Reproducibility of the Mayo and Schatzker classification systems in proximal ulna fractures - 14/12/23

Doi : 10.1016/j.otsr.2023.103790

Fabio Carminati ⁎ , Laurent Obert, François Saade, Camille Bouteille, Edouard Woussen, Zouhair Aouzal, Michaël Bourgeois, Harrison Haight, Inès Regas, Séverin Rochet, Daniel Lepage, Patrick Garbuio, François Loisel
Service de chirurgie orthopédique et traumatologique, chirurgie plastique, esthétique et reconstructrice, chirurgie de la main, hôpital Jean-Minjoz, 25030 Besançon, France

^⁎Corresponding author.

Sous presse. Épreuves corrigées par l'auteur. Disponible en ligne depuis le Thursday 14 December 2023
Cet article a été publié dans un numéro de la revue, cliquez ici pour y accéder

Abstract

Introduction

A fracture classification system should provide a reliable and reproducible means of communication between different parties. It should be logical and understandable, with few categories to memorize. The aim of this study was to determine the intra- and interobserver reliability of the Schatzker and Mayo classification systems for the assessment of proximal ulna fractures.

Materials and methods

Intra- and interobserver reliability studies were conducted on 39 X-rays of injured elbows drawn randomly from 74 cases previously used in a series on predictors of ulnohumeral osteoarthritis in proximal ulna fractures. Ten observers independently reviewed these X-rays on 2 separate occasions 3 months apart. The fracture type was assessed according to the Schatzker and Mayo classification systems during each reading session. Cohen's and Fleiss’ kappa were used to measure the intra- and interobserver reliability.

Results

The Schatzker classification had a fair interobserver reliability for the first (Schatzker R₁, Fleiss’ κ: 0.394) and second (Schatzker R₂, Fleiss’ κ: 0.351) readings. The mean intraobserver reliability value between the 10 reviewers for the Schatzker classification was rated as substantial (0.61). The Mayo classification had a fair interobserver reliability for the first (Mayo R₁, Fleiss’ κ: 0.278) and second (Mayo R₂, Fleiss’ κ: 0.292) readings. The mean intraobserver reliability value between the 10 reviewers for the Mayo classification was rated as fair (0.52).

Discussion

The classification systems for proximal ulna fractures showed poor reproducibility between the different observers since they had low interobserver agreement values. Nevertheless, their use remained reliable since the measured intraobserver agreement value was deemed substantial.