Using machine learning models to predict oxygen saturation following ventilator support adjustment in critically ill children: A single center pilot study

Abstract

Background: In an intensive care units, experts in mechanical ventilation are not continuously at patient's bedside to adjust ventilation settings and to analyze the impact of these adjustments on gas exchange. The development of clinical decision support systems analyzing patients' data in real time offers an opportunity to fill this gap.

Objective: The objective of this study was to determine whether a machine learning predictive model could be trained on a set of clinical data and used to predict transcutaneous hemoglobin oxygen saturation 5 min (5min SpO2) after a ventilator setting change.

Data sources: Data of mechanically ventilated children admitted between May 2015 and April 2017 were included and extracted from a high-resolution research database. More than 776,727 data rows were obtained from 610 patients, discretized into 3 class labels (< 84%, 85% to 91% and c92% to 100%).

Performance metrics of predictive models: Due to data imbalance, four different data balancing processes were applied. Then, two machine learning models (artificial neural network and Bootstrap aggregation of complex decision trees) were trained and tested on these four different balanced datasets. The best model predicted SpO2 with area under the curves < 0.75.

Conclusion: This single center pilot study using machine learning predictive model resulted in an algorithm with poor accuracy. The comparison of machine learning models showed that bagged complex trees was a promising approach. However, there is a need to improve these models before incorporating them into a clinical decision support systems. One potentially solution for improving predictive model, would be to increase the amount of data available to limit over-fitting that is potentially one of the cause for poor classification performances for 2 of the three class labels.

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Fig 1. Schematic description of the items involved and analysis process.

EMR: electronic Medical Record, FiO2: inspired fraction of Oxygen, Vt: tidal volume, PEEP: Positive end expiratory pressure, PS above PEEP: pressure support level Above PEEP, PC above PEEP: pressure control level above PEEP, I:E Ratio: inspiratory time over expiratory time, Measured RR: respiratory rate measured by the ventilator. 5minSpO2: SpO2 observed 5 min after PEEP, FiO2, tidal volume, PS above PEEP, PC above PEEP change, ML: machine learning. Heart and pulse rate were only used to validate the database SpO2 value (see below and S1 File).

Fig 2. ROC curve for each SpO2 prediction at 5 min following a ventilator setting change of the best predictive model (bootstrap aggregation of complex decision trees (BACDT) classifiers on Test Dataset 3).

Class 1: 5 minSpO2 < 84%, class 2: 5 minSpO2 between 85% and 91%, class 3: 5 minSpO2 between 92% and 100%. AUC: area under the curve, 95IC: 95% confidence interval.