Improving Pediatric Emergency Department Triage with Modality Dropout in Late Fusion Multimodal EHR Models

TL;DR

This study introduces a multimodal model with modality dropout that improves pediatric emergency triage predictions and generalizes better across demographics, especially when trained on adult data.

Contribution

It presents a late-fusion architecture with modality dropout that enhances cross-demographic generalization in clinical AI models.

Findings

Modality dropout improves zero-shot pediatric triage performance.

The model outperforms single-modality baselines.

30-40% modality dropout yields significant performance gains.

Abstract

Emergency department triage relies heavily on both quantitative vital signs and qualitative clinical notes, yet multimodal machine learning models predicting triage acuity often suffer from modality collapse by over-relying on structured tabular data. This limitation severely hinders demographic generalizability, particularly for pediatric patients where developmental variations in vital signs make unstructured clinical narratives uniquely crucial. To address this gap, we propose a late-fusion multimodal architecture that processes tabular vitals via XGBoost and unstructured clinical text via Bio_ClinicalBERT, combined through a Logistic Regression meta-classifier to predict the 5-level Emergency Severity Index. To explicitly target the external validity problem, we train our model exclusively on adult encounters from the MIMIC-IV and NHAMCS datasets and evaluate its zero-shot generalization on a traditionally overlooked pediatric cohort. Furthermore, we employ symmetric modality dropout during training to prevent the ensemble from overfitting to adult-specific clinical correlations. Our results demonstrate that the multimodal framework significantly outperforms single-modality baselines. Most notably, applying a 30-40% symmetric modality dropout rate yielded steep performance improvements in the unseen pediatric cohort, elevating the Quadratic Weighted Kappa to 0.351. These findings highlight modality dropout as a critical regularization technique for mitigating modality collapse and enhancing cross-demographic generalization in clinical AI.