Patient length of stay and mortality prediction: A survey

  title={Patient length of stay and mortality prediction: A survey},
  author={Aya Awad and Mohamed Bahy Bader-El-Den and James McNicholas},
  journal={Health Services Management Research},
  pages={105 - 120}
Over the past few years, there has been increased interest in data mining and machine learning methods to improve hospital performance, in particular hospitals want to improve their intensive care unit statistics by reducing the number of patients dying inside the intensive care unit. Research has focused on prediction of measurable outcomes, including risk of complications, mortality and length of hospital stay. The length of stay is an important metric both for healthcare providers and… 

Figures and Tables from this paper

Predicting Hospital Length of Stay for Accident and Emergency Admissions
A novel approach for the prediction of length of stay (LoS) prediction is investigated in this paper using only data based upon generic patient diagnoses, which demonstrates a classification accuracy of around 75%.
Length-of-stay and mortality prediction for a major hospital through interpretable machine learning
Though less accurate, simpler and interpretable models, such as decision trees, demonstrate very good predictive power, provide insights on discharge barriers and have been instrumental in interacting with care providers.
Machine learning in the prediction of medical inpatient length of stay
A review of machine learning in the prediction of total hospital inpatient LOS for medical inpatients was conducted using the databases PubMed, EMBASE and Web of Science, finding studies that had specifically evaluated machine learning LOS prediction in patients with diabetes and tuberculosis.
Predicting hospital mortality for intensive care unit patients: Time-series analysis
A thorough time-series analysis on the performance of different data mining methods during the first 48 h of intensive care unit admission showed that the discrimination power of the machine-learning classification methods after 6 h of admission outperformed the main scoring systems used in intensive care medicine.
A systematic review of the prediction of hospital length of stay: Towards a unified framework
An extensive review of the literature, examining the approaches employed for the prediction of LoS in terms of their merits and shortcomings and adopting a unified framework to better generalise the approaches that are being used to predict length of stay is proposed.
Associations between length of stay and municipal characteristics for a large Norwegian stroke unit
Although a regression model cannot prove causal relations, a reasonable interpretation is that the given predictors capture some aspects of a municipality’s capacity for home care, which makes it possible for the SU to discharge patients relatively early.
Predicting hospital mortality for ICU patients: time series analysis
A thorough timeseries analysis on the performance of different data mining methods during the first 48 hours of ICU admission showed that the discrimination power of the machine learning classification methods after 6 hours of admission outperformed the main scoring systems used in intensive care medicine after 48 hours.
Pursuing Optimal Prediction of Discharge Time in ICUs with Machine Learning Methods
A data-driven model is obtained to predict the discharge time of ICU patients, in a daily basis based on random forest technology, and results were obtained differentiating by patients’ gender and age, confirming the approach as a good means to achieve optimal performance when more data will be available.
Patient length of stay analysis with machine learning algorithms
  • S. Tomovic
  • Computer Science
    Comput. Sci. Inf. Syst.
  • 2021
The problem of measuring factor importance on patient length of stay in an emergency department is discussed and basic idea consists of identifying appropriate neighbourhood and measure distances between the new observation and its neighbourhood in the historical dataset with respect to each factor.


Healthcare Data Mining: Predicting Hospital Length of Stay (PHLOS)
This study provides insight into the underlying factors that influence hospital length of stay, using a multi-tiered data mining approach to form training sets and identifying patients who need aggressive or moderate early interventions to prevent prolonged stays.
Predicting Hospital Length of Stay (PHLOS): A Multi-tiered Data Mining Approach
This paper proposes a methodology that employs clustering to create the training sets to train different classification algorithms and consistently found that using clustering as a precursor to form the training set gives better prediction results as compared to non-clustering based training sets.
Predicting in-hospital mortality of ICU patients: The PhysioNet/Computing in cardiology challenge 2012
The focus of the PhysioNet/CinC Challenge 2012 is to develop methods for patient-specific prediction of in-hospital mortality by measuring performance of a binary classifier and a risk estimator.
Daily assessment of severity of illness and mortality prediction for individual patients
The prognosis of critically ill patients using a statistical model that incorporates the daily probabilities of hospital mortality during the first week of stay in the intensive care unit (ICU) is refined.
Modelling patient length of stay in public hospitals in Mexico
This research highlights the areas of opportunity and research gap from previous studies and practices, and proposes the use of finite mixture models to approximate the distribution of length of stay, as the foundation of more sophisticated models designed to include the internal and external factors associated with LoS.
Factors influencing hospital high length of stay outliers
Age, type of admission, and hospital type were significantly associated with high LOS outliers, and the proportion of high outliers does not seem to be related to their financial coverage; they should be studied in order to highlight areas for further investigation.
Predicting hospital mortality for patients in the intensive care unit: A comparison of artificial neural networks with logistic regression models
When sample size is adequate, LR and ANN models have similar performance, however, development sets of ≤800 were generally inadequate, which is concerning, given typical sample sizes used for individual ICU mortality prediction.
Real-time prediction of mortality, readmission, and length of stay using electronic health record data
Bayesian Networks can model EHRs to provide real-time forecasts for patient outcomes, which provide richer information than traditional independent point predictions of length of stay, death, or readmission, and can thus better support decision making.
Outcome prediction models on admission in a medical intensive care unit: do they predict individual outcome?
It is concluded that the estimation of risk on admission by the three methods investigated might be helpful for global comparisons of ICU populations, although the lack of disease specificity reduces their applicability for severity grading of a given illness.
Forecasting hospital bed requirements and cost of care using phase type survival trees
How phase type survival tree based clustering can be used for better forecasting of bed requirements and cost in a stroke unit by analyzing the effects of individual clusters and their interactions with the whole patient care unit is illustrated.