Analysis and prediction of unplanned intensive care unit readmission using recurrent neural networks with long shortterm memory

Yu Wei Lin, Yuqian Zhou, Faraz Faghri, Michael J. Shaw, Roy H. Campbell

Research output: Contribution to journalArticle

Abstract

Background Unplanned readmission of a hospitalized patient is an indicator of patients' exposure to risk and an avoidable waste of medical resources. In addition to hospital readmission, intensive care unit (ICU) readmission brings further financial risk, along with morbidity and mortality risks. Identification of high-risk patients who are likely to be readmitted can provide significant benefits for both patients and medical providers. The emergence of machine learning solutions to detect hidden patterns in complex, multi-dimensional datasets provides unparalleled opportunities for developing an efficient discharge decision-making support system for physicians and ICU specialists. Methods and findings We used supervised machine learning approaches for ICU readmission prediction. We used machine learning methods on comprehensive, longitudinal clinical data from the MIMIC-III to predict the ICU readmission of patients within 30 days of their discharge. We incorporate multiple types of features including chart events, demographic, and ICD-9 embeddings. We have utilized recent machine learning techniques such as Recurrent Neural Networks (RNN) with Long Short-Term Memory (LSTM), by this we have been able to incorporate the multivariate features of EHRs and capture sudden fluctuations in chart event features (e.g. glucose and heart rate). We show that our LSTM-based solution can better capture high volatility and unstable status in ICU patients, an important factor in ICU readmission. Our machine learning models identify ICU readmissions at a higher sensitivity rate of 0.742 (95% CI, 0.718-0.766) and an improved Area Under the Curve of 0.791 (95% CI, 0.782-0.800) compared with traditional methods. We perform in-depth deep learning performance analysis, as well as the analysis of each feature contribution to the predictive model. Conclusion Our manuscript highlights the ability of machine learning models to improve our ICU decision- making accuracy and is a real-world example of precision medicine in hospitals. These data-driven solutions hold the potential for substantial clinical impact by augmenting clinical decision-making for physicians and ICU specialists. We anticipate that machine learning models will improve patient counseling, hospital administration, allocation of healthcare resources and ultimately individualized clinical care.

Original languageEnglish (US)
Article numbere0218942
JournalPloS one
Volume14
Issue number7
DOIs
StatePublished - Jan 1 2019

Fingerprint

Intensive care units
Recurrent neural networks
artificial intelligence
Short-Term Memory
neural networks
Intensive Care Units
Learning systems
Data storage equipment
prediction
decision making
Patient Readmission
physicians
Long-Term Memory
Decision making
Decision Making
Medical Waste
support systems
counseling
Hospital Administration
Physicians

ASJC Scopus subject areas

  • Biochemistry, Genetics and Molecular Biology(all)
  • Agricultural and Biological Sciences(all)
  • General

Cite this

Analysis and prediction of unplanned intensive care unit readmission using recurrent neural networks with long shortterm memory. / Lin, Yu Wei; Zhou, Yuqian; Faghri, Faraz; Shaw, Michael J.; Campbell, Roy H.

In: PloS one, Vol. 14, No. 7, e0218942, 01.01.2019.

Research output: Contribution to journalArticle

@article{828b9d3b945541ec94b1c89913bea42a,
title = "Analysis and prediction of unplanned intensive care unit readmission using recurrent neural networks with long shortterm memory",
abstract = "Background Unplanned readmission of a hospitalized patient is an indicator of patients' exposure to risk and an avoidable waste of medical resources. In addition to hospital readmission, intensive care unit (ICU) readmission brings further financial risk, along with morbidity and mortality risks. Identification of high-risk patients who are likely to be readmitted can provide significant benefits for both patients and medical providers. The emergence of machine learning solutions to detect hidden patterns in complex, multi-dimensional datasets provides unparalleled opportunities for developing an efficient discharge decision-making support system for physicians and ICU specialists. Methods and findings We used supervised machine learning approaches for ICU readmission prediction. We used machine learning methods on comprehensive, longitudinal clinical data from the MIMIC-III to predict the ICU readmission of patients within 30 days of their discharge. We incorporate multiple types of features including chart events, demographic, and ICD-9 embeddings. We have utilized recent machine learning techniques such as Recurrent Neural Networks (RNN) with Long Short-Term Memory (LSTM), by this we have been able to incorporate the multivariate features of EHRs and capture sudden fluctuations in chart event features (e.g. glucose and heart rate). We show that our LSTM-based solution can better capture high volatility and unstable status in ICU patients, an important factor in ICU readmission. Our machine learning models identify ICU readmissions at a higher sensitivity rate of 0.742 (95{\%} CI, 0.718-0.766) and an improved Area Under the Curve of 0.791 (95{\%} CI, 0.782-0.800) compared with traditional methods. We perform in-depth deep learning performance analysis, as well as the analysis of each feature contribution to the predictive model. Conclusion Our manuscript highlights the ability of machine learning models to improve our ICU decision- making accuracy and is a real-world example of precision medicine in hospitals. These data-driven solutions hold the potential for substantial clinical impact by augmenting clinical decision-making for physicians and ICU specialists. We anticipate that machine learning models will improve patient counseling, hospital administration, allocation of healthcare resources and ultimately individualized clinical care.",
author = "Lin, {Yu Wei} and Yuqian Zhou and Faraz Faghri and Shaw, {Michael J.} and Campbell, {Roy H.}",
year = "2019",
month = "1",
day = "1",
doi = "10.1371/journal.pone.0218942",
language = "English (US)",
volume = "14",
journal = "PLoS One",
issn = "1932-6203",
publisher = "Public Library of Science",
number = "7",

}

TY - JOUR

T1 - Analysis and prediction of unplanned intensive care unit readmission using recurrent neural networks with long shortterm memory

AU - Lin, Yu Wei

AU - Zhou, Yuqian

AU - Faghri, Faraz

AU - Shaw, Michael J.

AU - Campbell, Roy H.

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Background Unplanned readmission of a hospitalized patient is an indicator of patients' exposure to risk and an avoidable waste of medical resources. In addition to hospital readmission, intensive care unit (ICU) readmission brings further financial risk, along with morbidity and mortality risks. Identification of high-risk patients who are likely to be readmitted can provide significant benefits for both patients and medical providers. The emergence of machine learning solutions to detect hidden patterns in complex, multi-dimensional datasets provides unparalleled opportunities for developing an efficient discharge decision-making support system for physicians and ICU specialists. Methods and findings We used supervised machine learning approaches for ICU readmission prediction. We used machine learning methods on comprehensive, longitudinal clinical data from the MIMIC-III to predict the ICU readmission of patients within 30 days of their discharge. We incorporate multiple types of features including chart events, demographic, and ICD-9 embeddings. We have utilized recent machine learning techniques such as Recurrent Neural Networks (RNN) with Long Short-Term Memory (LSTM), by this we have been able to incorporate the multivariate features of EHRs and capture sudden fluctuations in chart event features (e.g. glucose and heart rate). We show that our LSTM-based solution can better capture high volatility and unstable status in ICU patients, an important factor in ICU readmission. Our machine learning models identify ICU readmissions at a higher sensitivity rate of 0.742 (95% CI, 0.718-0.766) and an improved Area Under the Curve of 0.791 (95% CI, 0.782-0.800) compared with traditional methods. We perform in-depth deep learning performance analysis, as well as the analysis of each feature contribution to the predictive model. Conclusion Our manuscript highlights the ability of machine learning models to improve our ICU decision- making accuracy and is a real-world example of precision medicine in hospitals. These data-driven solutions hold the potential for substantial clinical impact by augmenting clinical decision-making for physicians and ICU specialists. We anticipate that machine learning models will improve patient counseling, hospital administration, allocation of healthcare resources and ultimately individualized clinical care.

AB - Background Unplanned readmission of a hospitalized patient is an indicator of patients' exposure to risk and an avoidable waste of medical resources. In addition to hospital readmission, intensive care unit (ICU) readmission brings further financial risk, along with morbidity and mortality risks. Identification of high-risk patients who are likely to be readmitted can provide significant benefits for both patients and medical providers. The emergence of machine learning solutions to detect hidden patterns in complex, multi-dimensional datasets provides unparalleled opportunities for developing an efficient discharge decision-making support system for physicians and ICU specialists. Methods and findings We used supervised machine learning approaches for ICU readmission prediction. We used machine learning methods on comprehensive, longitudinal clinical data from the MIMIC-III to predict the ICU readmission of patients within 30 days of their discharge. We incorporate multiple types of features including chart events, demographic, and ICD-9 embeddings. We have utilized recent machine learning techniques such as Recurrent Neural Networks (RNN) with Long Short-Term Memory (LSTM), by this we have been able to incorporate the multivariate features of EHRs and capture sudden fluctuations in chart event features (e.g. glucose and heart rate). We show that our LSTM-based solution can better capture high volatility and unstable status in ICU patients, an important factor in ICU readmission. Our machine learning models identify ICU readmissions at a higher sensitivity rate of 0.742 (95% CI, 0.718-0.766) and an improved Area Under the Curve of 0.791 (95% CI, 0.782-0.800) compared with traditional methods. We perform in-depth deep learning performance analysis, as well as the analysis of each feature contribution to the predictive model. Conclusion Our manuscript highlights the ability of machine learning models to improve our ICU decision- making accuracy and is a real-world example of precision medicine in hospitals. These data-driven solutions hold the potential for substantial clinical impact by augmenting clinical decision-making for physicians and ICU specialists. We anticipate that machine learning models will improve patient counseling, hospital administration, allocation of healthcare resources and ultimately individualized clinical care.

UR - http://www.scopus.com/inward/record.url?scp=85069291399&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85069291399&partnerID=8YFLogxK

U2 - 10.1371/journal.pone.0218942

DO - 10.1371/journal.pone.0218942

M3 - Article

C2 - 31283759

AN - SCOPUS:85069291399

VL - 14

JO - PLoS One

JF - PLoS One

SN - 1932-6203

IS - 7

M1 - e0218942

ER -