Compartir
array:24 [ "pii" => "S0210569123002085" "issn" => "02105691" "doi" => "10.1016/j.medin.2023.06.012" "estado" => "S300" "fechaPublicacion" => "2024-01-01" "aid" => "1907" "copyright" => "Elsevier España, S.L.U. y SEMICYUC" "copyrightAnyo" => "2023" "documento" => "article" "crossmark" => 1 "subdocumento" => "fla" "cita" => "Med Intensiva. 2024;48:3-13" "abierto" => array:3 [ "ES" => true "ES2" => true "LATM" => true ] "gratuito" => true "lecturas" => array:1 [ "total" => 0 ] "Traduccion" => array:1 [ "en" => array:19 [ "pii" => "S2173572723001303" "issn" => "21735727" "doi" => "10.1016/j.medine.2023.07.009" "estado" => "S300" "fechaPublicacion" => "2024-01-01" "aid" => "1907" "copyright" => "Elsevier España, S.L.U. and SEMICYUC" "documento" => "article" "crossmark" => 1 "subdocumento" => "fla" "cita" => "Med Intensiva. 2024;48:3-13" "abierto" => array:3 [ "ES" => true "ES2" => true "LATM" => true ] "gratuito" => true "lecturas" => array:1 [ "total" => 0 ] "en" => array:13 [ "idiomaDefecto" => true "cabecera" => "<span class="elsevierStyleTextfn">Original article</span>" "titulo" => "Predictors of mechanical ventilation and mortality in critically ill patients with COVID-19 pneumonia" "tienePdf" => "en" "tieneTextoCompleto" => "en" "tieneResumen" => array:2 [ 0 => "en" 1 => "es" ] "paginas" => array:1 [ 0 => array:2 [ "paginaInicial" => "3" "paginaFinal" => "13" ] ] "titulosAlternativos" => array:1 [ "es" => array:1 [ "titulo" => "Predictores de ventilación mecánica y mortalidad en pacientes críticos con neumonía por COVID-19" ] ] "contieneResumen" => array:2 [ "en" => true "es" => true ] "contieneTextoCompleto" => array:1 [ "en" => true ] "contienePdf" => array:1 [ "en" => true ] "resumenGrafico" => array:2 [ "original" => 0 "multimedia" => array:8 [ "identificador" => "fig0015" "etiqueta" => "Figure 3" "tipo" => "MULTIMEDIAFIGURA" "mostrarFloat" => true "mostrarDisplay" => false "figura" => array:1 [ 0 => array:4 [ "imagen" => "gr3.jpeg" "Alto" => 1594 "Ancho" => 3425 "Tamanyo" => 463599 ] ] "detalles" => array:1 [ 0 => array:3 [ "identificador" => "at0085" "detalle" => "Figure " "rol" => "short" ] ] "descripcion" => array:1 [ "en" => "<p id="spar0015" class="elsevierStyleSimplePara elsevierViewall">ICU Mortality Tree Predictors. The predictors appear in different branches attending to their significance in the predictive model. Values in bold letters represent the registries per branch. Values in red bold letters represent the percentage of registries with positive outcome. The variable named as “DAYS_SIMPTONS_ADMISION” is related with the number of days from first symptoms to ICU admission. The variable “linf_total”, is related to lymphocyte count per mm3. The variable named as “dosis_equiv_mpred_5d” is related with the corticosteroid dose, during the first five days of admission (mg of equivalent methylprednisolone dose). The variable named as “bbTot” is related with the total levels of bilirubin in blood. The variable names as “ldh” is related to the lactate dehydrogenase serum level. The variable DAYS_UNTIL_O2 is related to the number of days until the patient requires O2.</p>" ] ] ] "autores" => array:1 [ 0 => array:2 [ "autoresLista" => "Sergio Muñoz Lezcano, Miguel Ángel Armengol de la Hoz, Alberto Corbi, Fernando López, Miguel Sánchez García, Antonio Nuñez Reiz, Tomás Fariña González, Viktor Yordanov Zlatkov" "autores" => array:8 [ 0 => array:2 [ "nombre" => "Sergio" "apellidos" => "Muñoz Lezcano" ] 1 => array:2 [ "nombre" => "Miguel Ángel" "apellidos" => "Armengol de la Hoz" ] 2 => array:2 [ "nombre" => "Alberto" "apellidos" => "Corbi" ] 3 => array:2 [ "nombre" => "Fernando" "apellidos" => "López" ] 4 => array:2 [ "nombre" => "Miguel Sánchez" "apellidos" => "García" ] 5 => array:2 [ "nombre" => "Antonio Nuñez" "apellidos" => "Reiz" ] 6 => array:2 [ "nombre" => "Tomás Fariña" "apellidos" => "González" ] 7 => array:2 [ "nombre" => "Viktor Yordanov" "apellidos" => "Zlatkov" ] ] ] ] ] "idiomaDefecto" => "en" "Traduccion" => array:1 [ "en" => array:9 [ "pii" => "S0210569123002085" "doi" => "10.1016/j.medin.2023.06.012" "estado" => "S300" "subdocumento" => "" "abierto" => array:3 [ "ES" => true "ES2" => true "LATM" => true ] "gratuito" => true "lecturas" => array:1 [ "total" => 0 ] "idiomaDefecto" => "en" "EPUB" => "https://multimedia.elsevier.es/PublicationsMultimediaV1/item/epub/S0210569123002085?idApp=WMIE" ] ] "EPUB" => "https://multimedia.elsevier.es/PublicationsMultimediaV1/item/epub/S2173572723001303?idApp=WMIE" "url" => "/21735727/0000004800000001/v1_202401020427/S2173572723001303/v1_202401020427/en/main.assets" ] ] "itemSiguiente" => array:18 [ "pii" => "S0210569123001833" "issn" => "02105691" "doi" => "10.1016/j.medin.2023.06.003" "estado" => "S300" "fechaPublicacion" => "2024-01-01" "aid" => "1898" "copyright" => "Elsevier España, S.L.U. y SEMICYUC" "documento" => "article" "crossmark" => 1 "subdocumento" => "fla" "cita" => "Med Intensiva. 2024;48:14-22" "abierto" => array:3 [ "ES" => true "ES2" => true "LATM" => true ] "gratuito" => true "lecturas" => array:1 [ "total" => 0 ] "en" => array:13 [ "idiomaDefecto" => true "cabecera" => "<span class="elsevierStyleTextfn">Original article</span>" "titulo" => "Improvement in intubation success during COVID-19 pandemic with a simple and low-cost intervention: A quasi-experimental study" "tienePdf" => "en" "tieneTextoCompleto" => "en" "tieneResumen" => array:2 [ 0 => "en" 1 => "es" ] "paginas" => array:1 [ 0 => array:2 [ "paginaInicial" => "14" "paginaFinal" => "22" ] ] "titulosAlternativos" => array:1 [ "es" => array:1 [ "titulo" => "Mejoría del éxito en la intubación durante la pandemia covid-19 con una intervención simple y de bajo costo: estudio cuasi-experimental" ] ] "contieneResumen" => array:2 [ "en" => true "es" => true ] "contieneTextoCompleto" => array:1 [ "en" => true ] "contienePdf" => array:1 [ "en" => true ] "resumenGrafico" => array:2 [ "original" => 0 "multimedia" => array:8 [ "identificador" => "fig0010" "etiqueta" => "Figure 2" "tipo" => "MULTIMEDIAFIGURA" "mostrarFloat" => true "mostrarDisplay" => false "figura" => array:1 [ 0 => array:4 [ "imagen" => "gr2.jpeg" "Alto" => 1140 "Ancho" => 1675 "Tamanyo" => 112365 ] ] "detalles" => array:1 [ 0 => array:3 [ "identificador" => "at0040" "detalle" => "Figure " "rol" => "short" ] ] "descripcion" => array:1 [ "en" => "<p id="spar0010" class="elsevierStyleSimplePara elsevierViewall">Percentage of intubation on the first attempt, before and after the intervention, according to the experience of the physician who performed the procedure.</p>" ] ] ] "autores" => array:1 [ 0 => array:2 [ "autoresLista" => "Guillermo Jesús Mauro, Gustavo Armando, Lorena Natalia Cabillón, Santiago Tomás Benitez, Sabrina Mogliani, Amanda Roldan, Marisol Vilca, Ricardo Rollie, Gustavo Martins" "autores" => array:9 [ 0 => array:2 [ "nombre" => "Guillermo Jesús" "apellidos" => "Mauro" ] 1 => array:2 [ "nombre" => "Gustavo" "apellidos" => "Armando" ] 2 => array:2 [ "nombre" => "Lorena Natalia" "apellidos" => "Cabillón" ] 3 => array:2 [ "nombre" => "Santiago Tomás" "apellidos" => "Benitez" ] 4 => array:2 [ "nombre" => "Sabrina" "apellidos" => "Mogliani" ] 5 => array:2 [ "nombre" => "Amanda" "apellidos" => "Roldan" ] 6 => array:2 [ "nombre" => "Marisol" "apellidos" => "Vilca" ] 7 => array:2 [ "nombre" => "Ricardo" "apellidos" => "Rollie" ] 8 => array:2 [ "nombre" => "Gustavo" "apellidos" => "Martins" ] ] ] ] ] "idiomaDefecto" => "en" "EPUB" => "https://multimedia.elsevier.es/PublicationsMultimediaV1/item/epub/S0210569123001833?idApp=WMIE" "url" => "/02105691/0000004800000001/v1_202401020431/S0210569123001833/v1_202401020431/en/main.assets" ] "itemAnterior" => array:19 [ "pii" => "S0210569123002139" "issn" => "02105691" "doi" => "10.1016/j.medin.2023.07.001" "estado" => "S300" "fechaPublicacion" => "2024-01-01" "aid" => "1910" "copyright" => "Elsevier España, S.L.U. y SEMICYUC" "documento" => "simple-article" "crossmark" => 1 "subdocumento" => "edi" "cita" => "Med Intensiva. 2024;48:1-2" "abierto" => array:3 [ "ES" => true "ES2" => true "LATM" => true ] "gratuito" => true "lecturas" => array:1 [ "total" => 0 ] "es" => array:10 [ "idiomaDefecto" => true "cabecera" => "<span class="elsevierStyleTextfn">Editorial</span>" "titulo" => "Análisis avanzado de datos y medicina intensiva" "tienePdf" => "es" "tieneTextoCompleto" => "es" "paginas" => array:1 [ 0 => array:2 [ "paginaInicial" => "1" "paginaFinal" => "2" ] ] "titulosAlternativos" => array:1 [ "en" => array:1 [ "titulo" => "Advanced data analysis and intensive care medicine" ] ] "contieneTextoCompleto" => array:1 [ "es" => true ] "contienePdf" => array:1 [ "es" => true ] "autores" => array:1 [ 0 => array:2 [ "autoresLista" => "Federico Gordo Vidal, Natalia Gordo Herrera" "autores" => array:2 [ 0 => array:2 [ "nombre" => "Federico" "apellidos" => "Gordo Vidal" ] 1 => array:2 [ "nombre" => "Natalia" "apellidos" => "Gordo Herrera" ] ] ] ] ] "idiomaDefecto" => "es" "Traduccion" => array:1 [ "en" => array:9 [ "pii" => "S2173572723001352" "doi" => "10.1016/j.medine.2023.07.012" "estado" => "S300" "subdocumento" => "" "abierto" => array:3 [ "ES" => true "ES2" => true "LATM" => true ] "gratuito" => true "lecturas" => array:1 [ "total" => 0 ] "idiomaDefecto" => "en" "EPUB" => "https://multimedia.elsevier.es/PublicationsMultimediaV1/item/epub/S2173572723001352?idApp=WMIE" ] ] "EPUB" => "https://multimedia.elsevier.es/PublicationsMultimediaV1/item/epub/S0210569123002139?idApp=WMIE" "url" => "/02105691/0000004800000001/v1_202401020431/S0210569123002139/v1_202401020431/es/main.assets" ] "en" => array:22 [ "idiomaDefecto" => true "cabecera" => "<span class="elsevierStyleTextfn">Original article</span>" "titulo" => "Predictors of mechanical ventilation and mortality in critically ill patients with COVID-19 pneumonia" "tieneTextoCompleto" => true "paginas" => array:1 [ 0 => array:2 [ "paginaInicial" => "3" "paginaFinal" => "13" ] ] "autores" => array:1 [ 0 => array:4 [ "autoresLista" => "Sergio Muñoz Lezcano, Miguel Ángel Armengol de la Hoz, Alberto Corbi, Fernando López, Miguel Sánchez García, Antonio Nuñez Reiz, Tomás Fariña González, Viktor Yordanov Zlatkov" "autores" => array:8 [ 0 => array:4 [ "nombre" => "Sergio" "apellidos" => "Muñoz Lezcano" "email" => array:1 [ 0 => "smunozle@gmail.com" ] "referencia" => array:3 [ 0 => array:2 [ "etiqueta" => "<span class="elsevierStyleSup">a</span>" "identificador" => "aff0005" ] 1 => array:2 [ "etiqueta" => "<span class="elsevierStyleSup">1</span>" "identificador" => "fn0005" ] 2 => array:2 [ "etiqueta" => "*" "identificador" => "cor0005" ] ] ] 1 => array:3 [ "nombre" => "Miguel Ángel" "apellidos" => "Armengol de la Hoz" "referencia" => array:2 [ 0 => array:2 [ "etiqueta" => "<span class="elsevierStyleSup">b</span>" "identificador" => "aff0010" ] 1 => array:2 [ "etiqueta" => "<span class="elsevierStyleSup">1</span>" "identificador" => "fn0005" ] ] ] 2 => array:3 [ "nombre" => "Alberto" "apellidos" => "Corbi" "referencia" => array:1 [ 0 => array:2 [ "etiqueta" => "<span class="elsevierStyleSup">c</span>" "identificador" => "aff0015" ] ] ] 3 => array:3 [ "nombre" => "Fernando" "apellidos" => "López" "referencia" => array:1 [ 0 => array:2 [ "etiqueta" => "<span class="elsevierStyleSup">d</span>" "identificador" => "aff0020" ] ] ] 4 => array:3 [ "nombre" => "Miguel Sánchez" "apellidos" => "García" "referencia" => array:1 [ 0 => array:2 [ "etiqueta" => "<span class="elsevierStyleSup">e</span>" "identificador" => "aff0025" ] ] ] 5 => array:3 [ "nombre" => "Antonio Nuñez" "apellidos" => "Reiz" "referencia" => array:1 [ 0 => array:2 [ "etiqueta" => "<span class="elsevierStyleSup">f</span>" "identificador" => "aff0030" ] ] ] 6 => array:3 [ "nombre" => "Tomás Fariña" "apellidos" => "González" "referencia" => array:1 [ 0 => array:2 [ "etiqueta" => "<span class="elsevierStyleSup">g</span>" "identificador" => "aff0035" ] ] ] 7 => array:3 [ "nombre" => "Viktor Yordanov" "apellidos" => "Zlatkov" "referencia" => array:1 [ 0 => array:2 [ "etiqueta" => "<span class="elsevierStyleSup">a</span>" "identificador" => "aff0005" ] ] ] ] "afiliaciones" => array:7 [ 0 => array:3 [ "entidad" => "PhD Student of the Program in Computer Science, Universidad Internacional de La Rioja (UNIR), Avenida de La Paz, 137, 26006 Logroño, La Rioja, Spain" "etiqueta" => "a" "identificador" => "aff0005" ] 1 => array:3 [ "entidad" => "Big Data Department, PMC-FPS, Consejería de Salud y Consumo, Junta de Andalucía, Spain" "etiqueta" => "b" "identificador" => "aff0010" ] 2 => array:3 [ "entidad" => "Research Institute for Innovation & Technology in Education (iTED), Universidad Internacional de La Rioja (UNIR), Avenida de La Paz, 137, 26006 Logroño, La Rioja, Spain" "etiqueta" => "c" "identificador" => "aff0015" ] 3 => array:3 [ "entidad" => "Mathematical Analysis and Applied Mathematics Department, Faculty of Mathematics. Universidad Complutense de Madrid, Spain" "etiqueta" => "d" "identificador" => "aff0020" ] 4 => array:3 [ "entidad" => "Critical Care Department, Hospital Clínico San Carlos, Martín Lagos s/n, 28040 Madrid, Spain" "etiqueta" => "e" "identificador" => "aff0025" ] 5 => array:3 [ "entidad" => "Critical Care Department, Hospital Universitario Clínico San Carlos, Martín Lagos s/n, 28040 Madrid, Spain" "etiqueta" => "f" "identificador" => "aff0030" ] 6 => array:3 [ "entidad" => "Critical Care Department, Hospital Universitario Infanta Sofía, Spain" "etiqueta" => "g" "identificador" => "aff0035" ] ] "correspondencia" => array:1 [ 0 => array:3 [ "identificador" => "cor0005" "etiqueta" => "⁎" "correspondencia" => "Corresponding author." ] ] ] ] "titulosAlternativos" => array:1 [ "es" => array:1 [ "titulo" => "Predictores de ventilación mecánica y mortalidad en pacientes críticos con neumonía por COVID-19" ] ] "resumenGrafico" => array:2 [ "original" => 0 "multimedia" => array:8 [ "identificador" => "fig0015" "etiqueta" => "Figure 3" "tipo" => "MULTIMEDIAFIGURA" "mostrarFloat" => true "mostrarDisplay" => false "figura" => array:1 [ 0 => array:4 [ "imagen" => "gr3.jpeg" "Alto" => 1594 "Ancho" => 3425 "Tamanyo" => 463757 ] ] "detalles" => array:1 [ 0 => array:3 [ "identificador" => "at0085" "detalle" => "Figure " "rol" => "short" ] ] "descripcion" => array:1 [ "en" => "<p id="spar0015" class="elsevierStyleSimplePara elsevierViewall">ICU Mortality Tree Predictors. The predictors appear in different branches attending to their significance in the predictive model. Values in bold letters represent the registries per branch. Values in red bold letters represent the percentage of registries with positive outcome. The variable named as “DAYS_SIMPTONS_ADMISION” is related with the number of days from first symptoms to ICU admission. The variable “linf_total”, is related to lymphocyte count per mm3. The variable named as “dosis_equiv_mpred_5d” is related with the corticosteroid dose, during the first five days of admission (mg of equivalent methylprednisolone dose). The variable named as “bbTot” is related with the total levels of bilirubin in blood. The variable names as “ldh” is related to the lactate dehydrogenase serum level. The variable DAYS_UNTIL_O2 is related to the number of days until the patient requires O2.</p>" ] ] ] "textoCompleto" => "<span class="elsevierStyleSections"><span id="sec0005" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0105">Introduction</span><p id="par0005" class="elsevierStylePara elsevierViewall">Invasive mechanical ventilation (IMV) is a cornerstone of organ support in severe COVID-19 patients with acute respiratory distress syndrome (ARDS). As widely experienced in ICUs during the SARS-CoV-2 pandemic, IMV frequently causes complications.<a class="elsevierStyleCrossRefs" href="#bib0005"><span class="elsevierStyleSup">1,2</span></a> Hospital services were overwhelmed not only by the surge of patients, but also by scarce human resources and equipment, lack of sufficient mechanical ventilators being probably the most relevant. In surge scenarios, appropriate triage strategies are therefore needed to allocate IMV or alternatives such as high flow nasal prongs. These strategies should be based on the knowledge and understanding of specific potential predictors<a class="elsevierStyleCrossRef" href="#bib0015"><span class="elsevierStyleSup">3</span></a> that could help clinicians to personalize decisions regarding IMV.</p><p id="par0010" class="elsevierStylePara elsevierViewall">There is still considerable controversy regarding who and when to intubate. Several recent studies have addressed the subject,<a class="elsevierStyleCrossRef" href="#bib0020"><span class="elsevierStyleSup">4</span></a> although bias cannot be excluded in observational non-randomized trials. A retrospective study suggested that early intubation and IMV is associated with favorable outcomes but included only intubated patients instead of the whole population at risk.</p><p id="par0015" class="elsevierStylePara elsevierViewall">Previous studies have identified covid-19 progression predictors including age, comorbidities, renal function, or immunodeficiency<a class="elsevierStyleCrossRef" href="#bib0025"><span class="elsevierStyleSup">5</span></a> using traditional statistical approaches, where collinearity of data cannot be ruled out. Artificial intelligence (AI) is currently being used for COVID-19 risk stratification,<a class="elsevierStyleCrossRef" href="#bib0030"><span class="elsevierStyleSup">6</span></a> studying multiple clinical features to increase effectiveness and efficiency in diagnosis, treatment, and prognosis. Self-explainable Machine learning (ML) techniques can help with risk factor selection through ranking methodologies.<a class="elsevierStyleCrossRef" href="#bib0035"><span class="elsevierStyleSup">7</span></a> In this context, the utilization of artificial intelligence (AI) holds potential in facilitating the development of a conceptual model aimed at comparing the significance of variables. This can be achieved by employing regularization models<a class="elsevierStyleCrossRef" href="#bib0040"><span class="elsevierStyleSup">8</span></a> to enhance predictor selection, followed by the implementation of the Generalized Linear Mixed-effects Model (GLMM)<a class="elsevierStyleCrossRefs" href="#bib0045"><span class="elsevierStyleSup">9–11</span></a> to construct the said conceptual model. Such an approach becomes particularly relevant when assessing and comparing outcomes across different AI models, enabling a comprehensive evaluation of variable significance. This is a novel methodology, leveraging modern machine learning techniques to provide rigorous and applicable insight into relevant clinical questions when randomized clinical trials are not feasible. From here on, in this paper, we aim to determine if potential predictors for invasive mechanical ventilation (IMV) are also determinants for mortality in COVID-19-associated acute respiratory distress syndrome (C-ARDS) while comparing the significance of variables in both cases.</p></span><span id="sec0010" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0110">Patients and methods</span><span id="sec0015" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0115">Selection and description of patients</span><p id="par0020" class="elsevierStylePara elsevierViewall">In our retrospective observational study, we have collected and curated data from our electronic medical records (EMR) from March 3rd of 2020 through February 28th of 2021. We selected patients admitted to our ICU at San Carlos Hospital (HCSC) in Madrid (<a class="elsevierStyleCrossRef" href="#fig0005">Fig. 1</a>) but were initially not mechanically ventilated. The selection of patients considered just COVID-19 pneumonia patients, incidental COVID-19 was excluded. The age range for inclusion was restricted to individuals aged 18 years or older.</p><elsevierMultimedia ident="fig0005"></elsevierMultimedia><p id="par0025" class="elsevierStylePara elsevierViewall">The database comprises hourly data points for each patient during the first five days. Afterwards, we utilized multi-stage machine learning algorithms to assess the most significant variables in predicting invasive mechanical ventilation (IMV) and <span class="elsevierStyleBold">ICU</span> mortality (<a class="elsevierStyleCrossRef" href="#fig0010">Fig. 2</a>). It is worth noting that 28-day mortality, while frequently used in large studies like RECOVERY, may not be a suitable outcome measure in COVID-19 patients due to the possibility of delayed mortality.</p><elsevierMultimedia ident="fig0010"></elsevierMultimedia><p id="par0030" class="elsevierStylePara elsevierViewall">All data were registered in our electronic medical record (ICCA Philips). A total of 12,163 longitudinal sets of hourly clinical and lab data were gathered. Longitudinal sets are grouped in clustered events associated with patients. Each entry contains demographics data, first or second wave admission, time elapsed from start of symptoms to O2 therapy and ICU admission, APACHE II score, monitoring, blood gases and therapy-related data. We discarded variables with more than 33% of missing values for consistency. We used mode imputation or mean imputation to complete missing values of the remaining variables. <a class="elsevierStyleCrossRefs" href="#tbl0005">Tables 1 and 2</a> show the predictors that were finally used for the purposes of the study.</p><elsevierMultimedia ident="tbl0005"></elsevierMultimedia><elsevierMultimedia ident="tbl0010"></elsevierMultimedia><p id="par0035" class="elsevierStylePara elsevierViewall">Data were anonymized, excluding demographic or temporal information. The study protocol was approved by the local ethics committee (approval code 22/007-E), who waived the need for informed consent due to the retrospective non-interventional nature of the study.</p></span><span id="sec0020" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0120">Methods and techniques</span><p id="par0040" class="elsevierStylePara elsevierViewall">Data collected as described above were used to fit the model<a class="elsevierStyleCrossRef" href="#bib0060"><span class="elsevierStyleSup">12</span></a> following four steps for the whole process, as shown in <a class="elsevierStyleCrossRef" href="#fig0010">Fig. 2</a>. Considering that our data involve a concatenation of longitudinal data for each patient in different events, it was necessary to identify correlations within the cluster when trying to build an accurate prediction model.<a class="elsevierStyleCrossRef" href="#bib0050"><span class="elsevierStyleSup">10</span></a></p><p id="par0045" class="elsevierStylePara elsevierViewall">The different regression approaches to select potential predictors for IMV and ICU mortality risk tested were: LASSO,<a class="elsevierStyleCrossRef" href="#bib0065"><span class="elsevierStyleSup">13</span></a> Ridge,<a class="elsevierStyleCrossRef" href="#bib0070"><span class="elsevierStyleSup">14</span></a> Elastic-net,<a class="elsevierStyleCrossRef" href="#bib0075"><span class="elsevierStyleSup">15</span></a> Boruta<a class="elsevierStyleCrossRef" href="#bib0080"><span class="elsevierStyleSup">16</span></a> and R-Part.<a class="elsevierStyleCrossRef" href="#bib0085"><span class="elsevierStyleSup">17</span></a> LASSO, Ridge and Elastic-net perform an automatic predictor selection supported by L1 and L2 regularization terms<a class="elsevierStyleCrossRef" href="#bib0090"><span class="elsevierStyleSup">18</span></a> that minimizes the risk of overfitting, reducing variance and reaching an attenuation effect over the correlation between features. Boruta<a class="elsevierStyleCrossRef" href="#bib0095"><span class="elsevierStyleSup">19</span></a> is a feature selection model based on a Random Forest algorithm that selects all the risk predictors that are relevant for classification purposes defined as <span class="elsevierStyleItalic">all-relevant problems</span>. R-Part<a class="elsevierStyleCrossRef" href="#bib0085"><span class="elsevierStyleSup">17</span></a> builds a classification model based on binary trees. R-Part <span class="elsevierStyleItalic">varImp function</span><a class="elsevierStyleCrossRef" href="#bib0100"><span class="elsevierStyleSup">20</span></a> identifies the effect of model predictors based on the loss function mean squared error. In any case, potential predictors have been analyzed and confirmed or rejected based on clinical criteria.</p><p id="par0050" class="elsevierStylePara elsevierViewall">After identifying the optimal set of potential predictors (Figure 10–14 in Supplementary material), clustering effects by patient and temporal distribution, as well as cutoff points of the significant variables and their interactions were assessed with GLMM Trees.<a class="elsevierStyleCrossRefs" href="#bib0045"><span class="elsevierStyleSup">9–11</span></a> To build these trees, we took the entire dataset into account, grouping data by patient and data charting time as random variables to fit the model.<a class="elsevierStyleCrossRef" href="#bib0060"><span class="elsevierStyleSup">12</span></a> This fitting methodology avoids both over and underfitting effects that could impact the model’s performance.<a class="elsevierStyleCrossRef" href="#bib0105"><span class="elsevierStyleSup">21</span></a> Models were implemented based on a 10-fold cross validation strategy using a four-depth-of-layers (full, 5, 10 and 20) strategy. This means the fitting procedure was executed ten times per algorithm implementation. It’s necessary to remark that the positive class for the invasive mechanical ventilation (IMV) variable refers to cases where IMV is required, while the positive class for the <span class="elsevierStyleBold">ICU</span> mortality variable is related to cases where patients die. It is worth mentioning that the focus of the study is on identifying independent variables and their associated thresholds with IMV and <span class="elsevierStyleBold">ICU</span> mortality, without defining specific categories to predict.</p><p id="par0055" class="elsevierStylePara elsevierViewall">We used a GLMM Tree to build conceptual models that explain the association between the potential predictors and the two outcome variables. This algorithm accounts for data clusters and temporal characteristics of the dataset, utilizing a mixed-effect strategy to combine the potential predictors that influence the outcome variables. Additionally, the algorithm provides a cut-off value for variables, allowing for comparison with clinical experience.</p><p id="par0060" class="elsevierStylePara elsevierViewall">GLMM Tree performance metrics were Area Under the Curve of Sensibility-Specificity (AUC), the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC),<a class="elsevierStyleCrossRef" href="#bib0110"><span class="elsevierStyleSup">22</span></a> as well as the deviance, the likelihood statistical,<a class="elsevierStyleCrossRef" href="#bib0115"><span class="elsevierStyleSup">23</span></a> and the sensitivity and specificity parameters. All the regression and GLMM Tree models were fitted with the same subset of variables shown in <a class="elsevierStyleCrossRef" href="#tbl0005">Table 1</a>.</p><p id="par0065" class="elsevierStylePara elsevierViewall">We used both regressions and GLMM family trees to gain a wider understanding of potential predictors for IMV and <span class="elsevierStyleBold">ICU</span> mortality. This combined approach offers more intuitive decision-making compared to black-box modeling strategies. We assessed each predictor's effectiveness and used the same set of variables (<a class="elsevierStyleCrossRef" href="#tbl0010">Table 2</a>) to build an <span class="elsevierStyleBold">ICU</span> mortality model for the entire cohort. The study's anonymized database and scripts can be found on the associated GitHub repository.<a class="elsevierStyleCrossRef" href="#bib0120"><span class="elsevierStyleSup">24</span></a> The database will be published in PhysioNet<a class="elsevierStyleCrossRef" href="#bib0125"><span class="elsevierStyleSup">25</span></a> project in order to disseminate and exchange the anonymized clinical records looking for cooperative project replication.</p></span></span><span id="sec0025" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0125">Results</span><span id="sec0030" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0130">Patient characteristics</span><p id="par0070" class="elsevierStylePara elsevierViewall"><span class="elsevierStyleBold">The complete cohort consisted of 280 patients who were included in the study.</span> A total of 154 patients (55 %) required IMV after ICU admission (<a class="elsevierStyleCrossRef" href="#fig0005">Fig. 1</a>), 65 of 80 patients (81.2 %) during the first and 89 of 200 patients (44.5 %) during the second wave. ICU mortality of the whole cohort was 25.7% (72 of 280 patients), 33.7% (27 of 80 patients) during the first and 22.5% (45 of 200 patients) in the second wave. <a class="elsevierStyleCrossRef" href="#tbl0010">Table 2</a> shows IMV and <span class="elsevierStyleBold">ICU</span> mortality predictors for the whole patient’s cohort. Mean registers per patient was 43.4, for a total of 12,163 hourly registers in the whole database (Figure 12 in complementary material).</p></span><span id="sec0035" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0135">Significance of predictors</span><p id="par0075" class="elsevierStylePara elsevierViewall">R-Part classification achieves the best and most clinically plausible results in selecting the twelve most representative predictors for IMV and <span class="elsevierStyleBold">ICU</span> mortality from the whole group of available potential predictors (<a class="elsevierStyleCrossRef" href="#tbl0010">Table 2</a>). Concerning this subset of predictors, the final selection is based on decreasing order of importance, according to results reached by the loss function (mean squared error), scaled from 0 to 100 points. Taking into account this premise, the predictors are: days from first symptoms to ICU admission (100), the APACHE II score (92.25), the oxygenation index, ROX index (72.46), blood procalcitonin (69.59), serum lactic dehydrogenase (54.45), total serum bilirubin (36.54), the COVID-19 wave (31.18), the dose of corticosteroids administered during the first five days of admission (30.96), lymphocyte count (15.57), pH (13.29), BMI (12.76), C-reactive protein (12.74), time to oxygen therapy (12.42) and body temperature (10.82).</p></span><span id="sec0040" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0140">Modeling performance</span><p id="par0080" class="elsevierStylePara elsevierViewall">In <a class="elsevierStyleCrossRef" href="#tbl0015">Table 3</a>, the performance of the IMV model is presented. The R-part predictors Regression-GLMTREE pair achieved the highest performance with an AUROC of 0.87, as shown in Figure 8 in the Supplementary material. Additionally, the <span class="elsevierStyleBold">ICU</span> mortality model performed well, with an AUROC of 0.88, as demonstrated in Figure 9 in the Supplementary material. The IMV likelihood ratio (RV+ 3.16, RV- 0.177) suggests that the test result is moderately useful for identifying or discharge patients susceptible to being treated with IMV. Related to the CI (95%), the reached interval (0.918 and <span class="elsevierStyleBold">0.928</span>) suggests a high level of precision considering the sensitivity, specificity, and accuracy of the model. Related to <span class="elsevierStyleBold">ICU</span> mortality, the IMV likelihood ratio (RV+ 5,105, RV− 0.424) and CI (95%) interval (<span class="elsevierStyleBold">0.817</span> and 0.833), results are also moderately useful. <a class="elsevierStyleCrossRef" href="#fig0015">Fig. 3</a> illustrates the <span class="elsevierStyleBold">ICU</span> Mortality decision tree, while Figure 7 in the Supplementary material presents the IMV decision tree. The optimal cut-off point for the prediction model was determined based on the IMV and <span class="elsevierStyleBold">ICU</span> mortality AUC, using Youden's Index,<a class="elsevierStyleCrossRef" href="#bib0130"><span class="elsevierStyleSup">26</span></a> which identifies the point of maximum sum of sensitivity and specificity in ROC curve analysis.</p><elsevierMultimedia ident="tbl0015"></elsevierMultimedia><elsevierMultimedia ident="fig0015"></elsevierMultimedia><p id="par0085" class="elsevierStylePara elsevierViewall">The trees in Figures 6 and 7 of the Supplementary material indicate that oxygenation status (ROX index) has the most significant influence on IMV, with a threshold near 5.2. On the other hand, <span class="elsevierStyleBold">ICU</span> mortality is mainly influenced by comorbidities (APACHE II score) and LDH, as revealed by the same trees.</p></span></span><span id="sec0045" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0145">Discussion</span><p id="par0090" class="elsevierStylePara elsevierViewall">The results of the present study include some highly relevant clinical results. First, the variable sets predicting IMV, and <span class="elsevierStyleBold">ICU</span> mortality are different. Whereas oxygenation variables are independent predictors of IMV, <span class="elsevierStyleBold">ICU</span> mortality is associated with increased age and LDH and the presence of comorbidities. The latter variables may be considered markers of two processes: COVID-19-associated inflammation and ICU-acquired superinfection (see Figure 4 in the Supplementary material). Secondly, the characteristics of pharmacological therapy, including the administration of steroid drugs, has little influence on both the need for IMV and <span class="elsevierStyleBold">ICU</span> mortality, considering our results. We included in the analysis 64 patients not receiving steroids and 216 receiving this treatment, at the usual 6<span class="elsevierStyleHsp" style=""></span>mg dexamethasone or equivalent daily dose. This is a remarkable finding, because the effect of steroids on mortality identified in a previous trial<a class="elsevierStyleCrossRef" href="#bib0135"><span class="elsevierStyleSup">27</span></a> have influenced recommendations, as well as clinical practice, since its publication. It may be speculated that the decision to include and randomize or not at the discretion of the attending physicians, and based on undisclosed criteria, rendered different results by selecting a study subset of COVID-19 cases with different characteristics. In comparison, no inclusion-exclusion criteria for selection process were applied in our “pragmatic” type of cohort. Steroids were given to almost every patient unless a severe contraindication existed, after the results of the RECOVERY trial were made available.</p><p id="par0095" class="elsevierStylePara elsevierViewall">The present study applied a novel methodology (logistic regression with regularization plus GLMM Tree mixed models) to evaluate the relative importance of several variables as predictors of significant clinical events. Using machine learning and a fine-grained longitudinal multifaceted database, we have established relevant variable value thresholds to support clinical decisions. Although the model would perform quite well as predictor for IMV and <span class="elsevierStyleBold">ICU</span> mortality, with good positive predictive values, it is important to emphasize that this is not a predictive model in the classical sense, but an attempt to pinpoint the most important clinical events that represent turning points during the studied process (in this case, clinical management of patients not initially under IMV). In this sense, we should say that the inclusion of the likelihood ratio as an evaluation factor for comparing performance model was reach great results. However, following the premise of model explainability, we believe it is important to take this element into account as a final selection factor for the set of predictors that best fit daily clinical practice. This study demonstrates that predictor-ranking methodologies using self-explainable machine learning may support therapeutic decision-making using observational data, when randomized clinical trials are unfeasible or unethical.</p><p id="par0100" class="elsevierStylePara elsevierViewall">Regarding with the strengths of our study, we would like to mention the quantity and quality of the data set. Collected data have a high level of detail, leveraging the power of strategically devised electronic health records (EHR), which include relevant information in a highly structured and recoverable format. Every effort was made to configure our EHR to optimally gather all relevant information about COVID-19 patients. Also, our anonymized database is available in the repository along with the script we used for statistical analysis, is highly detailed and has been extensively curated to reflect temporal evolution and to improve data quality as much as possible. In any case, the collection of variables from Electronic Health Records (EHR) may be biased, affecting data quality. Age and gender biases are possible, as well as biases related to the selection and measurement of clinical variables. These biases can lead to incomplete or skewed representations of certain population groups and may impact the validity and generalizability of research findings and clinical decision-making. It is important to be aware of these biases to ensure proper interpretation and use of EHR data.</p><p id="par0105" class="elsevierStylePara elsevierViewall">On the other hand, the limitations of our study results relate mainly to its single-centered nature and require confirmation in a multicenter dataset to gain external validity. Our methodology would be perfectly suited for a multicenter study, including “center” as a random factor in the second (GLMM Tree) part of the process. We suggest that future research applying this methodology could focus on designing clinical studies using observational data to answer relevant clinical questions without the logistic requirements of a randomized clinical trial or for hypothesis-generating purposes. Furthermore, when considering the limitations of using generalized linear mixed effects models (GLMMs) for modeling causation in critical care medicine research, it is important to highlight the absence of explicit causality assumptions. GLMMs primarily focus on association or correlation analysis, lacking the ability to address the assumptions necessary for establishing causal relationships. Specifically, GLMMs do not provide frameworks for the identification of causal effects or account for unmeasured confounding variables, which are crucial considerations in causal inference. In contrast, causal inference methods, such as the potential outcomes framework, explicitly address these assumptions, offering a more comprehensive approach for investigating causality. Therefore, when establishing causal relationships between variables, researchers should carefully consider the limitations of GLMMs and opt for causal inference methods, which provide a more robust approach for investigating causality in critical care medicine research.</p><p id="par0110" class="elsevierStylePara elsevierViewall">In conclusion, different variables predict IMV and <span class="elsevierStyleBold">ICU</span> mortality in severe COVID-19 patients, suggesting that the therapeutic decision of when to use IMV has little impact on <span class="elsevierStyleBold">ICU</span> mortality. Our methodology is a valid option to assess therapeutic decisions using observational data when randomized clinical trials are not feasible or ethical.</p></span><span id="sec0050" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0150">Author's contribution</span><p id="par0115" class="elsevierStylePara elsevierViewall">SM, MA and AN conceived the presented idea. <span class="elsevierStyleBold">SM and MA contributed equally as first authors.</span> SM and MA developed the theory and performed the computations. AN conducted an independent literature search to identify potentially relevant studies. MS independently reviewed the search results to identify pertinent articles. MS, AN, TF and VY contributed to the interpretation of the results. SM, MA, AN, MS, FL and AC took the lead in writing the manuscript. All authors provided critical feedback and helped shape the research, analysis, and manuscript.</p></span><span id="sec0055" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0155">Funding</span><p id="par0120" class="elsevierStylePara elsevierViewall">This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.</p></span><span id="sec0060" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0160">Conflict of interest</span><p id="par0125" class="elsevierStylePara elsevierViewall">The authors declare that they have no conflict of interest.</p></span></span>" "textoCompletoSecciones" => array:1 [ "secciones" => array:13 [ 0 => array:3 [ "identificador" => "xres2066952" "titulo" => "Abstract" "secciones" => array:8 [ 0 => array:2 [ "identificador" => "abst0005" "titulo" => "Objective" ] 1 => array:2 [ "identificador" => "abst0010" "titulo" => "Design" ] 2 => array:2 [ "identificador" => "abst0015" "titulo" => "Setting" ] 3 => array:2 [ "identificador" => "abst0020" "titulo" => "Patients or participants" ] 4 => array:2 [ "identificador" => "abst0025" "titulo" => "Interventions" ] 5 => array:2 [ "identificador" => "abst0030" "titulo" => "Main variables of interest" ] 6 => array:2 [ "identificador" => "abst0035" "titulo" => "Results" ] 7 => array:2 [ "identificador" => "abst0040" "titulo" => "Conclusions" ] ] ] 1 => array:2 [ "identificador" => "xpalclavsec1763938" "titulo" => "Keywords" ] 2 => array:3 [ "identificador" => "xres2066953" "titulo" => "Resumen" "secciones" => array:8 [ 0 => array:2 [ "identificador" => "abst0045" "titulo" => "Objetivo" ] 1 => array:2 [ "identificador" => "abst0050" "titulo" => "Diseño" ] 2 => array:2 [ "identificador" => "abst0055" "titulo" => "Ámbito" ] 3 => array:2 [ "identificador" => "abst0060" "titulo" => "Pacientes o participantes" ] 4 => array:2 [ "identificador" => "abst0065" "titulo" => "Intervenciones" ] 5 => array:2 [ "identificador" => "abst0070" "titulo" => "Principales variables de interés" ] 6 => array:2 [ "identificador" => "abst0075" "titulo" => "Resultados" ] 7 => array:2 [ "identificador" => "abst0080" "titulo" => "Conclusiones" ] ] ] 3 => array:2 [ "identificador" => "xpalclavsec1763937" "titulo" => "Palabras clave" ] 4 => array:2 [ "identificador" => "sec0005" "titulo" => "Introduction" ] 5 => array:3 [ "identificador" => "sec0010" "titulo" => "Patients and methods" "secciones" => array:2 [ 0 => array:2 [ "identificador" => "sec0015" "titulo" => "Selection and description of patients" ] 1 => array:2 [ "identificador" => "sec0020" "titulo" => "Methods and techniques" ] ] ] 6 => array:3 [ "identificador" => "sec0025" "titulo" => "Results" "secciones" => array:3 [ 0 => array:2 [ "identificador" => "sec0030" "titulo" => "Patient characteristics" ] 1 => array:2 [ "identificador" => "sec0035" "titulo" => "Significance of predictors" ] 2 => array:2 [ "identificador" => "sec0040" "titulo" => "Modeling performance" ] ] ] 7 => array:2 [ "identificador" => "sec0045" "titulo" => "Discussion" ] 8 => array:2 [ "identificador" => "sec0050" "titulo" => "Author's contribution" ] 9 => array:2 [ "identificador" => "sec0055" "titulo" => "Funding" ] 10 => array:2 [ "identificador" => "sec0060" "titulo" => "Conflict of interest" ] 11 => array:2 [ "identificador" => "xack720017" "titulo" => "Acknowledgements" ] 12 => array:1 [ "titulo" => "References" ] ] ] "pdfFichero" => "main.pdf" "tienePdf" => true "fechaRecibido" => "2023-03-03" "fechaAceptado" => "2023-06-22" "PalabrasClave" => array:2 [ "en" => array:1 [ 0 => array:4 [ "clase" => "keyword" "titulo" => "Keywords" "identificador" => "xpalclavsec1763938" "palabras" => array:6 [ 0 => "Acute respiratory distress syndrome" 1 => "Invasive mechanical ventilation" 2 => "COVID-19" 3 => "Machine learning" 4 => "Artificial intelligence" 5 => "Predictors" ] ] ] "es" => array:1 [ 0 => array:4 [ "clase" => "keyword" "titulo" => "Palabras clave" "identificador" => "xpalclavsec1763937" "palabras" => array:6 [ 0 => "Síndrome de distrés respiratorio agudo" 1 => "Ventilación mecánica invasiva" 2 => "COVID-19" 3 => "Aprendizaje automático" 4 => "Inteligencia artificilal" 5 => "Predictores" ] ] ] ] "tieneResumen" => true "resumen" => array:2 [ "en" => array:3 [ "titulo" => "Abstract" "resumen" => "<span id="abst0005" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0010">Objective</span><p id="spar0055" class="elsevierStyleSimplePara elsevierViewall">To determine if potential predictors for invasive mechanical ventilation (IMV) are also determinants for mortality in COVID-19-associated acute respiratory distress syndrome (C-ARDS).</p></span> <span id="abst0010" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0015">Design</span><p id="spar0060" class="elsevierStyleSimplePara elsevierViewall">Single center highly detailed longitudinal observational study.</p></span> <span id="abst0015" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0020">Setting</span><p id="spar0065" class="elsevierStyleSimplePara elsevierViewall">Tertiary hospital ICU: two first COVID-19 pandemic waves, Madrid, Spain.</p></span> <span id="abst0020" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0025">Patients or participants</span><p id="spar0070" class="elsevierStyleSimplePara elsevierViewall">: 280 patients with C-ARDS, not requiring IMV on admission.</p></span> <span id="abst0025" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0030">Interventions</span><p id="spar0075" class="elsevierStyleSimplePara elsevierViewall">None.</p></span> <span id="abst0030" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0035">Main variables of interest</span><p id="spar0080" class="elsevierStyleSimplePara elsevierViewall">: Target: endotracheal intubation and IMV, mortality.</p><p id="spar0085" class="elsevierStyleSimplePara elsevierViewall">Predictors: demographics, hourly evolution of oxygenation, clinical data, and laboratory results.</p></span> <span id="abst0035" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0040">Results</span><p id="spar0090" class="elsevierStyleSimplePara elsevierViewall">The time between symptom onset and ICU admission, the APACHE II score, the ROX index, and procalcitonin levels in blood were potential predictors related to both IMV and mortality. The ROX index was the most significant predictor associated with IMV, while APACHE II, LDH, and DaysSympICU were the most with mortality.</p></span> <span id="abst0040" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0045">Conclusions</span><p id="spar0095" class="elsevierStyleSimplePara elsevierViewall">According to the results of the analysis, there are significant predictors linked with IMV and mortality in C-ARDS patients, including the time between symptom onset and ICU admission, the severity of the COVID-19 waves, and several clinical and laboratory measures. These findings may help clinicians to better identify patients at risk for IMV and mortality and improve their management.</p></span>" "secciones" => array:8 [ 0 => array:2 [ "identificador" => "abst0005" "titulo" => "Objective" ] 1 => array:2 [ "identificador" => "abst0010" "titulo" => "Design" ] 2 => array:2 [ "identificador" => "abst0015" "titulo" => "Setting" ] 3 => array:2 [ "identificador" => "abst0020" "titulo" => "Patients or participants" ] 4 => array:2 [ "identificador" => "abst0025" "titulo" => "Interventions" ] 5 => array:2 [ "identificador" => "abst0030" "titulo" => "Main variables of interest" ] 6 => array:2 [ "identificador" => "abst0035" "titulo" => "Results" ] 7 => array:2 [ "identificador" => "abst0040" "titulo" => "Conclusions" ] ] ] "es" => array:3 [ "titulo" => "Resumen" "resumen" => "<span id="abst0045" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0055">Objetivo</span><p id="spar0100" class="elsevierStyleSimplePara elsevierViewall">Determinar si las variables clínicas independientes que condicionan el inicio de ventilación mecánica invasiva (VMI) son los mismos que condicionan la mortalidad en el síndrome de distrés respiratorio agudo asociado con COVID-19 (C-SDRA).</p></span> <span id="abst0050" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0060">Diseño</span><p id="spar0105" class="elsevierStyleSimplePara elsevierViewall">Estudio observacional longitudinal en un solo centro.</p></span> <span id="abst0055" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0065">Ámbito</span><p id="spar0110" class="elsevierStyleSimplePara elsevierViewall">UCI, hospital terciario: primeras dos olas de COVID-19 en Madrid, España.</p></span> <span id="abst0060" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0070">Pacientes o participantes</span><p id="spar0115" class="elsevierStyleSimplePara elsevierViewall">280 pacientes con C-SDRA que no requieren VMI al ingreso en UCI.</p></span> <span id="abst0065" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0075">Intervenciones</span><p id="spar0120" class="elsevierStyleSimplePara elsevierViewall">Ninguna.</p></span> <span id="abst0070" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0080">Principales variables de interés</span><p id="spar0125" class="elsevierStyleSimplePara elsevierViewall">Objetivo: VMI y Mortalidad.</p><p id="spar0130" class="elsevierStyleSimplePara elsevierViewall">Predictores: demográficos, variables clínicas, resultados de laboratorio y evolución de la oxigenación.</p></span> <span id="abst0075" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0085">Resultados</span><p id="spar0135" class="elsevierStyleSimplePara elsevierViewall">El tiempo entre el inicio de los síntomas y el ingreso en la UCI, la puntuación APACHE II, el índice ROX y los niveles de procalcitonina en sangre eran posibles predictores relacionados tanto con la IMV como con la mortalidad. El índice ROX fue el predictor más significativo asociada con la IMV, mientras que APACHE II, LDH y DaysSympICU fueron los más influyentes en la mortalidad.</p></span> <span id="abst0080" class="elsevierStyleSection elsevierViewall"><span class="elsevierStyleSectionTitle" id="sect0090">Conclusiones</span><p id="spar0140" class="elsevierStyleSimplePara elsevierViewall">Según los resultados obtenidos se identifican predictores significativos vinculados con la VMI y mortalidad en pacientes con C-ARDS, incluido el tiempo entre el inicio de los síntomas y el ingreso en la UCI, la gravedad de las olas de COVID-19 y varias medidas clínicas y de laboratorio. Estos hallazgos pueden ayudar a los médicos a identificar mejor a los pacientes en riesgo de IMV y mortalidad y mejorar su manejo.</p></span>" "secciones" => array:8 [ 0 => array:2 [ "identificador" => "abst0045" "titulo" => "Objetivo" ] 1 => array:2 [ "identificador" => "abst0050" "titulo" => "Diseño" ] 2 => array:2 [ "identificador" => "abst0055" "titulo" => "Ámbito" ] 3 => array:2 [ "identificador" => "abst0060" "titulo" => "Pacientes o participantes" ] 4 => array:2 [ "identificador" => "abst0065" "titulo" => "Intervenciones" ] 5 => array:2 [ "identificador" => "abst0070" "titulo" => "Principales variables de interés" ] 6 => array:2 [ "identificador" => "abst0075" "titulo" => "Resultados" ] 7 => array:2 [ "identificador" => "abst0080" "titulo" => "Conclusiones" ] ] ] ] "NotaPie" => array:1 [ 0 => array:3 [ "etiqueta" => "1" "nota" => "<p class="elsevierStyleNotepara" id="npar0025">Co-first authors: Authors contributed equally as first authors.</p>" "identificador" => "fn0005" ] ] "apendice" => array:1 [ 0 => array:1 [ "seccion" => array:1 [ 0 => array:4 [ "apendice" => "<p id="par0140" class="elsevierStylePara elsevierViewall">The following is Supplementary data to this article:<elsevierMultimedia ident="upi0005"></elsevierMultimedia></p>" "etiqueta" => "Appendix A" "titulo" => "Supplementary data" "identificador" => "sec0070" ] ] ] ] "multimedia" => array:7 [ 0 => array:8 [ "identificador" => "fig0005" "etiqueta" => "Figure 1" "tipo" => "MULTIMEDIAFIGURA" "mostrarFloat" => true "mostrarDisplay" => false "figura" => array:1 [ 0 => array:4 [ "imagen" => "gr1.jpeg" "Alto" => 2293 "Ancho" => 1675 "Tamanyo" => 311205 ] ] "detalles" => array:1 [ 0 => array:3 [ "identificador" => "at0075" "detalle" => "Figure " "rol" => "short" ] ] "descripcion" => array:1 [ "en" => "<p id="spar0005" class="elsevierStyleSimplePara elsevierViewall">COVID-19 patients admitted during first and second pandemic waves. The cohort comprises 280 severe COVID-19 patients admitted to the ICU Department at HCSC in Madrid, Spain, between March 3, 2020, and February 28, 2021. During this time period, SARS-COV-2 wild-type and subsequently alpha variants were prevalent in Spain. Over the study time period 4229 covid-19 patients were admitted to HCSC, 405 of whom required ICU admission (first wave: 153, second wave: 252 patients).</p>" ] ] 1 => array:8 [ "identificador" => "fig0010" "etiqueta" => "Figure 2" "tipo" => "MULTIMEDIAFIGURA" "mostrarFloat" => true "mostrarDisplay" => false "figura" => array:1 [ 0 => array:4 [ "imagen" => "gr2.jpeg" "Alto" => 2509 "Ancho" => 3341 "Tamanyo" => 501946 ] ] "detalles" => array:1 [ 0 => array:3 [ "identificador" => "at0080" "detalle" => "Figure " "rol" => "short" ] ] "descripcion" => array:1 [ "en" => "<p id="spar0010" class="elsevierStyleSimplePara elsevierViewall">Methodology for fitting the machine learning algorithms. In a previous stage, Figure 5 in Supplementary material shows the complete workflow, from the cohort selection according to clinical needs to the implementation of the algorithms that have been included in the explanation. The first step involves the cohort selection as well as the initial group of variables considered in this study, The second step consists in the implementation of a statistical study of each variable. This step also involves correlation (Figure 6 in Supplementary material) imputation and transformations procedures in order to dispose of the most accurate data in the following steps. The third step analyzed the most significant predictors based on five <span class="elsevierStyleBold">Machine</span> Learning (ML) techniques linked with regression analysis based on 10-fold cross-validation regressions. The fourth and last step identifies the behavior of each predictor attending to different proposes. The first one is related to mechanical ventilation needs attending to different settings in the Generalized Linear Mixed Model (GLMM) Tree (depth of layers) looking for the best balance between performance (Akaike Information Criterion (AIC), Bayesian information criterion (BIC), Area Under the Roc Curve (ROC) and more parameters within the table III) and explainability of the model. The second one is related to the most representative mortality predictors but following the same balance objective.</p>" ] ] 2 => array:8 [ "identificador" => "fig0015" "etiqueta" => "Figure 3" "tipo" => "MULTIMEDIAFIGURA" "mostrarFloat" => true "mostrarDisplay" => false "figura" => array:1 [ 0 => array:4 [ "imagen" => "gr3.jpeg" "Alto" => 1594 "Ancho" => 3425 "Tamanyo" => 463757 ] ] "detalles" => array:1 [ 0 => array:3 [ "identificador" => "at0085" "detalle" => "Figure " "rol" => "short" ] ] "descripcion" => array:1 [ "en" => "<p id="spar0015" class="elsevierStyleSimplePara elsevierViewall">ICU Mortality Tree Predictors. The predictors appear in different branches attending to their significance in the predictive model. Values in bold letters represent the registries per branch. Values in red bold letters represent the percentage of registries with positive outcome. The variable named as “DAYS_SIMPTONS_ADMISION” is related with the number of days from first symptoms to ICU admission. The variable “linf_total”, is related to lymphocyte count per mm3. The variable named as “dosis_equiv_mpred_5d” is related with the corticosteroid dose, during the first five days of admission (mg of equivalent methylprednisolone dose). The variable named as “bbTot” is related with the total levels of bilirubin in blood. The variable names as “ldh” is related to the lactate dehydrogenase serum level. The variable DAYS_UNTIL_O2 is related to the number of days until the patient requires O2.</p>" ] ] 3 => array:8 [ "identificador" => "tbl0005" "etiqueta" => "Table 1" "tipo" => "MULTIMEDIATABLA" "mostrarFloat" => true "mostrarDisplay" => false "detalles" => array:1 [ 0 => array:3 [ "identificador" => "at0090" "detalle" => "Table " "rol" => "short" ] ] "tabla" => array:3 [ "leyenda" => "<p id="spar0025" class="elsevierStyleSimplePara elsevierViewall">This group of predictors will be applied in the selection procedure linked with the five regression algorithms: Ridge, LASSO, Elastic, Boruta and R-part Based on the reached results, the group of predictors are going to be reduced attending to its behavior related to IMV needs. Figures 7–11 (Supplementary material) shows the results from each regression procedure where R-Part was finally selected due to its good balance between model performance and explicability of results.</p><p id="spar0030" class="elsevierStyleSimplePara elsevierViewall">Data updated June 22, 2023.</p>" "tablatextoimagen" => array:1 [ 0 => array:2 [ "tabla" => array:1 [ 0 => """ <table border="0" frame="\n \t\t\t\t\tvoid\n \t\t\t\t" class=""><thead title="thead"><tr title="table-row"><th class="td-with-role" title="\n \t\t\t\t\ttable-head\n \t\t\t\t ; entry_with_role_colgroup " colspan="6" align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t" scope="col" style="border-bottom: 2px solid black">Dataset clinical and biochemical characteristics</th></tr><tr title="table-row"><th class="td-with-role" title="\n \t\t\t\t\ttable-head\n \t\t\t\t ; entry_with_role_colgroup " colspan="6" align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t" scope="col" style="border-bottom: 2px solid black"><span class="elsevierStyleItalic">Invasive Mechanical Ventilation (IMV)</span></th></tr><tr title="table-row"><th class="td" title="\n \t\t\t\t\ttable-head\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t" scope="col">Variable \t\t\t\t\t\t\n \t\t\t\t\t\t</th><th class="td" title="\n \t\t\t\t\ttable-head\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t" scope="col"><span class="elsevierStyleItalic">N</span> \t\t\t\t\t\t\n \t\t\t\t\t\t</th><th class="td" title="\n \t\t\t\t\ttable-head\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t" scope="col">Overall, <span class="elsevierStyleItalic">N</span><span class="elsevierStyleHsp" style=""></span>=<span class="elsevierStyleHsp" style=""></span>12,163<a class="elsevierStyleCrossRef" href="#tblfn0005"><span class="elsevierStyleSup">a</span></a> \t\t\t\t\t\t\n \t\t\t\t\t\t</th><th class="td-with-role" title="\n \t\t\t\t\ttable-head\n \t\t\t\t ; entry_with_role_colgroup " colspan="2" align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t" scope="col" style="border-bottom: 2px solid black">Invasive Mechanical Ventilation (IMV)</th><th class="td" title="\n \t\t\t\t\ttable-head\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t" scope="col">p-value<a class="elsevierStyleCrossRef" href="#tblfn0010"><span class="elsevierStyleSup">b</span></a> \t\t\t\t\t\t\n \t\t\t\t\t\t</th></tr><tr title="table-row"><th class="td" title="\n \t\t\t\t\ttable-head\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t" scope="col" style="border-bottom: 2px solid black"> \t\t\t\t\t\t\n \t\t\t\t\t\t</th><th class="td" title="\n \t\t\t\t\ttable-head\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t" scope="col" style="border-bottom: 2px solid black"> \t\t\t\t\t\t\n \t\t\t\t\t\t</th><th class="td" title="\n \t\t\t\t\ttable-head\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t" scope="col" style="border-bottom: 2px solid black"> \t\t\t\t\t\t\n \t\t\t\t\t\t</th><th class="td" title="\n \t\t\t\t\ttable-head\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t" scope="col" style="border-bottom: 2px solid black">No, <span class="elsevierStyleItalic">N</span><span class="elsevierStyleHsp" style=""></span>=<span class="elsevierStyleHsp" style=""></span>9093<a class="elsevierStyleCrossRef" href="#tblfn0005"><span class="elsevierStyleSup">a</span></a> \t\t\t\t\t\t\n \t\t\t\t\t\t</th><th class="td" title="\n \t\t\t\t\ttable-head\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t" scope="col" style="border-bottom: 2px solid black">Yes, <span class="elsevierStyleItalic">N</span><span class="elsevierStyleHsp" style=""></span>=<span class="elsevierStyleHsp" style=""></span>3070<a class="elsevierStyleCrossRef" href="#tblfn0005"><span class="elsevierStyleSup">a</span></a> \t\t\t\t\t\t\n \t\t\t\t\t\t</th><th class="td" title="\n \t\t\t\t\ttable-head\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t" scope="col" style="border-bottom: 2px solid black"> \t\t\t\t\t\t\n \t\t\t\t\t\t</th></tr></thead><tbody title="tbody"><tr title="table-row"><td class="td-with-role" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t ; entry_with_role_rowhead " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t">Age, years, Median (Q1-Q3) \t\t\t\t\t\t\n \t\t\t\t</td><td class="td" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t">12,163 \t\t\t\t\t\t\n \t\t\t\t</td><td class="td" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t">59 (51–68) \t\t\t\t\t\t\n \t\t\t\t</td><td class="td" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t">58 (50–67) \t\t\t\t\t\t\n \t\t\t\t</td><td class="td" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t">63 (54–70) \t\t\t\t\t\t\n \t\t\t\t</td><td class="td" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t"><0.001 \t\t\t\t\t\t\n \t\t\t\t</td></tr><tr title="table-row"><td class="td-with-role" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t ; entry_with_role_rowhead " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t">Gender, n (%) \t\t\t\t\t\t\n \t\t\t\t</td><td class="td" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t">12,163 \t\t\t\t\t\t\n \t\t\t\t</td><td class="td" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t"> \t\t\t\t\t\t\n \t\t\t\t</td><td class="td" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t"> \t\t\t\t\t\t\n \t\t\t\t</td><td class="td" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t"> \t\t\t\t\t\t\n \t\t\t\t</td><td class="td" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t"><0.001 \t\t\t\t\t\t\n \t\t\t\t</td></tr><tr title="table-row"><td class="td-with-role" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t ; entry_with_role_rowhead " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t"><span class="elsevierStyleHsp" style=""></span>Male \t\t\t\t\t\t\n \t\t\t\t</td><td class="td" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t"> \t\t\t\t\t\t\n \t\t\t\t</td><td class="td" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t">8032 (66) \t\t\t\t\t\t\n \t\t\t\t</td><td class="td" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t">5649 (62) \t\t\t\t\t\t\n \t\t\t\t</td><td class="td" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t">2383 (78) \t\t\t\t\t\t\n \t\t\t\t</td><td class="td" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t"> \t\t\t\t\t\t\n \t\t\t\t</td></tr><tr title="table-row"><td class="td-with-role" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t ; entry_with_role_rowhead " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t"><span class="elsevierStyleHsp" style=""></span>Female \t\t\t\t\t\t\n \t\t\t\t</td><td class="td" title="\n \t\t\t\t\ttable-entry\n \t\t\t\t " align="left" valign="\n \t\t\t\t\ttop\n \t\t\t\t"> \t\t\t\t\t\t\n \t\t\t\t</td><td class="td" title="\n