Emergency Department Decision Support using Clinical Pseudo-notes
This addresses the challenge of effectively using EHR data for clinical decision support in emergency departments, though it appears incremental as an extension of existing text-based representation methods.
The authors tackled the problem of representing multimodal EHR tabular data for emergency department decision support by converting it into text using pseudo-notes, which outperformed traditional machine learning, EHR-specific foundation models, and general LLMs across multiple hospital systems.
In this work, we introduce the Multiple Embedding Model for EHR (MEME), an approach that serializes multimodal EHR tabular data into text using pseudo-notes, mimicking clinical text generation. This conversion not only preserves better representations of categorical data and learns contexts but also enables the effective employment of pretrained foundation models for rich feature representation. To address potential issues with context length, our framework encodes embeddings for each EHR modality separately. We demonstrate the effectiveness of MEME by applying it to several decision support tasks within the Emergency Department across multiple hospital systems. Our findings indicate that MEME outperforms traditional machine learning, EHR-specific foundation models, and general LLMs, highlighting its potential as a general and extendible EHR representation strategy.