Effect of a Chatbot on Nursing Students' Self-Efficacy in Pediatric Medication Administration

Effect of a Chatbot-Supported Clinical Training on Nursing Students' Self-Efficacy in Pediatric Medication Administration: A Randomized Controlled Trial

This study looked at whether a chatbot can help nursing students feel more confident about giving medicines to children. Giving medicines to children is difficult because doses depend on the child's weight, and mistakes can be harmful. Third-year nursing students starting their pediatric clinical placement took part. They were divided by chance into two groups. Both groups received the usual training. One group also used a chatbot that helped with medicine dose calculations and answered questions about giving medicines to children, based on a trusted nursing textbook. The other group used only the usual references. Students' confidence in giving medicines to children was measured with a questionnaire before and after the placement. The study compared how much each group's confidence changed to see whether the chatbot added a benefit beyond the usual training.

This was a two-group, parallel, pre-test-post-test randomized controlled trial evaluating the effect of a chatbot-supported clinical training on nursing students' self-efficacy in pediatric medication administration. The trial was reported in accordance with the CONSORT statement.

All eligible third-year nursing students entering the pediatric clinical placement were invited (complete enumeration), and 106 students who consented were enrolled. Participants were allocated to an experimental group (n=53) and a control group (n=53) using fixed block randomization (block size = 6).

Both groups received the routine curriculum. The experimental group additionally used a web-based chatbot before and during the placement. The chatbot contained two modules: a dose calculator (computing required drug amount and volume from the child's weight, prescribed dose, and drug concentration) and a question-answer assistant. The chatbot's content was drawn solely from a validated pediatric nursing reference textbook rather than free AI-generated responses. The control group accessed standard clinical references only. After the study, the control group was also offered access to the chatbot.

Data were analyzed using independent- and paired-samples t-tests and analysis of covariance (ANCOVA) with the pre-test score as a covariate.

Inclusion Criteria:

• Third-year nursing student
• Entering the pediatric (child health and diseases) clinical placement for the first time
• Owning a smartphone or computer
• Providing voluntary informed consent to participate

Exclusion Criteria:

• Previous completion of a pediatric clinical placement
• Prior participation in a similar educational intervention