Skip to main content

Enhancing healthcare providers' advance care planning competence with large language models: protocol for the development of an AI chatbot and its evaluation in a randomised controlled trial.

Publication ,  Journal Article
Tan, M; Tang, S; Kan, S; Zhang, H; Wu, B; Ni, Z; Lin, CC; Ding, J
Published in: BMJ open
November 2025

Advance care planning (ACP) can support individuals to express their autonomy in the decision-making process for future care. Traditional ACP training for healthcare providers faces significant challenges related to interactivity, accessibility, scalability and sustainability. Cutting-edge generative artificial intelligence (AI) holds promise in enabling intelligent and interactive chatbots for ACP. The present protocol outlines the development and evaluation of a large language model (LLM)-based ACP chatbot for healthcare providers.The development of the LLM-based ACP chatbot will follow four stages: construction of dialogue data sets, fine-tuning, multi-LLM orchestration and ablation studies. A randomised controlled trial will then evaluate the LLM-based ACP chatbot's effectiveness in enhancing ACP competence among healthcare providers. A total of 66 healthcare providers will be recruited from China and randomly assigned (1:1) to either: (1) The LLM-based ACP chatbot intervention or (2) An ACP knowledge manual. The primary outcome will be ACP competence, while secondary outcomes will include (1) ACP knowledge, (2) Attitudes/beliefs, (3) Practice willingness, (4) Readiness, (5) Self-efficacy, (6) Processes of change, and (7) Decisional balance. Both primary and secondary outcomes will be assessed to evaluate the immediate impact (postintervention) and short-term impact (3-month follow-up and 6-month follow-up) of the chatbot on ACP.The research was approved by the Ethical Review Board, Xiangya School of Nursing, Central South University (E202442). Study modifications will be discussed among the research team members until a consensus is reached. Amendments reflecting study modifications will be submitted for institutional review board approval at all sites, updated on the clinical trial registry, and fully detailed and explained in the manuscript reporting the results of the study. All participants will provide written informed consent. The study will be conducted according to the principles outlined in the Declaration of Helsinki. The results of this study will be submitted for publication in peer-reviewed journals and presented at (inter)national conferences.ChiCTR2400091022.

Duke Scholars

Published In

BMJ open

DOI

EISSN

2044-6055

ISSN

2044-6055

Publication Date

November 2025

Volume

15

Issue

11

Start / End Page

e099226

Related Subject Headings

  • Randomized Controlled Trials as Topic
  • Large Language Models
  • Language
  • Humans
  • Health Personnel
  • Generative Artificial Intelligence
  • Decision Making
  • China
  • Artificial Intelligence
  • Advance Care Planning
 

Citation

APA
Chicago
ICMJE
MLA
NLM
Tan, M., Tang, S., Kan, S., Zhang, H., Wu, B., Ni, Z., … Ding, J. (2025). Enhancing healthcare providers' advance care planning competence with large language models: protocol for the development of an AI chatbot and its evaluation in a randomised controlled trial. BMJ Open, 15(11), e099226. https://doi.org/10.1136/bmjopen-2025-099226
Tan, Minghui, Siyuan Tang, Shichao Kan, Haojie Zhang, Bei Wu, Zhao Ni, Chia Chin Lin, and Jinfeng Ding. “Enhancing healthcare providers' advance care planning competence with large language models: protocol for the development of an AI chatbot and its evaluation in a randomised controlled trial.BMJ Open 15, no. 11 (November 2025): e099226. https://doi.org/10.1136/bmjopen-2025-099226.

Published In

BMJ open

DOI

EISSN

2044-6055

ISSN

2044-6055

Publication Date

November 2025

Volume

15

Issue

11

Start / End Page

e099226

Related Subject Headings

  • Randomized Controlled Trials as Topic
  • Large Language Models
  • Language
  • Humans
  • Health Personnel
  • Generative Artificial Intelligence
  • Decision Making
  • China
  • Artificial Intelligence
  • Advance Care Planning