Portal de Periódicos da CAPES

Machine-learning score to predict in-hospital outcomes in patients hospitalized in Cardiac Intensive Care Unit

2024; Oxford University Press; Linguagem: Inglês

10.1093/ehjdh/ztae098

2634-3916

Orianne Weizman, Kenza Hamzi, Patrick Henry, Guillaume Schurtz, Marie Hauguel Moreau, Antonin Trimaille, Marc Bédossa, Jean-Claude Dib, Sabir Attou, Tanissia Boukertouta, Franck Boccara, Thibaut Pommier, Pascal Lim, Thomas Bochaton, Damien Millischer, Benoît Mérat, Fabien Picard, Nissim Grinberg, David Sulman, Bastien Pasdeloup, Yassine El Ouahidi, Trecy Gonçalves, Éric Vicaut, Jean‐Guillaume Dillinger, Solenn Toupin, Théo Pezel, Victor Aboyans, Emeric Albert, Franck Albert, Sean Alvain, Nabil Amri, Stéphane Andrieu, Sabir Attou, Simon Auvray, Sonia Azzakani, Ruben Azencot, Marc Bédossa, Franck Boccara, Claude Boccara, Thomas Bochaton, Eric Bonnefoy‐Cudraz, Guillaume Bonnet, Guillaume Bonnet, Nabil Bouali, Océane Bouchot, Claire Bouleti, Tanissia Boukertouta, Jean Baptiste Brette, Marjorie Canu, Aurès Chaïb, C. Charbonnel, Anne Solène Chaussade, Alexandre Coppens, Yves Cottin, Arthur Darmon, Elena De Angelis, Clément Delmas, Laura Delsarte, Antoine Deney, Jean Claude Dib, Jean‐Guillaume Dillinger, Clémence Docq, Valentin Dupasquier, Meyer Elbaz, Antony El Hadad, Amine El Ouahidi, Nacim Ezzouhairi, Julien Fabre, Damien Fard, Charles Fauvel, Édouard Gerbaud, Martine Gilard, Marc Goralski, Nissim Grinberg, Alain Grentzinger, Marie Hauguel Moreau, Patrick Henry, Fabien Huet, Thomas Landemaine, Benoît Lattuca, Léo Lemarchand, Thomas Levasseur, Pascal Lim, Laura Maitre Ballesteros, Nicolas Mansencal, B. De Sainte Marie, D. Rodríguez Martínez, Benoît Mérat, Christophe Meune, Damien Millischer, Thomas Moine, Pascal Nhan, Nathalie Noirclerc, Patrick Ohlmann, Théo Pezel, Fabien Picard, Nicolas Piliero, Thibaut Pommier, Étienne Puymirat, Arthur Ramonatxo, Reza Rossanaly Vasram, François Roubille, Vincent Roule, Guillaume Schurtz, Mathilde Stevenard, David Sulman, Fédérico Swedsky, Victoria Tea, Eugénie Thevenet, Christophe Thuaire, Antonin Trimaille, Christophe Tron, Guillaume Viboud, Dominique Yomi, Cyril Zakine,

Artificial Intelligence in Healthcare and Education

Abstract Background Although some scores based on traditional statistical methods are available for risk stratification in patients hospitalized in cardiac intensive care units (CICU), the interest of machine learning (ML) methods for risk stratification in this field is not well established. We aimed to build an ML-model to predict in-hospital major adverse events (MAE) in patients hospitalized in CICU. Methods In April 2021, a French national prospective multicenter study involving 39 centers included all consecutive patients admitted to CICU. The primary outcome was in-hospital MAE, including death, resuscitated cardiac arrest or cardiogenic shock. Using 31 randomly assigned centers as an index cohort (divided into training and testing sets), several ML-models were evaluated to predict in-hospital MAE. The 8 remaining centers were used as an external validation cohort. Results Among 1,499 consecutive patients included (aged 64±15 years, 70% male), 67 had in-hospital MAE (4.3%). Out of 28 clinical, biological, ECG, and echocardiographic variables, seven were selected to predict MAE in the training set (N=844). Boosted cost-sensitive C5.0 technique showed the best performance compared with other ML-methods (receiver-operator-characteristic area-under-the-curve [AUROC]=0.90, precision-recall (PR)-AUC=0.57, F1 score=0.5). Our ML score showed a better performance than existing scores (AUROC: ML score=0.90 vs TIMI-score: 0.56, GRACE-score: 0.52, ACUTE-HF-score: 0.65; all p<0.05). ML score also showed excellent performance in the external cohort (AUROC=0.88). Conclusions This new ML score is the first to demonstrate improved performance in predicting in-hospital outcomes over existing scores in patients admitted to the ICU, based on seven simple and rapid clinical and echocardiographic variables. Trial Registration ClinicalTrials.gov Identifier: NCT05063097