Volume 16, Issue 2 (March & April 2026)
J Research Health 2026, 16(2): 195-202 |
Back to browse issues page
Ethics code: IR.MUMS.FHMPM.REC.1402.214
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
sabouri S, Salari M. Predicting Length of Stay in Cardiovascular Patients Using Count Regression and Machine Learning Approaches. J Research Health 2026; 16 (2) :195-202
URL: http://jrh.gmu.ac.ir/article-1-2713-en.html
URL: http://jrh.gmu.ac.ir/article-1-2713-en.html
1- Department of Biostatistics, School of Health, Mashhad University of Medical Sciences, Mashhad, Iran. & Social Determinants of Health Research Center, Mashhad University of Medical Sciences, Mashhad, Iran. , sabourism@mums.ac.ir
2- Department of Biostatistics, School of Health, Mashhad University of Medical Sciences, Mashhad, Iran. & Social Determinants of Health Research Center, Mashhad University of Medical Sciences, Mashhad, Iran.
2- Department of Biostatistics, School of Health, Mashhad University of Medical Sciences, Mashhad, Iran. & Social Determinants of Health Research Center, Mashhad University of Medical Sciences, Mashhad, Iran.
Abstract: (320 Views)
Cardiovascular diseases often involve hospital readmissions and prolonged length of stay (LOS), that is a key measure for evaluating efficiency and quality of hospital care. This cross-sectional study aimed to predict and identify factors associated with LOS using various count regression and machine learning (ML) approaches. A total of 12,752 patients with cardiovascular disease were included. They were admitted to 43 hospitals affiliated with Mashhad University of Medical Sciences between March 2020 and February 2021. The mean patient age was 62.7±13.6 years, with an average LOS of 4.4±5.7 days and a median of 3 days (IQR: 3). Features including age, sex, diagnosis type, and percutaneous coronary intervention (PCI) were used to predict LOS. The predictive performance of the models was evaluated using mean absolute error (MAE). Although zero-truncated negative binomial regression (ZTNBR) outperformed other count regression models, support vector machine (SVM) achieved the lowest MAE (2.6), compared to artificial neural network (ANN) (MAE: 2.9) and the ZTNBR (MAE: 3.0). The type of cardiovascular disease was the most significant predictor of LOS, followed by PCI, age, and sex. These findings highlight the importance of patient-specific factors in designing healthcare strategies to optimize hospital resource allocation and improve patient outcomes.
Keywords: Cardiovascular disease, Machine learning (ML), Count data, Overdispersion, Length of stay (LOS)
Type of Study: Short Communication |
Subject:
● Service Quality
Received: 2024/12/27 | Accepted: 2025/06/25 | Published: 2026/03/1
Received: 2024/12/27 | Accepted: 2025/06/25 | Published: 2026/03/1
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC-By-NC), which permits use, distribution, and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.
Contact Information
