Evaluating the impact of an AI-powered chatbot on epilepsy education and stigma reduction: A pre-post intervention study using EpiloBot.

OBJECTIVE

Effective epilepsy management requires accurate epilepsy knowledge, active patient engagement, and stigma reduction to improve health outcomes. Educational interventions have been shown to enhance patients' quality of life (QOL) and knowledge. This study evaluated the effectiveness of EpiloBot, an artificial intelligence (AI)-powered chatbot designed to provide epilepsy education and psychosocial support, through a pre-post intervention study measuring changes in knowledge, attitudes, and self-stigma among people with epilepsy (PWE) and their supporters (including caregivers, family members, friends, or others closely involved with someone with epilepsy).

METHODS

The study consisted of two phases: a preliminary pretest phase to refine chatbot content and a main intervention phase. In the main phase, 176 participants (13 PWE, 69 active supporters, 28 potential future supporters, and 66 others) registered to use EpiloBot via the smartphone app. Of these, 82 completed both pre- and post-intervention surveys. The surveys assessed epilepsy-related knowledge, workplace attitudes, and self-stigma before and after chatbot use. Paired t-tests and Wilcoxon signed-rank tests were used to analyze changes over time.

RESULTS

The chatbot intervention led to a significant improvement in attitudes toward workplace equality for PWE (p < 0.001) and in knowledge about epilepsy treatment (p = 0.022). A slight increase in epilepsy-related self-stigma was observed among participants with epilepsy (n = 9, p = 0.022). Non-significant results were removed to streamline the findings. These results highlight the chatbot's potential as a scalable digital tool for epilepsy education and psychosocial support, although the challenge of addressing self-stigma remains a critical issue for future applications.

SIGNIFICANCE

EpiloBot demonstrates effectiveness in enhancing epilepsy-related knowledge and workplace attitudes but also reveals the complexities of self-stigma dynamics. While educational interventions can promote inclusivity and knowledge gain, self-stigma may increase due to heightened awareness of epilepsy-related challenges. Future studies should explore long-term impacts, integrating educational and psychosocial support interventions to optimize chatbot-based learning in epilepsy care.

PLAIN LANGUAGE SUMMARY

This study tested EpiloBot, an AI-powered chatbot designed to educate people with epilepsy (PWE) and their supporters, including caregivers and family members. We measured changes in epilepsy-related knowledge, attitudes, and self-stigma before and after chatbot use. Results showed improved attitudes toward PWE in the workplace and increased understanding of epilepsy treatment. However, self-stigma slightly increased among PWE, possibly due to greater awareness of social challenges, highlighting the need for long-term psychosocial support alongside education.