Fine-Tuning Large Language Models for Effective Nutrition Support in Residential Aged Care: A Domain Expertise Approach
Open Access
- 17 October 2025
- journal article
- research article
- Published by MDPI AG in Healthcare
- Vol. 13 (20) , 2614
Abstract
Background: Malnutrition is a serious health concern among older adults in residential aged care (RAC), and timely identification is critical for effective intervention. Recent advancements in transformer-based large language models (LLMs), such as RoBERTa, provide context-aware embeddings that improve predictive performance in clinical tasks. Fine-tuning these models on domain-specific corpora, like nursing progress notes, can further enhance their applicability in healthcare. Methodology: We developed a RAC domain-specific LLM by training RoBERTa on 500,000 nursing progress notes from RAC electronic health records (EHRs). The model’s embeddings were used for two downstream tasks: malnutrition note identification and malnutrition prediction. Long sequences were truncated and processed in segments of up to 1536 tokens to fit RoBERTa’s 512-token input limit. Performance was compared against Bag of Words, GloVe, baseline RoBERTa, BlueBERT, ClinicalBERT, BioClinicalBERT, and PubMed models. Results: Using 5-fold cross-validation, the RAC domain-specific LLM outperformed other models. For malnutrition note identification, it achieved an F1-score of 0.966, and for malnutrition prediction, it achieved an F1-score of 0.687. Conclusions: This approach demonstrates the feasibility of developing specialised LLMs for identifying and predicting malnutrition among older adults in RAC. Future work includes further optimisation of prediction performance and integration with clinical workflows to support early intervention.Keywords
Funding Information
- ARC Infrastructure Grant (LE240100131)
- ARIIA Grant (R6GA00353)
This publication has 42 references indexed in Scilit:
- A disease-specific language representation model for cerebrovascular disease researchComputer Methods and Programs in Biomedicine, 2021
- Domain specific word embeddings for natural language processing in radiologyJournal of Biomedical Informatics, 2020
- Transfer Learning for Clinical Time Series Analysis Using Deep Neural NetworksJournal of Healthcare Informatics Research, 2019
- Artificial Intelligence for Mental Health and Mental Illnesses: an OverviewCurrent Psychiatry Reports, 2019
- Automatic extraction and assessment of lifestyle exposures for Alzheimer’s disease using natural language processingInternational Journal of Medical Informatics, 2019
- Malnutrition Screening and Assessment in Hospitalised Older People: A ReviewThe Journal of nutrition, health and aging, 2019
- Prescription extraction using CRFs and word embeddingsJournal of Biomedical Informatics, 2017
- Nutrition screening practices amongst australian Residential Aged Care FacilitiesThe Journal of nutrition, health and aging, 2016
- The quality of paper-based versus electronic nursing care plan in Australian aged care homes: A documentation audit studyInternational Journal of Medical Informatics, 2015
- Malnutrition in the elderly: A narrative reviewMaturitas, 2013