What is the clinical evidence from this study?

Study design: Observational. Population: Heart Failure (n=21031). Intervention: Word2vec embeddings on CUIs with XGBoost vs. One-hot encoding (baseline). Primary outcome: AUROC for 30-day readmission.

October 6, 2025JMIR Medical Informatics1 citationsOpen Access

Predicting 30-Days Hospital Readmission for Patients with Heart Failure Using Electronic Health Record Embeddings: Comparative Evaluation

PSPrabin ShakyaSeoul National University AKAyush KhanejaMassachusetts General Hospital KWKavishwar B. WagholikarPolitical Research Associates

Key Result

Word2vec embeddings trained on electronic health record data improved the prediction of 30-day heart failure readmissions, achieving an AUROC of 0.65 compared to 0.54 for baseline one-hot encoding.

Study Design

Type

Observational (n=21,031)

Multicenter

Structured PICO

Do embedding approaches improve the prediction of 30-day unplanned readmissions in patients with heart failure compared to one-hot encoding?

Population

21,031 patients with heart failure identified using least restrictive phenotyping methods from the MIMIC-IV dataset

Intervention

Embedding approaches including word2vec on terminology codes and concept unique identifier (CUIs) and BERT on descriptive text of concept, using logistic regression, XGBoost, and ANN models

Comparator

Baseline one-hot encoding

Outcome

Prediction of unplanned 30-day readmissions (evaluated by AUROC and F1-scores)

EHR-trained word2vec embeddings significantly improve the discrimination of 30-day heart failure readmissions compared to traditional one-hot encoding or pre-trained BERT models.

Main Result

Absolute Event Rate: 0.65% vs 0.54%

Limitations

Patient encounters were modeled as static collection of medical codes without considering temporal patterns.
The study did not evaluate the portability of the models as the dataset is limited to a single institution.
Evaluation focused on traditional machine learning approaches (logistic regression, XGBoost, and ANN), which may not fully exploit the representational capacity of embedding methods.

Abstract

Background: Heart failure (HF) is a public health concern with a wider impact on quality of life and cost of care. One of the major challenges in HF is the higher rate of unplanned readmissions and suboptimal performance of models to predict the readmissions. Hence, in this study, we implemented embeddings-based approaches to generate features for improving model performance. Objective: The objective of this study was to evaluate and compare the effectiveness of different feature embedding approaches for improving the prediction of unplanned readmissions in patients with heart failure. Methods: We compared three embedding approaches including word2vec on terminology codes and concept unique identifier (CUIs) and BERT on descriptive text of concept with baseline (one hot-encoding). We compared area under the receiver operating characteristic (AUROC) and F1-scores for the logistic regression, eXtream gradient-boosting (XGBoost) and artificial neural network (ANN) models using these embedding approaches. The model was tested on the heart failure cohort (N=21,031) identified using least restrictive phenotyping methods from MIMIC-IV dataset. Results: We found that the embedding approaches significantly improved the performance of the prediction models. The XGBoost performed better for all approaches. The word2vec embeddings (0.65) trained on the dataset outperformed embeddings from pre-trained BERT model (0.59) using descriptive text. Conclusions: Embedding methods, particularly word2vec trained on electronic health record data, can better discriminate HF readmission cases compared to both one-hot encoding and pre-trained BERT embeddings on concept descriptions making it a viable approach of automation feature selection. The observed AUROC improvement (0.65 vs 0.54) may support more effective risk stratification and targeted clinical interventions.

问 AI

Bookmark

View Full Paper