Improving Patients’ Length of Stay Prediction Using Clinical and Demographics Features Enrichment

Hamzah Osop; Basem Suleiman; Muhammad Johan Alibasa; Drew Wrigley; Alexandra Helsham; Anne Asmaro

doi:10.1007/978-3-031-36021-3_9

Improving Patients’ Length of Stay Prediction Using Clinical and Demographics Features Enrichment

Hamzah Osop, Basem Suleiman, Muhammad Johan Alibasa, Drew Wrigley, Alexandra Helsham, Anne Asmaro

Research output: Chapter in Book/Report/Conference proceeding › Conference Paper › Research › peer-review

Abstract

Predicting patients’ length of stay (LOS) is crucial for efficient scheduling of treatment and strategic future planning, in turn reduce hospitalisation costs. However, this is a complex problem requiring careful selection of optimal set of essential factors that significantly impact the accuracy and performance of LOS prediction. Using an inpatient dataset of 285k of records from 14 general care hospitals in Vermont, USA from 2013–2017, we presented our novel approach to incorporate features to improve the accuracy of LOS prediction. Our empirical experiment and analysis showed considerable improvement in LOS prediction with an XGBoost model RMSE score of 6.98 and R2 score of 38.24%. Based on several experiments, we provided empirical analysis of the importance of different feature sets and its impact on predicting patients’ LOS.

Original language	English
Title of host publication	Computational Science – ICCS 2023, 23rd International Conference Prague, Czech Republic, July 3–5, 2023 Proceedings, Part II
Editors	Jiří Mikyška, Clélia de Mulatier, Valeria V. Krzhizhanovskaya, Peter M.A. Sloot, Maciej Paszynski, Jack J. Dongarra
Place of Publication	Cham Switzerland
Publisher	Springer
Pages	120-128
Number of pages	9
ISBN (Electronic)	9783031360213
ISBN (Print)	9783031360206
DOIs	https://doi.org/10.1007/978-3-031-36021-3_9
Publication status	Published - 2023
Externally published	Yes
Event	International Conference on Computational Science 2023 - Prague, Czechia Duration: 3 Jul 2023 → 5 Jul 2023 Conference number: 23rd https://link.springer.com/book/10.1007/978-3-031-36021-3 (Proceedings) https://www.iccs-meeting.org/iccs2023/ (Website)

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer
Volume	14074
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	International Conference on Computational Science 2023
Abbreviated title	ICCS 2023
Country/Territory	Czechia
City	Prague
Period	3/07/23 → 5/07/23
Internet address	https://link.springer.com/book/10.1007/978-3-031-36021-3 (Proceedings) https://www.iccs-meeting.org/iccs2023/ (Website)

Keywords

electronic health records
length of stay
machine learning

Access to Document

10.1007/978-3-031-36021-3_9

Cite this

Osop, H., Suleiman, B., Alibasa, M. J., Wrigley, D., Helsham, A., & Asmaro, A. (2023). Improving Patients’ Length of Stay Prediction Using Clinical and Demographics Features Enrichment. In J. Mikyška, C. de Mulatier, V. V. Krzhizhanovskaya, P. M. A. Sloot, M. Paszynski, & J. J. Dongarra (Eds.), Computational Science – ICCS 2023, 23rd International Conference Prague, Czech Republic, July 3–5, 2023 Proceedings, Part II (pp. 120-128). (Lecture Notes in Computer Science; Vol. 14074). Springer. https://doi.org/10.1007/978-3-031-36021-3_9

Osop, Hamzah ; Suleiman, Basem ; Alibasa, Muhammad Johan et al. / Improving Patients’ Length of Stay Prediction Using Clinical and Demographics Features Enrichment. Computational Science – ICCS 2023, 23rd International Conference Prague, Czech Republic, July 3–5, 2023 Proceedings, Part II. editor / Jiří Mikyška ; Clélia de Mulatier ; Valeria V. Krzhizhanovskaya ; Peter M.A. Sloot ; Maciej Paszynski ; Jack J. Dongarra. Cham Switzerland : Springer, 2023. pp. 120-128 (Lecture Notes in Computer Science).

@inproceedings{a9c576170ec248e1b5fddd9a39e68437,

title = "Improving Patients{\textquoteright} Length of Stay Prediction Using Clinical and Demographics Features Enrichment",

abstract = "Predicting patients{\textquoteright} length of stay (LOS) is crucial for efficient scheduling of treatment and strategic future planning, in turn reduce hospitalisation costs. However, this is a complex problem requiring careful selection of optimal set of essential factors that significantly impact the accuracy and performance of LOS prediction. Using an inpatient dataset of 285k of records from 14 general care hospitals in Vermont, USA from 2013–2017, we presented our novel approach to incorporate features to improve the accuracy of LOS prediction. Our empirical experiment and analysis showed considerable improvement in LOS prediction with an XGBoost model RMSE score of 6.98 and R2 score of 38.24\%. Based on several experiments, we provided empirical analysis of the importance of different feature sets and its impact on predicting patients{\textquoteright} LOS.",

keywords = "electronic health records, length of stay, machine learning",

author = "Hamzah Osop and Basem Suleiman and Alibasa, \{Muhammad Johan\} and Drew Wrigley and Alexandra Helsham and Anne Asmaro",

note = "Publisher Copyright: {\textcopyright} 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.; International Conference on Computational Science 2023, ICCS 2023 ; Conference date: 03-07-2023 Through 05-07-2023",

year = "2023",

doi = "10.1007/978-3-031-36021-3\_9",

language = "English",

isbn = "9783031360206",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "120--128",

editor = "Ji{\v r}{\'i} Miky{\v s}ka and \{de Mulatier\}, Cl{\'e}lia and Krzhizhanovskaya, \{Valeria V.\} and Sloot, \{Peter M.A.\} and Maciej Paszynski and Dongarra, \{Jack J.\}",

booktitle = "Computational Science – ICCS 2023, 23rd International Conference Prague, Czech Republic, July 3–5, 2023 Proceedings, Part II",

address = "Switzerland",

url = "https://link.springer.com/book/10.1007/978-3-031-36021-3, https://www.iccs-meeting.org/iccs2023/",

}

Osop, H, Suleiman, B, Alibasa, MJ, Wrigley, D, Helsham, A & Asmaro, A 2023, Improving Patients’ Length of Stay Prediction Using Clinical and Demographics Features Enrichment. in J Mikyška, C de Mulatier, VV Krzhizhanovskaya, PMA Sloot, M Paszynski & JJ Dongarra (eds), Computational Science – ICCS 2023, 23rd International Conference Prague, Czech Republic, July 3–5, 2023 Proceedings, Part II. Lecture Notes in Computer Science, vol. 14074, Springer, Cham Switzerland, pp. 120-128, International Conference on Computational Science 2023, Prague, Czechia, 3/07/23. https://doi.org/10.1007/978-3-031-36021-3_9

Improving Patients’ Length of Stay Prediction Using Clinical and Demographics Features Enrichment. / Osop, Hamzah; Suleiman, Basem; Alibasa, Muhammad Johan et al.
Computational Science – ICCS 2023, 23rd International Conference Prague, Czech Republic, July 3–5, 2023 Proceedings, Part II. ed. / Jiří Mikyška; Clélia de Mulatier; Valeria V. Krzhizhanovskaya; Peter M.A. Sloot; Maciej Paszynski; Jack J. Dongarra. Cham Switzerland: Springer, 2023. p. 120-128 (Lecture Notes in Computer Science; Vol. 14074).

Research output: Chapter in Book/Report/Conference proceeding › Conference Paper › Research › peer-review

TY - GEN

T1 - Improving Patients’ Length of Stay Prediction Using Clinical and Demographics Features Enrichment

AU - Osop, Hamzah

AU - Suleiman, Basem

AU - Alibasa, Muhammad Johan

AU - Wrigley, Drew

AU - Helsham, Alexandra

AU - Asmaro, Anne

N1 - Conference code: 23rd

PY - 2023

Y1 - 2023

N2 - Predicting patients’ length of stay (LOS) is crucial for efficient scheduling of treatment and strategic future planning, in turn reduce hospitalisation costs. However, this is a complex problem requiring careful selection of optimal set of essential factors that significantly impact the accuracy and performance of LOS prediction. Using an inpatient dataset of 285k of records from 14 general care hospitals in Vermont, USA from 2013–2017, we presented our novel approach to incorporate features to improve the accuracy of LOS prediction. Our empirical experiment and analysis showed considerable improvement in LOS prediction with an XGBoost model RMSE score of 6.98 and R2 score of 38.24%. Based on several experiments, we provided empirical analysis of the importance of different feature sets and its impact on predicting patients’ LOS.

AB - Predicting patients’ length of stay (LOS) is crucial for efficient scheduling of treatment and strategic future planning, in turn reduce hospitalisation costs. However, this is a complex problem requiring careful selection of optimal set of essential factors that significantly impact the accuracy and performance of LOS prediction. Using an inpatient dataset of 285k of records from 14 general care hospitals in Vermont, USA from 2013–2017, we presented our novel approach to incorporate features to improve the accuracy of LOS prediction. Our empirical experiment and analysis showed considerable improvement in LOS prediction with an XGBoost model RMSE score of 6.98 and R2 score of 38.24%. Based on several experiments, we provided empirical analysis of the importance of different feature sets and its impact on predicting patients’ LOS.

KW - electronic health records

KW - length of stay

KW - machine learning

UR - https://www.scopus.com/pages/publications/85169663285

U2 - 10.1007/978-3-031-36021-3_9

DO - 10.1007/978-3-031-36021-3_9

M3 - Conference Paper

AN - SCOPUS:85169663285

SN - 9783031360206

T3 - Lecture Notes in Computer Science

SP - 120

EP - 128

BT - Computational Science – ICCS 2023, 23rd International Conference Prague, Czech Republic, July 3–5, 2023 Proceedings, Part II

A2 - Mikyška, Jiří

A2 - de Mulatier, Clélia

A2 - Krzhizhanovskaya, Valeria V.

A2 - Sloot, Peter M.A.

A2 - Paszynski, Maciej

A2 - Dongarra, Jack J.

PB - Springer

CY - Cham Switzerland

T2 - International Conference on Computational Science 2023

Y2 - 3 July 2023 through 5 July 2023

ER -

Osop H, Suleiman B, Alibasa MJ, Wrigley D, Helsham A, Asmaro A. Improving Patients’ Length of Stay Prediction Using Clinical and Demographics Features Enrichment. In Mikyška J, de Mulatier C, Krzhizhanovskaya VV, Sloot PMA, Paszynski M, Dongarra JJ, editors, Computational Science – ICCS 2023, 23rd International Conference Prague, Czech Republic, July 3–5, 2023 Proceedings, Part II. Cham Switzerland: Springer. 2023. p. 120-128. (Lecture Notes in Computer Science). doi: 10.1007/978-3-031-36021-3_9