about their testing choices but also provide a way for them to engage their partner or other support person in these decisions." 

Different prenatal genetic screens and diagnostic tests are available throughout pregnancy, and testing is optional. Prenatal genetic screening determines the likelihood a fetus might have a genetic condition, most commonly for chromosomal aneuploidies. That also includes prenatal carrier screening to determine if parents may be carriers for a recessive condition. Diagnostic tests can identify whether a fetus has a genetic condition, using cells collected from the amniotic fluid or the placenta.

Why an AI-based chatbot? 

Chatbots can serve as an important adjunct to the counseling and education provided by patients' healthcare teams. They are designed to pull from an information bank using AI to mimic a human conversation. As users ask the chatbot questions, it learns how to respond with more relevant information.

Cleveland Clinic introduced a chatbot during the coronavirus pandemic as a free, 24/7 online screening tool, guiding patients through questions to determine their risk for COVID. Engaging Patients in Prenatal Genetic Testing Decisions as a Pathway to Improve Obstetric Outcomes (OPUS) builds on that formative effort, investigating how this type of technology can be used to support healthcare delivery to patients during pregnancy.

Another benefit of chatbots is that patients can explore topics on their own, either before a visit or if they want to consider reproductive healthcare options outside of discussions with their clinician. If a patient is unsure about asking their provider a question or does not have the chance to during the visit, the chatbot can offer a private place to get accurate information to inform medical decisions between their prenatal care visits, Dr. Farrell says.

She added the chatbot is being built to run without users entering identifying information or being sent to other resources on the internet, so patients can ask the questions they want without worrying about their IP address being tracked online. 

The research team is partnering with MetroHealth and local health group Neighborhood Family Practice on the project. Working with other health systems also helps to understand the utility of the chatbot across diverse patient populations, incorporating questions and perspective from patients outside of Cleveland Clinic's system. 

This resource must meet the needs of diverse patient populations and healthcare providers, Dr. Farrell says, so thorough testing is necessary to ensure everything is working well before it's launched. 

"AI has the potential to be a great mechanism for patients to access health information. Yet, we must understand how patients want to use AI-based chatbots and similar technologies and make sure it meets their needs," Dr. Farrell says. "We also must be certain we don't increase the 'digital divide' seen in healthcare over the past few years." 

Ensuring equitable access 

Dr. Farrell emphasizes that this study will look at patients' access to information about their prenatal genetic testing choices through an equity lens. 

Cleveland Clinic has outstanding resources to educate pregnant patients about their options and help them make informed decisions about whether, when and how to use prenatal genetic screens and diagnostic tests. Yet, other healthcare systems may have different resources. 

"It doesn't matter where the patient receives prenatal care; all patients should have access to accurate, unbiased and timely information about their choices so that can make decisions that best match what they need and want for themselves and their families," Dr. Farrell says.

The team will also provide tablet PCs and, when needed, internet access. Users will also receive orientation and training, so that those who may be less familiar or experienced with this type of resource will feel more comfortable and confident using it.

The project will also gather vital information about how pregnant patients engage with AI-based technologies, which, in turn, help researchers tailor the resource for different patients and their individual needs. The team will engage with leaders in AI and AI ethics to study how best to use these technologies among different patient populations.

"Pregnancy is this unique time; each week brings more information and choices to our patients. We want to be certain we are ready to support them at every stage, no matter where they are in their pregnancy journey," Dr. Farrell says.