COVID-19 ICU mortality prediction: a machine learning approach using SuperLearner algorithm

Springer Science and Business Media LLC - 2021
Giulia Lorenzoni1, Nicolò Sella2, Annalisa Boscolo3, Danila Azzolina1, Patrizia Bartolotta1, Laura Pasin3, Tommaso Pettenuzzo3, Alessandro De Cassai3, Fabio Baratto4, Fabio Toffoletto5, Silvia De Rosa6, Giorgio Fullin7, Mario Peta8, Paolo Rosi9, Enrico Polati10, Alberto Zanella11,12, Giacomo Grasselli12,11, Antonio Artigas12,11, Paolo Navalesi3,2, Darío Gregori1, Martina Tocco, Chiara Pretto, Enrico Tamburini10, Davide Fregolent, Pier Francesco Pirelli, Davide Marchesin, Matteo Perona, Nicola Franchetti, Michele Della Paolera, Caterina Simoni, Tatiana Falcioni, Alessandra Tresin, Chiara Schiavolin, Aldo Schiavi, Sonila Vathi, Daria Sartori, Alice Sorgato, Elisa Pistollato, Federico Linassi, Sara Gianoli, Silvia Gaspari6, Francesco Gruppo, Alessandra Maggiolo, Elena Giurisato, Elisa Furlani, Alvise Calore, Eugenio Serra, Demetrio Pittarello, Ivo Tiberio, Ottavia Bond, Elisa Michieletto, Luisa Muraro, Arianna Peralta, Paolo Persona2,3, Enrico Petranzan10, Francesco Zarantonello, Alessandro Di Graziano3, E. Piasentini, Lorenzo Bernardi1, Roberto Pianon, D Mazzon, Daniele Poole, Flavio Badii, Enrico Bosco3, Moreno Agostini, Paride Trevisiol, Antonio Farnia12,11, Lorella Altafini, Mauro Antonio Calò, Marco Meggiolaro, Francesco Lazzari, Ivan Martinello, Francesco Papaccio, Guido Di Gregorio, Alfeo Bonato, Camilla Sgarabotto, Francesco Montacciani, Parnigotto Alessandra, Giuseppe Gagliardi, Gioconda Ferraro, Luigi Ongaro, Marco Baiocchi, Vinicio Danzi, Paolo Zanatta2,3, Katia Donadello, Leonardo Gottin, Ezio Sinigaglia, Alessandra Da Ros, Simonetta Marchiotto, Silvia Bassanini6, Massimo Zamperini, Ivan Daroui, Walter Mosaner
1Unit of Biostatistics, Epidemiology and Public Health, Department of Cardiac Thoracic Vascular Sciences and Public Health, University of Padova, Padova, Italy
2Department of Medicine (DIMED), Padova University Hospital, Padova, Italy
3Institute of Anaesthesia and Intensive Care Unit, Padova University Hospital, Padova, Italy
4Anaesthesia and Intensive Care Unit, Ospedale Riuniti Padova Sud, Schiavonia, Italy
5Anaesthesia and Intensive Care Unit, Ospedale di San Donà di Piave e Jesolo, San Donà di Piave, Italy
6Anaesthesia and Critical Care Unit, San Bortolo Hospital, Vicenza, Italy
7Anaesthesia and Intensive Care Unit, Ospedale Dell’Angelo, AULSS 3 Serenissima, Mestre, Italy
8Anaesthesia and Intensive Care Unit, Ospedale Ca’ Foncello, AULSS 2 Marca Trevigiana, Treviso, Italy
9Emergency Medical Services, Regional Department, AULSS 3, Venice, Italy
10Anaesthesia and Intensive Care Unit B, Department of Surgery, Dentistry, Gynaecology and Pediatrics, University of Verona, AOUI - University Hospital Integrated Trust, Verona, Italy
11Anaesthesia and Critical Care, Department of Pathophysiology and Transplantation, University of Milan, Milan, Italy
12Department of Anaesthesia, Intensive Care and Emergency Medicine, Fondazione IRCCS Ca’ Granda-Ospedale Maggiore Policlinico, Milan, Italy

Tóm tắt

Abstract Background Since the beginning of coronavirus disease 2019 (COVID-19), the development of predictive models has sparked relevant interest due to the initial lack of knowledge about diagnosis, treatment, and prognosis. The present study aimed at developing a model, through a machine learning approach, to predict intensive care unit (ICU) mortality in COVID-19 patients based on predefined clinical parameters. Results Observational multicenter cohort study. All COVID-19 adult patients admitted to 25 ICUs belonging to the VENETO ICU network (February 28th 2020-april 4th 2021) were enrolled. Patients admitted to the ICUs before 4th March 2021 were used for model training (“training set”), while patients admitted after the 5th of March 2021 were used for external validation (“test set 1”). A further group of patients (“test set 2”), admitted to the ICU of IRCCS Ca’ Granda Ospedale Maggiore Policlinico of Milan, was used for external validation. A SuperLearner machine learning algorithm was applied for model development, and both internal and external validation was performed. Clinical variables available for the model were (i) age, gender, sequential organ failure assessment score, Charlson Comorbidity Index score (not adjusted for age), Palliative Performance Score; (ii) need of invasive mechanical ventilation, non-invasive mechanical ventilation, O2 therapy, vasoactive agents, extracorporeal membrane oxygenation, continuous venous-venous hemofiltration, tracheostomy, re-intubation, prone position during ICU stay; and (iii) re-admission in ICU. One thousand two hundred ninety-three (80%) patients were included in the “training set”, while 124 (8%) and 199 (12%) patients were included in the “test set 1” and “test set 2,” respectively. Three different predictive models were developed. Each model included different sets of clinical variables. The three models showed similar predictive performances, with a training balanced accuracy that ranged between 0.72 and 0.90, while the cross-validation performance ranged from 0.75 to 0.85. Age was the leading predictor for all the considered models. Conclusions Our study provides a useful and reliable tool, through a machine learning approach, for predicting ICU mortality in COVID-19 patients. In all the estimated models, age was the variable showing the most important impact on mortality.

Từ khóa


Tài liệu tham khảo

Wynants L, Calster BV, Collins GS, Riley RD, Heinze G, Schuit E et al (2020) Prediction models for diagnosis and prognosis of covid-19: systematic review and critical appraisal. BMJ 369:m1328

Tonetti T, Grasselli G, Zanella A, Pizzilli G, Fumagalli R, Piva S et al (2020) Use of critical care resources during the first 2 weeks (February 24–March 8, 2020) of the Covid-19 outbreak in Italy. Ann Intensive Care 10:1–8

Lorenzoni G, Lanera C, Azzolina D, Berchialla P, Gregori D, Gregori D et al (2020) Is a more aggressive COVID-19 case detection approach mitigating the burden on ICUs? Some reflections from Italy. Crit Care 24:175

Cavarretta E, Biondi-Zoccai G, Frati G, Versaci F (2020) Veneto’s successful lesson for a world shocked by COVID-19: Think globally and act locally. J Cardiothor Vasc Anesth 34:2346–2348

Pasin L, Sella N, Correale C, Boscolo A, Rosi P, Saia M et al (2020) Regional COVID-19 network for coordination of SARS-CoV-2 outbreak in Veneto, Italy. J Cardiothor Vasc Anesth 34:2341–2345

Boscolo A, Sella N, Lorenzoni G, Pettenuzzo T, Pasin L, Pretto C et al (2021) Static compliance and driving pressure are associated with ICU mortality in intubated COVID-19 ARDS. Crit Care 2021;25(1):263

Van der Laan MJ, Polley EC, Hubbard AE (2007) Super Learner. Stat Appl Genet Mol Biol 6:25

Greenwell BM, Boehmke BC, McCarthy AJ (2018) A simple and effective model-based variable importance measure. arXiv:1805.04755[cs.stat] 2018.

Tezza F, Lorenzoni G, Azzolina D, Barbar S, Leone LAC, Gregori D (2021) Predicting in-hospital mortality of patients with COVID-19 using machine learning techniques. J Personal Med 11:343

Calabrese F, Pezzuto F, Fortarezza F, Boscolo A, Lunardi F et al (2021) Machine learning-based analysis of Alveolar and Vascular injury in SARS-CoV-2 acute respiratory failure. J Pathol 254:173–184

Karagiannidis C, Mostert C, Hentschker C, Voshaar T, Malzahn J, Schillinger G et al (2020) Case characteristics, resource use, and outcomes of 10 021 patients with COVID-19 admitted to 920 German hospitals: an observational study. Lancet Respir Med 8(9):853–862

Vaschetto R, Barone-Adesi F, Racca F, Pissaia C, Maestrone C, Colombo D et al (2021) Outcomes of COVID-19 patients treated with continuous positive airway pressure outside the intensive care unit. ERJ Open Res 7(1):00541–02020

Grasselli G, Zangrillo A, Zanella A, Antonelli M, Cabrini L, Castelli A et al (2020) Baseline characteristics and outcomes of 1591 patients infected with SARS-CoV-2 admitted to ICUs of the Lombardy Region, Italy. JAMA 323(16):1574–1581

Yang X, Yu Y, Xu J, Shu H, Xia J, Liu H et al (2020) Clinical course and outcomes of critically Ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study. Lancet Respir Med 8(5):475–481

Alhazzani W, Møller MH, Arabi YM, Loeb M, Gong MN, Fan E et al (2020) Surviving Sepsis Campaign: Guidelines on the Management of Critically Ill Adults with Coronavirus Disease 2019 (COVID-19). Crit Care Med 48(6):e440–e469

Thông tin
Thông tin xuất bản

Springer Science and Business Media LLC

Thông tin tác giả