JMIR Publications

ABSTRACT

Background:

Chronic conditions cause over two-thirds of global deaths. Effective self-management is vital but often hindered by limited patient knowledge and skills. Osteoporosis (OP) leads to fragility fractures and high healthcare costs, yet many patients remain unaware of necessary lifestyle modifications. While online education provides flexibility, usability issues reduce engagement. Chatbots, offering interactive and personalized communication, show promise for health education and behavior support. However, evidence on their effectiveness in improving OP self-management remains scarce. Chatbots Self-Management Education Osteoporosis AB STRACT

Objective:

This study evaluated the effectiveness of a chatbot (OPBot) for self-management education (SME) in osteoporosis patients. Design: This study conducted a single-blind randomized controlled trial. Setting: Participants were recruited from inpatients in the affiliated hospital of Guangdong Medical University, Guangdong, China. Method: We conducted an RCT to compare the effectiveness of an OPBot versus traditional health education methods for patient SME. Participants were diagnosed with osteoporosis. We assessed osteoporosis knowledge, the time nurses invest in health education, adherence to disease management at 1, 3 and 6 months, and the reliability of the OPBot responses in free conversations. Result: A total of 100 participants were randomized; 12 were excluded from the final analysis due to loss to follow-up, refusal to complete the second knowledge assessment, or death, leaving 88 participants included in the analysis (45 in the intervention group and 43 in the control group). Knowledge scores improved significantly in the OPBot group compared with the control group (mean difference (MD): 6.069; 95% CI: −0.894 to Inf, PNI = 0.048). Nurses in the OPBot group spent less time on verbal education than those in the control group (MD: -12.848, 95% CI: -Inf to -10.007, PNI < 0.001). OPBot also demonstrated better patient self-management adherence at 1 month (MD: 7.733, 95% CI: 0.177 to Inf, PNI = 0.029) and 6 months (MD: 17.765, 95% CI: 8.911 to Inf, PNI < 0.001), but not at 3 months (MD: 2.694, 95% CI: -5.463 to Inf, PNI = 0.227). In the Q&A module, the reliability of OPBot responses was as follows: highly reliable, 89.4%; mostly reliable, 1.2%; partly reliable, 1.2%; unreliable, 1.2%; and unable to respond, 7.0%. Conclusion: Chatbots are suitable for basic SME, but further research and exploration are needed in the area of personalized education.