Search AMIA.org

This study aimed to leverage a large language model (LLM) to improve the efficiency and thoroughness of medical record documentation. This study focused on aiding clinical staff in creating structured summaries with the help of an LLM and assessing the quality of these artificial intelligence (AI)-proposed records in comparison to those produced by doctors.This strategy involved assembling a team of specialists, including data engineers, physicians, and medical information experts, to [...]

Large language models (LLMs) face challenges in inductive thematic analysis, a task requiring deep interpretive, domain-specific expertise. We evaluated the feasibility of using LLMs to replicate expert-driven thematic analysis of social media data.

To evaluate the efficacy of digital twins developed using a large language model (LLaMA-3), fine-tuned with Low-Rank Adapters (LoRA) on intensive care units (ICU) physician notes, and to determine whether specialty-specific training enhances treatment recommendation accuracy compared to other ICU specialties or zero-shot baselines.

Time spent in the electronic health record (EHR) is an important measure of clinical activity. Vendor-derived EHR use metrics may not correspond to actual EHR experience. Raw EHR audit logs enable customized EHR use metrics, but translating discrete timestamps to time intervals is challenging. There are insufficient data available to quantify inactivity between audit log timestamps.This study aimed to develop and validate a computer vision-based model that (1) classifies EHR [...]

To develop a natural language processing (NLP) pipeline for unstructured electronic health record (EHR) data to identify symptoms and functional impacts associated with Long COVID in children.

Unstructured data, such as procedure notes, contain valuable medical information that is frequently underutilized due to the labor-intensive nature of data extraction. This study aims to develop a generative artificial intelligence (GenAI) pipeline using an open-source Large Language Model (LLM) with built-in guardrails and a retry mechanism to extract data from unstructured right heart catheterization (RHC) notes while minimizing errors, including hallucinations.

Assess time impact and provider perception of AI-generated encounter summaries.

The timely administration of postoperative antibiotics is crucial for preventing surgical site infections. Despite surgical ordering workflows designed to facilitate care across settings, delays in antibiotic administration posttransfer to the pediatric intensive care unit (PICU) were identified. We aimed to develop a clinical decision support (CDS) system to enhance timely order activation in a large pediatric health system. We hypothesized that the time to release signed and held orders by [...]

Drug interaction checking software is ubiquitous in clinical decision support systems (CDSS-DI) but patient relevance and accuracy are variable and the impact on patient outcomes is unproven. We compared the effectiveness of CDSS-DI with similar care without CDSS-DI.

Clinical decision support systems (CDSS) have been suggested to be helpful in detecting and preventing drug-related problems such as adverse drug events (ADEs). However, patient participation systems monitoring self-reported data, such as symptoms, are still sparsely described in the literature.This study aimed to investigate if the use of a patient participating CDSS (PCDSS) can facilitate early detection of ADEs, thereby contributing to safer drug treatment in older adults.A 1-year prospective [...]