Enhancing Antibiotic Stewardship using a Natural Language Approach for Better Feature Representation

TL;DR

The paper tackles the antibiotic resistance crisis by integrating electronic health records (EHRs) into decision-support workflows through a novel text-based representation of EHR data, termed pseudo-notes. By converting tabular EHR data into serialized narrative form and applying pretrained foundation models, the authors demonstrate improved antibiotic susceptibility predictions and enhanced interpretability. They compare multiple representations, with pseudo-notes and foundation-model inputs outperforming raw tabular and other text-based approaches, and reveal meaningful patient clusters via BERTopic. The work emphasizes interoperability and potential zero-shot applications, suggesting a practical, data-driven approach to strengthen antibiotic stewardship in emergency care settings. Overall, pseudo-notes provide a flexible, interpretable interface that can adapt to evolving foundation-model backbones while aiding clinicians in selecting appropriate antibiotics.

Abstract

The rapid emergence of antibiotic-resistant bacteria is recognized as a global healthcare crisis, undermining the efficacy of life-saving antibiotics. This crisis is driven by the improper and overuse of antibiotics, which escalates bacterial resistance. In response, this study explores the use of clinical decision support systems, enhanced through the integration of electronic health records (EHRs), to improve antibiotic stewardship. However, EHR systems present numerous data-level challenges, complicating the effective synthesis and utilization of data. In this work, we transform EHR data into a serialized textual representation and employ pretrained foundation models to demonstrate how this enhanced feature representation can aid in antibiotic susceptibility predictions. Our results suggest that this text representation, combined with foundation models, provides a valuable tool to increase interpretability and support antibiotic stewardship efforts.

Figures (3)

• Figure 1: Area under the Receiver Operating Characteristic (AUROC) curves for each antibiotic classification shows that, despite the nuances, BioMegatron performs the best.
• Figure 2: Area under the Precision Recall Curve (AUPRC) curves for each antibiotic classification shows that, despite the nuances, BioMegatron performs the best.
• Figure 3: Identifying clusters from our patient embeddings is indicated by the squares forming along the diagonal of our similarity matrix.