Clinical Application of Large Language Models for Intervention Plan Development in Speech-Language Pathology

Area of Science:

• Speech-language pathology
• Artificial intelligence in healthcare
• Clinical decision support systems

Background:

• Large language models (LLMs) are increasingly integrated into clinical writing tools.
• Evaluating the efficacy of AI-generated content for clinical applications is crucial.

Purpose of the Study:

• To assess the performance of six AI tools in generating speech and language intervention plans.
• To identify the applications and limitations of AI in speech-language pathology intervention planning.

Main Methods:

• A mixed-methods approach combining quantitative and qualitative analyses was employed.
• Six AI tools were evaluated using three fictional pediatric speech and language disorder cases.
• Two prompt types with varying specificity levels were used to generate AI outputs.

Main Results:

• AI-generated intervention plans were rated from 'Needs Improvement' to 'Meets Expectations' for clinical competence.
• Highly specific and structured prompts led to better AI output ratings compared to general prompts.
• AI tools exhibited distinct strengths and weaknesses in supporting intervention plan development.

Conclusions:

• Findings provide foundational data for responsible AI utilization in speech-language pathology.
• Clinicians, educators, and students can leverage these insights for integrating AI into intervention planning.
• Further research is needed to optimize AI tool performance and clinical integration.