JMIR Publications

ABSTRACT

Background:

Chatbots, a type of digital health intervention, are computer programs designed to simulate conversation with human users. Well-designed, evidence-based chatbots can provide patients with real-time, personalized health interactions and reliable support tailored to their individual needs and preferences. However, the development of chatbots that support specific disease states, such as breast cancer, is in its infancy, and more reliable evidence is needed for their use to successfully bridge the patient-provider communication gap and empower rural patients.

Objective:

The goal of the proposed project is to gather preliminary data on the modalities and applications in which chatbots have been utilized to support rural patients and to define the functionality and perceived usefulness of a chatbot design for rural Nebraskans with breast cancer. Our specific aims include 1) delineating rural breast cancer patients’ needs for nutrition, exercise, and education support for Chatbot content, and 2) developing a low-fidelity chatbot prototype for rural breast cancer patients to gather initial patient feedback.

Methods:

We will use a community-based participatory research study design, integrating with the user-centric Design Thinking Process. We will prospectively recruit a purposive sample of 33 participants, comprising 30 patients (15 English speakers and 15 Hispanic and Latino individuals) with breast cancer, and a sample of 3 patients with breast cancer living in rural areas to convene a community work group (CWG). Inclusion criteria are patients aged 19 years or older, English- and/or Spanish-literate, and living in rural Nebraska. The exclusion criteria comprise patients in hospice care. We will conduct interviews with study participants to understand their needs for nutrition, exercise, and educational support according to the phase/type of treatment and symptoms; their preferred communication methods; and their perceptions of receiving support through a chatbot. During the interviews, we will also administer the Breast Cancer Literacy Assessment Tool (Breast-CLAT) and the eHealth Literacy Scale (eHEALS) to measure participants’ health and technology literacy. With the CWG, we will conduct 4 brainstorming sessions to gather their feedback on study design adaptations, recruitment, and the chatbot content. Qualitative data will be recorded in their original language and will be analyzed using NVivo® software. The literacy questionnaires selected by the research team are standardized and validated in both languages, allowing meta-theme analysis and comparisons between the groups.

Results:

The results of this project inform the design, development, usability, and content validity of a chatbot for rural breast cancer patients in Nebraska and beyond for future study.

Conclusions:

There is a critical need to identify the unique challenges rural breast cancer patients face due to their remote geographic proximity to timely oncology care. This proposal addresses an important scientific knowledge gap by designing a low-fidelity chatbot prototype that meets the needs of rural patients with breast cancer.