Revising the modified Neer classification for distal clavicle fractures: Description and reliability

Abstract

BackgroundThe modified Neer classification is the most widely used classification system for distal clavicle fractures. However, it provides limited information for treatment decisions. The objective of this study was to revise the modified Neer classification to make it more suitable for treatment decision-making. HypothesisThe revised version of the modified Neer classification has good intra- and interobserver agreements and provides an instructive treating algorithm. Study designCohort study MethodsSix observers, including three experienced shoulder specialists and three junior orthopaedic residents, independently reviewed plain radiographs of 52 patients with distal clavicle fractures. They were asked to classify the fracture types according to the modified Neer classification and our revised new classification separately to determine treatment approaches for each patient. Images were mirrored and randomized to verify the intraobserver agreement. Reliabilities were measured using the Fleiss kappa values. ResultsBoth the modified Neer classification and our revised version had near perfect intraobserver agreement (κ values: 0.87–1.00), whereas our revised Neer classification had a better interobserver agreement (κ values: 0.78 vs. 0.70, z = 4.70, p < 0.01) and stronger relevance to treatment decisions (coefficient of contingency: 0.70 vs. 0.44). ConclusionOur study demonstrated a near-perfect intraobserver and substantial interobserver agreement of the revised new classification, indicating that our revised new classification was better than the modified Neer classification. Meanwhile, our revised classification brought few disputes in treatment selection. Clinical relevanceThe modified Neer classification was revised to make it more accurate and suitable for guiding treatment. Type of studyStudy of diagnostic test Level of studyLevel II