Cary cares about you

Menstrual health has historically been neglected in research despite evidence suggesting its significant impact on physical and psychological well-being. This neglect extends to the provision of personalized digital healthcare technologies (DHTs), such as healthcare chatbots, for menstruating individuals. To the best of our knowledge, no study has specifically explored how the menstrual cycle affects patient-chatbot engagement within healthcare settings. This study explored how the menstrual cycle influences patients’ engagement with empathic chatbots. We specifically evaluate the impact of empathic self-awareness (ESA) and empathic active listening (EAL) cues in text- and rule-based healthcare chatbots and how these cues affect perceptions of empathy, user engagement, and the patient-chatbot relationship in healthcare contexts.

Research Questions

How does user’s menstrual and mental health affect their perceptions and engagement with a healthcare chatbot?
How do distinct empathic cues affect users’ self-disclosure?

Approach

We developed four functional but fictitious healthcare chatbot prototypes engaging patients in a realistic anamnesis dialogue incorporating EAL and ESA cues. These cues were manipulated following a 2 x 2 full-factorial between-groups design (1: no empathic cues, 2: ESA only, 3: EAL only, 4: EAL+ESA). We aimed to recruit 1’000 participants with chronic conditions from the UK. Participants were invited via Prolific and randomly assigned to interact with one of these prototypes; 921 provided complete data (female: 50.27%; mean age = 42.4 years, SD = 14.1). Menstruating individuals were also asked for menstrual health and cycle-related questionnaires to examine the influence of characteristics such as menstrual cycle phase or menstrual health disorders on the patient-chatbot engagement.

Expected Outcomes

As we continue our analysis, we expect to find that patient-chatbot engagement and perceived empathy vary across different menstrual cycle phases and mental health conditions. These insights could reveal the need for personalized empathic responses in chatbots to address the fluctuating emotional and physical states of users. By understanding these variations, the Cary Project aims to establish guidelines for developing adaptive, inclusive, and empathetic chatbot communication strategies that respond sensitively to the diverse and changing needs of menstruating individuals.

Impact

The findings from the CARY Project will inform the design of DHTs that better serve the unique health needs of menstruating individuals, empowering them to engage in meaningful healthcare interactions. Through this research, we aim to contribute to the development of more equitable digital health tools that respect and adapt to the nuanced experiences of all users, fostering a more inclusive digital healthcare landscape.

Screenshots

Publications

Sou, D., Kuhlmeier, F., Kowatsch, T., von Wangenheim, F., Nißen, M., Towards Digital Empathy in Healthcare Chatbots: A Conceptual Framework and Empirical Study, In Proc. of the International Conference on Information Systems (ICIS), Bangkok, Thailand, Dec 15-18. https://aisel.aisnet.org/icis2024/ishealthcare/ishealthcare/5/

Sou, D., Budig, T., Kowatsch, T., Kuhlmeier, F., von Wangenheim, F., Nißen, M.K., The Impact of Menstrual and Mental Health on Patients’ Interactions with a Healthcare Chatbot, Menarche Menstruation Menopause and Mental Health (4M) Conference 2024, 20-21 June, University of Exeter, UK, Abstract & Poster. [PDF]