Improving social determinants of health documentation in French electronic health records using large language models - Scientific Reports

Social determinants of health (SDoH) significantly influence health outcomes, shaping disease progression, treatment adherence, and health disparities. However, their documentation in structured electronic health records (EHRs) is often incomplete or missing. This study presents an approach based on large language models (LLMs) for extracting 13 SDoH categories from French clinical notes. We trained Flan-T5-Large on annotated social history sections from clinical notes at Nantes University Hospital, France. We evaluated the model at two levels: (i) identification of SDoH categories and associated values, and (ii) extraction of detailed SDoH with associated temporal and quantitative information. The model performance was assessed across four datasets, including two that we publicly release as open resources. The model achieved strong performance for identifying well-documented categories such as living condition, marital status, descendants, job, tobacco, and alcohol use (F1 score > 0.80). Performance was lower for categories with limited training data or highly variable expressions, such as employment status, housing, physical activity, income, and education. Our model identified 95.8% of patients with at least one SDoH, compared to 2.8% for ICD-10 codes from structured EHR data. Our error analysis showed that performance limitations were linked to annotation inconsistencies, reliance on English-centric tokenizer, and reduced generalizability due to the model being trained on social history sections only. These results demonstrate the effectiveness of NLP in improving the completeness of real-world SDoH data in a non-English EHR system.

Pluralistic: Doctors’ union may yet save the NHS from Palantir (12 Feb 2026) – Pluralistic: Daily links from Cory Doctorow

Bluesky

Designing for Health: Interview with Farhan Ahmad and Jon Keevil, MD [Podcast]

On this podcast, hear how Farhan Ahmad and Jon Keevil, MD, balanced software expertise with clinical knowledge to create accessible clinical decision-making tools.

Ask the Expert: Dr. Craig Joseph, Chief Medical Officer at Nordic Global

Join us on The Doc Lounge Podcast as we delve into the world of healthcare, technology, and human-centered design with Dr. Craig Joseph. With over 25 years of experience in healthcare and technology, Dr. Joseph, Chief Medical Officer at Nordic, brings a wealth of knowledge and insights to our discussion. From his early days as a primary care pediatrician to his pivotal roles at Epic and now Nordic, Dr. Joseph has been at the forefront of transforming healthcare through innovative design and technology solutions. His expertise in human-centered design and health IT makes him a must-have guest for C-suite executives in hospital systems and healthcare payers. Tune in to gain valuable insights into the challenges and opportunities in healthcare technology, the role of clinicians in shaping health IT, and the future trends that will revolutionize the industry.

Big jump seen in EHR secure messaging. Is that a good thing?

One health system saw a 29% rise in EHR secure-message use in just six months. But does avoiding the phone offer a leg up for better care?

Want to improve patient safety? Up your EHR’s usability

A recent study discovered a significant link between the safety of EHR systems and the user experience of frontline clinicians. The frustration caused by poorly designed EHRs could affect patient safety, read more from Craig Joseph, MD, in this edition of the Checkup.