Feasibility of LLM-assisted post-discharge tuberculosis care: A comparative study of medication counseling, patient education, and follow-up planning.
Large Language Models gave more precise tuberculosis patient-education responses than physicians in this comparison
Feasibility of LLM-assisted post-discharge tuberculosis care: A comparative study of medication counseling, patient education, and follow-up planning.
Tuberculosis (tuberculosis) remains a major global health concern, with poor post-discharge adherence driving relapse and drug resistance.
To explore the feasibility of Large Language Models as auxiliary tools to complement tuberculosis specialists in post-discharge medication counseling, patient education, and follow-up planning.
This exploratory study compared two Large Language Models (ChatGPT-4o, DeepSeek-R1) with tuberculosis physicians using 17 standardized clinical cases.
Large Language Models gave more accurate patient-education answers than physicians
Conversely, no statistically significant differences were observed in objective metrics for medication counseling and follow-up planning (all P > .050), although Large Language Models demonstrated a tendency toward broader but more discordant coverage.
Subjectively, Large Language Models were rated significantly higher in suitability, understandability, and empathy across all tasks (all P < .050).
No significant differences were found regarding perceived safety or text readability ( P > .050).
This study suggests the potential of Large Language Models as auxiliary tools in tuberculosis post-discharge management.
While quantitative performance in high-stakes tasks was comparable to physicians, the complementary qualitative patterns support a hybrid human-AI approach.
These preliminary findings provide a basis for integrating Large Language Models into clinical workflows under professional supervision to enhance efficiency in resource-limited settings.