Prediction of Intensive Care Unit Length of Stay in the MIMIC-IV Dataset
Source abstract
Abstract
Accurately estimating the length of stay (LOS) of patients admitted to the intensive care unit (ICU) in relation to their health status helps healthcare management allocate appropriate resources and better plan for the future. This paper presents predictive models for the LOS of ICU patients from the MIMIC-IV database based on typical demographic and administrative data, as well as early vital signs and laboratory measurements collected on the first day of ICU stay. The goal of this study was to demonstrate a practical, stepwise approach to predicting patient’s LOS in the ICU using machine learning and early available typical clinical data. The results show that this approach significantly improves the performance of models for predicting actual LOS in a pragmatic framework that includes only data with short stays predetermined by a prior classification.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Machine Learning in Healthcare · Sepsis Diagnosis and Treatment · Hemodynamic Monitoring and Therapy
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Deptarment for Medical Data Science, Leipzig University Medical Center, 04107 Leipzig, Germany
Institute for Medical Informatics, Statistics and Epidemiology, Leipzig University, 04107 Leipzig, Germany
Faculty Applied Computer and Bio Sciences, Mittweida University of Applied Sciences, 09648 Mittweida, Germany
Funding
German Ministry for Research and Education · German Ministry of Health
References (69)
Show source-provided references
1. What is an intensive care unit? A report of the task force of the World Federation of Societies of Intensive and Critical Care Medicine10.1016/j.jcrc.2016.07.015 ↗
2. From Intensive Care to Critical Care Medicine: A Historical Perspective10.1164/rccm.201008-1341OE ↗
3. Machine learning identifies ICU outcome predictors in a multicenter COVID-19 cohort10.1186/s13054-021-03720-4 ↗
4. Using machine learning for predicting intensive care unit resource use during the COVID-19 pandemic in Denmark10.1038/s41598-021-98617-1 ↗
5. Prediction of Hospital Length of Stay
6. Stone, K., Zwiggelaar, R., Jones, P., and Parthaláin, N.M. (2022). A systematic review of the prediction of hospital length of stay: Towards a unified framework. PLOS Digit. Health, 1.10.1371/journal.pdig.0000017 ↗
7. Quality measurement at intensive care units: Which indicators should we use?10.1016/j.jcrc.2007.01.002 ↗
8. Lingsma, H.F., Bottle, A., Middleton, S., Kievit, J., Steyerberg, E.W., and Marang-van de Mheen, P.J. (2018). Evaluation of hospital outcomes: The relation between length-of-stay, readmission, and mortality in a large international administrative database. BMC Health Serv. Res., 18.10.1186/s12913-018-2916-1 ↗
9. Length of stay as quality indicator in emergency departments: Analysis of determinants in the German Emergency Department Data Registry (AKTIN registry)10.1007/s11739-021-02919-1 ↗
10. Is patient length of stay related to quality of care?
11. Length of in-hospital stay and its relationship to quality of care10.1136/qhc.11.3.209 ↗
12. Length of Stay: An Appropriate Quality Measure?10.1001/archsurg.142.5.461 ↗
13. Critical care medicine in the United States 2000–2005: An analysis of bed numbers, occupancy rates, payer mix, and costs10.1097/CCM.0b013e3181b090d0 ↗
14. Predicting Intensive Care Unit Length of Stay and Mortality Using Patient Vital Signs: Machine Learning Model Development and Validation10.2196/21347 ↗
15. The Physician as an Estimator of Hospital Stay10.1177/001872086600800302 ↗
16. Length of stay: Prediction and explanation
17. ICU physicians are unable to accurately predict length of stay at admission: A prospective study10.1093/intqhc/mzv112 ↗
18. Can the experienced ICU physician predict ICU length of stay and outcome better than less experienced colleagues?10.1007/s00134-003-2139-7 ↗
19. Machine learning in the prediction of medical inpatient length of stay10.1111/imj.14962 ↗
20. Prediction algorithm for ICU mortality and length of stay using machine learning10.1038/s41598-022-17091-5 ↗
21. Artificial intelligence in intensive care medicine10.1007/s00134-020-06203-2 ↗
22. Use of machine learning to analyse routinely collected intensive care unit data: A systematic review10.1186/s13054-019-2564-9 ↗
23. Artificial Intelligence in the Intensive Care Unit10.1186/s13054-020-2785-y ↗
24. Twelve key challenges in medical machine learning and solutions10.1016/j.ibmed.2022.100068 ↗
25. Multitask learning and benchmarking with clinical time series data10.1038/s41597-019-0103-9 ↗
26. Sridhar, S., Whitaker, B., Mouat-Hunter, A., and McCrory, B. (2022). Predicting Length of Stay using machine learning for total joint replacements performed at a rural community hospital. PLoS ONE, 17.10.1371/journal.pone.0277479 ↗
27. Predictors of in-hospital length of stay among cardiac patients: A machine learning approach10.1016/j.ijcard.2019.01.046 ↗
28. Associations between Short or Long Length of Stay and 30-Day Readmission and Mortality in Hospitalized Patients with Heart Failure10.1016/j.jchf.2017.03.012 ↗
29. Alturki, L., Aloraini, K., Aldughayshim, A., and Albahli, S. (2019, January 3–7). Predictors of Readmissions and Length of Stay for Diabetes Related Patients. Proceedings of the 2019 IEEE/ACS 16th International Conference on Computer Systems and Applications (AICCSA), Abu Dhabi, United Arab Emirates.10.1109/AICCSA47632.2019.9035280 ↗
30. Morton, A., Marzban, E., Giannoulis, G., Patel, A., Aparasu, R., and Kakadiaris, I.A. (2014, January 3–6). A Comparison of Supervised Machine Learning Techniques for Predicting Short-Term In-Hospital Length of Stay among Diabetic Patients. Proceedings of the 2014 13th International Conference on Machine Learning and Applications, Detroit, MI, USA.10.1109/ICMLA.2014.76 ↗
31. From Data to Optimal Decision Making: A Data-Driven, Probabilistic Machine Learning Approach to Decision Support for Patients with Sepsis10.2196/medinform.3445 ↗
32. Osuagwu, U.L., Xu, M., Piya, M.K., Agho, K.E., and Simmons, D. (2022). Factors associated with long intensive care unit (ICU) admission among inpatients with and without diabetes in South Western Sydney public hospitals using the New South Wales admission patient data collection (2014–2017). BMC Endocr. Disord., 22.10.1186/s12902-022-00933-8 ↗
33. Gentimis, T., Alnaser, A.J., Durante, A., Cook, K., and Steele, R. (2017, January 6–10). Predicting Hospital Length of Stay Using Neural Networks on MIMIC III Data. Proceedings of the 2017 IEEE 15th Intl Conf on Dependable, Autonomic and Secure Computing, 15th International Conference on Pervasive Intelligence and Computing, 3rd International Conference on Big Data Intelligence and Computing and Cyber Science and Technology Congress (DASC/PiCom/DataCom/CyberSciTech), Orlando, FL, USA.10.1109/DASC-PICom-DataCom-CyberSciTec.2017.191 ↗
34. Takekawa, D., Endo, H., Hashiba, E., and Hirota, K. (2022). Predict models for prolonged ICU stay using APACHE II, APACHE III and SAPS II scores: A Japanese multicenter retrospective cohort study. PLoS ONE, 17.10.1371/journal.pone.0269737 ↗
35. Intensive care unit length of stay: Benchmarking based on Acute Physiology and Chronic Health Evaluation (APACHE) IV10.1097/01.CCM.0000240233.01711.D9 ↗
36. Scalable and accurate deep learning with electronic health records10.1038/s41746-018-0029-1 ↗
37. Rocheteau, E., Liò, P., and Hyland, S. (2021, January 8–10). Temporal pointwise convolutional networks for length of stay prediction in the intensive care unit. Proceedings of the Conference on Health, Inference, and Learning, Virtual Event USA.10.1145/3450439.3451860 ↗
38. MIMIC-IV, a freely accessible electronic health record dataset10.1038/s41597-022-01899-x ↗
39. Johnson, A., Bulgarelli, L., Pollard, T., Horng, S., Celi, L.A., and Mark, R. (2023, June 03). MIMIC-IV. Available online: https://physionet.org/content/mimiciv/2.2/.
40. Vital signs10.1111/j.1479-6988.2004.00012.x ↗
41. World Health Organization (2015). International Statistical Classification of Diseases and Related Health Problems, World Health Organization.
42. Wang, W., Li, Y., and Yan, J. (2023, June 03). Touch: Tools of Utilization and Cost in Healthcare. Available online: https://cran.r-project.org/web/packages/touch/index.html.
43. Age-predicted maximal heart rate revisited10.1016/S0735-1097(00)01054-8 ↗
44. Respiratory rates in emergency department patients10.1016/0736-4679(89)90257-6 ↗
45. Target oxygen saturation range: 92–96% Versus 94–98%: Target oxygen saturation range10.1111/resp.12879 ↗
46. Mathew, T.M., and Sharma, S. (2023). High Altitude Oxygenation, StatPearls Publishing.
47. Normal Body Temperature: A Systematic Review10.1093/ofid/ofz032 ↗
48. Predicting Length of Stay Across Hospital Departments10.1109/ACCESS.2021.3066562 ↗
49. Chen, T., and Guestrin, C. (2016, January 13–17). XGBoost: A Scalable Tree Boosting System. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA.10.1145/2939672.2939785 ↗
50. Prediction and Analysis of Length of Stay Based on Nonlinear Weighted XGBoost Algorithm in Hospital10.1155/2021/4714898 ↗
51. Chen, R., Zhang, S., Li, J., Guo, D., Zhang, W., Wang, X., Tian, D., Qu, Z., and Wang, X. (2023). A study on predicting the length of hospital stay for Chinese patients with ischemic stroke based on the XGBoost algorithm. BMC Med. Inform. Decis. Mak., 23.10.1186/s12911-023-02140-4 ↗
52. Chicco, D., and Jurman, G. (2020). The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation. BMC Genom., 21.10.1186/s12864-019-6413-7 ↗
53. Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance10.3354/cr030079 ↗
54. Root mean square error (RMSE) or mean absolute error (MAE)?—Arguments against avoiding RMSE in the literature10.5194/gmd-7-1247-2014 ↗
55. The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation10.7717/peerj-cs.623 ↗
56. An R-squared measure of goodness of fit for some common nonlinear regression models10.1016/S0304-4076(96)01818-0 ↗
57. On the use of r and r squared in correlation and regression10.1111/1440-1703.1011 ↗
58. Wickham, H. (2016). ggplot2: Elegant Graphics for Data Analysis, Springer. [2nd ed.].10.1007/978-3-319-24277-4 ↗
59. Wickham, H., Vaughan, D., and Girlich, M. (2023, June 03). Tidyr: Tidy Messy Data. Available online: https://tidyr.tidyverse.org.
60. Wickham, H., François, R., Henry, L., Müller, K., and Vaughan, D. (2023, June 03). Dplyr: A Grammar of Data Manipulation. Available online: https://dplyr.tidyverse.org.
61. Resting Heart Rate in Cardiovascular Disease10.1016/j.jacc.2007.04.079 ↗
62. Effect of introducing the Modified Early Warning score on clinical outcomes, cardio-pulmonary arrests and intensive care utilisation in acute medical admissions: Forum10.1046/j.1365-2044.2003.03258.x ↗
63. The ups and downs of heart rate10.1097/CCM.0b013e318232e50c ↗
64. Age-predicted maximal heart rate in healthy subjects: The HUNT Fitness Study: Maximal heart rate in a population10.1111/j.1600-0838.2012.01445.x ↗
65. Reference values for the physical work capacity on a bicycle ergometer for men—A comparison with a previous study on women: Physical work capacity on a bicycle ergometer10.1046/j.1475-097X.2003.00491.x ↗
66. Elevated body temperature independently contributes to increased length of stay in neurologic intensive care unit patients10.1097/01.CCM.0000129484.61912.84 ↗
67. Brodersen, K.H., Ong, C.S., Stephan, K.E., and Buhmann, J.M. (2010, January 23–26). The Balanced Accuracy and Its Posterior Distribution. Proceedings of the 2010 20th International Conference on Pattern Recognition, Istanbul, Turkey.10.1109/ICPR.2010.764 ↗
68. A comparison of medical record with billing diagnostic information associated with ambulatory medical care10.2105/AJPH.71.2.145 ↗
69. A deep attention model to forecast the Length Of Stay and the in-hospital mortality right on admission from ICD codes and demographic data10.1016/j.jbi.2021.103778 ↗