Characterising risk of in-hospital mortality following cardiac arrest using machine learning: A retrospective international registry study

Abstract

BackgroundResuscitated cardiac arrest is associated with high mortality; however, the ability to estimate risk of adverse outcomes using existing illness severity scores is limited. Using in-hospital data available within the first 24 hours of admission, we aimed to develop more accurate models of risk prediction using both logistic regression (LR) and machine learning (ML) techniques, with a combination of demographic, physiologic, and biochemical information. Methods and findingsPatient-level data were extracted from the Australian and New Zealand Intensive Care Society (ANZICS) Adult Patient Database for patients who had experienced a cardiac arrest within 24 hours prior to admission to an intensive care unit (ICU) during the period January 2006 to December 2016. The primary outcome was in-hospital mortality. The models were trained and tested on a dataset (split 90:10) including age, lowest and highest physiologic variables during the first 24 hours, and key past medical history. LR and 5 ML approaches (gradient boosting machine [GBM], support vector classifier [SVC], random forest [RF], artificial neural network [ANN], and an ensemble) were compared to the APACHE III and Australian and New Zealand Risk of Death (ANZROD) predictions. In all, 39,566 patients from 186 ICUs were analysed. Mean (±SD) age was 61 ± 17 years; 65% were male. Overall inhospital mortality was 45.5%. Models were evaluated in the test set. The APACHE III and PLOS Medicine | https://doi.

Opening the Black Box: Interpretable Machine Learning for Geneticists

“…While the complex logic underlying the whole model may be too much for a surrogate model to learn, the logic for one instance or a group of similar instances (e.g., coexpressed genes), hence local, may be simple enough. For example, LIME was used to better understand why some patients were misclassified by a black box model predicting survival after cardiac arrest [59]. A LIME model for a patient that was mispredicted to survive showed that the black box model was too heavily influenced by certain features (e.g., healthy neurologic status, lack of chronic respiratory illness) and did not place sufficient weight on other features that are also important (e.g., elevated creatinine, advanced age).…”

Machine Learning for Clinical Outcome Prediction

“…The local interpretable model-agnostic explanation (LIME) [120] generates a local explanation of the model behaviour using a shallow model. It has been even used to explain ML models for the prediction of in-hospital mortality [121]. However, it has also been argued that linear models, rule-based models, and decision trees are not intrinsically interpretable [117].…”

Opening the Black Box: Interpretable Machine Learning for Geneticists

“…For example, LIME was used to better understand why some patients (i.e. instances) were misclassified by a black box model predicting survival after cardiac arrest [51]. A LIME model for a patient that was mis-predicted to survive showed that the black box model was too heavily influenced by certain features (e.g.…”