Empathetic AI chatbot transforms reproductive health conversations
Two of the biggest challenges with chatbots for healthcare are incorporating empathy and including evidence-based content. Chatbots need to provide trustworthy information while also sounding like a friend. Given that reproductive healthcare provision requires privacy, education, compassion, and personal decision making, it is a prime opportunity for leveraging chatbots.
Why reproductive health chatbots are appealing:
- They increase education: Fewer than half of US high school students receive education that meets the CDC’s standard of sexual education, including topics ranging from prevention of sexually transmitted infections, how to avoid or delay pregnancy, and how to use available health services, but 95% of American teens have a mobile phone that could access a chatbot on a web or mobile app.
- They are discrete: Users can quickly access the information that they need and close out the browser tab or delete the app, protecting anonymity of teens, women at risk for domestic violence, and those in litigious reproductive health states.
- They expand access: More than 2.2 million women of childbearing age live in maternity care deserts in the United States, and 4.7 million women live in areas with limited access. Worldwide, 218 million women have unmet family planning needs. Bots linked to care delivery can efficiently connect patients with in-person providers when needed, like Planned Parenthood’s Roo bot.
- They are accessible: Want answers about whether or not to take Plan B at 2am? Chatbots offer 24/7 access to information and support.
- They are cost effective: Many healthcare chatbots are free and can reduce administrative and financial costs for providers by providing quick answers to questions that would otherwise end up in their patient messaging portals.
Past attempts at healthcare chatbots have had mixed success:
The evidence is mixed when it comes to how acceptable and effective reproductive health chatbots are to people. A 2024 systematic review of 15 sources found that “chatbots may be experienced as acceptable, convenient, anonymous, and private, but also as incompetent, inconvenient, and unsympathetic.” Another 2021 qualitative research study shared that when compared to internet searches and conversations with healthcare providers, chatbots are perceived as inferior due to content constraints and lack of empathy. But a 2023 study noted that chatbots can also model safe and open conversations about sexual and reproductive health, and potentially improve health literacy and language skills when it comes to open conversations around those topics.
Clinician informed care brings together empathy and evidence
Kandyce Brennan, a certified nurse midwife and researcher at the UNC-Chapel Hill School of Nursing aims to empower individuals to take an active, informed role in managing their reproductive health with SARHAchat, a chatbot she developed. Users can ask questions about safe sex practices and reproductive health – all of which is private and does not require creating a log in or sharing any identifiable information. Users can also download an optional summary of the conversation if they want to use it as a reference for discussion with a healthcare provider. This work is funded through a Society for Family Planning Changemakers grant.
What makes SARHAchat special is the scientific rigor brought to the table by Brennan, who is conducting accuracy and acceptability research alongside product development. When I spoke with Brennan about how she is ensuring evidence-based content, she shared, “We integrate the 2024 CDC contraception recommendations and have established an automated review process to ensure accuracy, along with an expert review layer. We also trained the model using examples from real user conversations and applied principles from the patient-centered care model.”
Empathy is also built into the product. I asked Brennan about what sets SARHAchat apart when it comes to user experience, and she shared that, “Other chatbots I researched were developed using a dataset of frequently asked sexual and reproductive health questions and corresponding responses. When users ask questions outside of this predefined scope, the chatbot struggles to fully address their needs. This often results in responses perceived as unempathetic or robotic, exacerbated by technical limitations. In contrast, SARHAchat utilizes natural language processing (NLP) and machine learning to provide users with accurate, tailored responses that address their unique needs. Additionally, we have integrated a set of trusted sources for real-time responses, allowing our chatbot to go beyond standard FAQs and offer more comprehensive support.”
Data from their pilot study showed high satisfaction with respectfulness and personalization when using the chatbot.
I took SARHAchat for a spin and agree that empathy comes to the forefront of the user experience upon opening the chat. Language like “Our conversation will be guided by mutual respect for what you feel is best for you” invites trust. I found that the chatbot started a conversation exploring the best contraceptive method for me by inquiring about my personal tolerance of side effects and responding with appreciation for sharing my viewpoints. Further prompts went into health history concerns like chronic diseases, high blood pressure, and migraines, personalizing the experience further.
The final results surprised and delighted me (I’ve volunteered as a women’s health educator in the past.) They included images of different contraceptive methods (ex: uterus with an IUD inserted), pros and cons of options, and other important considerations like the need for an in-office procedure versus purchasing something over the counter, like condoms. As a woman on the other side of having two children, I found that the results from SARHAchat mirrored what I had previously heard from healthcare providers since I started my reproductive health journey.
I wrapped up my conversation with Brennan by asking about her vision for SARHAchat. She shared that she hopes it will “serve as an educational resource and sexual and reproductive health (SRH) triage tool for national health systems, telehealth services, and online pharmacies by providing accurate, patient-centered education and care, aiming to reduce health disparities, particularly in SRH care deserts.” SAHRAchat is an excellent example of how to engineer empathy and evidence into patient experiences.
Katie D. McMillan, MPH is the CEO of Well Made Health, LLC, a business strategy consulting firm for health technology companies. She is also a curious researcher and writer focusing on digital health evidence, healthcare innovation, and women’s health. Katie can be reached at katie@wellmadehealth.com or LinkedIn.