A Novel Survival Analysis Approach to Predict the Need for Intubation in Intensive Care Units

Abstract

Intubation for mechanical ventilation (MV) is a common procedure performed in Intensive Care Units (ICUs). Early prediction of the need for intubation may have a positive impact by providing timely alerts to clinicians and consequently avoiding high risk late intubations. In this work, we propose a new machine learning method to predict intubation for MV, based on the concept of cure survival models. We tested our approach and compared it to other predictive models on a dataset collected from a secondary care hospital (AZ Groeninge, Kortrijk, Belgium). The results corroborate that our approach can improve the prediction of the need for intubation forMVin critically ill patients by using routinely collected data within the first hours of admission in the ICU. Early warning of need for intubation may be used to help clinicians predicting the risk of intubation and ranking patients according to their expected time to intubation.