Have a personal or library account? Click to login
Machine learning to predict extubation success using the spontaneous breathing trial, objective cough measurement, and diaphragmatic contraction velocity: Secondary analysis of the COBRE-US trial Cover

Machine learning to predict extubation success using the spontaneous breathing trial, objective cough measurement, and diaphragmatic contraction velocity: Secondary analysis of the COBRE-US trial

Open Access
|Jan 2025

Figures & Tables

Machine Learning Methods for Predicting Success in Spontaneous Breathing Trial

ModelStudy VariableAccuracySensitivitySpecificityPPVNVP
k-meansSBT* training64,072,631,579,923,5
SBT * test63,072,035,772,738,5

Hierarchical ClusteringSBT * training52,753,350,780,222,4
SBT * test60,954,964,379,240,9

Decision TreesSBT * training77,399,91,1NI1,0
SBT * test69,699,91,0NI1,0

Support Vector MachinesSBT * training77,399,91,1NI1,0
SBT * test69,699,91,1NI1,0

Neural NetworksSBT * training77,399,91,0NI1,0
SBT * test69,699,91,0NI1,0

QualificationDescription
0No cough
1Audible movement of air through the endotracheal tube, but no audible cough
2Strong cough with movement of secretions into the endotracheal tube
3Strong cough with movement of secretions out (expulsion) of the endotracheal tube

Etiology of Respiratory Failure and Reason for Admission to Intensive Care

VariablesValues
Shock, n(%)52 (14,9)
Hypercapnia (pH < 7,25, CO2 elevated), n(%)23 (6,6)
Hypoxemia (PaO2 < 60, usual FiO2), n(%)261 (75)
Neuromuscular, n(%)2 (0,6)
Perioperative, n(%)10 (2,9)

Reason for ICU Admission, n (%)
Medical345 (94)
Surgical (post-surgical only)22 (6)

General Characteristics of the Population_

Variables n (%)Values
Male n (%)219 (59,7)
Age, median (Range)61 (18 – 88)
Weight in kg, median (IQR)70 (60 – 80)
Height in cm, mean (SD)163,6 (10)
Body Mass Index (BMI) in kg/m2,
median (IQR)25,3 (21,7 – 29,1)
Active smoking, n (%)33 (9)
Alcoholism n (%)22 (6)

Comorbidities, n (%)
Diabetes Mellitus113 (30,8)
Hypertension173 (47,1)
Asthma8 (2,2)
Pulmonary Fibrosis6 (1,6)
Chronic Kidney Disease69 (18,8)
Chronic Liver Disease17 (4,6)

Machine Learning Methods for Predicting Extubation Success

ModelStudy VariableAccuracySensitivitySpecificityPPVNVP
k-meansSBT* training63,474,435,174,734,7
SBT * test63,076,737,569,846,2

Hierarchical ClusteringSBT * training66,491,68,069,731,4
SBT * test65,290,018,867,550

Decision TreesSBT * training89,898,370,494,668,7
SBT * test95,799,987,599,968,7

Support Vector MachinesSBT * training85,999,056,095,955
SBT * test93,599,981,399,981,3

Neural NetworksSBT * training85,999,056,095,955
SBT * test93,599,981,399,981,3
DOI: https://doi.org/10.2478/jccm-2025-0009 | Journal eISSN: 2393-1817 | Journal ISSN: 2393-1809
Language: English
Page range: 70 - 77
Submitted on: Oct 22, 2024
Accepted on: Jan 26, 2025
Published on: Jan 31, 2025
Published by: University of Medicine, Pharmacy, Science and Technology of Targu Mures
In partnership with: Paradigm Publishing Services
Publication frequency: 4 issues per year

© 2025 Fabio Varón-Vega, Eduardo Tuta-Quintero, Adriana Maldonado-Franco, Henry Robayo-Amórtegui, Luis F Giraldo-Cadavid, Daniel Botero-Rosas, published by University of Medicine, Pharmacy, Science and Technology of Targu Mures
This work is licensed under the Creative Commons Attribution 4.0 License.