Non-Invasive Ventilation Failure in Pediatric ICU: A Machine Learning Driven Prediction
Maria Vittoria Chiaruttini,
Giulia Lorenzoni,
Marco Daverio,
Luca Marchetto,
Francesca Izzo,
Giovanna Chidini,
Enzo Picconi,
Claudio Nettuno,
Elisa Zanonato,
Raffaella Sagredini,
Emanuele Rossetti,
Maria Cristina Mondardini,
Corrado Cecchetti,
Pasquale Vitale,
Nicola Alaimo,
Denise Colosimo,
Francesco Sacco,
Giulia Genoni,
Daniela Perrotta,
Camilla Micalizzi,
Silvia Moggia,
Giosuè Chisari,
Immacolata Rulli,
Andrea Wolfler,
Angela Amigoni,
Dario Gregori
Affiliations
Maria Vittoria Chiaruttini
Unit of Biostatistics, Epidemiology and Public Health, Department of Cardiac, Thoracic, and Vascular Sciences and Public Health, University of Padova, Via Loredan 18, 35131 Padova, Italy
Giulia Lorenzoni
Unit of Biostatistics, Epidemiology and Public Health, Department of Cardiac, Thoracic, and Vascular Sciences and Public Health, University of Padova, Via Loredan 18, 35131 Padova, Italy
Marco Daverio
Pediatric Intensive Care Unit, Department of Women’s and Children’s Health, University Hospital of Padova, Via Giustiniani 3, 35128 Padova, Italy
Luca Marchetto
Pediatric Intensive Care Unit, Department of Women’s and Children’s Health, University Hospital of Padova, Via Giustiniani 3, 35128 Padova, Italy
Francesca Izzo
Pediatric Intensive Care Unit, Buzzi Children’s Hospital, Via Lodovico Castelvetro 32, 20154 Milan, Italy
Giovanna Chidini
Department of Anesthesia Resuscitation Emergency Care, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico di Milano, Via Francesco Sforza 35, 20122 Milan, Italy
Enzo Picconi
Pediatric Intensive Care Unit, Pediatric Trauma Center, Fondazione IRCCS Policlinico Universitario “A. Gemelli”, Largo Agostino Gemelli 8, 00136 Rome, Italy
Claudio Nettuno
Anaesthesia and Pediatric Resuscitation, AOU Alessandria, SS Antonio e Biagio e Cesare Arrigo Hospital, Spalto Marengo 43, 15121 Alessandria, Italy
Elisa Zanonato
Pediatric Intensive Care Unit, San Bortolo Hospital, Viale Ferdinando Rodolfi 37, 36100 Vicenza, Italy
Raffaella Sagredini
Anesthesia and Resuscitation Unit, IRCCS Burlo Garofolo, Via dell’Istria 65, 34137 Trieste, Italy
Emanuele Rossetti
Anaesthesia, Emergency and Pediatric Intensive Care Unit, Bambino Gesu’ Children Hospital IRCCS, Piazza di Sant’Onofrio 4, 00165 Rome, Italy
Maria Cristina Mondardini
IRCCS AOU di Bologna, Via Giuseppe Massarenti 9, 40138 Bologna, Italy
Corrado Cecchetti
Department of Emergency Acceptance, Bambino Gesù Children’s Hospital, Piazza di Sant’Onofrio 4, 00165 Rome, Italy
Pasquale Vitale
Pediatric and Neonatal Intensive Care Unit, Children’s Hospital Regina Margherita, Piazza Polonia 94, 10126 Turin, Italy
Nicola Alaimo
ARNAS G. di Cristina Hospital, 90127 Palermo, Italy
Denise Colosimo
Pediatric Intensive Care Unit, Children’s Hospital Meyer, IRCCS, Viale Gaetano Pieraccini 24, 50139 Florence, Italy
Francesco Sacco
Paediatric Intensive Care Unit, Azienda Ospedaliera Universitaria Integrata di Verona, Piazzale Aristide Stefani 1, 37126 Verona, Italy
Giulia Genoni
Neonatal and Pediatric Intensive Care Unit, Maggiore della Carità University Hospital, L.go Bellini, 28100 Novara, Italy
Pediatric and Neonatal Intensive Care Unit, IRCCS G Gaslini, Via Gerolamo Gaslini 5, 16147 Genoa, Italy
Silvia Moggia
Pediatric Intensive Care Unit, AORN Santobono-Pausilipon, Via della Croce Rossa 8, 80122 Naples, Italy
Giosuè Chisari
UOSD Pediatric Resuscitation, ARNAS Garibaldi PO Nesima, Piazza Santa Maria di Gesù 5, 95124 Catania, Italy
Immacolata Rulli
UOC Neonatal Pathology and TIN, AOU G MARTINO, Via Consolare Valeria 1, 98124 Messina, Italy
Andrea Wolfler
Department of Emergency, Division of Anesthesia IRCCS G Gaslini, Via Gerolamo Gaslini 5, 16147 Genoa, Italy
Angela Amigoni
Pediatric Intensive Care Unit, Department of Women’s and Children’s Health, University Hospital of Padova, Via Giustiniani 3, 35128 Padova, Italy
Dario Gregori
Unit of Biostatistics, Epidemiology and Public Health, Department of Cardiac, Thoracic, and Vascular Sciences and Public Health, University of Padova, Via Loredan 18, 35131 Padova, Italy
Background/Objectives: Non-invasive ventilation (NIV) has emerged as a possible first-step treatment to avoid invasive intubation in pediatric intensive care units (PICUs) due to its advantages in reducing intubation-associated risks. However, the timely identification of NIV failure is crucial to prevent adverse outcomes. This study aims to identify predictors of first-attempt NIV failure in PICU patients by testing various machine learning techniques and comparing their predictive abilities. Methods: Data were sourced from the TIPNet registry, which comprised patients admitted to 23 Italian Paediatric Intensive Care Units (PICUs). We selected patients between January 2010 and January 2024 who received non-invasive ventilation (NIV) as their initial approach to respiratory support. The study aimed to develop a predictive model for NIV failure, selecting the best Machine Learning technique, including Generalized Linear Models, Random Forest, Extreme Gradient Boosting, and Neural Networks. Additionally, an ensemble approach was implemented. Model performances were measured using sensitivity, specificity, AUROC, and predictive values. Moreover, the model calibration was evaluated. Results: Out of 43,794 records, 1861 admissions met the inclusion criteria, with 678 complete cases and 97 NIV failures. The RF model demonstrated the highest AUROC and sensitivity equal to 0.83 (0.64, 0.94). Base excess, weight, age, systolic blood pressure, and fraction of inspired oxygen were identified as the most predictive features. A check for model calibration ensured the model’s reliability in predicting NIV failure probabilities. Conclusions: This study identified highly sensitive models for predicting NIV failure in PICU patients, with RF as a robust option.
