Predicting in-hospital length of stay: a two-stage modeling approach to account for highly skewed data

Xu, Zhenhui; Zhao, Congwen; Scales, Charles D.; Henao, Ricardo; Goldstein, Benjamin A.

doi:10.1186/s12911-022-01855-0

Cited by 8 publications

(6 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As such, population-based LOS predictions are key enablers of organizational resource planning as well as the daily access and flow issues managed by the frontline staff. Recent study by Xu and colleagues 56 has shown good LOS predictive capability in all-inclusive elective surgery population. Hence, the purpose of prediction should guide the choice of procedure-specific vs population-specific models.…”

Section: Discussionmentioning

confidence: 95%

Hospital length of stay prediction for general surgery and total knee arthroplasty admissions: Systematic review and meta-analysis of published prediction models

et al. 2023

View full text Add to dashboard Cite

Objective Systematic review of length of stay (LOS) prediction models to assess the study methods (including prediction variables), study quality, and performance of predictive models (using area under receiver operating curve (AUROC)) for general surgery populations and total knee arthroplasty (TKA). Method LOS prediction models published since 2010 were identified in five major research databases. The main outcomes were model performance metrics including AUROC, prediction variables, and level of validation. Risk of bias was assessed using the PROBAST checklist. Results Five general surgery studies (15 models) and 10 TKA studies (24 models) were identified. All general surgery and 20 TKA models used statistical approaches; 4 TKA models used machine learning approaches. Risk scores, diagnosis, and procedure types were predominant predictors used. Risk of bias was ranked as moderate in 3/15 and high in 12/15 studies. Discrimination measures were reported in 14/15 and calibration measures in 3/15 studies, with only 4/39 externally validated models (3 general surgery and 1 TKA). Meta-analysis of externally validated models (3 general surgery) suggested the AUROC 95% prediction interval is excellent and ranges between 0.803 and 0.970. Conclusion This is the first systematic review assessing quality of risk prediction models for prolonged LOS in general surgery and TKA groups. We showed that these risk prediction models were infrequently externally validated with poor study quality, typically related to poor reporting. Both machine learning and statistical modelling methods, plus the meta-analysis, showed acceptable to good predictive performance, which are encouraging. Moving forward, a focus on quality methods and external validation is needed before clinical application.

show abstract

Section: Discussionmentioning

confidence: 95%

Hospital length of stay prediction for general surgery and total knee arthroplasty admissions: Systematic review and meta-analysis of published prediction models

et al. 2023

View full text Add to dashboard Cite

show abstract

“…In this systematic review of machine learning models related to the length of stay for COVID-19 patients, we identified and critically evaluated prediction models described in 10 studies. Nine [4,[23][24][25][26][27][28][29][30] of the prognostic models are qualitative and tried to predict the length of stay for COVID-19 patients in the shape of "Short" or" Long" and only one [31] of them predict the length of stay quantitatively. In the qualitative studies evaluation metrics for qualitative modeling such as accuracy, F1-score, specificity, sensitivity, and AUC have been reported and for the quantitative ones, the evaluation metrics were MAE, MSE, and MRE.…”

Section: Discussionmentioning

confidence: 99%

“…According to the modeling approach presented in the studies included in this study, four articles [24,26,27,31] have studied the general department of the hospital, three studies [23,28,29] have targeted the general department of the hospital and the ICU simultaneously, one study [4] related to the emergency department and two studies [25,30] were related to the ICU department. The only study [31] with quantitative modeling was associated with the hospital's general department. Among machine learning algorithms, Ensemble algorithms such as Random Forest, Gradient Boosting, and XGBoost have been used the most.…”

Section: Discussionmentioning

confidence: 99%

A Systematic Review on Machine Learning Algorithms to Predict the Length of Stay for COVID-19 Patients

Sayadi,

Sadeghian Yazdeli,

Asaadi Vaskas

et al. 2023

Front Health Inform

View full text Add to dashboard Cite

Introduction: Managing resources is one of the most important challenges that healthcare providers worldwide face during the COVID-19 pandemic. In recent years, machine learning has been developed to provide valuable help in predicting disease and estimating the duration of their stay. This study aimed to identify the machine learning models for predicting length of stay in COVID-19.Material and Methods: Online databases, including Scopus, PubMed, Web of Science, and Science Direct, were searched, and a hand search through Google Scholar and grey literature was done up to August 2023 and updated in December 2023 to identify articles to find all relevant studies. To manage the process and check the quality of included articles PRISMA guidelines and CASP checklist were used and data was extracted using a data extraction form.Results: Among all 489 research articles, 10 studies met the inclusion criteria. The best models reported in the included articles were random forest (n=3), gradient boosting (n=2), XGBoost (n=2), SVM (n=1), KNN (n=1), and DataRobot (n=1). Except one of the studies that used quantitative modeling and reported MSE and MAE as evaluation criteria, other studies used qualitative modeling and reported accuracy, specificity, and F1-score. The focus of the included articles was on the general and ICU departments as the important resources in the hospital and emphasized the use of machine learning to predict the length of stay.Conclusion: The results of this systematic review showed that a data mining approach and using a machine learning algorithm can help to manage the critical resources of the hospital especially when we are faced with a pandemic disease like COVID-19.

show abstract

“…One major hurdle involves determining the appropriate time window for data collection, ranging from early-stage predictions within 12-48 hours [33, 34, 35] to encompassing the entire duration of admission [36],or retrospective analysis [37, 38]. Additionally, the scarcity of examples within certain data classes can impede the modeling of specific clinical targets [39, 40]. For example, while predicting outcomes for hospital stays exceeding seven days may be feasible, it becomes impractical for stays surpassing 30 days due to insufficient data samples.…”

Section: Introductionmentioning

confidence: 99%

EHR-ML: A generalisable pipeline for reproducible clinical outcomes using electronic health records

Ramakrishnaiah,

Macesic,

Webb

et al. 2024

Preprint

View full text Add to dashboard Cite

The healthcare landscape is experiencing a transformation with the integration of Artificial Intelligence (AI) into traditional analytic workflows. However, this advancement encounters challenges due to variations in clinical practices, resulting in a crisis of generalisability. Addressing this issue, our proposed solution, EHR-ML, offers an open-source pipeline designed to empower researchers and clinicians. By leveraging institutional Electronic Health Record (EHR) data, EHR-ML facilitates predictive modelling, enabling the generation of clinical insights. EHR-ML stands out for its comprehensive analysis suite, guiding researchers through optimal study design, and its built-in flexibility allowing for construction of robust, customisable models. Notably, EHR-ML integrates a dedicated two-layered ensemble model utilising feature representation learning. Additionally, it includes a feature engineering mechanism to handle intricate temporal signals from physiological measurements. By seamlessly integrating with our quality assurance pipelines, this utility leverages its data standardization and anomaly handling capabilities. Benchmarking analyses demonstrate EHR-ML's efficacy, particularly in predicting outcomes like inpatient mortality and the Intensive Care Unit (ICU) Length of Stay (LOS). Models built with EHR-ML outperformed conventional methods, showcasing its generalisability and versatility even in challenging scenarios such as high class-imbalance. We believe EHR-ML is a critical step towards democratising predictive modelling in healthcare, enabling rapid hypothesis testing and facilitating the generation of biomedical knowledge. Widespread adoption of tools like EHR-ML will unlock the true potential of AI in healthcare, ultimately leading to improved patient care.

show abstract

Predicting in-hospital length of stay: a two-stage modeling approach to account for highly skewed data

Cited by 8 publications

References 30 publications

Hospital length of stay prediction for general surgery and total knee arthroplasty admissions: Systematic review and meta-analysis of published prediction models

Hospital length of stay prediction for general surgery and total knee arthroplasty admissions: Systematic review and meta-analysis of published prediction models

A Systematic Review on Machine Learning Algorithms to Predict the Length of Stay for COVID-19 Patients

EHR-ML: A generalisable pipeline for reproducible clinical outcomes using electronic health records

Contact Info

Product

Resources

About