Development of a Rule-Based Chatbot for Supporting Family Caregivers After Psychiatric Hospital Discharge: An Exploratory Qualitative Study in Indonesia

Purpose: Family caregivers play a crucial role in supporting patients with mental disorders after hospital discharge. However, caregivers often experience difficulties related to limited information, caregiving burden, and restricted access to mental health services. This study aimed to explore the informational and skill-related needs of caregivers of psychiatric patients after hospital discharge and to develop a prototype chatbot-based consultation system tailored to these needs in Indonesia. Materials and Methods: An exploratory qualitative design guided by Social Cognitive Theory was employed to explore caregiver needs and support the development of a preliminary rule-based chatbot prototype. Data were collected through semi-structured interviews with eight participants (five mental health professionals and three family caregivers), supported by analysis of 67 anonymized WhatsApp consultation records and five discharge education documents. Data were analyzed using inductive thematic analysis following Braun and Clarke’s framework and organized using NVivo. Trustworthiness was enhanced through triangulation, member checking, and peer debriefing. The identified themes were translated into chatbot features including psychoeducation, caregiving guidance, relapse support, and referral navigation, followed by expert validation involving six multidisciplinary experts. Results: Thematic analysis identified nine domains of caregiver needs: patient condition understanding, treatment and medication management, home care practices, community support, caregiver burden and coping strategies, supporting health services, barriers in online consultation systems, expectations for digital media, and future challenges in chatbot implementation. Caregivers emphasized the need for continuous psychoeducation, clear guidance for relapse prevention, and accessible consultation channels after discharge. Based on these findings, the identified caregiver needs were translated into a preliminary rule-based decision tree chatbot prototype (KJOL-ITEUNGBOT) designed to support caregiving information and referral navigation. Conclusion: Family caregivers of psychiatric patients in Indonesia face complex informational and psychosocial challenges after hospital discharge. A chatbot-based consultation system may serve as a potentially useful tool for supporting caregiver access to information and post-discharge guidance. Future research should evaluate the usability, effectiveness, and scalability of the chatbot in broader clinical settings.