Uniform and reliable assessment of bone union on radiographs in osteotomies around the knee: a novel classification system

Abstract

BACKGROUND: To evaluate the inter- and intra-rater reliability of the Unified Bone Union (UBU) classification for assessing time-dependent bone healing on radiographs in osteotomies around the knee, including negative union signs. Secondary aims included assessing union progression over time, applicability across osteotomy types, and correlation between radiographic and CT-based UBU scores. METHODS: The UBU classification assesses bone healing on anteroposterior radiographs in three anatomical zones, graded from phase 0 (no callus) to phase 3 (bridging callus), including radiological negative union signs. Radiographs (n = 110) from 38 medial opening-wedge high tibial osteotomy patients were retrospectively reviewed twice by three independent raters. Inter- and intra-rater reliability were assessed using quadratic-weighted kappa (κ). Percent agreement was calculated for classification modifiers. Time-dependent changes in union were analyzed. Reliability was also tested across osteotomy types. Correlation between 6-month radiographic and CT-based UBU scores was determined using Spearman's rho. RESULTS: Interrater reliability was substantial (κ 0.74-0.79), while intra-rater reliability showed almost perfect agreement (κ 0.78-0.98). Modifier agreement was good (inter-rater: 91-98 %; intra-rater: 89-95 %). The UBU score increased over time. The UBU showed substantial interrater reliability (κ = 0.75) across various osteotomy types. A strong correlation was found between radiographic and CT-based UBU scores (r = 0.82, p < 0.01). CONCLUSION: The UBU classification provides a reliable and standardized method for evaluating bone union after osteotomies around the knee. It incorporates negative union signs and demonstrates strong inter- and intra-rater agreement, as well as strong correlation with CT imaging. Further research should validate its diagnostic accuracy and clinical utility.