Myrthe A. van den Berg MSc, Fleur Boel MSc, Michiel M.A. van Buuren MD, Noortje S. Riedstra MD PhD, Jinchi Tang MD PhD, Harbeer Ahedi PhD, Nigel K. Arden MD, Sita M.A. Bierma-Zeinstra PhD, Cindy G. Boer PhD, Flavia M. Cicuttini PhD, Timothy F. Cootes PhD, David T. Felson MD, Willem Paul Gielis MD PhD, Joshua Heerey PhD, Graeme Jones MD, Stefan Kluzek PhD, Nancy E. Lane MD, Claudia Lindner PhD, Joyce B.J. van Meurs PhD, Andrea Mosler PhD, Amanda E. Nelson MD, Michael C. Nevitt PhD, Edwin H. Oei PhD, Jos Runhaar PhD, Harrie Weinans PhD, Jesse H. Krijthe PhD, Rintje Agricola MD PhD
This study aims to develop hip morphology-based radiographic hip osteoarthritis (RHOA) risk prediction models and investigates the added predictive value of hip morphology measurements and the generalizability to different populations.
We combined data from nine prospective cohort studies participating in the World COACH consortium. RHOA grades were harmonized, and incident RHOA was defined as hips without definite RHOA at baseline that developed definite RHOA within 4-8 years. Baseline hip morphology was quantified with automatically and uniformly determined lateral center edge angle and alpha angle measurements on anteroposterior radiographs. Discriminative performance of generalized linear mixed model (GLMM) definitions with and without hip morphology measurements was determined with stratified cross-validation. With leave-one-cohort-out cross-validation, the generalizability to unseen populations of hip morphology-based GLMMs and Random Forest (RF) models was evaluated.
From the included 35,984 hips without definite RHOA at baseline, 4.7% developed incident RHOA within 4-8 years. The GLMM with cohort-specific intercept, considering baseline demographics, RHOA grade, and hip morphology measurements, showed a mean AUC of 0.80 (±0.01) in stratified cross-validation. Using a marginal intercept decreased performance by 0.1 in AUC. Similar results were found for a GLMM without hip morphology measurements. Leave-one-cohort-out cross-validation showed comparable discrimination (AUC between 0.56-0.88) and calibration performance for hip morphology-based GLMMs and RF models.
In hips free of definite RHOA, our AUCs for the incident RHOA models showed good predictive performance in similar populations. However, the added predictive value of the morphology measurements was small, and model performance was heterogeneous in leave-one-cohort-out cross-validation.
Read full text: https://doi.org/10.1002/acr.25629
