Predicting hospital readmission for lupus patients: An RNN-LSTM-based deep-learning methodology

Bhargava K. Reddy, Dursun Delen

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

Hospital readmission is one of the critical metrics used for measuring the performance of hospitals. The HITECH Act imposes penalties when patients are readmitted to hospitals if they are diagnosed with one of the six conditions mentioned in the Act. However, patients diagnosed with lupus are the sixth highest in terms of rehospitalization. The heterogeneity in the disease and patient characteristics makes it very hard to predict rehospitalization. This research utilizes deep learning methods to predict rehospitalization within 30 days by extracting the temporal relationships in the longitudinal EHR clinical data. Prediction results from deep learning methods such as LSTM are evaluated and compared with traditional classification methods such as penalized logistic regression and artificial neural networks. The simple recurrent neural network method and its variant, gated recurrent unit network, are also developed and validated to compare their performance against the proposed LSTM model. The results indicated that the deep learning method RNN-LSTM has a significantly better performance (with an AUC of.70) compared to traditional classification methods such as ANN (with an AUC of 0.66) and penalized logistic regression (with an AUC of 0.63). The rationale for the better performance of the deep learning method may be due to its ability to leverage the temporal relationships of the disease state in patients over time and to capture the progression of the disease—relevant clinical information from patients’ prior visits is carried forward in the memory, which may have enabled the higher predictability for the deep learning methods.

Original languageEnglish
Pages (from-to)199-209
Number of pages11
JournalComputers in Biology and Medicine
Volume101
DOIs
StatePublished - 1 Oct 2018
Externally publishedYes

Fingerprint

Patient Readmission
Learning
Logistics
Area Under Curve
Recurrent neural networks
American Recovery and Reinvestment Act
Logistic Models
Aptitude
Deep learning
Neural networks
Data storage equipment

Keywords

  • Deep learning
  • LSTM
  • Lupus
  • Machine learning
  • Predictive analytics
  • Readmission

Cite this

@article{cd156a9c60cf4738844ddf4661a9c47e,
title = "Predicting hospital readmission for lupus patients: An RNN-LSTM-based deep-learning methodology",
abstract = "Hospital readmission is one of the critical metrics used for measuring the performance of hospitals. The HITECH Act imposes penalties when patients are readmitted to hospitals if they are diagnosed with one of the six conditions mentioned in the Act. However, patients diagnosed with lupus are the sixth highest in terms of rehospitalization. The heterogeneity in the disease and patient characteristics makes it very hard to predict rehospitalization. This research utilizes deep learning methods to predict rehospitalization within 30 days by extracting the temporal relationships in the longitudinal EHR clinical data. Prediction results from deep learning methods such as LSTM are evaluated and compared with traditional classification methods such as penalized logistic regression and artificial neural networks. The simple recurrent neural network method and its variant, gated recurrent unit network, are also developed and validated to compare their performance against the proposed LSTM model. The results indicated that the deep learning method RNN-LSTM has a significantly better performance (with an AUC of.70) compared to traditional classification methods such as ANN (with an AUC of 0.66) and penalized logistic regression (with an AUC of 0.63). The rationale for the better performance of the deep learning method may be due to its ability to leverage the temporal relationships of the disease state in patients over time and to capture the progression of the disease—relevant clinical information from patients’ prior visits is carried forward in the memory, which may have enabled the higher predictability for the deep learning methods.",
keywords = "Deep learning, LSTM, Lupus, Machine learning, Predictive analytics, Readmission",
author = "Reddy, {Bhargava K.} and Dursun Delen",
year = "2018",
month = "10",
day = "1",
doi = "10.1016/j.compbiomed.2018.08.029",
language = "English",
volume = "101",
pages = "199--209",
journal = "Computers in Biology and Medicine",
issn = "0010-4825",
publisher = "Elsevier Ltd",

}

Predicting hospital readmission for lupus patients : An RNN-LSTM-based deep-learning methodology. / Reddy, Bhargava K.; Delen, Dursun.

In: Computers in Biology and Medicine, Vol. 101, 01.10.2018, p. 199-209.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Predicting hospital readmission for lupus patients

T2 - An RNN-LSTM-based deep-learning methodology

AU - Reddy, Bhargava K.

AU - Delen, Dursun

PY - 2018/10/1

Y1 - 2018/10/1

N2 - Hospital readmission is one of the critical metrics used for measuring the performance of hospitals. The HITECH Act imposes penalties when patients are readmitted to hospitals if they are diagnosed with one of the six conditions mentioned in the Act. However, patients diagnosed with lupus are the sixth highest in terms of rehospitalization. The heterogeneity in the disease and patient characteristics makes it very hard to predict rehospitalization. This research utilizes deep learning methods to predict rehospitalization within 30 days by extracting the temporal relationships in the longitudinal EHR clinical data. Prediction results from deep learning methods such as LSTM are evaluated and compared with traditional classification methods such as penalized logistic regression and artificial neural networks. The simple recurrent neural network method and its variant, gated recurrent unit network, are also developed and validated to compare their performance against the proposed LSTM model. The results indicated that the deep learning method RNN-LSTM has a significantly better performance (with an AUC of.70) compared to traditional classification methods such as ANN (with an AUC of 0.66) and penalized logistic regression (with an AUC of 0.63). The rationale for the better performance of the deep learning method may be due to its ability to leverage the temporal relationships of the disease state in patients over time and to capture the progression of the disease—relevant clinical information from patients’ prior visits is carried forward in the memory, which may have enabled the higher predictability for the deep learning methods.

AB - Hospital readmission is one of the critical metrics used for measuring the performance of hospitals. The HITECH Act imposes penalties when patients are readmitted to hospitals if they are diagnosed with one of the six conditions mentioned in the Act. However, patients diagnosed with lupus are the sixth highest in terms of rehospitalization. The heterogeneity in the disease and patient characteristics makes it very hard to predict rehospitalization. This research utilizes deep learning methods to predict rehospitalization within 30 days by extracting the temporal relationships in the longitudinal EHR clinical data. Prediction results from deep learning methods such as LSTM are evaluated and compared with traditional classification methods such as penalized logistic regression and artificial neural networks. The simple recurrent neural network method and its variant, gated recurrent unit network, are also developed and validated to compare their performance against the proposed LSTM model. The results indicated that the deep learning method RNN-LSTM has a significantly better performance (with an AUC of.70) compared to traditional classification methods such as ANN (with an AUC of 0.66) and penalized logistic regression (with an AUC of 0.63). The rationale for the better performance of the deep learning method may be due to its ability to leverage the temporal relationships of the disease state in patients over time and to capture the progression of the disease—relevant clinical information from patients’ prior visits is carried forward in the memory, which may have enabled the higher predictability for the deep learning methods.

KW - Deep learning

KW - LSTM

KW - Lupus

KW - Machine learning

KW - Predictive analytics

KW - Readmission

UR - http://www.scopus.com/inward/record.url?scp=85052738184&partnerID=8YFLogxK

U2 - 10.1016/j.compbiomed.2018.08.029

DO - 10.1016/j.compbiomed.2018.08.029

M3 - Article

C2 - 30195164

AN - SCOPUS:85052738184

VL - 101

SP - 199

EP - 209

JO - Computers in Biology and Medicine

JF - Computers in Biology and Medicine

SN - 0010-4825

ER -