Analysis and prediction of unplanned intensive care unit readmission using recurrent neural networks with long short-term memory.

Analysis and prediction of unplanned intensive care unit readmission using recurrent neural networks with long short-term memory.

Publication date: Jun 27, 2018

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.

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.

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.

Open Access PDF

Lin, Y.W., Zhou, Y., Faghri, F., Shaw, M.J., and Campbell, R.H. Analysis and prediction of unplanned intensive care unit readmission using recurrent neural networks with long short-term memory. 04962. 2018 PLoS One (14):7.

Concepts Keywords
Counseling Recurrent neural networks
Deep Learning Counseling
Demographic Healthcare quality
Financial Risk Hospital readmission
Glucose Articles
Healthcare Intensive care unit
Heart Rate Intensive care medicine
Hospital Predictive modelling
Intensive Care Unit Long short-term memory
Manuscript
Morbidity
Mortality
Multivariate
Predictive Model
Recurrent Neural Networks
Short Term Memory
Supervised Learning
Volatility

Semantics

Type Source Name
disease MESH ESRD
disease MESH renal
disease DOID hypoglycemia
disease MESH hypoglycemia
disease DOID congestive heart failure
gene UNIPROT GAN
disease MESH risk factors
disease DOID pneumonia
disease MESH pneumonia
disease DOID renal failure
disease MESH renal failure
disease MESH comorbidities
gene UNIPROT SLC26A5
gene UNIPROT THOP1
gene UNIPROT NR1H4
gene UNIPROT ADRB2
gene UNIPROT BFAR
gene UNIPROT PPRC1
drug DRUGBANK Saquinavir
gene UNIPROT NCOA5
gene UNIPROT ASF1A
gene UNIPROT NT5E
gene UNIPROT SET
gene UNIPROT MEA1
drug DRUGBANK Cysteamine
gene UNIPROT TTLL5
gene UNIPROT ALPK3
gene UNIPROT MAK
gene UNIPROT REST
gene UNIPROT SOAT1
gene UNIPROT CHL1
gene UNIPROT FBLIM1
drug DRUGBANK Oxygen
disease MESH coma
gene UNIPROT DEPP1
gene UNIPROT GOPC
gene UNIPROT CRAT
gene UNIPROT CAT
gene UNIPROT GLYAT
gene UNIPROT AMACR
disease MESH chronic diseases
gene UNIPROT COPE
gene UNIPROT KCNK3
drug DRUGBANK Aspartame
gene UNIPROT PTPN5
gene UNIPROT CLU
disease MESH heart failure
gene UNIPROT PRAC1
gene UNIPROT CD2AP
drug DRUGBANK Trihexyphenidyl
drug DRUGBANK Coenzyme M
gene UNIPROT GTF2IRD1
gene UNIPROT RALA
gene UNIPROT ERBB2
gene UNIPROT NEU1
gene UNIPROT NEURL1
pathway BSID Aging
disease MESH Aging
gene UNIPROT CD2BP2
gene UNIPROT IMPACT
drug DRUGBANK Tropicamide
drug DRUGBANK D-glucose
drug DRUGBANK Dextrose unspecified form
gene UNIPROT GNPTAB
disease MESH multiple
disease MESH multi

Similar

Original Article

Leave a Comment

Your email address will not be published. Required fields are marked *