Prediction of Extubation Readiness in Extreme Preterm Infants by the Automated Analysis of CardioRespiratory Behavior

Study Summary

The investigators hypothesize that machine learning methods using a combination of novel, quantitative measures of cardio-respiratory variability can accurately predict the optimal time to extubate extreme preterm infants. In this multicenter prospective study, cardiorespiratory signals will be recorded from 250 extreme preterm infants who are eligible for extubation. Automated signal analysis algorithms will compute a variety of metrics for each infant describing the cardiorespiratory state. Machine learning methods will then be used to find the optimal combination of these statistical measures and clinical features that provide the best overall predictor of extubation readiness. Finally, investigators will develop an Automated system for Prediction of EXtubation (APEX) that will integrate the software for data acquisition, signal analysis, and outcome prediction into a single application suitable for use by medical personnel in the Neonatal Intensive Care Unit (NICU). The performance of APEX will later be clinically validated in 50 additional infants prospectively.

Primary Outcome

Infants will be considered to have failed extubation if they meet one or more of the following criteria within 72 hours of extubation: 1. Fraction of inspired oxygen (FiO2) \> 0.5 in order to maintain oxygen saturation (SpO2) \> 88% or PaO2 \> 45 mmHg (for 2 consecutive hours) 2. PaCO2 \> 55-60 mmHg with a pH \< 7.25 in two consecutive blood gases done 1-2 hours apart 3. 1 episode of apnea requiring positive pressure ventilation with bag and mask 4. Multiple episodes of apnea (≥ 6 episodes / 6

Conditions Studied

Interventions

• OTHER Cardiorespiratory signal acquisition

Study Locations (5)