Abstract Background A peripherally inserted central catheter (PICC) for vesicant or long-term chemotherapy is recommended for safe and sustainable drug delivery. However, maintaining its benefits requires regular and careful self-management. Although medical staff provide education and telephone consultation, proactive support accessible at any time or location remains limited. Therefore, we developed a rule-based chatbot to support PICC self-management. Objective This study aimed to evaluate the feasibility of a chatbot designed to support PICC self-management by examining chatbot use rate, usability, and user experience. Methods A mixed methods study was conducted from September to December 2022, adhering to the GRAMMS (Good Reporting of a Mixed Methods Study) guideline. Patients with cancer scheduled for PICC insertion and their caregivers were recruited, as PICC care is commonly performed by patients or cohabiting caregivers. All participants provided written informed consent. The chatbot was designed to provide structured responses based on prespecified dialog trees and to recognize users’ intent using natural language processing. It was delivered through KakaoTalk and accessed on participants’ personal mobile phones without requiring a separate app installation. Participants received face-to-face training at enrollment and were asked to voluntarily use the chatbot for 1 month. Baseline and postintervention surveys assessing usability were administered using paper-based questionnaires. Usage logs were collected from a secure researcher dashboard and analyzed for inquiry topics, free-text inputs, and fallback situations. Semistructured interviews were conducted approximately 1 month after the intervention during outpatient visits, with invitations by telephone, to explore participants’ experiences regarding chatbot use. Quantitative data were analyzed descriptively to summarize participant characteristics, chatbot use, and usability outcomes, while qualitative interview data were analyzed using thematic analysis. Results A total of 56 participants were included in the final analysis (mean age 55.4 years, SD 13.7; female: n=39, 70%). Among them, 28 (50%) used the chatbot at least once. Chatbot users were younger than nonusers (51.1 vs 59.6 y; P =.02). Of the 25 users who agreed to log analysis, 347 inquiries were recorded; frequent topics included catheter care (126 observations), managing daily life (85 observations), symptoms (72 observations), and heparin use (55 observations). Among the 23 users who completed the usability survey, 20 (87%) reported that the chatbot was helpful for PICC-related issues. Qualitative interviews (N=56) identified 3 major benefits—information accessibility, effective guidance, and psychosocial support—while also revealing unmet needs related to conversational issues, user experience issues, and lack of personalization. Conclusions A rule-based chatbot designed to support PICC self-management demonstrates potential to enhance information accessibility, provide practical guidance, and offer psychosocial support. However, limitations related to conversational flexibility, interface usability, and personalization highlight the need for future development incorporating large language models. Longitudinal and multisite studies are warranted to assess sustained user engagement and clinical outcomes.
Jo et al. (Wed,) studied this question.