ROAD2H: Development and evaluation of an open‐source explainable artificial intelligence approach for managing co‐morbidity and clinical guidelines

J Domínguez; D Prociuk; B Marović; K Čyras; O Cocarascu; F Ruiz; E Mi; E Mi; C Ramtale; A Rago; A Darzi; F Toni; V Curcin; B Delaney

doi:10.1002/lrh2.10391

ROAD2H: Development and evaluation of an open‐source explainable artificial intelligence approach for managing co‐morbidity and clinical guidelines

J Domínguez, D Prociuk, B Marović, K Čyras, O Cocarascu, F Ruiz, E Mi, E Mi, C Ramtale, A Rago, A Darzi, F Toni, V Curcin, B Delaney

School of Engineering, Computing and Mathematics

Research output: Contribution to journal › Article › peer-review

Abstract

AbstractIntroductionClinical decision support (CDS) systems (CDSSs) that integrate clinical guidelines need to reflect real‐world co‐morbidity. In patient‐specific clinical contexts, transparent recommendations that allow for contraindications and other conflicts arising from co‐morbidity are a requirement. In this work, we develop and evaluate a non‐proprietary, standards‐based approach to the deployment of computable guidelines with explainable argumentation, integrated with a commercial electronic health record (EHR) system in Serbia, a middle‐income country in West Balkans.MethodsWe used an ontological framework, the Transition‐based Medical Recommendation (TMR) model, to represent, and reason about, guideline concepts, and chose the 2017 International global initiative for chronic obstructive lung disease (GOLD) guideline and a Serbian hospital as the deployment and evaluation site, respectively. To mitigate potential guideline conflicts, we used a TMR‐based implementation of the Assumptions‐Based Argumentation framework extended with preferences and Goals (ABA+G). Remote EHR integration of computable guidelines was via a microservice architecture based on HL7 FHIR and CDS Hooks. A prototype integration was developed to manage chronic obstructive pulmonary disease (COPD) with comorbid cardiovascular or chronic kidney diseases, and a mixed‐methods evaluation was conducted with 20 simulated cases and five pulmonologists.ResultsPulmonologists agreed 97% of the time with the GOLD‐based COPD symptom severity assessment assigned to each patient by the CDSS, and 98% of the time with one of the proposed COPD care plans. Comments were favourable on the principles of explainable argumentation; inclusion of additional co‐morbidities was suggested in the future along with customisation of the level of explanation with expertise.ConclusionAn ontological model provided a flexible means of providing argumentation and explainable artificial intelligence for a long‐term condition. Extension to other guidelines and multiple co‐morbidities is needed to test the approach further.

Original language	English
Article number	e10391
Number of pages	0
Journal	Learning Health Systems
Volume	8
Issue number	2
Early online date	12 Sept 2023
DOIs	https://doi.org/10.1002/lrh2.10391
Publication status	Published - Apr 2024

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1002/lrh2.10391

Cite this

Domínguez, J., Prociuk, D., Marović, B., Čyras, K., Cocarascu, O., Ruiz, F., Mi, E., Mi, E., Ramtale, C., Rago, A., Darzi, A., Toni, F., Curcin, V., & Delaney, B. (2024). ROAD2H: Development and evaluation of an open‐source explainable artificial intelligence approach for managing co‐morbidity and clinical guidelines. Learning Health Systems, 8(2), Article e10391. https://doi.org/10.1002/lrh2.10391

@article{472f137a2b6041069d11288248c748ef,

title = "ROAD2H: Development and evaluation of an open‐source explainable artificial intelligence approach for managing co‐morbidity and clinical guidelines",

abstract = "AbstractIntroductionClinical decision support (CDS) systems (CDSSs) that integrate clinical guidelines need to reflect real‐world co‐morbidity. In patient‐specific clinical contexts, transparent recommendations that allow for contraindications and other conflicts arising from co‐morbidity are a requirement. In this work, we develop and evaluate a non‐proprietary, standards‐based approach to the deployment of computable guidelines with explainable argumentation, integrated with a commercial electronic health record (EHR) system in Serbia, a middle‐income country in West Balkans.MethodsWe used an ontological framework, the Transition‐based Medical Recommendation (TMR) model, to represent, and reason about, guideline concepts, and chose the 2017 International global initiative for chronic obstructive lung disease (GOLD) guideline and a Serbian hospital as the deployment and evaluation site, respectively. To mitigate potential guideline conflicts, we used a TMR‐based implementation of the Assumptions‐Based Argumentation framework extended with preferences and Goals (ABA+G). Remote EHR integration of computable guidelines was via a microservice architecture based on HL7 FHIR and CDS Hooks. A prototype integration was developed to manage chronic obstructive pulmonary disease (COPD) with comorbid cardiovascular or chronic kidney diseases, and a mixed‐methods evaluation was conducted with 20 simulated cases and five pulmonologists.ResultsPulmonologists agreed 97% of the time with the GOLD‐based COPD symptom severity assessment assigned to each patient by the CDSS, and 98% of the time with one of the proposed COPD care plans. Comments were favourable on the principles of explainable argumentation; inclusion of additional co‐morbidities was suggested in the future along with customisation of the level of explanation with expertise.ConclusionAn ontological model provided a flexible means of providing argumentation and explainable artificial intelligence for a long‐term condition. Extension to other guidelines and multiple co‐morbidities is needed to test the approach further.",

author = "J Dom{\'i}nguez and D Prociuk and B Marovi{\'c} and K {\v C}yras and O Cocarascu and F Ruiz and E Mi and E Mi and C Ramtale and A Rago and A Darzi and F Toni and V Curcin and B Delaney",

year = "2024",

month = apr,

doi = "10.1002/lrh2.10391",

language = "English",

volume = "8",

journal = "Learning Health Systems",

issn = "2379-6146",

publisher = "John Wiley & Sons Inc.",

number = "2",

}

Domínguez, J, Prociuk, D, Marović, B, Čyras, K, Cocarascu, O, Ruiz, F, Mi, E, Mi, E, Ramtale, C, Rago, A, Darzi, A, Toni, F, Curcin, V & Delaney, B 2024, 'ROAD2H: Development and evaluation of an open‐source explainable artificial intelligence approach for managing co‐morbidity and clinical guidelines', Learning Health Systems, vol. 8, no. 2, e10391. https://doi.org/10.1002/lrh2.10391

TY - JOUR

T1 - ROAD2H: Development and evaluation of an open‐source explainable artificial intelligence approach for managing co‐morbidity and clinical guidelines

AU - Domínguez, J

AU - Prociuk, D

AU - Marović, B

AU - Čyras, K

AU - Cocarascu, O

AU - Ruiz, F

AU - Mi, E

AU - Ramtale, C

AU - Rago, A

AU - Darzi, A

AU - Toni, F

AU - Curcin, V

AU - Delaney, B

PY - 2024/4

Y1 - 2024/4

N2 - AbstractIntroductionClinical decision support (CDS) systems (CDSSs) that integrate clinical guidelines need to reflect real‐world co‐morbidity. In patient‐specific clinical contexts, transparent recommendations that allow for contraindications and other conflicts arising from co‐morbidity are a requirement. In this work, we develop and evaluate a non‐proprietary, standards‐based approach to the deployment of computable guidelines with explainable argumentation, integrated with a commercial electronic health record (EHR) system in Serbia, a middle‐income country in West Balkans.MethodsWe used an ontological framework, the Transition‐based Medical Recommendation (TMR) model, to represent, and reason about, guideline concepts, and chose the 2017 International global initiative for chronic obstructive lung disease (GOLD) guideline and a Serbian hospital as the deployment and evaluation site, respectively. To mitigate potential guideline conflicts, we used a TMR‐based implementation of the Assumptions‐Based Argumentation framework extended with preferences and Goals (ABA+G). Remote EHR integration of computable guidelines was via a microservice architecture based on HL7 FHIR and CDS Hooks. A prototype integration was developed to manage chronic obstructive pulmonary disease (COPD) with comorbid cardiovascular or chronic kidney diseases, and a mixed‐methods evaluation was conducted with 20 simulated cases and five pulmonologists.ResultsPulmonologists agreed 97% of the time with the GOLD‐based COPD symptom severity assessment assigned to each patient by the CDSS, and 98% of the time with one of the proposed COPD care plans. Comments were favourable on the principles of explainable argumentation; inclusion of additional co‐morbidities was suggested in the future along with customisation of the level of explanation with expertise.ConclusionAn ontological model provided a flexible means of providing argumentation and explainable artificial intelligence for a long‐term condition. Extension to other guidelines and multiple co‐morbidities is needed to test the approach further.

AB - AbstractIntroductionClinical decision support (CDS) systems (CDSSs) that integrate clinical guidelines need to reflect real‐world co‐morbidity. In patient‐specific clinical contexts, transparent recommendations that allow for contraindications and other conflicts arising from co‐morbidity are a requirement. In this work, we develop and evaluate a non‐proprietary, standards‐based approach to the deployment of computable guidelines with explainable argumentation, integrated with a commercial electronic health record (EHR) system in Serbia, a middle‐income country in West Balkans.MethodsWe used an ontological framework, the Transition‐based Medical Recommendation (TMR) model, to represent, and reason about, guideline concepts, and chose the 2017 International global initiative for chronic obstructive lung disease (GOLD) guideline and a Serbian hospital as the deployment and evaluation site, respectively. To mitigate potential guideline conflicts, we used a TMR‐based implementation of the Assumptions‐Based Argumentation framework extended with preferences and Goals (ABA+G). Remote EHR integration of computable guidelines was via a microservice architecture based on HL7 FHIR and CDS Hooks. A prototype integration was developed to manage chronic obstructive pulmonary disease (COPD) with comorbid cardiovascular or chronic kidney diseases, and a mixed‐methods evaluation was conducted with 20 simulated cases and five pulmonologists.ResultsPulmonologists agreed 97% of the time with the GOLD‐based COPD symptom severity assessment assigned to each patient by the CDSS, and 98% of the time with one of the proposed COPD care plans. Comments were favourable on the principles of explainable argumentation; inclusion of additional co‐morbidities was suggested in the future along with customisation of the level of explanation with expertise.ConclusionAn ontological model provided a flexible means of providing argumentation and explainable artificial intelligence for a long‐term condition. Extension to other guidelines and multiple co‐morbidities is needed to test the approach further.

U2 - 10.1002/lrh2.10391

DO - 10.1002/lrh2.10391

M3 - Article

SN - 2379-6146

VL - 8

JO - Learning Health Systems

JF - Learning Health Systems

IS - 2

M1 - e10391

ER -

ROAD2H: Development and evaluation of an open‐source explainable artificial intelligence approach for managing co‐morbidity and clinical guidelines

Abstract

UN SDGs

Access to Document

Fingerprint

Cite this