Mathijs P Hendriks1, Xander A A M Verbeek2, Thijs van Vegchel2, Maurice J C van der Sangen3, Luc J A Strobbe4, Jos W S Merkus5, Harmien M Zonderland6, Carolien H Smorenburg7, Agnes Jager8, Sabine S Siesling2,9. 1. Northwest Clinics, Alkmaar, the Netherlands. 2. Netherlands Comprehensive Cancer Organization, Utrecht, the Netherlands. 3. Catharina Hospital, Eindhoven, the Netherlands. 4. Canisius Wilhelmina Hospital, Nijmegen, the Netherlands. 5. Haga Teaching Hospital, The Hague, the Netherlands. 6. University of Amsterdam, the Netherlands. 7. Netherlands Cancer Institute, Amsterdam, the Netherlands. 8. Erasmus MC Cancer Institute, Rotterdam, the Netherlands. 9. University of Twente, Enschede, the Netherlands.
Abstract
PURPOSE: The essence of guideline recommendations often is intertwined in large texts. This impedes clinical implementation and evaluation and delays timely modular revisions needed to deal with an ever-growing amount of knowledge and application of personalized medicine. The aim of this project was to model guideline recommendations as data-driven clinical decision trees (CDTs) that are clinically interpretable and suitable for implementation in decision support systems. METHODS: All recommendations of the Dutch national breast cancer guideline for nonmetastatic breast cancer were translated into CDTs. CDTs were constructed by nodes, branches, and leaves that represent data items (patient and tumor characteristics [eg, T stage]), data item values (eg, T2 or less), and recommendations (eg, chemotherapy), respectively. For all data items, source of origin was identified (eg, pathology), and where applicable, data item values were defined on the basis of existing classification and coding systems (eg, TNM, Breast Imaging Reporting and Data System, Systematized Nomenclature of Medicine). All unique routes through all CDTs were counted to measure the degree of data-based personalization of recommendations. RESULTS: In total, 60 CDTs were necessary to cover the whole guideline and were driven by 114 data items. Data items originated from pathology (49%), radiology (27%), clinical (12%), and multidisciplinary team (12%) reports. Of all data items, 101 (89%) could be classified by existing classification and coding systems. All 60 CDTs could be integrated in an interactive decision support app that contained 376 unique patient subpopulations. CONCLUSION: By defining data items unambiguously and unequivocally and coding them to an international coding system, it was possible to present a complex guideline as systematically constructed modular data-driven CDTs that are clinically interpretable and accessible in a decision support app.
PURPOSE: The essence of guideline recommendations often is intertwined in large texts. This impedes clinical implementation and evaluation and delays timely modular revisions needed to deal with an ever-growing amount of knowledge and application of personalized medicine. The aim of this project was to model guideline recommendations as data-driven clinical decision trees (CDTs) that are clinically interpretable and suitable for implementation in decision support systems. METHODS: All recommendations of the Dutch national breast cancer guideline for nonmetastatic breast cancer were translated into CDTs. CDTs were constructed by nodes, branches, and leaves that represent data items (patient and tumor characteristics [eg, T stage]), data item values (eg, T2 or less), and recommendations (eg, chemotherapy), respectively. For all data items, source of origin was identified (eg, pathology), and where applicable, data item values were defined on the basis of existing classification and coding systems (eg, TNM, Breast Imaging Reporting and Data System, Systematized Nomenclature of Medicine). All unique routes through all CDTs were counted to measure the degree of data-based personalization of recommendations. RESULTS: In total, 60 CDTs were necessary to cover the whole guideline and were driven by 114 data items. Data items originated from pathology (49%), radiology (27%), clinical (12%), and multidisciplinary team (12%) reports. Of all data items, 101 (89%) could be classified by existing classification and coding systems. All 60 CDTs could be integrated in an interactive decision support app that contained 376 unique patient subpopulations. CONCLUSION: By defining data items unambiguously and unequivocally and coding them to an international coding system, it was possible to present a complex guideline as systematically constructed modular data-driven CDTs that are clinically interpretable and accessible in a decision support app.
Authors: Ryan P Merkow; Deborah Korenstein; Rubaya Yeahia; Peter B Bach; Shrujal S Baxi Journal: JAMA Intern Med Date: 2017-05-01 Impact factor: 21.873
Authors: Elvira L Vos; Sabine Siesling; Margreet H A Baaijens; Cornelis Verhoef; Agnes Jager; Adri C Voogd; Linetta B Koppert Journal: Breast Cancer Res Treat Date: 2017-04-07 Impact factor: 4.872
Authors: Lotte Keikes; Stephanie Medlock; Daniel J van de Berg; Shuxin Zhang; Onno R Guicherit; Cornelis J A Punt; Martijn G H van Oijen Journal: J Clin Transl Res Date: 2018-07-27
Authors: Mathijs P Hendriks; Xander A A M Verbeek; Jeannette G van Manen; Sannah E van der Heijden; Shirley H L Go; Gea A Gooiker; Thijs van Vegchel; Sabine Siesling; Agnes Jager Journal: Breast Cancer Res Treat Date: 2020-07-06 Impact factor: 4.872
Authors: Kees C W J Ebben; Melle S Sieswerda; Ernest J T Luiten; Joan B Heijns; Carmen C van der Pol; Maud Bessems; Aafke H Honkoop; Mathijs P Hendriks; Janneke Verloop; Xander A A M Verbeek Journal: JCO Clin Cancer Inform Date: 2020-04
Authors: Lotte Keikes; Milan Kos; Xander A A M Verbeek; Thijs Van Vegchel; Iris D Nagtegaal; Max J Lahaye; Alejandra Méndez Romero; Sandra De Bruijn; Henk M W Verheul; Heidi Rütten; Cornelis J A Punt; Pieter J Tanis; Martijn G H Van Oijen Journal: Int J Qual Health Care Date: 2021-04-03 Impact factor: 2.038
Authors: Kees C W J Ebben; Mathijs P Hendriks; Lieke Markus; Milan Kos; Ignace H J T De Hingh; Jorg R Oddens; Joost Rothbarth; Hans De Wilt; Luc J A Strobbe; Maud Bessems; Carsten T Mellema; Sabine Siesling; Xander A A M Verbeek Journal: Int J Qual Health Care Date: 2022-03-19 Impact factor: 2.038