Design and evaluate a bilingual chatbot for healthcare communication improvement

Thesis Abstract

Effective communication between healthcare providers and patients is critical for ensuring accurate diagnosis, adherence to treatment, and improved health outcomes. However, language barriers and linguistic diversity pose significant challenges in many healthcare settings, often leading to miscommunication, patient dissatisfaction, and compromised care quality. This study aims to design, develop, and evaluate a bilingual chatbot tailored to facilitate healthcare communication for both English and Spanish-speaking patients within urban primary healthcare clinics. The specific objectives include examining the chatbot’s usability, accuracy, and patient satisfaction, as well as assessing its impact on communication clarity and health literacy. Employing a mixed-methods research design, the study integrates quantitative and qualitative approaches to generate comprehensive insights. The target population comprises adult patients visiting five urban primary care clinics, totaling approximately 1,000 individuals. A stratified random sampling technique is used to select a sample size of 300 participants to ensure diverse representation across age, gender, and language proficiency. Data collection instruments include structured questionnaires to measure usability, satisfaction, and perceived communication improvements, as well as semi-structured interviews to capture participants’ experiences and perceptions in-depth. The bilingual chatbot, developed using natural language processing (NLP) algorithms aligned with the Common Sense Model of Health Communication, is employed as an intervention over a three-month period. Quantitative data will be analyzed using descriptive statistics, t-tests, and multiple regression analyses to evaluate the chatbot’s effectiveness in improving communication quality, health literacy, and patient satisfaction. Thematic analysis will be conducted on qualitative interview data to identify common themes related to user experience, cultural appropriateness, and perceived barriers. Data triangulation will support the validity of findings, while the integration of quantitative and qualitative results will provide a comprehensive assessment of the chatbot’s performance. Expected findings suggest that the bilingual chatbot will significantly enhance communication clarity, patient engagement, and satisfaction levels, especially among non-English speakers who previously experienced language barriers. It is anticipated that users will report improved understanding of medical instructions and health information, leading to greater adherence to prescribed treatments. The study also expects to identify linguistic and cultural factors influencing user interaction with chatbot technology and pinpoint areas for linguistic refinement and interface optimization. This research contributes to the existing body of knowledge by providing empirical evidence on the feasibility, effectiveness, and user acceptance of artificial intelligence-driven communication tools in healthcare contexts, particularly those supporting multilingual populations. It advances theoretical understanding by integrating models such as the Health Belief Model and the Technology Acceptance Model within the framework of NLP-based interventions for health communication. The practical implications include guiding healthcare providers and developers in designing culturally sensitive, accessible, and user-friendly chatbot applications, with potential for scalability across diverse healthcare systems. In conclusion, the study affirms that a well-designed bilingual chatbot can serve as a scalable solution to mitigate language barriers, improve health literacy, and promote patient-centered care. Recommendations highlight the importance of incorporating user feedback, cultural considerations, and ethical standards into chatbot development processes. Future research directions include exploring long-term impacts on health outcomes, integrating multimedia interfaces, and expanding to other language pairs and healthcare settings to enhance inclusivity and accessibility in health communication strategies.