Free Counseling Chatbot in a Communication Application For Improving Patient Satisfaction with Indwelling Ureteric Stents After Ureterorenoscopic Lithotripsy: A Retrospective Study (Preprint)

Abstract

Health education is important for improving patients’ adherence to treatment, thereby reducing morbidity. Face-to-face communication is not sufficient nowadays, and online interaction can improve patient–physician communication and education.

We designed a chatbot for patients who received indwelling double-J ureteric stents (DJs) after ureterorenoscopic lithotripsy (URSL) and evaluated the efficacy of this chatbot in improving patient satisfaction in clinical practice.

We designed a chatbot, based on the free module provided by the communication application Line©, which described the associated symptoms with DJs and the self-care of DJs after discharge and emphasized the importance of timely DJ removal. Patients could interact with the chatbot for any concerns regarding their DJs after discharge. We prospectively included patients who received indwelling DJs after URSL at our hospital from August 1st, 2019 to November 30th, 2019. Patient education on DJ-related information was conducted either by medical staff before discharge or by using the chatbot, based on patients’ preference. Patients were asked to rate the severity of their DJ-related symptoms and their satisfaction with using the free chatbot on a five-point scale before DJ removal. Fisher’s exact test was used to evaluate the effect of the chatbot on the severity of DJ-related symptoms and the possible factors associated with the satisfaction with this chatbot.

A total of 70 patients were included. Twenty patients received routine education by medical staff while 50 patients elected to have additional interaction through the chatbot. The patients in the chatbot group were significantly younger (age <60 years: 74% versus 15%, P < .001), had a higher education level (40% versus 5%, P = .004), and reported more severe gross hematuria (66% versus 15%, P < .001) than those in the medical-staff group. No differences were observed for other DJ-associated symptoms. On multivariate analysis, severe gross hematuria was significantly associated with age younger than 60 years (odds ratio 6.704, P = .003, 95% CI 1.898–23.673) and the use of the chatbot (odds ratio 6.63, P = .02, 95% 1.374–31.989). All 50 patients in the chatbot group reported being satisfied (32%) or very satisfied (68%) with the chatbot tool. Patients older than 60 years were significantly more satisfied with the chatbot (35.5% versus 6.3%, P = .04). Education level, the severity of DJ-associated symptoms, and the recognition of the necessity of DJ removal were not significantly associated with the degree of satisfaction.

The use of a chatbot resulted in high satisfaction of the patients, especially elderly patients. Younger patients with higher education levels were more likely to adopt this new form of communication, which helped improve their knowledge of DJ-associated symptoms.