A machine learning approach reveals features related to clinicians' diagnosis of clinically relevant knee osteoarthritis

Qiuke Wang, Jos Runhaar, Margreet Kloppenburg, Maarten Boers, Johannes W J Bijlsma, Jaume Bacardit, Sita M A Bierma-Zeinstra, the CREDO expert group

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

OBJECTIVES: To identify highly ranked features related to clinicians' diagnosis of clinically relevant knee OA. METHODS: General practitioners (GPs) and secondary care physicians (SPs) were recruited to evaluate 5-10 years follow-up clinical and radiographic data of knees from the CHECK cohort for the presence of clinically relevant OA. GPs and SPs were gathered in pairs; each pair consisted of one GP and one SP, and the paired clinicians independently evaluated the same subset of knees. A diagnosis was made for each knee by the GP and SP before and after viewing radiographic data. Nested 5-fold cross-validation enhanced random forest models were built to identify the top 10 features related to the diagnosis. RESULTS: Seventeen clinician pairs evaluated 1106 knees with 139 clinical and 36 radiographic features. GPs diagnosed clinically relevant OA in 42% and 43% knees, before and after viewing radiographic data, respectively. SPs diagnosed in 43% and 51% knees, respectively. Models containing top 10 features had good performance for explaining clinicians' diagnosis with area under the curve ranging from 0.76-0.83. Before viewing radiographic data, quantitative symptomatic features (i.e. WOMAC scores) were the most important ones related to the diagnosis of both GPs and SPs; after viewing radiographic data, radiographic features appeared in the top lists for both, but seemed to be more important for SPs than GPs. CONCLUSIONS: Random forest models presented good performance in explaining clinicians' diagnosis, which helped to reveal typical features of patients recognized as clinically relevant knee OA by clinicians from two different care settings.

Original languageEnglish
Pages (from-to)2732-2739
Number of pages8
JournalRheumatology (Oxford, England)
Volume62
Issue number8
Early online date19 Dec 2022
DOIs
Publication statusPublished - 1 Aug 2023

Bibliographical note

Funding Information:
Disclosure statement: J.R. and M.K. received research grants from the Dutch Arthritis Society; M.K. reports fee for consultancy (Abbvie, Pfizer, Levicept, GlaxoSmithKline, Merck-Serono, Kiniksa, Flexion, Galapagos, Jansen, CHDR, Novartis, UCB) and local investigator of industry-driven trial (Abbvie), from Wolters Kluwer (UptoDate), Springer Verlag (Reumatologie en klinische immunologie), board member for OARSI, president of the Dutch Society Rheumatology and member of the EULAR Council. Acknowledgements

Funding Information:
This work was supported by the Dutch Arthritis Society (Project ID 15–1-301); Q.W. was financed by China Scholarship Council (CSC) (grant number: 201906230308).

Publisher Copyright:
© 2022 The Author(s). Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved.

Fingerprint

Dive into the research topics of 'A machine learning approach reveals features related to clinicians' diagnosis of clinically relevant knee osteoarthritis'. Together they form a unique fingerprint.

Cite this