How a 50-year-old medical template is training the next generation of AI
In short: The SOAP note has long been the cognitive backbone of clinical reasoning. Now, researchers are using this structured format to help large language models better identify hidden drug reactions and to refine the skills of medical residents through innovative peer-led training.
Developed nearly 50 years ago by Dr. To assist doctors in organising care for patients with various ailments, Lawrence Weed developed the SOAP note—an acronym representing Subjective, Objective, Assessment, and Plan—as part of the problem-oriented medical record. Today, it serves as a standardized communication tool used by everyone from physicians and veterinarians to emergency medical technicians to document patient progress and reduce confusion across healthcare disciplines. The format is used for both initial visits and monitoring follow-up care. It is a common feature in electronic medical records, providing a template for physicians to guide their documentation.
The structure of these notes is increasingly vital for the advancement of medical AI. Research involving Manitoba Primary Care Research Network electronic records demonstrated that NLP models achieved higher accuracy in detecting adverse drug events when clinical documentation was restructured into a SOAP format. Specifically, a model known as GritLM-SOAP achieved a mean F1-score of 72.53%, outperforming models trained on original, unstructured clinical text. In a study of 1,085 annotated notes, researchers found that BioBERT-SOAP weighted achieved the highest mean recall at 76.34%. This research specifically focused on patients aged 55 years and older.
Beyond machine learning, the SOAP format is central to human clinical training. In a prospective study conducted between January 2024 and June 2025, researchers tested an integrated training model for obstetrics and gynecology residents. By using 'Students as Standardized Patients' and structured peer assessment, the study found that residents showed significantly greater satisfaction in areas like clinical reasoning development and improved objective performance in SOAP note writing. The study included 51 residents in the experimental group and 54 in the control group. Results indicated that the new model specifically improved higher-order cognitive skills like data interpretation.
Opens the Intellect+ feed with this card first. Swipe up for the next one.
Sources
- Summary by Intellect · source: Using Natural Language Processing to Identify Adverse Drug Events Characterized by Medication Replacement in Primary Care Electronic Medical Records: Algorithm and Validation Study.
- Also: SOAP note
- Also: Integrated model of Student Standardized Patients and peer assessment enhances SOAP note proficiency in obstetrics and gynecology residency.
To cite this card, link to https://intellectplus.co.uk/app/health-soap-note.
How we source, check and correct cards: Editorial standards.
General information, not professional advice. Spotted an error? Flag the card or see corrections and complaints.
Intellect Plus is a visual educational discovery platform for exploring fascinating facts and ideas. Brain Food, Not Brain Rot. What is Intellect Plus?
Related
More in First aid · More in Health · Latest · A–Z · All topics
