Reliability of the Masada Classification for Forearm Involvement in Patients with Hereditary Multiple Osteochondromas (HMO)

Background: Classifications describing forearm lesions in patients with Hereditary Multiple Osteochondromatosis (HMO) have been used to recommend surgical intervention and stratify outcomes; however, there is no consensus on which classification offers greater reliability. The purpose of this study was to determine the reliability of the Masada classification and newer classifications among pediatric hand surgeons. Methods: One hundred one patients with HMO between June 2014 and October 2019 were enrolled in the Congenital Upper Limb Differences (CoULD) Registry. Of those, 67 patients with 101 forearms were included. Four pediatric hand surgeons from the CoULD study group undertook an online evaluation. Each rater classified radiographs according to the Masada classification. Six weeks later, raters were asked to reclassify images according to the Masada, Gottschalk, and Jo classifications. Rater agreement for these classifications was assessed by estimating Fleiss kappa along with a 95% CI. Results: Interrater agreement for Masada classification after the first reading was poor (κ=0.35; 95% CI=0.30-0.41) across all raters. Interrater agreement across the 4 raters decreased for the Masada classification from the first to the second reading (κ=0.35 vs 0.21; P<0.001). Intrarater agreement for the Masada classification ranged from 0.32 to 0.63 from the first to the second study reading. Gottschalk and Jo classifications yielded significantly better interrater agreement compared with Masada (κ=0.43 vs 0.21; P<0.001). Unclassifiable cases were highest in the Masada classification (34% to 44%) and lower in the Jo (17%) and Gottschalk (14%) classifications. Conclusion: Despite wide use, the Masada classification was found to have low reliability when classifying forearm deformities in HMO. Gottschalk offered more options for location, yet lacked deformity description including radial head dislocation. Jo classification offered more locations than Masada and incorporated radial head dislocation in some patterns. Based on the shortcomings in all 3 classification systems, the development of a more inclusive and reliable classification is warranted. Level of Evidence: Level II; Diagnostic.