Premium
Interobserver agreement of N eer and AO classifications for proximal humeral fractures
Author(s) -
Papakonstantinou Maritsa K.,
Hart Melissa J.,
Farrugia Richard,
Gabbe Belinda J.,
Kamali Moaveni Afshin,
van Bavel Dirk,
Page Richard S.,
Richardson Martin D.
Publication year - 2016
Publication title -
anz journal of surgery
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.426
H-Index - 70
eISSN - 1445-2197
pISSN - 1445-1433
DOI - 10.1111/ans.13451
Subject(s) - medicine , kappa , classification scheme , machine learning , mathematics , computer science , geometry
Background The classification of proximal humeral fractures remains challenging. The two main classification systems used, the N eer and the AO classification, have both been shown to have less than ideal interobserver agreement. Agreement in classification is required, however, to guide fracture management. Method Data from the V ictorian O rthopaedic T rauma O utcomes R egistry were collected and the X‐rays of 104 proximal humeral fractures were reviewed by three orthopaedic consultants. They classified the fractures according to the N eer and AO classifications, as well as their simplified versions. Interobserver agreement was then assessed using kappa statistics. Results Interobserver agreement was better overall in the N eer classification, which was moderate (kappa = 0.40–0.58), than the AO classification, which was fair to moderate (kappa = 0.31–0.54). When simplified, the N eer and AO classification interobserver agreement remained similar. Conclusion The classification of proximal humeral fractures with both the N eer and the AO systems remains difficult with minimal improvements seen when reducing the number of categories in each classification system. From these results, the N eer classification system would appear slightly more useful in clinical practice to guide treatment.