Predicting Intensive Care Unit Readmission with Machine Learning Using Electronic Health Record Data

Abstract

A machine learning approach to predicting intensive care unit readmission was significantly more accurate than previously published algorithms in both our internal validation and the Medical Information Mart for Intensive Care-III cohort. Implementation of this approach could target patients who may benefit from additional time in the intensive care unit or more frequent monitoring after transfer to the hospital ward.

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

“…One barrier to this approach is the challenge of harmonising the data, especially when combining data from different sources. This is one reason that we did not include diagnosis codes or severity of illness scores in this study, although they have previously been shown to be predictive of adverse events following discharge [11,12]. During a patient's stay in ICU, many of their physiological parameters are controlled by clinical intervention, and their expected physiological state is dependent on their medical history (see, for example, guidelines on acceptable levels of Hb in different patient types [38]).…”

“…These tools range from criteria to evaluate discharge readiness [22,42], to guidelines for discharge planning and education [6]. Additionally, a number of risk models have been developed to predict adverse outcomes following ICU discharge [11][12][13]15,43]. In particular Badawi and Breslow demonstrated that mortality and readmission should be modelled independently as separate outcomes [12].…”

“…This work considers the ICU discharge process, which has accessible data from routine collection and requires a simple but important binary decision that could benefit from an evidence-based approach. Indeed, several statistical models have recently been developed to predict the risk of adverse events following intensive care discharge [11][12][13][14][15]. Such risk models are invaluable tools for clinical decision making and, in the context of ICU discharge, can provide information with which to plan complex deescalations of care.…”

Developing a Machine Learning prediction model for bedside decision support by predicting readmission or death following discharge from the Intensive Care unit

“…The decision to discharge a patient from the ICU is usually based on clinical judgment using experience, clinical intuition, physiological parameters and scores to assess severity of illness and specific criteria, such as the need for mechanical ventilation or vasoactive medication [9][10][11][12].…”

“…Given the complexity of the decision and taking into account the vast amount of data gathered routinely in the ICU, machine learning is particularly suited for this task. Indeed, recent years have witnessed several attempts to develop these models [10,[15][16][17][18][19][20][21][22][23][24][25][26][27]. However, none of these models seem to have been specifically developed with the intention to directly implement them within the context of existing electronic health records to deliver real time decision support at the bedside.…”