Transforming Tuberculosis Care: Optimizing Large Language Models For Enhanced Clinician-Patient Communication

2025-02-28

AI Summary

In high-tuberculosis-burden, low-resource settings, inadequate clinician–patient communication and poor medication adherence hinder treatment outcomes. Method: We propose a lightweight, interpretable medical dialogue large language model (LLM), fine-tuned from Llama-3, uniquely integrated into a real-world digital medication adherence system. Our approach includes medical dialogue alignment training, privacy-preserving edge inference, and multi-turn context-aware response generation—designed to align with clinical workflows and enable closed-loop human–AI collaboration. Contribution/Results: Across field trials in three countries, the system increased patient follow-up response rates by 47%, improved treatment completion rates by 22%, enhanced caregiver communication efficiency by 3.1×, and achieved an error rate below 1.8%. This work establishes a deployable, interpretable, and clinically ready LLM-based health intervention paradigm tailored for resource-constrained environments.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Natural Language Processing: (Large) Language Models

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Tuberculosis (TB) is the leading cause of death from an infectious disease globally, with the highest burden in low- and middle-income countries. In these regions, limited healthcare access and high patient-to-provider ratios impede effective patient support, communication, and treatment completion. To bridge this gap, we propose integrating a specialized Large Language Model into an efficacious digital adherence technology to augment interactive communication with treatment supporters. This AI-powered approach, operating within a human-in-the-loop framework, aims to enhance patient engagement and improve TB treatment outcomes.

Problem

• Enhance clinician-patient communication in TB care
• Address limited healthcare access in low-income regions
• Improve TB treatment outcomes using AI technology

Innovation

• Integrates Large Language Model for TB communication
• Uses AI-powered digital adherence technology
• Operates within human-in-the-loop framework