Towards a decision support tool for intensive care discharge: machine learning algorithm development using electronic health care data from MIMIC-III and Bristol, UK.

Abstract

ObjectiveThe primary objective is to develop an automated method for detecting patients that are ready for discharge from intensive care. DesignWe used two datasets of routinely collected patient data to test and improve upon a set of previously proposed discharge criteria. SettingBristol Royal Infirmary general intensive care unit (GICU). PatientsTwo cohorts derived from historical datasets: 1870 intensive care patients from GICU in Bristol, and 7592 from MIMIC-III (a publicly available intensive care dataset). ResultsIn both cohorts few successfully discharged patients met all of the discharge criteria. Both a random forest and a logistic classifier, trained using multiple-source cross-validation, demonstrated improved performance over the original criteria and generalised well between the cohorts. The classifiers showed good agreement on which features were most predictive of readiness-fordischarge, and these were generally consistent with clinical experience. By weighting the discharge criteria according to feature importance from the logistic model we showed improved performance over the original criteria, while retaining good interpretability. ConclusionsOur findings indicate the feasibility of the proposed approach to ready-for-discharge classification, which could complement other risk models of specific adverse outcomes in a future decision support system. Avenues for improvement to produce a clinically useful tool are identified. Strengths and Limitations of this study: Training data from multiple source domains is leveraged to produce general classifiers.  The restrictive feature representation tested could be expanded to better exploit the richness of available data and boost performance.  Our approach has the potential to streamline the discharge process in cases where patient physiology makes them clear candidates for a de-escalation of care.  High-risk patients would require additional levels of decision support to facilitate complex discharge planning.

Enhancing the value to users of machine learning-based clinical decision support tools: A framework for iterative, collaborative development and implementation

“…Given its importance, researchers and practitioners have sought to understand and predict bed capacity in order to smooth patient flow and assist hospital managers by adjusting approaches to scheduling and allocation (Litvak et al, 2008;Van Houdenhoven et al, 2008); estimating the need for ICU beds in a given population using techniques like discrete event simulation (Zhu et al, 2012), queuing models (McManus et al, 2004;Shmueli et al, 2003), and time series evaluations (Angelo et al, 2017); or both (Ryckman et al, 2009). ML algorithms have been developed to predict readiness for discharge from the ICU (Barnes et al, 2016;McWilliams et al, 2019) and to predict transfer into the ICU (Cheng et al, 2020); to our knowledge, no studies of ML-based CDS tools have focused on predicting low ICU bed capacity, and none progressed beyond retrospective analysis.…”