Articles producció científica> Medicina i Cirurgia

Machine learning to predict major bleeding during anticoagulation for venous thromboembolism: possibilities and limitations

  • Identification data

    Identifier: imarina:9295150
    Handle: https://hdl.handle.net/20.500.11797/imarina9295150
    Authors:
    Mora DMateo JNieto JABikdeli BYamashita YBarco SJimenez DDemelo-Rodriguez PRosa VYoo HHBSadeghipour PMonreal MAdarraga MDAlberich-Conesa AAlonso-Carrillo JAgudo PAmado CAmorós SArcelus JIBallaz ABarba RBarbagelata CBarrón MBarrón-Andrés BBlanco-Molina ABotella ECarrero-Arribas RCasado IChasco LCriado Jdel Toro JDe Ancos CDe Juana-Izquierdo CDemelo-Rodríguez PDíaz-Brasero AMDíaz-Pedroche MCDíaz-Peromingo JADíaz-Simón RDubois-Silva AEscribano JCEspósito FFalgá CFarfán-Sedano AIFernández-Aracil CFernández-Capitán CFernández-Jiménez BFernández-Muixi JFernández-Reyes JLFont CFrancisco IGaleano-Valle FGarcía MAGarcía de Herreros MGarcía-Bragado FGarcía-González CGarcía-Ortega AGavín-Sebastián OGil-De Gómez MGil-Díaz AGómez-Cuervo CGómez-Mosquera AMGonzález-Martínez JGrau EGuirado LGutiérrez JHernández-Blasco LJaras MJJiménez DJou IJoya MDLacruz BLainez-Justo SLecumberri RLobo JLLópez-De la Fuente MLópez-Jiménez LLópez-Miguel PLópez-Núñez JJLópez-Reyes RLópez-Ruiz ALópez-Sáez JBLorente MALorenzo ALumbierres MMadridano OMaestre AMarcos MMartín-Guerra JMMartín-Martos FMas-Maresma LMellado MMena EMercado MIMoisés JMonreal MMuñoz-Blanco AMuñoz-Gamito GNieto JANúñez-Fernández MJOsorio J
    Abstract:
    Predictive tools for major bleeding (MB) using machine learning (ML) might be advantageous over traditional methods. We used data from the Registro Informatizado de Enfermedad TromboEmbólica (RIETE) to develop ML algorithms to identify patients with venous thromboembolism (VTE) at increased risk of MB during the first 3 months of anticoagulation. A total of 55 baseline variables were used as predictors. New data prospectively collected from the RIETE were used for further validation. The RIETE and VTE-BLEED scores were used for comparisons. External validation was performed with the COMMAND-VTE database. Learning was carried out with data from 49 587 patients, of whom 873 (1.8%) had MB. The best performing ML method was XGBoost. In the prospective validation cohort the sensitivity, specificity, positive predictive value and F1 score were: 33.2%, 93%, 10%, and 15.4% respectively. F1 value for the RIETE and VTE-BLEED scores were 8.6% and 6.4% respectively. In the external validation cohort the metrics were 10.3%, 87.6%, 3.5% and 5.2% respectively. In that cohort, the F1 value for the RIETE score was 17.3% and for the VTE-BLEED score 9.75%. The performance of the XGBoost algorithm was better than that from the RIETE and VTE-BLEED scores only in the prospective validation cohort, but not in the external validation cohort.

  • Others:

    Author, as appears in the article.: Mora D; Mateo J; Nieto JA; Bikdeli B; Yamashita Y; Barco S; Jimenez D; Demelo-Rodriguez P; Rosa V; Yoo HHB; Sadeghipour P; Monreal M; Adarraga MD; Alberich-Conesa A; Alonso-Carrillo J; Agudo P; Amado C; Amorós S; Arcelus JI; Ballaz A; Barba R; Barbagelata C; Barrón M; Barrón-Andrés B; Blanco-Molina A; Botella E; Carrero-Arribas R; Casado I; Chasco L; Criado J; del Toro J; De Ancos C; De Juana-Izquierdo C; Demelo-Rodríguez P; Díaz-Brasero AM; Díaz-Pedroche MC; Díaz-Peromingo JA; Díaz-Simón R; Dubois-Silva A; Escribano JC; Espósito F; Falgá C; Farfán-Sedano AI; Fernández-Aracil C; Fernández-Capitán C; Fernández-Jiménez B; Fernández-Muixi J; Fernández-Reyes JL; Font C; Francisco I; Galeano-Valle F; García MA; García de Herreros M; García-Bragado F; García-González C; García-Ortega A; Gavín-Sebastián O; Gil-De Gómez M; Gil-Díaz A; Gómez-Cuervo C; Gómez-Mosquera AM; González-Martínez J; Grau E; Guirado L; Gutiérrez J; Hernández-Blasco L; Jaras MJ; Jiménez D; Jou I; Joya MD; Lacruz B; Lainez-Justo S; Lecumberri R; Lobo JL; López-De la Fuente M; López-Jiménez L; López-Miguel P; López-Núñez JJ; López-Reyes R; López-Ruiz A; López-Sáez JB; Lorente MA; Lorenzo A; Lumbierres M; Madridano O; Maestre A; Marcos M; Martín-Guerra JM; Martín-Martos F; Mas-Maresma L; Mellado M; Mena E; Mercado MI; Moisés J; Monreal M; Muñoz-Blanco A; Muñoz-Gamito G; Nieto JA; Núñez-Fernández MJ; Osorio J
    Department: Medicina i Cirurgia
    URV's Author/s: Porras Ledantes, Jose Antonio
    Keywords: Venous thrombosis Pulmonary embolism Outcomes Machine learning Haemorrhage
    Abstract: Predictive tools for major bleeding (MB) using machine learning (ML) might be advantageous over traditional methods. We used data from the Registro Informatizado de Enfermedad TromboEmbólica (RIETE) to develop ML algorithms to identify patients with venous thromboembolism (VTE) at increased risk of MB during the first 3 months of anticoagulation. A total of 55 baseline variables were used as predictors. New data prospectively collected from the RIETE were used for further validation. The RIETE and VTE-BLEED scores were used for comparisons. External validation was performed with the COMMAND-VTE database. Learning was carried out with data from 49 587 patients, of whom 873 (1.8%) had MB. The best performing ML method was XGBoost. In the prospective validation cohort the sensitivity, specificity, positive predictive value and F1 score were: 33.2%, 93%, 10%, and 15.4% respectively. F1 value for the RIETE and VTE-BLEED scores were 8.6% and 6.4% respectively. In the external validation cohort the metrics were 10.3%, 87.6%, 3.5% and 5.2% respectively. In that cohort, the F1 value for the RIETE score was 17.3% and for the VTE-BLEED score 9.75%. The performance of the XGBoost algorithm was better than that from the RIETE and VTE-BLEED scores only in the prospective validation cohort, but not in the external validation cohort.
    Thematic Areas: Medicina iii Medicina ii Medicina i Interdisciplinar Hematology General medicine Farmacia Engenharias iii Ciências biológicas iii Ciências biológicas ii Ciências biológicas i Biotecnología
    licence for use: https://creativecommons.org/licenses/by/3.0/es/
    Author's mail: joseantonio.porras@urv.cat
    Author identifier: 0000-0001-6418-1822
    Record's date: 2025-02-19
    Paper version: info:eu-repo/semantics/publishedVersion
    Paper original source: British Journal Of Haematology. 201 (5): 971-981
    APA: Mora D; Mateo J; Nieto JA; Bikdeli B; Yamashita Y; Barco S; Jimenez D; Demelo-Rodriguez P; Rosa V; Yoo HHB; Sadeghipour P; Monreal M; Adarraga MD; Alb (2023). Machine learning to predict major bleeding during anticoagulation for venous thromboembolism: possibilities and limitations. British Journal Of Haematology, 201(5), 971-981. DOI: 10.1111/bjh.18737
    Licence document URL: https://repositori.urv.cat/ca/proteccio-de-dades/
    Entity: Universitat Rovira i Virgili
    Journal publication year: 2023
    Publication Type: Journal Publications

  • Keywords:

    Hematology
    Venous thrombosis
    Pulmonary embolism
    Outcomes
    Machine learning
    Haemorrhage
    Medicina iii
    Medicina ii
    Medicina i
    Interdisciplinar
    Hematology
    General medicine
    Farmacia
    Engenharias iii
    Ciências biológicas iii
    Ciências biológicas ii
    Ciências biológicas i
    Biotecnología

  • Documents:

    DocumentPrincipal

  • Cerca a google

    Search to google scholar