Respiratory physiology after resupination following prone ventilation to predict 28-day mortality in mechanically ventilated patients: a machine learning analysis

The clinical significance of resupination parameters following prone positioning remains largely unknown. This study employed machine learning to predict the survival of patients receiving mechanical ventilation (MV) by analyzing oxygenation and respiratory mechanics after resupination. Data were extracted from the COVID-Predict Dutch Data Warehouse. Patients receiving MV who underwent the supine–prone–supine sequence were selected, and the variables related to respiratory physiology within 4 h before proning and after resupination were recorded. Machine learning models were trained on the features selected using LASSO regression to predict the 28-day mortality. Patients who did not survive (157/522, 30.1%) had lower PaO2/FiO2 values, higher ventilatory ratios, increased physiological dead space, higher driving pressure, and lower static and dynamic lung compliance values at resupination. The predictive performance of the individual clinical parameters for 28-day mortality was generally modest, and FiO2, PaO2/FiO2, physiological dead space, and dynamic lung compliance were the best predictors of mortality. Overall, XGBoost showed the best recall (0.732) and maintained the highest AUC (0.719), while its F1-score (0.500) was the best for predicting mortality despite a low precision (0.380). Survival after prone positioning in patients receiving MV can be stratified by physiological responses after resupination.

Authors

Institutions

Publication Details

Journal
Scientific Reports
Published
2026-02-10
DOI
https://doi.org/10.1038/s41598-026-39336-3
Primary Topic
Respiratory Support and Mechanisms
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Respiratory physiology after resupination following prone ventilation to predict 28-day mortality in mechanically ventilated patients: a machine learning analysis

Robert Stefan Willem Entjes, Thijs C.D. Rettig, Bosman Rj, Tariq A. Dam et al.
Scientific Reports
Respiratory Support and Mechanisms
article

Respiratory physiology after resupination following prone ventilation to predict 28-day mortality in mechanically ventilated patients: a machine learning analysis

Robert Stefan Willem Entjes, Thijs C.D. Rettig, Bosman Rj, Tariq A. Dam, Age D. Boelens, Evelien Oostdijk, R. J. M. Nowitzky, Marlijn J. A. Kamps, Gert B. Brunnekreef, Lada Lijović, Dharmanand Ramnarain, Moon Seong Baek, Tae Wan Kim, Won-Young Kim, Sesmu Arbous, Remko M de Jong, Gyungah Kim, Arjan Cloïn, Olaf L. Cremer, D.Jannet Mehagnoul-Schipper, Judith Lens, Lucas M. Fleuren, Alexander D. Cornet, Harald J. Faber, Evert-Jan Wils, Dave A. Dongelmans, Cornelis P.C. de Jager, Auke C. Reidinga, Stefaan H.A. Hendriks, Ellen G.M. Smit, Diederik Gommers, The Dutch ICU Data Sharing Collaborators, Barbara Festen-Spanjer, Tim Frenzel, Paul W. G. Elbers, Tom Dormans, Marco A.A. Peters, Walter van denTempel, A. Karakus, Louise C. Urlings-Strop, Paul de Jong, Sefanja Achterberg, Peter Koetsier
article en

Abstract

The clinical significance of resupination parameters following prone positioning remains largely unknown. This study employed machine learning to predict the survival of patients receiving mechanical ventilation (MV) by analyzing oxygenation and respiratory mechanics after resupination. Data were extracted from the COVID-Predict Dutch Data Warehouse. Patients receiving MV who underwent the supine–prone–supine sequence were selected, and the variables related to respiratory physiology within 4 h before proning and after resupination were recorded. Machine learning models were trained on the features selected using LASSO regression to predict the 28-day mortality. Patients who did not survive (157/522, 30.1%) had lower PaO2/FiO2 values, higher ventilatory ratios, increased physiological dead space, higher driving pressure, and lower static and dynamic lung compliance values at resupination. The predictive performance of the individual clinical parameters for 28-day mortality was generally modest, and FiO2, PaO2/FiO2, physiological dead space, and dynamic lung compliance were the best predictors of mortality. Overall, XGBoost showed the best recall (0.732) and maintained the highest AUC (0.719), while its F1-score (0.500) was the best for predicting mortality despite a low precision (0.380). Survival after prone positioning in patients receiving MV can be stratified by physiological responses after resupination.

Scientific ReportsVol. 16(1)
Radboud University Nijmegen (NL), Catharina Ziekenhuis (NL), Medisch Spectrum Twente (NL), Erasmus MC (NL), Medisch Centrum Leeuwarden (NL), Radboud University Medical Center (NL), University Medical Center Utrecht (NL), Sint Franciscus Gasthuis (NL), Albert Schweitzer Ziekenhuis (NL), Slingeland Ziekenhuis (NL), Medisch Centrum Haaglanden (NL), Reinier de Graaf Hospital (NL), Spaarne Ziekenhuis (NL), Northwest Hospital (US), St. Antonius Ziekenhuis (NL), Ikazia Ziekenhuis (NL), OLVG (NL), Amphia Ziekenhuis (NL), Diakonessenhuis hospital (NL), Amsterdam Neuroscience (NL), Maasstad Ziekenhuis (NL), Martini Ziekenhuis (NL), Canisius-Wilhelmina Ziekenhuis (NL), VieCuri Medisch Centrum (NL), Ziekenhuis Gelderse Vallei (NL), Zuyderland Medisch Centrum (NL), IJsselland Ziekenhuis (NL), Haga Hospital (NL), Ziekenhuis Groep Twente (NL), Univé (Netherlands) (NL), Admiraal De Ruyter Ziekenhuis (NL), Amsterdam University Medical Centers (NL), Laurentius Ziekenhuis (NL), Chung-Ang University Hospital (KR), Jeroen Bosch Ziekenhuis (NL), Spaarne Gasthuis (NL)
Amsterdam University Medical Centers, Universiteit van Amsterdam, Universiteit Utrecht, Radboud Universiteit, Technische Universiteit Eindhoven, Erasmus Medisch Centrum, Universitair Medisch Centrum Utrecht, Maastricht Universitair Medisch Centrum
Good health and well-being
Openalex Percentile: Top 63%
Respiratory Support and Mechanisms
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.